THE WASSERMANN TEST THE WASSEBMANN TEST BY CHARLES F. CRAIG, M.D., M.A., F.A.C.S. LIEUTENANT COLONEL, MEDICAL CORPS, UNITED STATES ARMY; PROFESSOR OF BACTERIOLOGY, PARASITOLOGY AND PREVENTIVE MEDICINE, AND DIRECTOR OF LABORATORIES, ARMY MEDICAL SCHOOL, WASHINGTON, D. C. FORMERLY CURATOR, ARMY MEDICAL MUSEUM, AND COMMAND- ING OFFICER, YALE ARMY LABORATORY SCHOOL SECOND EDITION Revised and Enlarged ILLUSTRATED WITH COLORED PLATES, HALFTONE PLATES, AND SIXTY-ONE TABLES ST. LOUIS C. V. MOSBY COMPANY 1921 Copyright, 1918, 1921, By C. V. Mosby Company Press of C. V. Mosby Company St. Louis PREFACE TO SECOND EDITION In preparing the second edition of this work, it has been the writer's aim to incorporate everything of real value relating to the subject that has been published since the appearance of the first edition, and it is believed that in the new edition will be found much additional matter which will prove of great value to the laboratory worker and prac- titioner of medicine. In order to include so much new mate- rial it has been necessary to practically rewrite portions of the work and thanks are due the publishers for their en- thusiastic assistance in making it possible to present the work in its present form, a not altogether easy task with the present high cost of production and shortage of print paper. It will be noted that several important changes have been made in the technic of the test, as originally published by the writer, the most important being the addition of blood grouping in the directions concerning the preparation of the blood cell suspension used in the test. The importance of selecting donors for the blood cell suspension from in- dividuals belonging to Group IV, of the Moss classification, first thoroughly demonstrated by Williams, can, in the writer's opinion, hardly be overestimated, and the adoption of this procedure adds greatly to the accuracy of the technic when the antihuman hemolytic system is employed, and obviates many of the objections which have been raised, from time to time, to the use of such a system. The writer is greatly pleased that a second edition of this work should have been called for within two years of its appearance, as it is thus evident that it has proved of serv- ice to the medical profession, and it is his earnest hope that this edition will prove still more useful to both the lab- oratory worker and the general practitioner. 9 10 PREFACE Much of the data appearing in the work has been ob- tained from the writer's experience with the complement- fixation test for syphilis in army laboratories, and are pub- lished with the permission of the Surgeon General of the Army, but it should be understood that he is in no way re- sponsible for any of the opinions expressed, or the conclu- sions drawn, from these data. In this edition there is included, at the request of Dr. A. J. Pacinni, Chief of Laboratory Section, Public Health Service, the modification adopted by the Hygienic Labo- ratory of that Service in performing the complement-fixa- tion test for syphilis. This method, as well as the writer's modification, is authorized for use in the hospitals con- trolled by the United States Public Health Service. In conclusion, the writer regrets that it seemed necessary to retain the name "Wassermann Test" in describing the complement-fixation test for syphilis, but this name has be- come so fixed in the literature and in the minds of the med- ical profession and the layman, that it has seemed unwise to change it. In our present state of ignorance as to the exact nature of the reaction, even the term "complement fixa- tion" may be an entirely erroneous conception, as applied to this very important test, but in this work the terms "Wassermann test or reaction" and "complement-fixation test or reaction" are synonymous, and both are used in the discussion of various phases of the subject. Charles F. Craig. Army Medical School Washington, D. C. January 1, 1921 PREFACE TO FIRST EDITION The following work upon the Wassermann test has been written at the solicitation of friends who have urged me to place upon record the experience gained from nearly ten years of personal use of this test in the diagnosis and as a control of the treatment of syphilis. I have tried to include in the work all of the most important data that have been published regarding this very valuable diagnostic measure, and the conclusions that I feel may rightly be drawn from its use in the diagnosis, control of treatment, and as an in- dex of the prevalence of this disease. The work has been largely prepared since the outbreak of the present war and, for this reason, is not as exhaus- tive as I had originally intended it to be, as owing to official duties it has been impossible to spend as much time in its preparation as would have been necessary to make it an ex- haustive treatise, and it has also been impossible for me to consult much of the very extensive literature that has accumulated during recent years in regard to the test. However, it is believed that the work contains all of the essential and really valuable facts regarding the test which have been reported in the literature, and if there have been any omissions I would deem it a favor to have them called to my attention. I have quoted quite liberally from some of the more re- cent investigators, as Noguchi, Nichols, Vedder, and Kol- mer, and have also used much data previously published by myself in various medical journals, and it is a pleasure to tender my thanks to the editors of the Journal of the Amer- ican Medical Association, the Journal of Experimental Medicine, the Journal of Infectious Diseases, and the Amer- ican Journal of Syphilis, for permission to avail myself of the data previously published in these journals. 11 12 PREFACE From personal experience, I believe that there is still a great deal of misunderstanding and confusion among the members of the medical profession regarding the exact value and limitations of the Wassermann test, both in the diagnosis of syphilis, and when used as a control of the treatment of the disease, and if this work will help in clear- ing up this confusion it will be a source of great gratifica- tion. Much of this misunderstanding rests upon the shoul- ders of laboratory workers, for it must be admitted that too often the performance of the Wassermann test has been delegated to poorly trained or careless assistants, and the results obtained with the test have thus been erroneous and unsatisfactory. I can not urge too strongly upon the pro- fession the necessity for submitting material for this test to well-qualified serologists if reliable results are to be ob- tained. A standard technic for the test is much to be de- sired but all efforts in this direction have failed, owing largely to the difficulty of securing a standard antigen, so that at the present time several methods of performing the test are in use, all of which are reliable in the hands of experienced serologists. The method recommended in this work has stood the test of time and has been used by many different workers in thousands of syphilitic infections, and it is believed that it is as simple in technic and as accurate in results as any method of performing the Wassermann test that has been devised. My thanks are due the publishers for their courtesy and assistance in many ways and for the very excellent manner in which they have reproduced the color plates pub- lished in the work. Charles F. Craig. Department Laboratory, Central Department, U. S. Army, Fort Leavenworth, Kansas. June 1, 1918. CONTENTS CHAPTER I PAGE The Discovery of the Wassermann Test 17 CHAPTER II General Description of the Wassermann Test. Nature of the Reaction 30 CHAPTER III Factors Which Influence the1 Result of the Test 40 CHAPTER IV Preparation and Titration of the Reagents Used in the Wasser- mann Test 61 CHAPTER V The Technic of the Writer's Modification of the Wassermann Test. The Test Upon the Cerebrospinal Fluid. The Origi- nal Wassermann Technic. Other Modifications of the Wassermann Test 99 CHAPTER VI Complement Fixation in Syphilis with Antigens Prepared from Pure Cultures of Treponema Pallidum 146 CHAPTER VII The Results of the Wassermann Test in the Various Stages of Syphilis. Time of Appearance of the Reaction. The Spec- ificity of the Wassermann Test 157 CHAPTER VIII The Interpretation of the Results of the Wassermann Test. The Wassermann Test as an Index of the Prevalence of Syph- ilis in Communities 184 13 14 CONTENTS CHAPTER IX The Effect of Treatment upon the Wassermann Reaction. The Wassermann Test as a Control of the Treatment of Syph- ilis. The Provocative Wassermann Reaction 205 CHAPTER X The Wassermann Test upon the Cerebrospinal Fluid. Interpreta- tion of Results. The Colloidal Gold Test. The Cell Count and Tests for Increase in Globulins in the Cerebro- spinal Fluid 242 BIBLIOGRAPHY Bibliography 269 ILLUSTRATIONS PAGE Plate I. (1) Titration of complement. (2) Titration of amboceptor pa- pers. (3) Titration of antigen 88 Plate II. (1) Reading of results of Wassermann test. (2) Positive reac- tion. (3) Negative reaction 112 Plate III. Paretic reaction. Tabetic reaction 256 FIG. 1. Showing hemolytic and bacteriolytic systems 24 2. Schematic representation of complement fixation in syphilis or the Wassermann test' 32 3. Luer syringes 64 4. Electric centrifuge 70 5. Electric centrifuge with shaker head attached 76 6. Small incubator 82 7. Water-bath suitable for Wassermann test incubations 84 8. Amboceptor eutter. Front view, showing gauge 92 9. Amboceptor cutter. Rear view, showing scale and set screw . . 94 10.- Amboceptor markers 96 15 THE WASSERMANN TEST CHAPTER I THE DISCOVERY OF THE WASSERMANN TEST The discovery of the Wassermann test for the diagnosis of syphilis was the direct result of the study of immunity and especially of the discovery of certain facts of basic importance in that process; i. e., bacteriolysis, hemolysis, and complement fixation. Bacteriolysis In 1888, Nuttall1 demonstrated that freshly drawn blood serum and defibrinated blood had a marked destructive effect upon bacteria, an effect that could be definitely esti- mated, and which was undoubtedly due to some substance or substances in the blood serum and defibrinated blood. This effect consisted in the dissolving of the bacteria and the process was called bacteriolysis, the substance causing the destruction of the bacteria being known as a bacter- iolysin. The bacteriolytic effect of the blood serum or de- fibrinated blood could be demonstrated in the following manner: Definite amounts of the blood serum or defibrinated blood were mixed with definite amounts of the bacterium experi- mented with, and incubated at 37° C. for several hours, after which plates were poured and the number of colonies of bacteria developing compared with the number upon plates containing the same bacterium but to which no blood serum or defibrinated blood had been added. It was ob- 17 18 THE WASSERMANN TEST served that the plates prepared from the mixtures of bac- teria and blood serum and defibrinated blood showed very many less colonies of the bacterium than those to which no blood serum or defibrinated blood had been added, and, in some of the experiments, the plates were almost sterile. Nuttall's experiments were confirmed by von Fodor2 and Buchner,3 and it was the study and discussion of this sub- ject that introduced into the science of medicine the "hu- moral" theories regarding the nature of immunity. Nuttall had shown that the bactericidal power of blood serum was very unstable and that it was only present in fresh blood or blood serum, and that heating the serum at 55 to 56° C., or allowing it to stand for any considerable time destroyed its capacity for dissolving the bacteria. Buchner,4 in explaining the phenomenon of bacteriolysis, evolved what is now known as the humoral explanation, i. e., that the capacity of the blood or serum for dissolving the bacteria resided in a constituent of the fresh blood serum which he termed ' ' alexin; ' ' which he believed to be active against all bacteria; and which he compared to the action of a ferment. Buchner further stated that this "alexin" might be of cellular origin, either a product of the tissue cells or leucocytes. Alexin is identical with the "complement" of Ehrlich and his followers, and in this work the term "complement" will be used in preference to "alexin." In 1894, Pfeiffer,5 by his classical experiment upon chol- era immunity in the guinea pig, demonstrated that the process of bacteriolysis is a specific one and that the injec- tion of a specific bacterium into an animal is followed by the production of a specific bacteriolysin in the blood serum of that animal for the bacterium injected. His experiment, now a common classroom demonstration of bacteriolysis, was briefly as follow'S: If a guinea pig, which has recovered from cholera, be in- jected in the peritoneal cavity with a suspension containing cholera spirilla, it will be found, upon examination of the DISCOVERY 19 exudate from the peritoneal cavity of the animal, that the spirilla undergo rapid destruction, all stages of the process being visible in samples of the exudate taken at graduated intervals of time after the injection. The spirilla will be observed to become swollen, distorted in shape and gran- ular in appearance, and soon entirely disappear, being dis- solved in the peritoneal exudate. That this process is pro- tective in nature is shown by the fact that such an animal will survive doses of the spirilla that are lethal to animals that have not been protected by a previous attack of chol- era, and that it is specific by the fact that other bacteria, when injected into the peritoneal cavity of the cholera pig, are unaffected. In other words, the experiment proved that the pig, after recovery from cholera, had present in the peritoneal ex- udate some protective substance, specific in nature, which was capable of dissolving the cholera spirilla and incapable of acting upon other bacteria. This substance was the same as that demonstrated by Nuttall in fresh blood serum and defibrinated blood and to which the name "bacterio- lysin" had been given. A very significant phenomenon observed by Pfeiffer was that if the peritoneal cavity of a normal guinea pig was injected with a mixture of cholera spirilla and some blood serum from the immunized pig, the normal pig also escaped infection. That is, it was possible thus to transfer the bac- teriolytic power possessed by the peritoneal exudate of the immunized pig to the peritoneal cavity of an unprotected pig by mixing the cholera spirilla with the blood serum of the protected animal, thus proving that the bacteriolysin also existed in the blood serum. It was also found that while the bactericidal powTer of the blood serum of the im- munized animal was destroyed, in vitro, by heating, the heated blood serum is just as effectual in producing im- munity in a normal guinea pig when injected with the chol- era spirilla as the unheated. This fact Pfeiffer thought to be due to the necessity of some living tissue taking part 20 THE WASSERMANN TEST in the reaction and stated that the endothelial cells of the peritoneum were probably the cells concerned in this par- ticular experiment. His conclusions, in this respect, lead him to believe that bacteriolysis, so far as it occurred with the cholera spirillum, could only take place in the living body. The researches of Pfeiffer stimulated a great deal of re- search upon the subject and to Bordet we owe the true ex- planation of the process of bacteriolysis. Pfeiffer's con- ception that it could only occur in the living animal was disproved by Bordet6 who produced bacteriolysis of the cholera spirillum in hanging drop preparations, and showed, further, that while the bacteriolytic action of im- mune serum is destroyed by heating the serum at 56° C., the addition to this so-called ''inactivated'' serum of a small amount of fresh normal blood serum, will "reac- tivate" or restore the bacteriolytic property of the serum. Thus, while normal blood serum may contain no bacte- riolytic substances against the organism experimented with, when added to an immune serum which has been heated and its bacteriolytic activity destroyed, it has the power of again bringing the bacteriolytic properties of the immune serum into play and bacteriolysis results. From this experiment it is evident that in the unheated immune serum the dissolving of the bacterium must de- pend upon two substances, one rendered powerless by heat, while the other is unaffected. The substance rendered in- active by heat is said to be thermolabile, while the unaf- fected substance is said to be thermostable. It is also evi- dent that the one can not act without the other upon the bacterium, and that the thermolabile body is present in normal, unheated serum, while the thermostable body is only present in the immune serum. Bordet's researches also proved the specific nature of the bacteriolysins and that an animal immunized to several specific bacteria will produce in its blood specific bacteriolysins against the va- rious bacteria injected during the process of immunization. DISCOVERY 21 From the researches mentioned it is evident that the process of bacteriolysis depends upon the production in the blood of an animal of a specific protective substance which, in the presence of fresh normal blood serum, is ca- pable of dissolving the specific bacterium against which the animal has been immunized. For bacteriolysis to occur there must be present three factors; one, the bacterium used in immunization; the second, the blood serum contain- ing the protective substance produced by immunization; and third, the substance which is present in fresh normal blood serum. In immunological literature certain terms have been given the factors mentioned and in this work those used by Ehrlich and his followers will be used. Thus, the bac- terium injected in order to produce the bacteriolytic sub- stance is called an antigen, and this name may be applied to any substance, which when injected into an animal, results in the formation of an antibody. The bacteriolytic sub- stance, itself, is called an antibody or amboceptor; while the substance which is present in fresh normal blood serum and which, when brought into contact with the antigen and amboceptor, causes bacteriolysis is called complement. This substance is identical with Buchner's "alexin." Definitions.-As these terms are used, throughout this work, in describing the Wassermann reaction, the follow- ing definitions of them are inserted for convenience, as it is essential that they be understood if one desires to have a clear conception of the test: Antigen.-Any substance which, when injected into suit- able animals, will result in the formation of specific anti- bodies. Antigens may be bacterial, cellular, or chemical in nature. Amboceptor.-The specific antibody produced by the in- jection of an antigen. Amboceptors may be normally pres- ent in certain blood sera but can be greatly increased by the injection of the suitable antigen. Amboceptors are thermostable, retaining their activity after heating at 22 THE WASSERMANN TEST 56° C. for one half hour. A bacteriolysin is a bacterio- lytic amboceptor. A hemolysin is a hemolytic amboceptor. Complement.-The substance present in all fresh nor- mal sera, which when added to a mixture containing anti- gen and amboceptor, results in the production of bacterio- lysis or hemolysis. Complement is rendered inactive by heating for one half hour at 56° C. and is, therefore, ther- molabile. It also loses strength rapidly on standing at room temperature. Bacteriolytic System.-The combination of antigen (bacterium), amboceptor (immune serum), and comple- ment (normal serum) is called a bacteriolytic system. Hemolysis During his researches upon the nature of bacteriolysis Bordet7 discovered that if an animal be injected with the red blood corpuscles of another animal a substance is pro- duced in the blood serum of the injected animal capable of taking the red corpuscles injected. To this substance the name "hemolysin" is generally applied and the process of laking is known as hemolysis. Bordet's discovery of hem- olysis was announced in 1898 and at the same time Belfanti and Carbone8 also described their experiments demonstrat- ing the same phenomenon, so that to these investigators belongs equally the credit for the discovery of one of the most useful and important phenomena of immunity. Al- though foreshadowed by the work of Landois, who, in 1875, observed that the blood serum of certain animals produced laking of the blood of other species when the two were added together, the real explanation of the process was made clear by the work of Bordet, who proved that, while hemolysins are sometimes present in the freshly drawn blood of certain species against the cells of other species of animals, they can be produced in large amount by the injection of the washed erythrocytes of one species into another, and that the hemolysins thus produced are abso- lutely specific in nature. DISCOVERY 23 Thus, if a rabbit be injected with the washed red blood corpuscles of man, the hemolysin thus produced in the rabbit's blood serum will cause taking of human erythro- cytes but will be powerless against erythrocytes of other species; if injected with the erythrocytes of a sheep, the hemolysin produced will only act upon sheep's erythro- cytes; if injected with the erythrocytes of an ox, the rab- bit's blood serum will only hemolyze ox corpuscles. In other words, hemolysins are absolutely specific against the kind of erythrocytes injected in immunizing an animal and are powerless against other erythrocytes. As will be seen, the hemolysins are strictly analogous to the bacteriolysins, being produced by the injection of an antigen, the erythrocytes, the only difference being that in the one case bacteria are dissolved while in the other the hemoglobin is dissolved from the red blood corpuscles em- ployed during immunization. In both instances the reac- tion will not occur except in the presence of complement, which is furnished by the addition of fresh normal blood serum, and in both instances heating for one half hour at 56° C. destroys the activity of the immune serum, as it de- stroys the complement present in the serum. Natural hem- olysins occur in the blood of one species of animal against the erythrocytes of another, to a limited extent, just as nat- ural bacteriolysins against certain bacteria occur in the blood of certain animals, but these may be greatly in- creased in amount by the injection of suitable erythrocytes. To all bacteriolysins and hemolysins the group term '1 cytolysins ' ' has been applied, owing to the fact that their injurious action is upon the cells injected as antigen in immunization. In the case of the bacteriolysins the bacte- rium injected constitutes the cell used as antigen, and the injection is followed by the production in the blood serum of a bacteriolysin capable of destroying the bacterium in- jected; while in the case of the hemolysins, the erythrocyte injected constitutes the antigen and is followed by the pro- duction in the animal's blood serum of hemolysins capable 24 THE WASSERMANN TEST of laking the erythrocytes of the species of animal from which the injected erythrocytes were obtained. Hemolytic System.-The combination of antigen (ery- throcytes), amboceptor (immune serum) and complement (normal serum) is called a 'hemolytic system. The various theories that have been advanced in expla- nation of the cause of bacteriolysis and hemolysis and the relative part played in the process by antigen, amboceptor, and complement, will not be discussed here, as it is unnec- essary in the consideration of the Wassermann test. Complement Fixation In 1906, Bordet and Gengou9 published their paper de- scribing complement fixation and as the Wassermann test A A Hemolytic system. Bacteriolytic system B B i Fig. 1. ii I--A. Hemolytic system; a, red blood corpuscle, or antigen; b, immune rabbit serum, or amboceptor; c, guinea pig serum, or complement. B. Showing union of antigen, a; amboceptor, b; and complement, c, with resulting hemolysis. II--A. Bacteriolytic system; a, bacterium or antigen; b, immune rabbit serum, or ambo- ceptor; c, guinea pig serum, or complement. B. Showing union of antigen, a; amboceptor, b; and complement, c, with resulting bacteriolysis. is based upon this reaction it is essential that it be thor- oughly understood before the technic of the test can be considered. In experiments upon bacteriolysis and hemolysis it will be remembered that neither process can occur without the presence of three factors; i. e., antigen, amboceptor, and DISCOVERY 25 complement. For instance, if the red corpuscles of a sheep be added to the blood serum of a rabbit which has been immunized to these corpuscles, nothing occurs, but if to this mixture a little fresh normal blood serum be added, the hemoglobin of the red cells is dissolved from them and hemolysis occurs; in the same way, if the erythrocytes and complement, or fresh normal blood serum, be placed to- gether, nothing will occur until the immune serum, or am- boceptor be added, when hemolysis results. Furthermore, in neither case is the complement in the normal blood se- rum fixed by either antigen or amboceptor alone, but can only be fixed, or used up, when both antigen and ambocep- tor are combined with it. In the reaction the amboceptor, contained in the immune serum, apparently acts as a con- necting link, uniting the antigen and the complement and making hemolysis possible. This fact led Bordet and Gengou to the discovery of com- plement fixation as a specific test for the differentiation of bacteria, for they found that by using a bacteriolytic and hemolytic system together it is possible to discover the presence of a specific antigen or specific amboceptor in a mixture. In complement-fixation tests the hemolytic sys- tem is used as an indicator, as hemolysis is visible to the unaided eye, w'hile bacteriolysis is not. In order that complement fixation, as applied to the dem- onstration of specific immune bodies in the blood serum may be understood by the reader, the following hypothet- ical experiment will be described. The complement-fixa- tion test is not used in the diagnosis of typhoid fever prac- tically, because it is complicated and the agglutination test for the disease is more practical and easier of application, but the experiment described can be carried out if desired. Problem: A patient is supposed to be suffering from typhoid fever and it is desired to determine the diagnosis, if possible, by means of complement fixation. Procedure.-In a test tube containing the necessary amount of normal salt 26 THE WASSERMANN TEST solution there is placed a certain amount of the patient's serum, which has been inactivated by heating at 56° C. for one half hour; some normal guinea pig serum or complement, and a certain amount of a suspension of typhoid bacilli, or antigen. This mixture is allowed to stand in a water-bath at 37° C. for one-half hour, after which there are added a certain amount of a suspension of human red blood corpuscles and some serum from a rabbit which has been immunized against these erythrocytes. This mixture is allowed to stand in the water-bath at 37° C. for one hour, placed in the ice box for one hour, and the results then noted. Result.-If it is remembered that complement is absolutely necessary in order that hemolysis occur and that complement can not be fixed or used up unless both antigen and amboceptor be present with it, the result of this experiment can be easily understood. If the patient is really suffering from typhoid fever and immune substances or amboceptor be present in the blood serum, the complement in the fresh guinea pig serum which is added to the typhoid bacillus suspension, or antigen, and the patient's serum which contains amboceptor against the typhoid bacillus, will be used up or fixed and bacteriolysis will occur. If now the erythrocytes and the rabbit's serum containing amboceptor against them be added to the mixture, hemolysis will not occur, because the complement has already been used up or fixed by the bacteriolytic system and is no longer available for union with the hemolytic antigen and amboceptor. The test would, therefore, be positive for typhoid, as it would demonstrate that the patient's blood serum contained amboceptor or immune bodies for the typhoid bacillus. On the other hand, if the patient did not have typhoid the complement would not be fixed, so that when the erythrocytes and immune rabbit serum were added, hemolysis would occur, and the test would be called a negative one. From this experiment it will be seen that the principle of complement fixation can be utilized in the diagnosis of dis- eases due to a specific microorganism and this fact was quickly recognized and investigators endeavored to work out complement-fixation tests for many of the acute spe- cific infections. Since the discovery of complement fixation the test has been applied to pertussis, gonorrheal infec- tions, tuberculosis, glanders, certain helminthic infections, and several other conditions, but its greatest practical im- portance was reached in the discovery by Wassermann, Neisser and Bruck of the complement-fixation test for syphilis, or, as it is commonly known, the Wassermann re- action. Discovery of the Wassermann Complement-Fixation Test for Syphilis.-Many students of the subject, after the dis- DISCOVERY 27 co very of complement fixation by Bordet and Gengou, were able to show that it was not necessary to employ the whole bacteria as antigen in making tests but that watery extracts of the bacteria, when used as antigen, gave equally good results. This fact caused Wassermann and Bruck to con- clude that tuberculins should act well as antigen in a com- plement-fixation test for tuberculosis, and it was while working upon this problem that Wassermann conceived the idea that complement fixation could be applied to the diagnosis of syphilis. The discovery of Treponema palli- dum, by Schaudinn, had shown that the disease was due to an organism that is present in large numbers in the tissues of certain organs, and to these organs Wassermann turned for his antigenic extract, as Treponema pallidum had not then been grown in pure culture. Wassermann, working in conjunction with Neisser and Bruck,10 attempted to produce complement fixation in syph- ilis by using the blood serum of syphilitic monkeys as am- boceptor and watery extracts of syphilitic organs as anti- gen, and, as expected, it was found that such a combination resulted in complement fixation. It was but a step from this experiment to the application of the test to man, and in 1906, Wassermann, Neisser, Bruck, and Schucht11 pub- lished a paper giving their results in the diagnosis of syph- ilis in man by the complement-fixation test. Their paper was antedated a short time by that of Detre,12 but as Was- sermann, Neisser and Bruck, in their previous publication describing the test upon syphilitic monkeys antedated Detre's paper, the credit for the discovery remains with Wassermann and his colleagues. Wassermann considered the complement-fixation reac- tion in syphilis as a true antigen-antibody reaction, the fixation being due to specific antibodies or amboceptors present in the patient's blood, produced by Treponema pal- lidum, the etiological agent of the disease. This conception of the reaction was held by all observers until 1907, when several investigators demonstrated that complement fixa- 28 THE WASSERMANN TEST tion in syphilis occurred not only when an extract of tissues containing Treponema pallidum was used as antigen, but also when nonspecific substances were used, as extracts of normal tissues and even certain chemical reagents. In 1907, Weygandt13 obtained a positive reaction in tabes by using an antigen obtained by extracting a normal spleen, and Marie and Levaditi14 obtained positive reactions in paresis with an aqueous extract of normal liver. Later, Plant,15 Landsteiner,16 and others obtained positive results in complement fixation in syphilis by using aqueous ex- tracts of normal organs, both of man and animals, as anti- gen, and finally, the publication of the paper by Land- steiner, Muller and Potzl,17 in which they demonstrated that as good results could be obtained in the diagnosis of syphilis by the complement-fixation test with alcoholic ex- tracts of guinea pig heart as antigen as when aqueous ex- tracts of foetal syphilitic liver were employed, demon- strated that the reaction could no longer be considered spe- cific in the sense that it is a true antigen-antibody reaction. The work of the last named investigators was soon con- firmed by others and it was soon shown that alcoholic ex- tracts of tissues were as valuable as antigens in the Was- sermann test as watery extracts, and that normal organs of man or animals were apparently as rich in complement- fixing substances as were organs containing Treponema pallidum. Noguchi18 added greatly to our knowledge of the subject by proving that the substances which fix complement in the Wassermann reaction are acetone insoluble, thus indicating that they consisted largely of lipoids, and a further anal- ysis showed that such antigenic extracts contained lecithin, cholesterin, and sodium oleate. The fact that the addition of a certain amount of cholesterin to antigens used in the diagnosis of syphilis greatly increased their antigenic value was first noted by Browning and Cruickshank19 and was soon confirmed by other workers, notably, in this coun- DISCOVERY 29 try, by Walker and Swift,20 and at the present time such cholesterinized antigens are very generally used and, in the experience of the writer and most authorities who have used them, are perfectly reliable when properly controlled and are more valuable in the specific diagnosis of syphilis by the Wassermann test than are antigens which have not been thus fortified. The discovery that the antigens used in the Wassermann test need not be prepared from tissues containing Trepo- nema pallidum but that alcoholic extracts of normal tissues are just as effective, and that certain chemicals, as choles- terin, when added to antigens prepared from normal tis- sues, increase their antigenic value in the diagnosis of the disease, demonstrated, beyond question, that the early con- ception of Wassermann that complement fixation in syph- ilis is a specific antigen-antibody reaction, is erroneous, and that this reaction can no longer be considered a specific one in the strict sense in which this term is employed in immunology. Although this is undoubtedly true, the prac- tical application of the test in many hundreds of thousands of cases of syphilis has proved that, except in a very few easily recognized conditions, the test is specific for syphilis and that, clinically, it is probably even more valuable than it would be if it were a true antigen-antibody reaction, for it is doubtful if as large a percentage of infections would react positively to the test if this were true, judging from the results of complement fixation in other diseases which are true antigen-antibody reactions. Although, as generally performed, the Wassermann test is not a true specific reaction, the work of Noguchi21 and of Nichols and the writer22 has proved that, with antigens prepared from pure cultures of Treponema pallidum com- plement fixation can be obtained with syphilitic sera, and that in such instances the reaction is really a specific one, due to antibodies in the patient's blood serum against the treponema. This subject will be discussed in a later chap- ter. (Chapter VI.) CHAPTER II GENERAL DESCRIPTION OF THE WASSERMANN TEST. NATURE OF THE REACTION As stated in the preceding chapter, Wassermann and his colleagues considered the complement-fixation test for syphilis as a true antigen-antibody reaction, the aqueous extract of tissue containing Treponema pallidum furnish- ing the specific antigen, while the patient's blood serum contained the specific antibody or amboceptor. This con- ception of the reaction had to be abandoned when it was demonstrated that aqueous or alcoholic extracts of normal tissues could serve equally well as an antigen in the test. Therefore, it follows that the terms "antigen" and "ambo- ceptor" when used in reference to this test, except when applied to the hemolytic system employed, are not strictly accurate, inasmuch as the antigenic extracts are not true antigens, their injection into animals not being followed by the development of antibodies; while the substance or sub- stances in the syphilitic patient's serum which, in the pres- ence of complement and the antigenic extract, fix the for- mer, are not true amboceptors, for they are not produced by injection of the antigenic extracts employed in the test. However, as the reaction agrees in all essentials with really specific complement-fixation reactions, and as the antigenic extracts and the syphilitic patient's serum act in the same manner as do true antigens and antibodies, it is convenient to retain these terms in speaking of the reaction. As already stated the discovery of complement fixation by Bordet and Gengou was quickly utilized in the diagnosis of infectious diseases and the Wassermann test is simply an adaptation of the principle to the diagnosis of syphilis. In performing the test a hemolytic system is used as an 30 GENERAL DESCRIPTION 31 indicator and what corresponds with a bacteriolytic system is used as the diagnostic system. The hemolytic system consists of a suspension of erythrocytes, the antigen; an immune serum containing antibodies to the erythrocytes used as antigen, the amboceptor; and fresh normal blood serum, the complement. The bacteriolytic system of true specific complement-fixation reactions is replaced by the syphilitic system, as it may be called, consisting of an ex- tract of syphilitic or normal tissues, the antigen; the syph- ilitic patient's serum, corresponding to the amboceptor; and complement, or fresh normal blood serum. These various reagents are employed in exactly the same manner as are the hemolytic and bacteriolytic systems in specific complement-fixation reactions, as will be apparent from the general description of the test which follows: It will be remembered that in neither a hemolytic nor bac- teriolytic experiment will either hemolysis or bacteriolysis be brought about unless all three factors necessary for this to occur are brought together; i. e., antigen, amboceptor and complement must be placed in the same mixture before either hemolysis or bacteriolysis can occur. The immune serum contains the amboceptor and this body acts as an intermediary or connecting link between the antigen and complement, and when this union occurs either hemolysis or bacteriolysis results, as the case may be. In the Was- sermann test the patient's blood serum, if he is syphilitic, acts as the amboceptor, the serum containing some sub- stance or substances capable of uniting with antigen (the extract employed) and complement and binding the latter. In order to demonstrate the presence of this' complement binding substance in the patient's serum the following pro- cedure is employed, and it is this procedure that is known as the complement-fixation test for syphilis or the Wasser- mann test: In a test tube containing a suitable amount of normal salt solution there is placed a certain amount of the blood serum from the patient to be tested; a certain amount of 32 THE WASSERMANN TEST fresh normal blood serum, generally from a guinea pig, which contains complement; and a certain amount of an ex- tract of syphilitic or normal tissues, the antigen. This mix- ture is allowed to stand in a water-bath at 37° C. for one half hour, at the end of which time there are added to the A B c Fig. 2. Schematic representation of complement fixation in syphilis or the Wassermann test. A. The reagents entering into the reaction; a, human red blood corpuscles or hemolytic antigen; b, hemolytic amboceptor, or immune rabbit serum; c, complement, or guinea pig serum; d, patient's blood serum, which may, or may not contain syphilitic antibody; e, syphilitic antigen or antigenic extract. B. Representing a positive reaction. The patient's blood serum, d, containing syph- ilitic amboceptor, has caused the union of complement, c, and syphilitic antigen, e, so that when the hemolytic antigen, a, and the hemolytic amboceptor, b, are added to the mixture, no hemolysis results, as the complement has already been fixed or absorbed, as shown in the figure. C. _ Representing a negative reaction. The patient's blood serum, d, containing no syphilitic amboceptor, does not fix the complement with the syphilitic antigen, e, so that when the hemolytic antigen, a, and the hemolytic amboceptor, b, are added, union occurs between them and the complement, and hemolysis results. mixture a certain amount of a suspension of human or other red blood corpuscles and a certain amount of the blood serum of a rabbit immunized to the blood corpuscles GENERAL DESCRIPTION 33 employed in the suspension. The mixture is again incu- bated in a water-bath at 37° C., this time for one hour, and then placed in an ice box, and the result noted at the end of an hour or two. If the patient is suffering from syphilis, the result will be that no hemolysis occurs, and the test is called positive, while, if syphilis is not present, hemolysis will occur, and the test is said to be negative. The positive result is due to the fact that during the first incubation the substance in the patient's blood serum, corresponding to a specific am- boceptor, unites the antigen, or extract used in the test and the complement, thus binding the latter, so that when later the erythrocytes and immune serum or amboceptor against them is added, there is no complement left to unite with them and therefore, hemolysis can not occur. In case that the result is negative, this is due to the fact that the patient's blood serum contained no substance capable of uniting the antigenic extract and complement, so that when later the erythrocytes and the immune serum to them were added, the amboceptor in the immune serum caused the union of the erythrocytes and the complement and hemolysis resulted, the complement having been left free in the mixture, owing to nonunion during the first incuba- tion. From this brief description of the Wassermann test it will be noted that it corresponds in every way as regards technic and results with specific complement-fixation tests, but that instead of a specific bacteriolytic antigen there is substituted a nonspecific tissue extract which serves as an antigen and which acts in the same manner as does the specific antigen employed in bacteriolytic experiments, while the complement-fixing substance or substances in the syphilitic patient's blood serum act in the same manner as does the amboceptor in the immune serum in bacteriolytic experiments. When, however, an extract of a pure cul- ture of Treponema pallidum is used as an antigen there is no difference between the Wassermann test and other com- 34 THE WASSERMANN TEST plement fixation reactions, as in this case the complement fixation is the result of a true antigen-antibody reaction. Unfortunately, the antigens prepared from pure cultures of Treponema pallidum are very unreliable in practice, so that they can not be used for this purpose, and the nonspe- cific antigens have to be depended upon in the test. Nature of the Wassermann Reaction The exact nature of the Wassermann reaction is still a mystery, despite an immense amount of work that has been done in the endeavor to solve the problem. The fact that ex- tracts of tissues which do not contain Treponema pallidum serve equally well as antigens as those that do, and even better than antigens prepared from pure cultures of the organism, proves that the reaction is not a true antigen- antibody reaction. The experiments of Noguchi23 demon- strating that the active antigenic principles in the normal tissue extracts were lipoids and his conclusion that the re- action depends upon lipotropic substances in the patient's blood serum, showed that the cause of the reaction is closely connected with the chemistry of the lipoids, and this is about all that we can claim to know regarding its exact nature to date, although many ingenious and inter- esting theories have been advanced in explanation of the phenomenon. Citron24 adopted the hypothesis that the antibody pro- ducing antigen is a toxolipoid and has proposed the name "Luesargine" for the syphilitic antibodies. In conjunc- tion with Miinck he claims to have demonstrated that the true syphilitic antigen is contained only in aqueous ex- tracts of syphilitic tissues and that only rabbits inoculated with such extracts produce antibodies that react positively with alcoholic extracts of normal organs, but his work has not been confirmed. Schmidt25 considers that the reaction is colloidal in na- ture, depending upon the union of the colloids in the anti- NATURE OF THE REACTION 35 genic extracts and the serum globulins in the patient's blood serum, which are increased in amount and changed in nature in syphilitic patients. Levaditi and Yamanouchi26 believe that the reaction de- pends upon the union of two colloidal factors, one present in the patient's blood serum and nonproteid in character, the other consisting of lipoidal substances in the extracts used as antigen. Weil and Braun27 conclude that during the progress of the disease certain tissue products, lipoidal in nature, are absorbed, and give rise to antibodies which will react in vitro with the lipoids present in the antigenic extracts, the so-called autoantibody theory. This theory is somewhat similar to Citron's, both resting upon the belief that in the tissues of the patient there are produced antibodies to the lipoids in the antigenic extracts, these antibodies being pro- duced by the absorption in the body of albumin-lipoid sub- stances produced by the growth of Treponema pallidum, and the subsequent production of antibodies against these substances. Bergel28 considers that the Wassermann reaction may be due to the action of a specific lipase which is produced in the syphilitic patient against "lues-lipoids." This theory can not hold in the light of the facts demonstrating that the reaction occurs with extracts prepared from normal or- gans, when used as antigen. Manwaring29 thinks that in guinea pig blood serum there is a proteolytic ferment which may weaken or destroy com- plement and that the fixation reaction is due to substances present both in the antigenic extracts and syphilitic pa- tient's serum which increase this effect. Bruck and Stern30 believe that the reaction is due to the interaction of the lipoid bodies in the antigenic extracts and similar bodies in the patient's blood serum; while Peritz31 concludes that it is due to some substance in the blood serum that has an affinity for lecithin, which is always contained in the extracts. 36 THE WASSERMANN TEST None of the various theories advanced to account for this reaction have been accepted as the true explanation of the process of complement fixation in syphilis and none of them rest upon sufficient experimental evidence to be ac- cepted as the explanation. The most that can be said re- garding the reaction is that it is due to an interaction be- tween lipotropic bodies present in the blood serum of syph - ilitics and lipoids in the antigenic extracts. A mixture of a lipoid extract and syphilitic blood will absorb a large amount of complement, so that it follows that the substance or substances in the serum must be lipotropic in nature. This lipotropic substance is more constantly present in the blood of syphilitics than in other diseases, and is prac- tically specific for that disease. Researches have conclu- sively proved that the lipotropic substance or substances are due to unknown changes in the tissues or blood serum brought about by the actual presence in the tissues of liv- ing Treponema pallidum, and that the reaction disappears with the cure of the infection, so that its presence is proof of the existence of syphilis even though months or years may have passed without definite clinical symptoms having been observed. Although the Wassermann reaction, as the test is usually performed, is not a true antigen-antibody reaction, there is proof that in the blood of syphilitics true antibodies against Treponema pallidum are formed which will fix complement in the presence of antigens prepared from pure cultures of the organism. The first experiments along this line were initiated by Schereschewsky,32 who obtained positive reactions in syph- ilitics by using antigens made from an impure culture of what he regarded as Treponema pallidum. Owing to the doubt attending the impurity of the culture and as to the exact species of treponema with which he worked, Scheres- chewsky's results were not conclusive, and it was not until Noguchi33 cultivated in pure culture the organism of syph- ilis and employed antigens prepared from these pure cul- NATURE OF THE REACTION 37 tures, that it was proved that true complement fixation did occur in syphilis. Noguchi34 demonstrated, that by using aqueous extracts of pure cultures of Treponema pallidum as antigen, it was possible to secure complement fixation in certain cases of the disease, but that the reaction did not occur as consistently as when nonspecific antigens were employed nor did it occur at all stages of the disease. Noguchi immunized rabbits with the pure cultures, using the aqueous pallidum extracts for injection, and found that their blood bound complement when the pallidum extract was used as antigen but that negative results were ob- tained when extracts of normal tissues were used as anti- gens. These results proved that the blood serum of the immunized rabbits contained a real antibody against the syphilitic toxin contained in the pallidum extract used in immunization. He found, however, that while the blood serum of rabbits with an active syphilitic orchitis, gave positive complement fixation with the extracts of normal tissue, rich in lipoids, it did not bind complement when the pallidum extract was used. He explained this result as follows :35 The findings show that the immune serums contain the specific antibodies for the pallidum but not for the lipoids, and that in the serums of syphilitic rabbits (active stage) there is an abundance of the lipotropic substance (probably an enzyme) capable of rendering the complement inactive (so- called fixation) in the presence of certain lipoids, but too little specific anti- bodies to bind complement with the pallidum antigen. When pure pallidum antigens were used with human syphilitic blood serum, Noguchi obtained similar results, the early stages of the disease, primary and secondary, seldom reacting with the specific antigen, while in the ter- tiary, latent, and late congenital cases the blood serum fre- quently gave a positive result. Because of these diverse results in the various stages of the disease, when the pure pallidum antigen was used, Noguchi concluded that a posi- tive result with the pallidum antigen indicated the forma- tion of true antibodies in the blood of the patient, and was, 38 THE WASSERMANN TEST therefore, of favorable omen, while a negative reaction in- dicated their absence. In 1912, the writer and Nichols,36 using antigens made by extracting pure cultures of Treponema pallidum with alco- hol, obtained complement fixation in all stages of syphilis, but positive reactions were also obtained with syphilitic blood serum with antigens made from pure cultures of Treponema pertenue, Treponema microdentium, and Spi- rocheta refringens, while antigens made from pure cul- tures of organisms like Bacillus typhosus and Spirillum cholera gave negative results. We concluded that the pos- itive reactions with the antigens prepared from the organ- isms mentioned was in the nature of a group reaction, and that the practical value of the test when the pallidum anti- gen was used was small, owing to the fact that many cases of the disease gave a negative reaction, even during the late stages of the infection. Our results have since been confirmed by Kolmer, Williams and Laubaugh37 and other investigators. While it may be considered as proved that antigens pre- pared from pure cultures of Treponema pallidum are unre- liable and of little value in the practical application of the Wassermann test, their use has demonstrated that a spe- cific fixation of complement may be obtained in syphilis when such antigens are used in the complement-fixation test. This fact renders it very probable that antigens pre- pared from fetal syphilitic liver, rich in treponemas, may give positive results in rare instances where extracts of normal tissues will fail, and the use of the two kinds of antigen in practice has shown that this is the case. The writer is, therefore, of the opinion that the Wassermann reaction, when an extract of fetal syphilitic liver is used as antigen, is of a dual nature, in some instances, consist- ing of a true antigen-antibody reaction between specific antibodies in the patient's blood serum and specific sub- stances in the antigen; and a nonspecific reaction between the lipotropic substance in the patient's serum and lipoids NATURE OE THE REACTION 39 in the antigen. When an extract of normal tissues is used as antigen the reaction is due to the lipoids in the antigen and lipotropic substances in the patient's blood serum, and, in such instances, the possible reaction between true anti- bodies and the specific substance in the antigen is lost, and the reaction may be negative instead of positive. While, in practice, in the writer's experience, this result very rarely occurs, it is considered best, if possible, to make use of both varieties of antigen when performing the Wasser- mann test. CHAPTER III FACTORS WHICH INFLUENCE THE RESULT OF THE TEST The Wassermann test is quantitative in nature and un- less all of the reagents are carefully titrated at frequent intervals and accurately used, erroneous results are sure to occur. The quantity of blood serum to be tested; its proper inactivation; the dose of complement and its source; and the amount of amboceptor and antigen to be used; are all important factors in the test and variations from the proper amount of each will be followed by faulty results, as will be shown in the chapter devoted to technic. However, aside from the reagents employed in the test there are certain factors which have a very marked influ- ence upon its result and which should be understood by physicians using this valuable diagnostic measure. These factors are: The influence of the ingestion of alcohol; the influence of certain bacteria in the blood serum; and varia- tions in the amount of complement inhibiting substance in the blood serum to be tested. The Influence of the Ingestion of Alcohol.-In 1911, Major Nichols, of the Medical Corps of the Army, and the writer38 called attention to the effect that the inges- tion of considerable amounts of alcohol has upon the re- sults of the Wassermann test. It was found that the in- gestion of alcohol in the form of beer or whiskey, and in considerable amounts, might convert a strongly positive serum into a negative one, if the test were made within twenty-four hours after the ingestion of the alcohol, and that the blood might remain negative for as long as three days, although it generally became positive again within twenty-four hours. This fact was discovered by the writer 40 FACTORS INFLUENCING RESULT OF TEST 41 while making routine examinations of the blood of a cer- tain individual who was being treated for syphilis by Major Nichols, and who had presented several strong positive reactions before the first negative one which was due to the ingestion of alcohol, and further research by Nichols and myself proved that this drug has a very po- tent influence in weakening or destroying the lipotropic substances in the patient's blood serum which apparently cause the Wasserrnann reaction. The case in which this phenomenon was first observed is of special interest in this connection and the clinical his- tory is here given: Male, 41 years of age. Infected with syphilis in 1904. Now has palmar lesions. February 25, 1911, 0.5 gm. of salvarsan was administered intra- venously, the Wasserrnann test before administration giving a plus reaction. Between this date and April 7th, 1911, his blood was tested five times, the reaction four times being double-plus, and the last test, on April 7th, being plus. (Double-plus here equals four-plus of most authorities.) April 8th his blood was again tested and was found to be negative. The result was unusual and on inquiry it was learned that on the night of April 7th he drank about 700 c.c. of Munich beer. His blood was tested again on April 12th, and gave a double-plus (four-plus of some authors) reaction. Be- lieving that the beer might have had something to do with the sudden disappearance of the positive reaction in this case, we requested that he repeat the beer on the night of April 14th, which he did, and on the morn- ing of April 15th the reaction was again negative. On the 17th of April it had become plus and on the 22nd again was double-plus. As it was evident that the negative result of the Wasser- mann test was due, in this case, to the ingestion of the beer, we decided to repeat the experiments on other cases. We selected nine syphilitic patients, who before the adminis- tration of alcohol gave a double-plus reaction, and Table I gives the results obtained in these patients. It will be re- membered that in this table, the double-plus reaction means absolute inhibition of hemolysis, and is equivalent to the four-plus reaction of most writers. Of the 9 cases included in this table, 3 were in the sec- ondary stage of the disease and 6 were in the tertiary stage. The quantity of alcohol administered varies some- 42 THE WASSERMANN TEST CASE NO. stage of DISEASE Before alcohol Alcohol given Date and result after alcohol REACTION DATE 1911 DATE 1911 AMOUNT DATE 1911 Time RESULT 1 -. 1st test, + Apr. 1 Apr. 7 700 c.c. Munich beei Apr. 8 9 A.M. in evening Apr. 12 9 A.M. + + 2nd test, + + Apr. 12 Apr. 14 Same amount of Apr. 15 9 A.M. beer, same time Apr. 17 9 A.M. + Apr. 22 9 A.M. + + 2 2nd + + May 19 May 20 90 c.c. of 95% al- May 20 7:30 p.m. + cohol, 7 a.m. to 7 P.M. May 21 10 A.M. May 22 10 A.M. + - May 23 10 A.M. + + May 24 10 A.M. + + 3 3rd + + one hour before May 29 May 29 240 c.c. whiskey be- May 29 6 P.M. + alcohol tween 10 a.m. and May 30 6 A.M. 5 P.M. May 31 10 A.M. - June 1 10 A.M. + + June 2 10 A.M. + + 4 3rd + + one hour before June 3 June 3 240 c.c. whiskey be- June 3 8 P.M. alcohol tween 10 a.m. and June 4 7 A.M. 6 P.M. June 5 10 A.M. + Tune 6 10 A M. + + 5 2nd + + day before al- June 4 June 5 180 c.c. whiskey be- June 5 8 P.M. cohol tween 9 a.m. and Tune 6 8 A.M. - 6 P.M. June 7 8 A.M. + + Tune 8 8 A.M. + + 6 2nd + + one hour before June 22 June 22 240 c.c. whiskey be- June 23 9 A.M. alcohol tween 9 a.m. and Tune 23 8 P.M. - 7 P.M. June 24 9 A.M. _j_ Tune 25 9 A.M. + + 7 3rd + + one hour before June 29 June 29 240 c.c. whiskey be- June 26 6 P.M. + + alcohol tween 10 a.m. and June 30 6 A.M. 5 P.M. July 1 6 A.M. -I- July 2 6 A.M. + + 8 3rd + + one hour before July 7 July 7 240 c.c. whiskey be- July 7 8 P.M. alcohol tween 12 M. and July 8 1 P.M. 5 P.M. July 9 1 P.M. - 9 3rd + + one hour before July 7 July 7 240 c.c. whiskey be- July 8 1 P.M. + + alcohol tween 12 m. and July 9 1 P.M. - 7 P.M. July 10 1 P.M. - July 11 1 P.M. + - July 12 1 P.M. + Ttilv 13 1 P.M 1- 4 Table I Illustrating the Effect of the Ingestion of Alcohol upon the Result of the Wassermann Test FACTORS INFLUENCING RESULT OF TEST 43 what but in every case the positive reaction disappeared after its administration. In most of the cases the blood was tested one hour before the administration of the al- cohol was begun as well as afterward. A study of the ta- ble shows that in 5 of the cases the reaction was found to be negative one hour after the last dose of the alcoholic liquor had been administered, and that in all the cases the reaction remained negative for several hours. In cases 4, 5, and 6 the reaction remained negative for over 24 hours while in case 9 it remained negative for three days and did not become double-plus again until over four days had elapsed from the time the alcohol was administered. Two of the patients had received salvarsan before the admin- istration of the alcohol, but in one the drug had been given so long before that it could not have had any effect upon the reaction, while in the other the fact that the reaction became double-plus again after becoming negative from the alcohol, proves that the salvarsan had nothing to do with its disappearance. That the disappearance of the reaction after the inges- tion of alcohol is not due to the alcohol itself, which may be present in the blood serum, is proved by the fact that a much larger quantity of alcohol is necessary to produce hemolysis, as proved by experiment, than could possibly be present in the blood serum of the patients tested. Thus, we found that it requires at least 0.15 c.c. of absolute alco- hol to produce complete hemolysis of 1 c.c. of a 1 per cent suspension of human red blood corpuscles within the time used in the Wassermann test for incubation of the hem- olytic system, whereas only 0.08 c.c. of the patient's blood serum was used in each test. It is thus evident that if the amount of blood serum used had been actually absolute al- cohol the quantity would have been insufficient to have pro- duced hemolysis. From these experiments, which have been confirmed by many investigators since our original publication, it is evident that alcohol may render inert the sub- 44 THE WASSERMANN TEST stance or substances in the blood serum of syphilitics which react with lipoids in the antigenic extracts, and thus a strongly positive serum may give a negative result. For this to occur, the alcohol must be taken in considerable quantity and probably within twenty-four hours, or at most three days before the test is made, but it should be remembered that smaller amounts of alcohol may ren- der weak reactions negative, so that cases which should present a single plus reaction will often react nega- tively after even moderate amounts of alcoholic liquors have been ingested. It is thus evident that very careful inquiry should be made regarding the use of alcoholic liq- uors before taking blood for a Wassermann test from any patient, and if there is a history of indulgence in such liq- uors within from twenty-four to thirty-six hours previ- ously, the taking of the specimen should be postponed and alcoholics forbidden until at least two days have passed. The neglect of this precaution has often resulted in errone- ous results and cases that should have been strongly posi- tive have been reported as negative or doubtful. When the test is used as a control of treatment alcohol, even in small amounts, should be forbidden for some days before the blood is collected, or the weak reactions often observed in treated cases will be entirely missed, thus obscuring the real condition of the patient as regards his serological status. From what has been said, it follows that a negative Was- sermann reaction, after the ingestion of alcoholic liquors, possesses no value whatever, and an immense amount of harm has doubtless been done by lack of care in ascertain- ing whether patients have indulged in such liquors within a short time before the blood was collected. Our observa- tions, as well as those of others, both in this country and abroad, have proved that the effect of alcohol upon the Wassermann reaction is very important from a practical standpoint, and one which should not be forgotten when FACTORS INFLUENCING RESULT OF TEST 45 patients come to the physician for the collection of blood for the test. The Influence of Certain Bacteria, when Developing in the Blood Serum, upon the Results of the Test.-It is a well- known fact that, upon standing for varying lengths of time, some human blood sera develop anticomplementary sub- stances, which cause inhibition of hemolysis in both the tubes containing antigen and the control tubes used in the Wassermann test. Most of these bodies are destroyed by heating the blood serum for one-half hour at 56° C., but, in some instances, they still persist after heating, and in such sera one is unable to read the result of the test owing to inhibition occurring in both the antigen and control tubes. From these observations it is evident that these anticomplementary bodies are of two kinds, thermolabile and thermostable, and that the inactivation of the blood serum by heating enables us to destroy only the thermo- labile group. While the exact nature of these anticomplementary bodies is still doubtful, the writer has shown39 that some of them are produced by the growth of bacteria in the blood serum to be tested, if the serum is kept for some time. Not only will certain bacteria produce anticomplementary bodies that cause inhibition of hemolysis in both the anti- gen and control tubes of the test, but certain species also possess the power of causing inhibition of hemolysis in the antigen tube alone, thus giving rise to a nonspecific reac- tion in the sera containing them. It is these species that are of special importance in the application of the Wasser- mann test and which will here be discussed. While carrying on some comparative tests of blood se- rum with another laboratory it was noted that in one in- stance which gave a positive reaction in the writer's lab- oratory and a negative one at another laboratory situated five days' journey distant, the blood serum when received was cloudy in appearance and had a bad odor. As it was known to be infected and as it was thought possible that 46 THE WASSERMANN TEST bodies developed in the serum by the growth of the bac- teria might be the cause of the difference in results, this serum was plated out upon agar-agar and cultures of the following organisms were isolated: Staphylococcus aureus; a large Staphylococcus albus; a small Staphylococcus albus, a large spore-bearing bacillus of the subtilis type; and a very minute diplobacillus. In order to determine whether any of these bacteria were responsible for the reaction obtained with this blood serum the following experiments were undertaken to- gether with experiments with a stock culture of Staphy- lococcus aureus, Staphylococcus citreus, and Streptococcus pyogenes. The same technic was followed in these exper- iments as was used in performing the Wassermann test, the mixtures of serum and bacteria being substituted for the patient's blood serum. Experiment 1.-An emulsion in normal salt solution was made from a pure culture of each organism isolated from the blood serum, and 0.08 c.c. of this emulsion was placed in 1 c.c. of a 1 per cent suspension of human red blood corpuscles, together with 0.08 c.c. of an inactivated normal serum. Table II gives the result of this experiment. Table II Results of Complement-Fixation Tests in Experiment No. 1 NAME OF BACTERIUM Results Controls ORIGIN OF BACTERIUM ANTIGEN TUBE CONTROL TUBE BACTERIA ALONE SERUM ALONE S. albus, large - - Isolated S. albus, small - - - - from S. aureus, - - - serum in Bacillus (Spore) - - - all cases Diplobacillus - - - - Combined bacteria - - - - From these results it is evident that merely adding an emulsion of the cultures of these bacteria to normal blood serum is without effect upon the result of the complement- fixation test, as all of the mixtures gave a negative reaction. Experiment 2.-This experiment was identical with Experiment 1, except that the mixtures of serum and bacteria were inactivated at 56° C. for one-half FACTORS INFLUENCING RESULT OF TEST 47 hour before being tested. The mixtures of bacteria and serum in this experi- ment all gave a negative reaction with the exception of the mixture of blood serum and the diplobacillus, which resulted in a plus reaction, the heating resulting in the liberation of enough inhibitory substances to cause this type of reaction in both the antigen and control tubes. Of course this reaction would have caused no confusion so far as the diagnosis of syphilis is concerned, as the fact that the inhibition occurred in the control tube as well as in the antigen tube would have thrown the test out entirely and another specimen would have been requested. Experiment 3.-Specimens of normal serum were inoculated with each of the bacteria isolated from the serum in question as well as with the stock cul- tures and allowed to stand at room temperature for one week before being tested. Table III gives the results of this experiment: Table III Results of Complement-Fixation Tests in Experiment No. 3 NAME OF BACTERIUM Result CONTROL, SERUM ALONE ORIGIN OF BACTERIUM ANTIGEN TUBE CONTROL TUBE 8. albus, large + - - Isolated from serum <8. albus, small - - - 44 4 4 4 4 8. aureus, - - - 4 4 4 4 44 Bacillus (spore) - - 44 44 44 Diplobacillus , + + + - 44 44 44 8. aureus + + - - Stock culture 8. citreus + + + + - 4 4 4 4 8. pyogenes - - - 4 4 4 4 It is evident, by consulting the table, that after incuba- tion in human blood serum for a week at room temperature the large S. albus isolated from the blood serum under discussion gave a plus-minus reaction in the antigen tube, while the diplobacillus isolated from the same serum gave a plus reaction in the antigen tube and a double-plus re- action in the control tube. The stock S. aureus gave a double-plus reaction in the antigen tube, while the stock culture of S. citreus caused total inhibition of hemolysis in both antigen and control tubes. From this experiment it is evident that the stock S. aureus, growing in normal blood serum at room tempera- ture for one week produced some substance or substances in the serum which gave a false positive reaction. Experiment 4.-This experiment was identical with Experiment 3, except that the mixtures of bacteria and serum were incubated at 37° C. for 24 hours 48 THE WASSERMANN TEST and then allowed to stand at room temperature for one week, before being tested. Table IV gives the results of this experiment: Table IV Results of Complement-Fixation Tests in Experiment No. 4 NAME OF BACTERIUM Result CONTROL SERUM ALONE ORIGIN OF BACTERIUM ANTIGEN TUBE CONTROL TUBE S. albus, large + + - - isolated from serum S. albus, small - - - IC C C Cl S. aureus + + Mi - IC C C Cl Bacillus (spore) - - - 11 ll ll Diplobacillus + + + + - ic c i c c S. aureus + + - - Stock culture S. citreus + + + + - I I c c S. pyogenes + + - - C C It From this experimental record it will be noted that no less than four of the organisms tested gave a double-plus (four-plus) positive reaction in the antigen tube alone, while two inhibited hemolysis completely in both the anti- gen and control tubes. It would appear that the 24 hours at incubator temperature lead to an increase in the inten- sity of the reaction, due to a richer development of the bacteria in the sera. Considering the experiments as a whole it is perfectly evident that a positive complement-fixation reaction may be obtained in a normal serum if certain species of bacteria are enabled to develop within it. That every strain of a bacterial species will not produce this result is siiown by the fact that while a double-plus reaction was obtained in the serum infected with the stock S. aureus and with the aureus isolated from the serum under discussion, a nega- tive result was later obtained with an S. aureus isolated from a blood serum that gave a negative reaction. It is also evident that temperature has much to do with the ability of the bacteria to produce the positive reaction, for in the mixtures of serum and bacteria kept for 24 hours at incubator temperature, the reactions were most marked, while two of the organisms which did not give a reaction when growing in the blood serum at room tern- FACTORS INFLUENCING RESULT OF TEST 49 perature, gave a strongly positive reaction when incubated. Inactivation also appears to favor the production of the false positive reaction in sera which are contaminated with certain bacteria. As regards the serum under discussion, it is evident that the positive result could have been caused by either the large S. albus or the S. aureus, or both, according as the conditions at the time of examination were favorable to the production of inhibitory substances by one or both of these organisms. The writer's observations regarding the action of bacte- ria upon the results of the complement-fixation test in syph- ilis have recently been confirmed by Kyntoku,139 who found that certain bacteria caused the development of both ther- molabile and thermostabile anticomplementary bodies in human blood serum which may seriously interfere with the test, rendering it unreliable as regards positive and nega- tive results. Kyntoku found that these anticomplementary bodies were especially apt to be caused by the presence in the serum of staphylococci, but he also found that Bacillus typhosus, Bacillus coli, and Bacillus subtilis produced such bodies in human serum. He states that in sterile blood se- rum only thermolabile anticomplementary bodies develop and that the presence of thermostabile anticomplementary bodies are evidence that the blood serum is not sterile. If his observations are correct it follows that blood serum col- lected under aseptic precautions, if not contaminated by bacteria, can always be used, provided it is first inactivated by heating it for one-half hour at 56° C. An interesting phenomenon first demonstrated by Kyn- toku is that both thermolabile and thermostabile anticom- plementary bodies may be removed from blood serum by filtration through Kitasato filters and that such filtration does not remove the substances causing complement fixa- tion from syphilitic blood serum. 50 THE WASSERMANN TEST The fact that, under certain conditions, such common bacteria as staphylococci and streptococci, when growing in normal blood serum, may give rise to a positive Was- sermann reaction, is of great practical importance, and one might conclude from these experiments that only fresh blood serum should be used in making complement-fixation tests. However, if proper precautions are taken in the collection of specimens, there is no danger of bacterial contamination, and it is our experience that, unless such contamination occurs, a positive result will not be obtained in normal serum, even when kept for a considerable period of time. Personal experiments have demonstrated that blood sera may be kept at room temperature for as long as one month without danger of a negative serum becom- ing positive, provided there is no bacterial contamination in the serum. While, in such specimens, anticomplemen- tary bodies frequently develop, after they have been kept for a week or more, these bodies cause inhibition of hem- olysis in both the antigen and control tubes of the serum, and thus while destroying the value of the test, this type of reaction does not cause any confusion in the diagnosis of the disease. However, these observations emphasize the importance of collecting specimens for the Wassermann test under aseptic precautions if the sera can not be ex- amined within 24 hours after collection. The Influence of Variations in the Strength of the Re- action in Untreated Cases of Syphilis upon the Results of the Test.-During the routine examinations of blood sera for the Wassermann test, the writer has frequently ob- served cases in which the reaction varied from a strongly suspicious one to a plus-minus or negative one within periods of one or two weeks, while, in rare instances, a positive reaction would be found negative upon a second examination within the same period of time, no treatment having been administered at any time. In all such in- stances subsequent tests resulted in a positive reaction, so FACTORS INFLUENCING RESULT OF TEST 51 that it was evident that the variations in results, provided the technic of the test was the same, must have depended upon the reduction or absence of the body or bodies in the patient's serum which produce the reaction. The fact that the blood serum of an undoubted syphilitic patient may, during certain intervals of time, give a nega- tive reaction in the absence of specific treatment, and when previous and subsequent tests are positive, is of practical importance. While in the writer's experience such anom- alous results have not been very numerous, their occur- rence, together with the criticisms of the test by some authors who have had a similar experience, rendered an experimental investigation of the phenomenon of value, and in 1914, the writer40 published certain observations upon the subject which will be briefly discussed at this time. The observations proved that marked daily varia- tions occur in the strength of the complement-fixation re- action in the blood of syphilitics. While, a priori, there would appear to be no reason why the strength of the re- action should not vary, we have grown to consider the Wassermann reaction, if once obtained, as a stable quan- tity, the discordant results sometimes observed being im- puted to the serologist or considered as proof of the unre- liability of the test. Ten patients suffering from undoubted syphilis were selected for the observations. Most of these men were prisoners at the U. S. Military Prison, at Fort Leaven- worth, and thus all chance of the reaction being influenced by the use of alcohol was obviated, the men being under a rigid discipline and on a plain, wholesome diet. Of the ten patients, two were in the primary stage of the disease, four in the secondary stage, and four in the latent stage. No treatment had been administered for over a year in any case, and then only to those who were in the latent stage. The patients in the primary stage had typical chancres and have since developed secondary eruptions; 52 THE WASSERMANN TEST the secondary cases all presented eruptions and mucous patches; while the four in the latent stage were free from obvious symptoms. All gave double-plus (four-plus) Was- sermann reactions at the time of the experiment. The blood was collected every day and tested upon the same day it was collected, daily titrations of the strength of the complement fixation being made. The technic used was that employed by the writer for several years and which is described in this work. The possible influence of variations in the strength of the hemolytic system from day to day, which might explain the results obtained in the titration of the complement-fixing strength of these sera, was controlled by titration of the complement before each series of tests and careful pre- liminary titration of the amboceptor and the antigen. In addition, in all of the titrations, both of the reagents used in the tests and of the sera tested, the blood corpuscles of the same individual were used in preparing the blood cell suspension, thus obviating any chance that isohemolysins present in the sera tested influenced the results of the titra- tions. Had the blood corpuscles of different individuals been used in preparing the blood cell suspension, from day to day, there might have been a chance that isohemolysins present in some of the sera tested would have produced negative results with blood cells from certain individuals, but as the same erythrocytes were used throughout the experiments this possible source of error was eliminated. Each patient's blood serum was titrated every day for one week, the quantities of serum employed being 0.02 c.c.; 0.04 c.c.; 0.06 c.c.; 0.08 c.c.; and 0.1 c.c. The quantity of blood serum used with the technic for diagnostic purposes is 0.1 c.c. so that this quantity may be taken as the diag- nostic amount in the tables that follow. A control of 0.1 c.c. of the serum titrated was always employed. In the tables it will be understood that the sign, + +, in- dicates absolute inhibition of hemolysis, or a positive re- FACTORS INFLUENCING RESULT OF TEST 53 action; the sign, +, anything between absolute inhibition and 50 per cent of inhibition of hemolysis; the sign, +-, anything between 50 per cent of inhibition and total hem- olysis, or practically a negative reaction; and the sign, -, total hemolysis or a negative result. Results in the Primary Stage of Syphilis.-Owing to lack of material, only two cases were tested in the primary stage of the disease. Both presented typical chancres and both gave a double-plus, or positive, Wassermann reaction when first tested. Tables V and VI give the results of the daily titrations of the blood serum in these two cases. Table V Results of Titration of Blood Serum for Complement-fixing Strength in Case 1-Primary Syphilis DATE OF TEST AMOUNT OF BLOOD SERUM IN C.C/S CONTROL. 0.1 C.C. 0.02 0.04 0.06 0.08 0.1 Nov. 19 + - + + 4 4 - Nov. 20 + - + - + 44 44 - Nov. 21 + - + + + 44 44 - Nov. 22 + - 4- 4 4 4 - Nov. 23 + 44 4 + 44 44 - Nov. 24 + - + 4 4 4 - Nov. 25 44 44 44 44 44 - Table VI Results of Titration of Blood Serum for Complement-fixing Strength in Case 2-Primary Syphilis DATE OF TEST AMOUNT OF BLOOD SERUM. C.C. CONTROL. 0.1 C.C. 0.02 0.04 0.06 0.08 0.1 Nov. 19 + + + + + + + + + - Nov. 20 + - + + + + + + + + - Nov. 21 - + - + + + - Nov. 22 - + - + - + + - Nov. 23 + + + + + + + + + - Nov. 24 + + + + + + + + + - Nov. 25 + + + + + + + + + - A study of these tables will show that considerable varia- tions occurred in the complement-fixing power of the blood serum in both these cases of primary syphilis, especially in the smaller amounts employed, but that in neither case 54 THE WASSERMANN TEST did the reaction become negative with the diagnostic amount of serum; i. e., 0.1 c.c. However, in one case a plus or doubtful reaction was obtained upon three of seven days, while in the other the same result was obtained upon two of the seven days. These results are very significant when it is remembered that in this stage of syphilis the Wassermann test is so often doubtful, or negative, and in- dicates that frequent examinations would give a much higher percentage of positive results than are usually re- corded. Both of these cases, if tested upon certain days of the week, would have been reported serologically doubt- ful, but, fortunately, in most primary infections Trepo- nema pallidum may be demonstrated with the dark-field, thus avoiding the confusion that might be caused by varia- tions in the strength of the Wassermann reaction. Results in the Secondary Stage of Syphilis.-The blood serum of four typical cases of secondary syphilis, Table VII Results of Titration of Serum for Complement-fixing Strength in Case 3-Secondary Syphilis DATE OF TEST AMOUNT OF BLOOD SERUM. C.C. CONTROL C.C. 0.02 0.04 0.06 0.08 0.1 0.1 Nov. 26 + + + + + + + + + + - Nov. 27 + + + + + + + + + + - Nov. 28 + + + + + + + + + + - Nov. 29 + + + + + + + + + + - Nov. 30 + + + + + + + + + + * Dec. 1 + + + + + + + + + + - Dec. 2 + + + + + + + + + + - Table VIII Results of Titration of Serum for Complement-fixing Strength in Case 4-Secondary Syphilis DATE OF TEST AMOUNT OF BLOOD SERUM. C.C. CONTROL C.C. 0.02 0.04 0.06 0.08 0.1 0.1 Jan. 3 + + + + + + + + + + - Jan. 4 + + + + + + + + + + - Jan. 5 + + + + + + + + + + - Jan. 6 + + + + + + + + - Jan. 7 + + + + + + + + + - Jan. 8 + + + + + + + + + Jan. 9 + + + + + + + + + + - FACTORS INFLUENCING RESULT OF TEST 55 Table IX Results of Titration of Serum for Complement-fixing Strength in Case 5-Secondary Syphilis DATE OF TEST AMOUNT OF BLOOD SERUM. C.C. CONTROL C.C. 0.02 0.04 0.06 0.08 0.1 0.1 Dec. 24 + - + - + + + + - Dec. 25 - - + - - + - - Dec. 26 - + - + - + + - Dec. 27 - + - + + - + - Dec. 28 + - + + + + + + -. Dec. 29 - + - + - + + + - Dec. 30 - + - + + + + - Table X Results of Titration of Serum for Complement-fixing Strength in Case 6-Secondary Syphilis DATE OF TEST AMOUNT OF BLOOD SERUM. C.C. CONTROL C.C. 0.02 0.04 0.06 0.08 0.1 0.1 Dec. 5 + - + - + + + + - Dec. 6 + + + + + + + + + + - Dec. 7 + + + + + + + + + + - Dec. 8 + + + + + + + + + + - Dec. 9 + + + + + + + + + + - Dec. 10 + - + + + + - Dec. 11 + - + - + + + - all showing lesions of the disease, was titrated in the same manner as described and Tables VII, VIII, IX, and X give the results obtained. In Case 3 the strength of the blood serum did not vary at all during the time it was titrated, a double-plus (four- plus) reaction being obtained, even with the smallest amount of serum used. The case is of interest as showing how strong the blood serum is in complement-fixing power in some instances and how little it varies in strength even when very small amounts of serum are used. The patient presented clinically a very marked secondary eruption over the chest, abdomen, and back. In Case 4, likewise, there was practically no variation in the strength of the reaction. In this case the patient pre- sented mucous patches in the mouth and a well marked secondary eruption. In Case 5, the variations in the strength of the comple- ment-fixing power of the blood serum varied markedly, 56 THE WASSERMANN TEST especially when amounts of serum of less than 0.1 c.c. wTere used. With the diagnostic amount of serum; i. e., 0.1 c.c., the serum gave a plus-minus, or practically negative result upon one day and a plus, or doubtful reaction upon two of the seven days during which it was tested. With 0.08 c.c. of the serum, an amount which was formerly used in the writer's laboratory as the diagnostic amount, a negative reaction was obtained upon one day and plus-minus upon one day and a plus, or doubtful reaction, upon two days. In this case the Wassermann reaction was practically nega- tive on one day and doubtful upon two days, although very marked secondary symptoms were present. In Case 6 the variation in the strength of the reaction was not very marked, w'hen the diagnostic amount of blood serum was used, the serum giving absolute inhibition of hemolysis upon all but two days, when it gave a plus or doubtful reaction. The results of the titration of the blood serum in second- ary cases demonstrate that variations do occur daily in cases showing well marked symptoms of the disease and that a case giving absolute inhibition of hemolysis upon one day may be negative or doubtful upon a succeeding day. When small amounts of blood serum are used the variations in the strength of the reaction are very marked, in some instances, as in Case 5, while in others the strength of the reaction remains unchanged. (Case 3.) Results in the Latent Stage of Syphilis.-Four pa- tients in the latent stage of syphilis were tested, all of whom gave a history of definite symptoms of the disease, but in none had symptoms been present within one year. All of the men had received one or two injections of sal- varsan and all gave a double-plus (four-plus) Wassermann reaction shortly before the titrations of the blood serum were commenced. Tables XI, XII, XIII, and XIV give the results of the titrations. FACTORS INFLUENCING RESULT OF TEST 57 Table XI Results of Titration of Serum for Complement-fixing Strength in Case 7-Latent Syphilis DATE OF TEST AMOUNT OF BLOOD SERUM. C.C. CONTROL C.C. 0.02 0.04 0.06 0.08 0.1 0.1 Dec. 19 + - + + + + + + - Dec. 20 - + - + + + - Dec. 21 Mi + + + + + + + - Dec. 22 - + - + + + - Dec. 23 + + + + + + + + + + - Dec. 24 + + + + + + + + + + Dec. 25 + - + + + + - Table XII Results of Titration of Serum for Complement-fixing Strength in Case 8-Latent Syphilis DATE OF tEST AMOUNT OF BLOOD SERUM. C.C. CONTROL C.C. 0.02 0.04 0.06 0.08 0.1 0.1 Jan. 12 + + + + + + + + + - Jan. 13 + + + + + + + + + - Jan. 14 + + + + + + + + + - Jan. 15 + - + + + + + + + - Jan. 16 + - + + + + + + + - Jan. 17 + - + + + + + + + Jan. 18 + - + + + + + - Table XIII Results of Titration of Serum for Complement-fixing Strength in Case 9-Latent Syphilis DATE OF TEST AMOUNT OF BLOOD SERUM. C.C. CONTROL C.C. 0.02 0.04 0.06 0.08 0.1 0.1 Jan. 15 - + - + + + + + + - Jan. 16 - + - + + + - Jan. 17 + - + + + + ++ - Jan. 18 - - + - + + + - Jan. 19 - - - - - Jan. 20 - - + - + + + - Jan. 21 - - - + - + - - Table XIV Results of Titration of Serum for Complement-fixing Strength in Case 10-Latent Syphilis DATE OF TEST AMOUNT OF BLOOD SERUM. C.C. CONTROL C.C. 0.02 0.04 0.06 0.08 0.1 0.1 Jan. 15 - - + - + - Jan. 16 - + - + + + + - Jan. 17 - + - + + + - Jan. 18 - - + - + + - Jan. 19 - + - + - + - + - - Jan. 20 + - + - + + + Jan. 21 + - + + ++ + + ++ - 58 THE WASSERMANN TEST In Case 7 the strength of the complement-fixation re- action, using the diagnostic amount of serum, i. e., 0.1 c.c., was only sufficient upon three of the seven days that the serum was tested to give a plus or doubtful reaction, while with smaller amounts the variations in the strength were very marked. Upon these three days the test would have been reported as doubtful and if further tests had not been made, the diagnosis, owing to the absence of symptoms, would have been missed. In Case 8 there was no variation from a positive reaction with the diagnostic amount of serum during the seven days the titrations were made, although some variation occurred with smaller amounts of the serum. In Case 9 there were very marked variations in the strength of the reaction with the diagnostic amount of se- rum, a negative reaction being obtained upon one day, a plus-minus upon another, and a plus or doubtful reaction upon a third. A double-plus, or positive reaction, was ob- tained upon only four of the seven days that the serum was tested. With amounts of serum smaller than the diag- nostic amount of 0.1 c.c. the variations were very great. In Case 10 there was also great variation, from day to day, in the strength of the complement-fixation reaction, the diagnostic amount of serum giving a plus-minus reaction upon one day, a plus reaction upon three days, and a double- plus or positive reaction upon three days of the week. With smaller amounts of serum the variations were still more marked. General Discussion.-The most important practical point brought out by these observations on the daily titration of the blood serum of untreated syphilitics is that great varia- tions may occur in the complement-binding power of the se- rum of patients uninfluenced by any kind of treatment, and that these variations occur from day to day, so that one neg- ative examination, or even more, in a suspected case, is of absolutely no value in excluding the disease. In several FACTORS INFLUENCING RESULT OF TEST 59 of the cases mentioned the blood serum, if tested upon cer- tain days only, would have given a negative, or practically negative, result, although serum from the same case had been previously positive, and again became positive within a day or two. If this be true of untreated cases, some of them showing severe lesions of the disease, it will certainly be found true of a greater number of latent and treated in- fections, the class of cases in which a negative Wassermann reaction is so often considered decisive as to the absence of the disease. The results of these tests indicate the great value of repeated examinations, when a negative Wasser- mann is reported, before a patient is considered as cured, or before he is assured that he is not suffering from a syphilitic infection. O'Leary176 has recently confirmed the writer's results as to the daily variations in the strength of the Wassermann reaction. He performed Wassermann tests upon thirty pa- tients for seven successive days without any provocative medication, and found that in 30 per cent of the patients positive results were obtained as well as negative results, the patients all being negative at the beginning of the ex- periment, and he concludes that "The fact that 30 per cent of the patients thus examined yielded positive results in the course of the series furnishes convincing evidence that though a provocative effect may exist as such, much of the clinical value of the so-called provocative test lies in the series of Wassermann reactions which it calls for rather than in the results of an injection of arsphenamine. " A series of seven Wassermann tests appears to afford an ex- cellent means of demonstrating the fluctuations that are to be expected in the Wassermann reaction in the opinion of this writer. As to the cause of the variations in the strength of the reaction observed from day to day, it is evident that the antibodies and lipotropic substances present in the pa- tient's serum must vary from day to day, or such results 60 THE WASSERMANN TEST would not be obtained. It may be accepted as proved that there exist in the blood of syphilitics antibodies and lipo- tropic substances that are capable of fixing complement in the presence of a proper antigenic extract, and that these bodies depend for their origin on the presence in the body of living treponemas, for when the latter disappear from the lesions in experimental animals, as the result of treat- ment, the complement-fixation reaction also disappears. Whether these substances are produced by the reaction of the tissues to endotoxins liberated by the breaking down of dead treponemas; to toxins produced by the living trep- onemas ; or to the presence in the blood of the toxins them- selves; are all mooted questions, but these experiments prove that there must be a certain amount of them in the blood serum before a positive Wassermann reaction can be obtained, and that this amount varies considerably from day to day. It is but just to call attention to the importance of these results in explaining, possibly, the discrepancies between Wassermann reports from various laboratories where spec- imens of blood from the same individual were examined at different times, for it is evident that unless the same spec- imen of blood be examined no reliable conclusions can be drawn regarding the reports of different laboratories, a fact that should be borne in mind when it is desired to obtain a report from more than one laboratory upon a suspected individual. The factors which have been mentioned in this chapter as markedly influencing the results of the Wassermann test; i. e., the effect of alcohol; of bacteria growing in the blood serum to be tested: and normal variations in the power of the blood serum to bind complement, are all of practical importance and should all be considered in per- forming the test and in judging-of its results in any case where there is a question of the accuracy of the reaction. If this were done, there would be much less criticism of the Wassermann test than there is, and a much better un- derstanding of its reliability and its limitations. CHAPTER IV PREPARATION AND TITRATION OF THE REAGENTS USED IN THE WASSERMANN TEST The following list of apparatus and reagents needed in performing the Wassermann test is here inserted and it will be found to include all that is really essential for this work: List of Apparatus 1 Luer syringe, capacity 1 e.c. 1 Luer syringe, capacity 5 c.e. 1 Luer syringe, capacity 20 c.c. 6 needles, for collecting blood, (special needles for this purpose can be obtained of any good surgical supply house). 1 electric centrifuge, International Instrument Co., Size 1, Type A, with one four 15 c.c. tube head and metal holders, and one four 50 c.c. tube head and metal holders. Type of current and voltage must be specified. 20 centrifuge tubes, plain, 15 c.c. 10 centrifuge tubes, graduated, 15 c.c. 10 centrifuge tubes, 50 c.c., plain. 1 incubator, either electric, gas, or oil, with proper thermoregu- lators, to run at 56° C. 1 water-bath, with thermometer, burner and thermoregulator, large enough to accommodate the required number of Wassermann racks, to run at 37° C. 1 set of 20 copper test tube racks, made by Topham, Wash- ington, D. C. 1 amboceptor cutter, made by Topham, Washington, D. C. 1 set of 3 amboceptor markers, 4 mm., 5 mm., and 6 mm., made by Topham, Washington, D. C. 1 paraffin bath to fit incubator mentioned above, to hold tubes during inactivation of blood serum. 100 or more test tubes, without lip, 100 by 12.5 millimeters. 4 test tube baskets, 6 inches high by 7.5 inches in diameter. 10 Mohr pipettes, 10 c.c. capacity, graduated in tenths. 10 Mohr pipettes, 5 c.c. capacity, graduated in tenths. 61 62 THE WASSERMANN TEST 20 pipettes, serologic, 1 c.c. capacity, graduated in hundredths to tip. These pipettes should be of small bore and the graduations far enough apart to render accurate read- ings possible for quantities as small as one-one-hundredth of a c.c. 100 or more pipettes, serologic, capacity 1 c.c., graduated in tenths. These are used for measuring the sera to be tested. 12 Erlenmeyer flasks, capacity 250 c.c. each. 6 animal cages. 20 pounds soft paraffin, melting point 43° C. 1 bottle, 500 gm. sodium chloride, C. P. 1 bottle, 30 gm. cholesterin, Merck's. 1 bottle, 500 gm. absolute alcohol. 6 triangular files, 3 inches long. Guinea pigs and rabbits must be kept for complement and amboceptor production and a small and suitable room or animal house should be set aside to shelter them and to allow for breeding. Preparation of Reagents The proper preparation of the reagents used in the Was- sermann test is of great importance and explicit directions are given in the following pages regarding this subject, which, if carefully adhered to, will result in the prepara- tion of reagents that will give accurate results. The Wassermann test as performed in the writer's lab- oratory and in most of the laboratories connected with the Medical Department of the Army, is a modification of both the test as originally recommended by Wassermann, and the modification of that test recommended by Noguchi. The technic was originally worked out by the writer and has since been standardized, so far as the apparatus is con- cerned, by Lieut. Colonel Vedder, of the Medical Corps. A human hemolytic system is employed instead of the sheep system recommended by Wassermann; alcoholic ex- tracts of both fetal syphilitic liver, and cholesterinized alcoholic extracts of normal human heart muscle, are used as antigen; and the blood serum is inactivated at 56° C. for PREPARATION AND TITRATION OF REAGENTS 63 one-half hour before it is tested. The following reagents are employed in the test: 1. Complement. The fresh blood serum of guinea pigs. 2. Hemolytic amboceptor. The blood serum of rabbits immunized to human red blood corpuscles. 3. Hemolytic antigen. A suspension of human red blood corpuscles. 4. Syphilitic antigen. An alcoholic extract of fetal syphilitic liver and a cholesterinized alcoholic extract of normal human heart muscle. 5. Syphilitic amboceptor. The patient's blood serum. This may or may not contain the lipotropic substance or substances necessary for complement fixation with the antigens mentioned. Preparation of Normal Salt Solution.-For preparing the erythrocyte suspension used in the complement-fixation test for syphilis and for making all dilutions of the re- agents employed, a normal salt solution should be used. This solution is prepared by dissolving 8.5 gms. of C. P. sodium chloride in 1000 c.c. of distilled water, thus making a 0.85 per cent solution. The solution should be made in a chemically clean glass flask and sterilized in an Arnold sterilizer. If an autoclave is used for sterilization there is danger that the loss of water in steam may concentrate the salt and render the solution unsuitable for use. After ster- ilization the solution should be tested by adding a drop or two of washed red blood corpuscles to 5 c.c. of the solution, and if no hemolysis has occurred at the end of half an hour, the solution is suitable for use in the test. The careful preparation of the saline solution is an essential part of the technic and the writer has many times been consulted regarding errors in the results of the test that were en- tirely due to the use of improperly prepared salt solutions. Preparation of Complement.-Complement is a constit- uent of the fresh blood serum of all animals but certain an- imals furnish better complement for complement-fixation 64 THE WASSERMANN TEST tests and for the Wassermann test than do others, and it has been demonstrated that for the Wassermann test the blood serum of the guinea pig supplies a better complement than does that of other animals. Noguchi and Bronfen- brenner tested the blood serum of the sheep, dog, ox, guinea pig, hog, and rabbit, and found that the serum of the guinea pig gave the best results when used as complement in the Wassermann test, so that the blood serum of this animal has been generally adopted for this purpose. The recent researches of Kolmer, Matsunami and Trist140 have shown conclusively that guinea pig blood serum is superior to that of the blood serum of other animals or man as a source of complement for the Wassermann test, and that such serum is generally free from natural hemo- lysins for human red blood corpuscles, but the work of Wil- liams141 has demonstrated that the blood serum of some guinea pigs contains natural hemolysins for the blood cor- puscles of individuals belonging to blood groups I and II, and in sufficient quantities to give rise to erroneous results as regards complement fixation if used as complement. In view of this fact, if the blood group of the individual from whom the blood corpuscles used in preparing the blood cell suspension for the test is unknown, the serum of each guinea pig to be used in the pooled complement should be tested for hemolytic properties by adding 0.2 c.c. of a 50 per cent dilution of the serum to 0.1 c.c. of a 5 per cent suspension of the red blood corpuscles in 0.9 c.c. of normal salt solution, and incubated for one hour in the water bath at 37° C. If there is any trace of hemolysis present at the end of this time the serum tested should not be used for complement. As will appear later, any chance of error from natural hemolysins which might be present in the complement or isohemolysins in the sera tested can be easily obviated by selecting an individual belonging to blood group IV as Fig- 3. Luer syringes. Set of seven, suitable for use in the Wassermann test, and for serologic work in general. PREPARATION AND TITRATION OF REAGENTS 65 donor of the blood corpuscles used in preparing the blood cell suspension. The blood serum of different guinea pigs varies consid- erably in complementary activity, even in health, and in unhealthy animals, and when the temperature is either very hot or cold, marked variations may occur. For this reason it is essential that the blood serum be titrated for its complementary power each time that it is used, and it is always best, in practice, to employ the serum from two or three guinea pigs, mixed, rather than to depend upon that of a single pig, which might be found to be so weak as to be inconvenient for use in the test. The guinea pig may be either bled from the heart or its throat cut. It is the practice, in the writer's laboratory, to bleed the pigs from the heart, as by this method the animal is not sacrificed and can be used over and over again for bleeding, provided the bleedings be from three to four weeks apart. Several lots of pigs can be rotated and thus considerable expense saved, as the animals usually stand the bleeding well and it is seldom that any are lost if the bleeding is properly done. The operation should be per- formed under ether and consists simply in entering the heart with a hypodermic needle attached to a syringe hold- ing 10 c.c. and nearly this amount of blood can be aspi- rated from the heart without danger to the animal. It is well to connect the needle of the syringe to the syringe with a flexible rubber tube, as it is much easier to make the puncture and handle the syringe afterward if this is done. The operation requires' considerable practice but when the technic is once mastered it is by far the most satisfactory way of obtaining complement for the Wassermann test, especially if only a small number of tests are to be made at one time. After bleeding, the blood should be ejected from the syringe into a sterile Petrie dish, allowed to stand at room temperature for two hours and then placed in the ice box 66 THE WASSERMANN TEST overnight. In the morning the clear serum is pipetted off and if it is to be used at once, diluted with one and one-half parts of normal salt solution- (0.85%). If it is not to be used for several hours it should be placed at once in the ice box, as the complementary power of blood serum is weak- ened by exposure to room temperature for any great length of time. The serum should be free from blood corpuscles and almost colorless or but very slightly blood stained. If a very large number of tests are to be made it may be better to sacrifice the pigs by cutting their throats, each pig being held over a Petrie dish, the throat cut quickly with a sharp scalpel or razor, after shaving the neck, and the blood allowed to run into the dish. Great care should be taken not to cut too deeply, so as to penetrate the esoph- agus or trachea, as the secretions from the stomach have a marked inhibitory effect upon complement and this is also true to a lesser degree of the secretions of the larynx or trachea. After collecting, the blood should be treated in the same manner as after puncture of the heart. The syringe, needles, Petrie dishes, and all apparatus used in collecting the blood for complement should be ster- ile and perfectly dry, with the exception of the syringe, which should be washed out wTith normal salt solution be- fore the puncture is made. If the blood serum, after be- ing pipetted off, is much blood stained, it should not be used for complement. If it contains blood corpuscles these may be removed by centrifuging and pipetting off the clear serum above the layer of corpuscles at the bottom of the centrifuge tube. Complement, even after being diluted one-half with nor- mal salt solution, may be kept in the ice box without much loss of strength for about three days, but usually upon the third day the titration of it will show such loss of complementary power that it will be useless for the test. Complement appears to be at the height of its com- plementary power about thirty-six hours after removal PREPARATION AND TITRATION OF REAGENTS 67 from the pig, but is much weaker immediately after with- drawal of the blood and grows gradually weaker from day to day after this period. Preservation of Complement.-Many attempts have been made to preserve complement by the addition of various substances, but none of the methods so far advocated have been generally adopted. Noguchi41 attempted to preserve it by impregnating filter paper with it, a method very suc- cessful with amboceptor serum, but without satisfactory results. Kolmer42 recommends chemically pure sodium chloride as the best preservative, adding 0.425 gram so- dium chloride to each 10 c.c. of sera obtained from several guinea pigs. He states that complement preserved in this manner will maintain its hemolytic and fixing properties for several weeks. An extensive series of experiments regarding the preser- vation of complement have been reported by Kolmer, Mat- sunami and Trist.142 They tried the various methods that have been recommended and found that whatever method be adopted the preserved complement should be kept as near the freezing point as possible and never above 40° C. In all they tested sixteen different methods and found that preservation with sodium chloride is most efficient. If the complement to be used is to be diluted 1-20, the addition of 0.17 grams of sodium chloride to each cubic centimeter and keeping it at a low temperature will preserve it for about three weeks. If a dilution of 1-10 or less is to be used Thompson's method of diluting the complement with equal parts of salt solution containing 4.1 per cent of sodium chloride or Neill's method, in which each cubic centimeter of complement is preserved with 0.1 c.c. of a saturated so- dium chloride solution in distilled water (36 per cent), were suitable if the mixtures were kept below 40° C. How- ever, with any of these methods they state that the hemo- lytic unit will be found only about one-half as active at the end of two weeks. 68 THE WASSERMANN TEST The writer has never obtained satisfactory results with any method of preserving complement and believes that for accurate work the complement should never be more than forty-eight hours old when it is used in the test. Preparation of Hemolytic Amboceptor.-The hemolytic serum or amboceptor used in the Wassermann test as per- formed by the writer is prepared by immunizing rabbits to human red blood corpuscles. Noguchi43 was the first to call to the attention of serologists the advantage of a hu- man hemolytic system in Wassermann work and the writer believes that it is the system best suited for complement- fixation reactions upon the blood of man. Its adoption ob- viates the danger that is present in human blood serum, of the action of the natural antisheep amboceptor normally present in such blood serum, and which, when the sheep hemolytic system is employed, should be removed from such serum before the test is made. It has not been prac- ticable, owing to the immense number of Wassermann tests made in army laboratories, to use any method which re- moves this natural anti sheep amboceptor from blood se- rum, and, therefore, aside from the superiority of the human hemolytic system for complement-fixation work, it was necessary to adopt it from practical considerations. The only objection that can be urged against the human system is that it is more difficult to immunize rabbits against human erythrocytes than against sheep corpuscles and that it requires a more careful technic both in collect- ing the blood and injecting it into the animals, but this objection is of no weight when the great advantages of the human hemolytic system are considered. An objection sometimes urged against the antihuman hemolytic system is that the blood serum of rabbits immun- ized to human red cells sometimes contains agglutinins to these cells which may interfere with hemolysis. This ob- jection is just as valid when other hemolytic systems are used, and, as will be seen later, such agglutinins may be PREPARATION AND TITRATION OF REAGENTS 69 rendered practically of no importance by impregnating the immune rabbit serum on filter paper. Before considering the method of preparing the hemo- lytic amboceptor serum there are certain general principles regarding amboceptors, either bacteriolytic or hemolytic, that will be mentioned. It will be remembered that amboceptors are produced in the blood serum of an animal in response to the injection of an antigen, and that these amboceptors are specific for the antigen injected. If the blood serum of the animal con- taining them is added to a mixture of the antigen and of fresh blood serum, or complement, the amboceptor will act as an intermediary body, allowing the complement to act upon the antigen, and bacteriolysis or hemolysis will result. Amboceptors are thermostable; i. e., their activity is not destroyed by heating the serum containing them for one half hour at 56° C. They are also resistant to a consider- able degree to acids and alkalies, and to drying. A good hemolytic serum can be preserved upon filter paper, in a dry condition, for many months with very little loss of strength and this fact is made use of in the Wassermann test, as will be noted later. Amboceptors appear to act as a chemical linking body, connecting the antigen with the complement, as the comple- ment is unable to act upon an antigen except through the agency of the amboceptor. Hemolytic amboceptors occur in very small amounts in the blood of normal animals and such amboceptors are sometimes present in an amount suf- ficient to cause trouble in complement-fixation reactions, as in the case of the normal antisheep amboceptor present in human blood serum. In producing hemolytic amboceptor for human red blood corpuscles in rabbits several methods have been recom- mended which will be described. The writer has obtained the best results with the method first described but the other methods are given in order that a choice may be made if desired, according to circumstances. 70 THE WASSERMANN TEST Well grown, perfectly healthy rabbits are selected, pref- erably pure white ones, and are given repeated injections of washed human red blood corpuscles. The blood corpus- cles are prepared for the injections in the following man- ner : The requisite amount of blood, allowing at least twice the amount of blood as the amount of corpuscles to be in- jected, is withdrawn from one of the large veins of the arm with a glass syringe, which has been sterilized and washed out with sodium citrate solution. After drawing the blood, it is at once ejected into a flask containing from 200 to 250 c.c. of normal salt solution (0.85%), and dis- tributed into large centrifuge tubes (50 c.c.) and centri- fuged until the erythrocytes are deposited and the super- natant fluid is practically colorless. The supernatant fluid is now poured or pipetted off and the tubes again filled with the salt solution, well shaken so as to distribute the red cells throughout the solution, and again centrifuged. This process is repeated four times, when the supernatant salt solution is tested for albumin; if the merest trace is demonstrated, the washing is again repeated and continued until no reaction for albumin is obtained. Almost invari- ably four washings will be found sufficient. Care should be taken not to centrifuge for so long that the corpuscles become packed at the bottom of the tubes, for the force necessary to dislodge them will break up a certain per- centage and render them unsuitable for injection. Care should also be taken to have each centrifuge tube weigh the same as all others, and this can be done by weighing each in a balance at the time they are filled. Unless this is done the centrifuge will run unevenly and is liable to breakage or accident. After the corpuscles are washed in the manner described they are mixed with as little normal salt solution, (0.85) which must be sterile, as is necessary to secure a suspen- sion that will pass through the needle of the injection syr- inge, and are at once injected, either intravenously or subcutaneously, with a sterilized syringe. Fig. 4. Electric centrifuge. PREPARATION AND TITRATION OF REAGENTS 71 Methods of Securing a Good Amboceptor Serum.-The methods of securing a good amboceptor serum vary some- what but the writer has found that the most rapid method of immunizing the rabbits and securing a good hemolytic serum is to give an intravenous injection, in the ear vein, of 1 c.c. of washed erythrocytes every other day until eight or ten injections have been given, when the rabbit's blood serum should be titrated and will generally be found strong enough in hemolytic properties to be used for the test. If it is not, two or three more intravenous injections of the same amount of red corpuscles should be given. Rarely, animals are found that will not develop a sufficiently strong serum, no matter how many injections may be given, but usually from eight to ten injections will be found sufficient. By this method it is very seldom that an animal is lost, as the frequent injections prevent the development of ana- phylaxis, and a good serum for use as amboceptor in the Wassermann test can be secured within a comparatively short period of time. Another good method is to give one subcutaneous injection of 5 c.c. of washed human erythrocytes in the abdomen, followed by from three to four intravenous in- jections in the marginal vein of the ear of from 3 to 4 c.c. of washed erythrocytes, at intervals of from five to six days. This method is much slower than the preceding, and not infrequently an animal is lost through anaphylactic phenomena. Noguchi44 recomTnends the following method of produc- ing hemolytic amboceptor for the Wassermann test: Five injections of washed human erythrocytes are given in- traperitoneally, the first of 5 c.c.; the second, 8 c.c.; the third, 12 c.c.; the fourth, 15 c.c.; and the fifth, 20 c.c.; at five day intervals. Nine days after the last injec- tion the rabbit is bled for a small amount of blood from the ear vein and the serum titrated. This method is an excel- lent one but consumes much more time than the method of 72 THE WASSERMANN TEST intravenous injections made every other day, recommended above. A method which has uniformly given excellent results is that in use at the Army Medical School. The rabbits are given daily injections, for three days, of 3 c.c. of the blood cell suspension intravenously; they are then allowed to rest for twenty-one days, and then given daily intravenous injections of 0.5 c.c. of the blood cell suspension for five days. At the end of from seven to nine days their serum is tested for hemolytic strength and generally will be found suitable for use. If not, five more intravenous injections of 0.5 c.c. of blood cell suspension are given, when almost invariably a good amboceptor serum will result. Williams has personally stated to the writer that he has never ob- served the death of an animal when this method of immuni- zation was employed. Another excellent method employed at the Army Medical School is to give a daily intravenous injection of 0.5 c.c. of the blood cell suspension for a period of twelve days. The blood serum of the rabbit is titrated five or six days after the last injection and will almost always be found to be suitable for use as amboceptor. If the injections are prop- erly made this is a very safe method for the rabbits, as it is very rare to lose an animal when it is employed. Preparation of Anti-Sheep Amboceptor.-While the writer does not recommend an antisheep hemolytic system for the complement-fixation test for syphilis, the following method of obtaining antisheep amboceptor is given for the convenience of those who desire to use this system. The rabbits are given five intravenous injections of 4 c.c. of the sheep cell suspension, which should be prepared as already described for the human cell suspension, at intervals of four days, and the serum tested for hemolytic strength from seven to nine days after the last injection. The in- jections should be given very slowly, as the rabbits not infrequently die from embolism unless this is done. In PREPARATION AND TITRATION OF REAGENTS 73 fact, all intravenous injections, no matter what species of blood cell is used, should be given slowly and carefully, and the utmost care should be taken that the suspension in- jected is free from foreign particles and blood cell clumps. Whatever method is used for immunizing the rabbits, the blood serum should not be tested for its hemolytic proper- ties until from 7 to 9 days have elapsed from the date of the last injection of red corpuscles. Preservation of the Amboceptor Serum.-The ambocep- tor serum may be preserved in ampules holding from 1 to 5 c.c. in the ice box, and used as required, but the writer has employed for nearly ten years the method of preserv- ing amboceptor serum for the Wassermann test recom- mended by Noguchi with the most satisfactory results and this method is used in most of the army laboratories. It consists in impregnating suitable filter paper with the blood serum, drying, and storing in a glass container in a dark closet or drawer. The method has entirely replaced the use of liquid amboceptor in our laboratories and pre- served in this manner the serum keeps much better, and preserves its strength for a much longer time, than when it is kept in liquid form, and the danger* of bacterial con- tamination, always present with the • liquid serum, is avoided. An objection to the use of amboceptor serum in liquid form is that the blood serum of rabbits used in the produc- tion of amboceptor may contain agglutinins to human red blood corpuscles which may interfere with hemolysis. These agglutinins are active in liquid serum but seldom interfere with hemolysis when the amboceptor serum is impregnated in filter paper and dried. Sands and West143 have definitely proved that drying antihuman amboceptor upon filter paper destroys very largely the hemagglutinins which may be present without injuring the hemolysins and they advocate this method of preparing amboceptor when an antihuman hemolytic system is employed. They also 74 THE WASSERMANN TEST found that filtration of antihuman hemolytic amboceptor through Kitasato and Chamberland filters removed a large amount of hemagglutinin with little or no loss of the hemo- lytic activity of the filtered serum. If filter paper which has been impregnated with anti- human serum be used within two weeks after impregna- tion, hemagglutinins, if present, may prove slightly trou- blesome, but if the paper is not used until the end of that time the writer has never observed any interference with hemolysis through the action of agglutinins. Before impregnating the filter paper, a preliminary ti- tration of the rabbit's blood serum should be made in order to determine whether it is strong enough in hemolytic amboceptor to use in the test. For this purpose a small amount of the rabbit's blood is collected from the ear vein in a capillary pipette or Wright tube and the serum allowed to separate. It is then separated from the clot and inac- tivated by heating it for one-half hour at 56° C. in the paraffin bath or water-bath, and one drop of it, from a capillary pipette, is mixed with 39 drops of sterile normal salt solution (0.85) and titrated for hemolytic strength. The method of titration is illustrated in Table XV. Table XV Preliminary Titration of Amboceptor Serum for Wassermann Test* TUBE NO. AMOUNT OF SALT SOLU- TION. 0.85%. c.c. NUMBER OF COMPLEMENT UNITS AMOUNT OF BLOOD SUS- PENSION. 5%. C.C. AMOUNT OF AMBOCEPTOR SERUM, DI- LUTED 1-40. DROPS Incubate i n water-bath at 37° C. for one hour or in incu- 1 0.9 2 0.1 1 bator at same 2 0.9 2 0.1 2 temperature for 3 0.9 2 0.1 3 two hours and 4 0.9 2 0.1 4 read results. 5 0.9 2 0.1 5 6 0.9 2 0.1 None *If the unit of complement has not been determined by a previous titration, use 0.1 c.c. of a 1:1.5 dilution in salt solution. After incubating in the water-bath at 37° C. for one hour, or in the incubator for two hours, the titration is PREPARATION AND TITRATION OF REAGENTS 75 read, and if either of the tubes containing one or two drops of the rabbit's serum shows complete hemolysis, the serum is strong enough in hemolytic amboceptor for use in the test and the rabbit should be bled at once. The control tube, No. 6, should show complete inhibition of hemolysis. Bleeding is accomplished most quickly by cutting the carotid vessels while holding the rabbit over a large glass dish, into which the blood is allowed to flow. The neck of the animal, where the incision is to be made, should be shaved and washed carefully but no disinfectant should be used. The glass dish to contain the blood should be ster- ilized before use and after the blood has been collected it should be kept at room temperature for about two hours and then placed in the ice box overnight. In the morning the clear serum should be pipetted off into small test tubes, inactivated by heating it for one-half hour at 56° C. in the paraffin or water-bath, and then kept in the ice box until one is ready to impregnate the filter paper. It does no harm if the blood serum is slightly tinged with hemoglobin, as the amount used in the test is so small that it will not interfere with the reading of the reaction. Impregnating the Filter Paper.-The filter paper used for this purpose is Schleicher and Schull's No. 597, but any other make of paper of approximately the same < texture will serve equally as well. The paper is cut into squares measuring 10 by 10 centimeters, and actual experience has shown that it will take about 1.5 c.c. of the rabbit serum to saturate one of these squares of paper, so that it is an easy matter to calculate how many squares will be required for any given amount of the serum. This number is placed, one by one, in a suitable Petrie dish or other glass dish containing the serum, and the process'continued until all the serum is absorbed by the paper. The squares of sat- urated paper are then carefully lifted with a pair of for- ceps, thoroughly drained, and drawn across the edge of the dish a few times to remove any excess of serum, and 76 THE WASSERMANN TEST then placed upon a piece of unbleached muslin, which should be large enough to contain all the slips of paper. The muslin containing the paper is now placed under an electric fan and the paper rapidly dried, after which it should be placed in air-tight glass containers and kept at room temperature in a dark and dry place. Before use, it should be titrated as will be described. The amboceptor paper thus prepared will retain its working strength for many months, but it should be ti- trated at least once a month in order to avoid any sudden loss of strength, although the writer has never observed such losses in the many years that he has used the ambo- ceptor paper. Paper over a year old has been repeatedly used and in practically the same dose as when it was first prepared and paper two years old was found to have lost only one-eighth of its original strength at the end of that time. Preparation of Hemolytic Antigen or Blood Cell Suspen- sion.-The hemolytic antigen used in the complement-fixa- tion test for syphilis recommended by the writer consists of a 5 per cent suspension of human red blood corpuscles, the corpuscles being preferably obtained from an individ- ual belonging to Blood Group IV, as regards hemolytic and agglutinative properties. The importance of using only blood from this group has only recently been recognized, although several investigators have mentioned it from a theoretical standpoint. The recent researches of Wil- liams144 have clearly shown that erroneous results may be obtained in some cases if the blood cell suspension is pre- pared with cells from individuals belonging to Blood Groups I and II, especially, and less often with cells from individuals belonging to Blood Group III, and that accuracy in the test is greatly augmented by using only Blood Group IV cells. His work, which the writer personally ob- served, has shown that an excess of amboceptor may result from natural hemolysins present in the guinea pig serum, Fig. 5. Same centrifuge illustrated in Fig. 4, with shaker head attached, for use in making antigenic extracts. PREPARATION AND TITRATION OF REAGENTS 77 used as complement, to human red blood corpuscles or to isohemolysins which may be present in the blood sera tested, and that this excess may be avoided by attention to proper grouping in selecting the red blood corpuscles used in making the blood cell suspension. Williams found that the blood serum of certain guinea pigs contained natural hemolysins to human red blood cor- puscles in quantities sufficient to cause complete hemolysis of 0.1 c.c. of a 5 per cent suspension of the cells, the usual amount of blood cell suspension used in the test, but that in such instances it was always found that the blood cell sus- pension was prepared with blood corpuscles from individ- uals belonging to Groups I and II, and that isohemolysins to Group I cells were frequently present in the blood sera submitted for examination and sometimes for Groups II and III. His experiments prove that if the blood cell suspension is prepared with the corpuscles of individuals belonging to Group IV, erroneous results, due to these hemolysins and isohemolysins, do not occur and that for the most accurate results when an antihuman hemolytic system is employed the blood cell suspension should be prepared with the cells of this group. If an antisheep hemolytic system is used, Williams has shown that the sheep used for securing blood cells for the blood cell suspension should be tested as regards the pres- ence in its blood corpuscles of receptors for hemolysins which may be present in the guinea pig serum used as com- plement, for he found that some sheep tested furnished blood corpuscles that were hemolyzed by guinea pig serum. From what has been said it is evident that the first step in the preparation of the human blood cell suspension used in the complement-fixation test for syphilis consists in ob- taining blood from an individual belonging to Group IV. Fortunately, this is easily accomplished, as individuals be- longing to this group constitute about 45 per cent of all 78 THE WASSERMANN TEST individuals, and the technic of blood grouping, which fol- lows, is very simple. Technic of Blood Grouping.-The classification of blood groups adopted is that of Moss, who places all individuals in one of four groups, as regards the hemolytic and agglu- tinative relations of their blood serum to one another. These groups are called Groups I, II, III and IV. The relation of these groups to one another is illustrated as follows: GROUP I GROUP II GROUP III GROUP IV I 0 + + + II 0 0 + + III 0 + 0 + IV 0 0 0 0 GROUP AGGLUTINATES BLOOD CELLS OF 0 = no agglutination. + = agglutination. For practical purposes, so far as the complement-fixation test for syphilis is concerned, it may be stated that agglu- tination is equivalent to hemolysis, so that in the prepara- tion of the human red blood cell suspension only blood from an individual belonging to Group IV should be used. The relative incidence of the four groups is approximately as follows: Group I, 5 per cent of individuals; Group II, 40 per cent; Group III, 10 per cent; and Group IV, 45 per cent. To determine the group to which an individual belongs the simple test devised by Vincent is recommended. Known blood serum from individuals belonging to Groups II and III are employed and the only apparatus required is a glass microscopic slide, a needle for making the puncture, and two very small glass rods. The known Group II and III sera can be obtained from any laboratory furnishing biologic supplies. The test is performed as follows: Place a small drop of Group II serum near one end of the glass slide, which should be clean and dry, and a drop of Group III serum near the other end, being very careful that the two sera do PREPARATION AND TITRATION OF REAGENTS 79 not come into contact. Puncture the lobe of the ear or the finger of the individual to be tested and transfer in turn to each of the drops of serum about one-third of a drop of blood by means of the end of the glass rod, using a separate rod for each transfer. Mix thoroughly the serum and the blood which should result in a uniform suspension. If agglutination of the blood cells occurs it will be noted that flocculi, visible to the naked eye, appear within a moment, as a rule, while if no agglutination occurs the suspension remains uniform in appearance. Great care should be taken to see that no mixture occurs between the two sus- pensions upon the slide. As a rule the reaction occurs within a minute but may be delayed slightly. If the results are questionable the blood should not be used for comple- ment fixation. If the blood of the individual tested is agglutinated by both Group II and Group III sera, it belongs to Group I; if it is agglutinated by Group III serum alone, it belongs to Group II; if it is agglutinated by Group II serum alone, it belongs to Group III; while if it is not agglutinated by either Group II or Group III serum it belongs to Group IV, and is suitable for use in making the blood cell suspension. As approximately 45 per cent of all individuals belong to Group IV, in which the blood serum does not agglutinate or hemolyze the corpuscles of individuals belonging to any of the other groups, the testing of the personnel of any laboratory will result in the finding of one or more individ- uals belonging to this group and from these individuals the blood used in preparing the blood cell suspension for the complement-fixation test for syphilis should be obtained. If, for any reason, the cells used in preparing the sus- pension are not from a known Group IV individual, the suspension should be tested with the serum of each guinea pig used for complement in order to ascertain whether any of the sera used for complement contain natural hemoly- 80 THE WASSERMANN TEST sins to the blood cells in the suspension. This may be done by adding to a tube containing 0.9 c.c. of normal salt solu- tion, 0.1 c.c. of the blood cell suspension and 0.2 c.c. of the guinea pig serum which has been diluted with one and one- half parts of salt solution. Incubate in the water-bath for one hour and if no hemolysis is present the guinea pig se- rum may be used for complement. In the Wassermann test, as recommended by the writer, each tube contains 0.1 c.c. of a 5 per cent suspension of human red blood corpuscles in 0.85 per cent salt solution. The quantity of blood required for a series of tests will depend, of course, upon their number. One cubic centi- meter of blood will furnish enough suspension for nearly one hundred tests, and unless one is doing a very large number of tests, enough blood may be easily secured by constricting the middle finger, after forcibly swinging the arm about for a few times, with a rubber tube, and prick- ing the finger just above the root of the nail with a sterile needle or glass point. The blood is allowed to drop into sterile graduated centrifuge tubes, each tube being filled to the 9 c.c. mark with sterile normal salt solution (0.85%), and enough blood is added to fill the tube to the 10 c.c. mark. The tubes are then shaken well and centrifuged until all the blood corpuscles are collected at the bottom. The su- pernatant liquid is then carefully poured, or pipetted off, the tubes again filled with salt solution, and again centri- fuged until the cells are at the bottom of the tube. This process of washing is repeated until the corpuscles have been washed four times, which will be found sufficient. The supernatant fluid is poured or pipetted off for the last time, and the amount of corpuscles at the bottom of the tube noted, and the tube filled with enough salt solution to make a 5 per cent suspension. Thus, if there are 0.5 c.c. of red corpuscles at the bottom of the tube, it would require 9.5 c.c. of salt solution to make a 5 per cent suspension. If only a small amount of blood is needed for a few tests, PREPARATION AND TITRATION OF REAGENTS 81 one centrifuge tube will be all that is required for the washing but if many tests are to be performed, the wash- ing should be done with a greater number of tubes. In the test, 0.1 c.c. of this suspension is added to each tube, which contains 0.9 c.c. of normal salt solution, thus making a half of one per cent suspension. This makes a suspension sufficiently strong to render the reading of the test easy and is much more saving of complement and am- boceptor than the use of a stronger suspension. Preparation of Antigens for Wassermann Test.-The syphilis antigen, as it may be called, as already stated, is not a true antigen, in the strict use of the term, as it is an extract containing lipoids which react with lipotropic sub- stances in the blood serum of the syphilitic patient, but as it acts like a true antigen in the Wassermann test, it is convenient to retain the name, but it should be understood that it is not a real antigen, as are the red blood corpuscles used in the hemolytic system just described. It is certainly true that in all alcoholic extracts of nor- mal tissues used as antigens in the Wassermann test, the active principle is contained within lipoids in the extracts and that there is nothing specific as regards Treponema pallidum, in these extracts. When alcoholic extracts of fetal syphilitic liver, and especially watery extracts of such livers, are used as antigens, the extracts, besides con- taining the lipoids present in the normal extracts, may also contain specific antigenic substances due to Treponema pallidum, as has already been shown. However, even in such extracts, the greater part of the reaction is undoubt- edly due to the lipoids present in them, although it can not be denied that in some instances, where the extracts of normal tissue give a negative or doubtful result, the ex- tracts of fetal syphilitic liver and of cultures of Trepo- nema pallidum give a positive one. Accordingly, the writer advises the use of both a cholesterinized alcoholic extract of normal human heart muscle and an alcoholic 82 extract of fetal syphilitic liver in performing the test. If both can not be obtained the best results will be secured by using cholesterinized alcoholic extracts of normal tis- sues, preferably human heart muscle. The extracts of pure cultures of Treponema pallidum, while giving interesting results from a scientific standpoint, are useless as antigen in routine Wassermann work because of their unreliability. In practical work the writer recommends the antigens al- ready mentioned, and, in addition, Noguchi's acetone- insoluble lipoid antigen, where the other antigens fail in apparently true syphilitic infections, but only in such in- fections. Alcoholic Extract of Fetal Syphilitic Liver.-A fetal liver, rich in Treponema pallidum is selected, washed free from blood, and any fat present removed. One hundred grams is weighed out, all blood removed by washing, and cut into very small pieces. The material is then placed in a suitable bottle, or other container, and 1000 grams of absolute alcohol, or 95% alcohol, added, and the whole placed in an incubator at 37° C. and allowed to remain for 10 days, being thoroughly shaken three times a day during that time, when it will be ready to titrate. If a shaking machine is available, the material, after adding the alco- hol, is placed in a suitable bottle or bottles, and shaken for 24 hours at a good rate of speed, at the end of which time it will be found that the tissue is thoroughly extracted. After extraction by either method is completed, the mate- rial is filtered through filter paper and the filtrate titrated. If the first titration is not satisfactory, the filtrate should be evaporated to two-thirds its original volume and again titrated. The antigenic extract should always be kept in the ice box except when in use. Cholesterinized Human Heart Extract.-The same method is used in preparing the alcoholic extract of heart muscle, the material being washed free from blood, all fat removed, and 100 grams of the heart muscle being cut into small pieces and extracted with alcohol in the same pro- THE WASSERMANN TEST Fig. 6. Small incubator, to run at 56° C., for inactivating sera. PREPARATION AND TITRATION OF REAGENTS 83 portions (1000 grams of alcohol to 100 grams of muscle). After extraction one-half of the material is filtered through paper, and 0.4% of cholesterin added, and titrated. If the titration does not result satisfactorily as regards the anti- genic strength of the extract, the remaining half is evapo- rated to two-thirds its original volume, again filtered through paper, and the cholesterin added, after which it is titrated. The amount of cholesterin added gives prac- tically a saturated solution and there is generally an ex- cess of cholesterin left after standing, and in using the antigen care should be taken not to get any of this excess into the pipette used in the work and thus in the diluted solution of the antigen. After the addition of the choles- terin the antigenic extract is placed in glass-stoppered bot- tles and kept in an ice box, except when in use. Noguchi's Acetone Insoluble Lipoid Antigen.-This type of antigen was recommended by Noguchi when he evolved his modification of the Wassermann test and it is an excel- lent antigen, especially for use in cases in which the ordinary antigens described give doubtful results, as it appears to be more delicate than the extracts already de- scribed. However, it is true that sometimes this antigen appears to give positive reactions in cases that are nega- tive with other antigens and in which syphilis can not be stated to exist upon clinical grounds, so that, in the writ- er's opinion, the use of this antigen should be restricted to the class of cases mentioned and as a control of other anti- gens, when doubtful results are obtained. It is prepared in the following manner: Normal organs are used and may be from man or the lower animals. Extracts made from heart muscle, liver, or kidney of man or the same organs of the guinea pig or rabbit are satisfactory and the extraction is done in the same manner as for the antigens already described, 10 parts of absolute alcohol being used for each part of tis- sue. After extraction, the material is filtered, and the 84 filtrate evaporated to dryness by the aid of an electric fan. The residue is taken up with ether and the solution thus obtained is allowed to stand overnight in a cool place. In the morning the supernatant ether will be found perfectly clear while the insoluble portion of the residue has settled to the bottom of the receptacle. The clear etheral portion is decanted off and evaporated to a small quantity and mixed with ten volumes of pure acetone. The precipitate which forms is allowed to settle and the supernatant liquid poured off. The precipitate is light brown in color and forms the material from which the antigen suspension is made, 0.3 gram of this acetone insoluble fraction being dis- solved in 1 c.c. of ether and then mixed with 9 c.c. of methyl alcohol. This mixture may be kept for a long time and the antigen suspension used in making the test is prepared by adding 1 c.c. of this stock solution to 9 c.c. of normal salt solution, or if less is needed at the time of making the tests, the same proportion of antigen and salt is preserved. Wassermann's Aqueous Extract Antigen.-The antigen employed by Wassermann in his method of performing the complement-fixation test for syphilis was prepared as follows: A fetal syphilitic liver, rich in treponemas, is cut into very small pieces and weighed. Four times its weight of 0.5 per cent phenol in physiologic salt solution is added and the mixture placed in a brown bottle and shaken in a shaking machine at room temperature for twenty-four hours. It is then filtered through gauze and placed in a brown bottle in the ice box for several days, after which it is titrated to determine its antigenic strength. A sediment forms in the bottle after standing in the ice box and one should be careful not to disturb this in using the antigen. This antigen is not satisfactory, as it varies very greatly in strength when made from different livers, and is very prone to lose strength rapidly, especially if it is left for any length of time at room temperature. It is not nearly as satisfactory as the antigens heretofore described. THE WASSERMANN TEST Fig. 7. Water-bath, suitable for Wassermann test incubations. PREPARATION AND TITRATION OF REAGENTS 85 Various other methods of preparing antigenic extracts for the Wassermann test have been described by Marie and Levaditi,45 Landsteiner, Muller, and Pdtzl,46 Porges and Meier,47 Browning, Cruickshank, and McKenzie,48 and others, but none of them are better in practical use than those which have been treated of in the foregoing para- graphs. Cholesterinized Antigens.-Before concluding this por- tion of our subject it may be well to discuss briefly the sub- ject of the use of cholesterinized antigens in the Wasser- mann test. The writer advocates their use, or, at least, the use of an extract of human heart muscle, cholesterinized, be- cause he has found by long experience that such an extract is perfectly reliable and does not give false positive reac- tions, as shown by the results in the many thousands of tests in which he has used such an extract. Some author- ities claim that, in their experience, cholesterinized ex- tracts are so very sensitive that they may give false posi- tive results but this has not been the writer's experience. Very rarely there has been observed a plus or plus-minus reaction in apparently normal serum, but the writer has never observed a distinctly positive reaction with the cholesterinized heart extracts used in his method in a se- rum which could be proved to be normal, but, on the other hand, numerous instances have been observed in which ordinary extracts gave doubtful or negative results but in which the cholesterinized heart extract gave positive re- actions. This is very apt to be the case in cases of syphilis that have been treated and, in the writer's opinion, the cholesterinized extract is far superior, as a control of treatment, than are extracts of normal tissues or of foetal syphilitic liver, to which cholesterin has not been added. Many authorities have recommended antigens prepared from the heart muscle of various animals, as the ox, sheep, dog, rabbit and guinea pig. The writer has always be- lieved that the best results are obtained with antigens pre- 86 THE WASSERMANN TEST pared from human heart muscle and has found that anti- gens prepared from guinea pig heart are not reliable. His observations regarding guinea pig heart alcoholic extracts, when used as antigen, have recently been confirmed by Ruediger,145 who tested 50 clinically nonsyphilitic cases, using antigens prepared from the heart muscle of man, ox, dog, rabbit and guinea pig. The antigen prepared from guinea pig heart gave strong positive results in 16, or 32 per cent of the cases tested, and positive results were ob- tained in the same number with antigens prepared from dog heart. Therefore, it is evident that antigens prepared from either guinea pig or dog heart should not be used in the complement-fixation test for syphilis. Titration of Reagents Used in the Wassermann Test. The accurate titration of the various reagents used in the Wassermann test is of fundamental importance, as the test is essentially a quantitative one and the smallest vari- ation in the amount of the reagents used may cause very great variations in the result of the test. If the amount of complement or amboceptor be excessive the strongest positive serum may give a negative result, while if the amount of these reagents is too small, normal sera will give positive results. The same is true of the amount of antigen, for an excess of antigen will result in a positive result in normal sera or in the blood sera from other dis- eases than syphilis, while too little will result in negative reactions in what should be found to be strongly positive sera. It is easily seen, therefore, that the accurate titration of complement, amboceptor, and antigen is of the greatest practical importance, and the neglect of such titrations will be sure to lead to erroneous results and to great injustice being done the patient who has entrusted the testing of his blood to the laboratory. Titration of the Complement.-The amount of comple- ment to be used in the Wassermann test is of vital impor- PREPARATION AND TITRATION OF REAGENTS 87 tance and is determined by a titration of the guinea pig serum each day before the tests are made. Some author- ities employ exactly one unit of complement with one unit of amboceptor for each test but this allows of no inhibitory effect by either the blood serum or antigen, and, is there- fore, a dangerous proceeding unless very carefully con- trolled. If the titrations are made with the amount of patient's serum added to the complement and amboceptor the fallacy residing in the possible anticomplementary action of the blood serum would be overcome, but this is a cumbersome procedure, especially when a large number of tests are to be made, and, in addition, there is still the factor of the natural inhibition of hemolysis which is pres- ent in almost all antigens. It is, therefore, much better to use an excess of complement in making the tests and it has been found perfectly safe in practice to double the unit of complement found by titration, when actually do- ing the test. In other words, two units of complement are used for each tube employed in the test. In order to titrate the complement it is necessary that a standard amount of amboceptor and a standard suspension of human red blood corpuscles be adopted and strictly ad- hered to, although the strength of the blood suspension may be varied, provided the complement be titrated with it before each series of tests. Standard doses of ambo- ceptor are adopted because this reagent changes little in strength during long periods of time, while if the ambo- ceptor were to be titrated before each series of tests, a standard dose of complement would have to be adopted and the latter varies so greatly in strength that it is pref- erable to titrate the complement with a standard dose of amboceptor. In the titration of the complement two units of ambo- ceptor, as determined by titration, are employed, or the amount of amboceptor that is actually used in the test. The writer formerly used one unit of amboceptor in titrat- PLATE I Fig. 1.-Titration of Complement. Each of the tubes contains the amount of complement marked above the tube. In the titration, as illustrated, hemo- lysis is complete in the fourth tube, containing 0.05 c.c. of complement. Hence 0.05 c.c. is the complement unit and twice this amount is used in the actual test, or 0.1 c.c. The tenth tube is a control tube containing no complement and it should show complete inhibition of hemolysis, as illustrated. Fig. 2.-Titration of Amboceptor Papers. Each of the first five tubes con- tains a piece of amboceptor paper of the size indicated above the tubes. In the titration illustrated the first tube showing complete hemolysis is the third tube, containing a piece of paper measuring 5 by 3 mm. Hence the unit of ambo- ceptor paper is a piece measuring 5 by 3 mm., and twice this amount is used in the test, or a piece measuring 5 by 6 mm. The sixth tube is a control tube containing no amboceptor paper and should show complete inhibition of hemolysis, as illustrated. Fig. 3.-Titration of Antigen. Tubes 1, 2, 3, and 4 contain respectively 0.05, 0.1, 0.15, and 0.2 c.c. of a 1 in 10 dilution of the antigenic extract, with 0.1 c.c. of a known syphilitic serum. Tube 5 contains 0.1 c.c. of the syphilitic serum without antigen. Tube 6 contains 0.1 c.c. of a normal blood serum with 0.2 c.c. of the diluted antigen. Tube 7 contains 0.1 c.c. of the normal serum without antigen. Tube 8 contains complement without amboceptor, while Tube 9 contains the hemolytic system, acting as a control of the latter. A good antigen should give the results illustrated in this titration, causing complete inhibition of hemolysis in the presence of 0.05 c.c. of a syphilitic serum, as shown in Tube 1, and no inhibition of hemolysis wfith a normal serum in amounts as large as 0.2 c.c., as shown in Tube 6. Complete inhibition of hemolysis should occur in Tube 8 and complete hemolysis in 5, 6, 7, and 9. O'OZcc. o-o5cc. o-oHc.c. o-o5cc o-obcc oo7c.c o o8cjc o o9cc o lcc Control Fig.I. Titration of Complement 5x1mm. 5x2mm. 5x3mm. 5x^mm. 5x5mm. Control Fig.2. Titration of Amboceptor Papers. 12 3 4 5 6 7 d 9 Fig.3. Titration of Antigen. Plate I. PREPARATION AND TITRATION OF REAGENTS 89 ing complement for his modification of the complement- fixation test for syphilis but experience proved that with the dose of complement doubled, which was ascertained by titrating with only one unit of amboceptor, weak positive cases were missed, owing to the sera giving partial reac- tions, and that this was especially apt to occur in treated cases, thus destroying the value of the test as an index of the efficiency of treatemnt or its diagnostic value in this class of cases. By using two units of amboceptor in the titration of complement the complement unit obtained, when doubled, as it is for the test, is never so weak as to give false positive results; neither is it so strong as to ren- der a weak positive serum negative. The complement, am- boceptor and blood cells should be all added at once, for if the amboceptor and the blood cells are left in contact, without the complement, for any length of time, the cells will become "sensitized" and a smaller amount of com- plement will produce hemolysis. In making the titration of the complement, either a 5 or 10 per cent suspension of human erythrocytes, in normal salt solution, may be used, but the writer prefers the 5 per cent suspension as it is more economic of amboceptor and is of sufficient strength to render the readings of the reac- tion easy. It should be remembered that the stronger the blood suspension is the greater will be the amount of am- boceptor and complement necessary to produce hemolysis, and the more extravagant the expenditure of these re- agents, so that it is foolish to employ a stronger blood sus- pension for the test than is necessary to enable one to readily read the reactions, as the hemolytic system is merely used as an indicator of the presence in the patient's blood of the substances that absorb or bind complement. In titrating the complement, test tubes measuring 100 by 12.5 millimeters are used. These tubes should be per- fectly clean and, while they do not need to be sterile, the best technic is to have them sterilized. The complement 90 THE WASSERMANN TEST is diluted with one and one-half parts of normal salt solu- tion (0.85 per cent) and graduated amounts are added to tubes containing the blood suspension and amboceptor, as illustrated in Table XVI. Table XVI Titration of the Complement TUBE NO. AMOUNT OF SALT SOLU- TION c.c. AMOUNT OF BLOOD SUS- PENSION, 5% C.C. NUMBER OF AMBOCEPTOR UNITS AMOUNT OF COMPLEMENT C.C. Incubate i n water-bath a t 37° C. for one 1 0.9 0.1 2 0.02 hour and read. 2 0.9 0.1 2 0.03 The tube show- 3 0.9 0.1 2 0.04 i n g complete 4 0.9 0.1 2 0.05 hemolysis con- 5 0.9 0.1 2 0.06 tains one unit 6 0.9 0.1 2 0.07 of complement. 7 0.9 0.1 2 0.08 Two units are 8 0.9 0.1 2 0.09 used in making 9 0.9 0.1 2 0.10 the test. 10 0.9 0.1 0 0.10 After incubating in the water-bath at 37° C. for one hour, or in the ordinary bacteriological incubator, for two hours at the same temperature, the titration is read, and the tube that contains the smallest amount of complement which shows complete hemolysis is noted and this amount is called one unit of complement. For instance, if it is found that Tube No. 4, containing 0.05 c.c. of complement, is the first of the tubes to show complete hemolysis, the tubes preceding it showing only partial hemolysis, then 0.05 c.c. of complement constitutes one unit of complement, and in performing the actual test twice this amount or 0.1 c.c. will be used for the complement dose for each tube, or two units of complement. The amount of complement is doubled in order to provide for the anticomplementary action of the patient's blood serum and of the antigen, and has been found in practice to be sufficient to answer this purpose and at the same time not to cause fallacies in the test. In reading the titration if there is the slightest cloud- iness in any of the tubes hemolysis can not be said to be PREPARATION AND TITRATION OF REAGENTS 91 complete and in selecting the tube which represents one complement unit, the first that shows an absolutely crystal clear solution should be noted. The control tube, Tube No. 10, which contains no amboceptor, should show complete inhibition of hemolysis. If it does not, there is something wrong with the complement or the salt solution used in making the titration, and the titration should be repeated with new reagents. This Titration of the Complement Must be Made Before Each Series of Tests or Very Serious Errors Will De- velop.-The blood serum of guinea pigs varies greatly in complementary strength, so that it is excessively danger- ous to use a stated amount of the serum for complement. A common observation, upon titrating different samples of guinea pig serum, is to find differences in complementary strength all the way from a unit equalling 0.03 c.c. to one of 0.08 c.c. and sometimes a serum will be found much stronger or weaker than these limits, so that the danger of using any definite amount is apparent. The complement must be carefully titrated before each series of tests. Titration of the Hemolytic Amboceptor.-As already stated, the hemolytic serum or amboceptor used in the Wassermann test as recommended by the writer is pre- pared by immunizing rabbits to human red blood corpus- cles. The human hemolytic system, as it is called, was adopted in order to obviate the danger arising from the natural antisheep amboceptor present in human blood se- rum, and which is sometimes present in quantities suffi- cient to render a positive syphilitic serum negative unless special precautions are taken to guard against such an occurrence. While this result is not frequent in practice, when the sheep hemolytic system is used, the fact that it may occur, and that, in order to be perfectly safe, it is necessary to remove from each patient's blood the anti- sheep amboceptor that may be present, renders the use of the human hemolytic system preferable, both from a scientific and practical point of view, and it is the system 92 THE WASSERMANN TEST the writer has always urged for complement-fixation tests upon human blood serum. The hemolytic amboceptor, having been prepared by impregnating suitable filter paper with it, as already de- scribed, must be titrated in order to determine the ambo- ceptor unit. After the paper is thoroughly dry a strip is cut from one of the sheets measuring 5 millimeters in width and varying lengths of this strip are used in the titration, as shown in Table XVII. The outer edge of the saturated sheet of filter paper should be trimmed off be- fore cutting the strip referred to, in order to avoid any concentration of the serum which might occur at that por- tion of the paper. Table XVII Titration of Amboceptor Paper* TUBE NO. AMOUNT OF SALT SOLU- TION c.c. AMOUNT OF COMPLEMENT UNITS AMOUNT OF BLOOD SUS- PENSION C.C. AMOUNT OF AMBOCEPTOR PAPER MM. Incubate i n water-bath a t 37° C. for one hour or in incu- bator for two hours at same t e m p e rature, and read re- sults. 1 2 3 4 5 6 0.9 0.9 0.9 0.9 0.9 0.9 2 2 2 2 2 2 0.1 0.1 0.1 0.1 0.1 0.1 5 by 1 5 by 2 5 by 3 5 by 4 5 by 5 None *If the unit of complement has not been determined against a known amboceptor by titration, use 0.1 c.c. of a 1:1.5 dilution of the complement as the unit. It may be neces- sary to carry the titration on further by using larger amounts of the amboceptor paper, but a good paper should give hemolysis in a piece 5 by 5 millimeters. After incubating in the water-bath at 37° C. for one hour, or in the incubator, at the same temperature, for two hours, the tubes being shaken every fifteen minutes in or- der to facilitate the liberation of the amboceptor serum, the titration is read and the first tube to show complete hemolysis is noted and the amount of the paper contained in that tube is called one amboceptor unit. A good ambo- ceptor paper should show complete hemolysis in tubes Nos. 2 or 3, or with pieces of paper measuring 5 by 2 or 3 millimeters, but paper measuring 5 by 5 millimeters can be used in the test. In many instances, the first tube will show complete hemolysis and, when this occurs, the paper Fig. 8. Amboceptor cutter. Front view showing gauge. Devised by Lieut. Colonel F. B. Vedder, Medical Corps, U. S. Army, for cutting amboceptor paper. (From Bulletin No. 8, War Department, Washington, D. C., Office of the Surgeon General, U. S. Army, 1915.) PREPARATION AND TITRATION OF REAGENTS 93 may be reduced in width, or the strength of the blood sus- pension may be increased, a ten per cent suspension being used instead of a five per cent suspension. The control tube, No. 6, should not show any hemolysis and if it does it indicates that something is wrong with the blood sus- pension or the complement, and the titration will have to be repeated with new reagents. In making the actual Wassermann test hvo units of amboceptor paper should be used in order to allow for the anticomplementary qualities of human blood serum and of the antigen. Thus, if one unit of the paper was found to be a piece measuring 5 by 2 mm. a piece 5 by 4 mm. should be used in making the test. Amboceptor paper, being very stable, does not need to be titrated very frequently, once a month being amply suf- ficient, and such paper will keep, with little loss of strength, for several months. Titration of the Syphilitic Antigen.-The antigens used in the Wassermann test must be carefully titrated and such titrations should be made at frequent intervals, as all antigens used in the test lose strength with more or less rapidity. To be on the safe side, the antigen should be titrated at least once in two weeks unless it be found that the particular antigens that are employed are very stable, when monthly titrations may be sufficient. A really good antigen loses strength slowly and the writer has often worked with extracts that were practically as strong in antigenic qualities after several months as when first pre- pared, but the reverse is often true, and an antigen will be found that rapidly loses strength, thus making frequent titrations advisable. If possible, it is best to titrate the antigen before each series of tests. An antigen must be titrated with a known syphilitic se- rum in order to determine its antigenic qualities; with a normal serum in order to demonstrate that it will not in- hibit hemolysis when used with such a serum; and in the 94 THE WASSERMANN TEST absence of either a syphilitic or normal serum, in order to determine whether or not it is anticomplementary or hemo- lytic of itself. Before titrating, the antigen should be properly diluted, which is accomplished by adding one part of the extract to nine parts of normal salt solution (0.85%). Any antigenic extract prepared in the manner recom- mended by the writer and diluted so as to produce an emulsion containing one part of the extract to nine parts of salt solution, that is neither anticomplementary nor hemolytic of itself or in the presence of a normal serum in a dose of 0.2 c.c. and that gives inhibition of hemolysis in the presence of a syphilitic serum in a dose of 0.05 or 0.1 c.c., will be found satisfactory in practice. Table XVIII Titration of Antigen for Hemolytic Properties TUBE NO. AMOUNT OF SALT SOLU- TION c.c. UNITS OF COMPLEMENT AMOUNT OF ANTIGEN EMULSION C.C. AMOUNT OF BLOOD SUS- PENSION. 5% C.C. Incubate water-bath. 37° C. for i n a t one 1 0.9 2 0.05 0.1 hour or in in- 2 0.9 2 0.10 0.1 cubator for two 3 0.9 2 0.15 0.1 ► hours, and read 4 0.9 2 0.20 0.1 results. 5 0.9 2 None 0.1 Table XVIII illustrates the titration of antigenic ex- tracts for the purpose of determining their hemolytic properties. The quantities of antigenic solution tested have been determined by a large number of experimental tests to yield satisfactory results. In this titration none of the tubes should show hemolysis. If hemolysis occurs in all, including Tube 5, the control tube without antigen, it would demonstrate that hemolytic substances were present either in the complement or blood suspension, or both; but if hemolysis should occur in one or more of the tubes containing the antigen, the control tube remaining unhemolyzed, it would demonstrate that the antigenic extract is hemolytic; and, in this event, it should be discarded unless, upon being further diluted with Fig- 9. Amboceptor cutter. Rear view, showing scale and set screw. Devised by Dieut. Colonel E. B. Vedder, Medical Corps, U. S. Army. (From Bulletin No. 8, War De- partment, Office of the Surgeon General, U. S. Army, Washington, D. C.) PREPARATION AND TITRATION OF REAGENTS 95 salt solution, it is found to give satisfactory results. If none of the tubes containing antigen show hemolysis, the antigen is not hemolytic, and therefore, so far as hemo- lytic properties are concerned, it is safe to use in making the Wassermann test. In order to determine whether antigens are anticomple- mentary or not, they are titrated as shown in Table XIX. Table XIX Titration of Antigen for Anticomplementary Properties A o £ . Eh % OF C EMUL- Incubate in water- bath at 37° C. for one- 55" 0 S 55 M ft P O M . ' AMBO- APER Incubate in water- bath at 37° C. for one d p o h O 55 M ° o g Eh 2 half hour or Eh p 0 «2 ri © hour or in 55 TO ^8 in incubator 55 ft . to © incu b a t 0 r M n AMOl SALT TION EH ~ o g g for one hour 6 0 0 Eh £h 5 Ph for two E £ O P O 3 1? 2 <! <! 02 3 <5 m 10 g a P Q hours at same t e m- perature and read results. 1 2 3 0.9 0.9 0.9 2 2 2 0.05 0.10 0.15 0.1 0.1 0.1 2 2 2 4 0.9 2 0.20 0.1 2 5 0.9 2 None 0.1 2 As the result of this titration there should be complete hemolysis in all of the tubes. If Tube 5, the control tube, should show inhibition of hemolysis, it would prove that either the complement, amboceptor, or blood suspension, was inhibitory, but if any of the tubes containing antigen show inhibition, the control tube being hemolyzed, it would demonstrate that the antigen is anticomplementary, and, unless further dilution will remove this property without injuring its antigenic value, the extract should not be used for the test. Having tested an antigen for its hemolytic and anticom- plementary properties, it is now necessary to titrate it against a known syphilitic blood serum to determine its antigenic value, and with a known normal blood serum to demonstrate that it will not give a positive reaction with nonsyphilitic sera. This titration is illustrated in Table XX. 96 THE WASSERMANN TEST TUBE NO. AMOUNT OF SALT SOLU- TION C.C. AMOUNT OF LUETIC SE- RUM INACTI- VATED C.C. UNITS OF COMPLE- MENT AMOUNT OF ANTIGENIC EMULSION C.C. AMOUNT OF BLOOD SUS- PENSION, 5% C.C. UNITS OF AMBOCEPTOR PAPER 1 2 0.9 0.9 0.1 0.1 2 2 0.05 0.10 Incubate in wa- 0.1 0.1 2 2 Incubate in wa- ter-bath at 37° C. 3 0.9 0.1 2 0.15 ter-bath at 37° C. 0.1 2 for one hour or in 4 0.9 0.1 2 0.20 for one-half hour 0.1 2 incubator for two 5 0.9 0.1 2 None or in incubator for 0.1 2 hours at same tern- 6 0.9 Normal 2 0.20 one hour. 0.1 2 perature and read 7 8 9 0.9 0.9 0.9 serum 0.1 Normal serum 0.1 None None 2 2 2 None None None 0.1 0.1 0.1 2 None 2 results. Table XX Titration of Antigen to Determine Antigenic Properties Fig. 10. Amboceptor markers. Devised by Lieut. Colonel E. B. \ edder, Medical Corps, U. S. Army, for accurately measuring units of amboceptor paper. (From Bulletin ho. 8, War Department, Office of the Surgeon General, U. S. Army, Washington, D. C.) PREPARATION AND TITRATION OF REAGENTS 97 As the result of this titration all of the tubes containing antigen should show complete inhibition of hemolysis in which syphilitic blood serum is also present and complete hemolysis in the tubes in which normal blood serum is present. Thus Tubes 1, 2, 3, and 4 should show absolute inhibition of hemolysis, while Tube 6 should show complete hemolysis. If Tube 6 shows any trace of inhibition, the antigen is not suitable and should be discarded; and the same is true if there is less than complete inhibition of hemolysis in Tube 2. There may be a slight trace of hem- olysis in Tube 1 containing only 0.05 c.c. of the antigenic emulsion, without harm; but it will always be found better in practice to use only those antigenic extracts that give complete inhibition of hemolysis with 0.05 of the diluted antigen. Tube 7, which is a control of the normal serum used, should show complete hemolysis, and if it does not, another serum should be tested. Tube 8, the control of the complement, should show complete inhibition of hemolysis; and Tube 9, the control of the hemolytic system, should show complete hemolysis. If these results are obtained, they demonstrate that the antigen, as diluted, is suitable for use in the Wassermann test as recommended by the writer, in that it produces inhibition of hemolysis in the presence of syphilitic blood serum in a dose as small as 0.05 c.c., and that it does not inhibit hemolysis in the presence of a normal serum in four times that dose, or 0.2 c.c. Such an antigen will give excel- lent results and even one that gives inhibition of hemolysis with a syphilitic blood serum in a dose of 0.1 will be found satisfactory in practice, provided it is otherwise all right. The amount of antigenic emulsion necessary to produce complete inhibition of hemolysis in the presence of a syph- ilitic blood serum is known as an antigen unit. The anti- gen unit is not doubled in making the test as is the unit of complement and amboceptor. 98 All blood sera used in titrating the antigen should be inactivated by heating for one-half hour at 56° C. and a neglect of this will produce most erroneous results. Antigens should be kept in the ice box when not in use as exposure to room temperature for any length of time will result in their rapidly losing strength. THE WASSERMANN TEST CHAPTER V THE TECHNIC OF THE WRITER'S MODIFICATION OF THE WASSERMANN TEST. THE TEST UPON THE CEREBROSPINAL FLUID. THE ORIGINAL WASSERMANN TECHNIC. OTHER MODIFI- CATIONS OF THE WAS- SERMANN TEST. The technic of the writer's modification of the Wasser- mann test was first published in 1913, in Bulletin No. 3, War Department, Office of the Surgeon General, Washing- ton, D. C. This method of performing the test has been employed by the writer since 1910 and since 1913 it has been the standard method adopted in nearly all of the large army laboratories and has been recommended for use in all army laboratories by Lieut. Colonel Vedder, of the Medical Corps, who has devised special apparatus for use with the method. This modification of the Wassermann test has been so thoroughly tried in the army, and so many thousands of tests have been made with it, that there is no question as regards its accuracy and value in the diagnosis of syphilis. It has been employed by many laboratory workers, and their results with it have been in agreement as to its accu- racy and as, in army practice, many repeated examinations are made in the same case in order to control the treatment of the disease, any irregularities in the results of the test would have soon been discovered. In the hands of differ- ent workers the results have been almost identical with the same character of cases, as instanced by Vedder,49 and it is believed that this modification is as well adapted to civil- 99 100 THE WASSERMANN TEST ian laboratories as to those of the army and will be found at once simple and accurate. The technic of the writer is a modification of that of Wassermann and of that of Noguchi, following the original Wassermann technic in inactivating the patient's blood se- rum and the technic of Noguchi, in using a human hemo- lytic system instead of the sheep system employed by Was- sermann. The human hemolytic system was adopted in order to avoid the fallacy that natural antisheep amboceptor fre- quently present in the patient's blood serum may cause in the results of the test. Despite the repeated denials of those who employ the antisheep hemolytic system in per- forming the Wassermann test, that the amount of antisheep hemolysin in the sera tested is not sufficient to produce erroneous results, the observations of numerous careful investigators prove conclusively that in many human blood sera the antisheep hemolysin present is sufficient to render weak positive sera negative and that, in some instances, the amount is sufficient to cause a negative result in strongly positive sera. Noguchi146 has analyzed 326 specimens of human sera as to their antisheep hemolysin content and found that while the majority contained less than four antisheep hemolytic units, 39 contained four units; 27, five units; 15, six units; 8, seven units; while five contained as many as ten units of antisheep hemolysin. He states that in weak positive syph- ilitic sera the presence of from 4 to 10 antisheep ambocep- tor units may produce a negative result if an antisheep hemolytic system is employed in the test. Simon147 states that he has encountered normal antisheep amboceptor in sera he has tested in sufficient amount to render a positive syphilitic serum partially or completely negative so frequently as to render it necessary to remove this source of error. He recommends diluting the amount of serum to be tested with five times the amount of blood TECHNIC OF THE WRITER'S MODIFICATION 101 cell suspension (sheep cells) and incubating for 10 minutes at 37° C. The mixture is then centrifuged, the supernatant serum removed, and tested as usual. The recent observations of Kolmer and Rule148 demon- strate the importance of natural antisheep amboceptor in human blood serum in influencing the results of the com- plement-fixation test for syphilis. In their experiments they found that when from 0.1 to 0.2 c.c. of strongly posi- tive syphilitic sera was used with an antisheep hemolytic system the results were not influenced by the antisheep he- molysin in the sera, but that negative results occurred in weakly positive syphilitic sera, in their experience from three to five per cent of such sera giving negative results. The use of an antisheep hemolytic system in quantitative complement-fixation tests with human sera, especially when the sera contained small amounts of syphilitic antibodies, gave erroneous results, and they conclude that this type of hemolytic system is not well adapted to any quantitative complement-fixation test based upon the use of varying amounts of the patient's blood serum. The experiments of numerous investigators have shown that, when active serum is tested, an antihuman hemolytic system should always be employed, and enough evidence has been cited to prove that the antihuman hemolytic sys- tem is preferable in all complement-fixation tests in which human blood serum is tested. For this reason, the writer urges the use of the antihuman hemolytic system and be- lieves that both experimental and practical results are in accord in showing its superiority to the antisheep hemolytic system when employed in the complement-fixation test for syphilis. Technic.-The methods of preparing the reagents used in the test have already been described, as well as their titration and the exact technic of the test will now be con- sidered. The glassware used in preparing the various reagents 102 THE WASSERMANN TEST for the test and in the test should be perfectly clean. No chemicals should be used in cleaning the glassware but only hot water and soap, followed by a thorough rinsing with hot water. All glassware, except the test tubes used in the test, should be sterilized by dry heat and one should be careful that a temperature of 180° C. is not exceeded, as a higher temperature than this is apt to so heat the ends of the small 1 c. c. pipettes as to close them. The test tubes do not need to be sterilized, and after washing, should not be plugged, but placed open end downward in a test tube basket and covered with a clean towel when not in use. Brown and Kolmer149 have shown that erroneous results in the test may be due to glassware containing traces of acid or alkali. They found that minute amounts of sodium hydroxide, as small as 1 c.c. of a 1-400 dilution of a N/l solution, in 2 to 3 c.c. of fluid may be anticomplementary and that 1 c.c. of a 1-100 dilution produced hemolysis. A N/l solution of hydrochloric acid is anticomplementary in doses of 1 c.c. of a 1-500 dilution and hemolytic in doses of 1 c.c. of a 1-300 dilution. The ordinary bichromate fluid used in cleaning glassware was found to be anticomple- mentary in doses as small as 1 c.c. of a 1-900 dilution and hemolytic in doses of 1 c.c. of a 1-600 dilution. They say: "Minute amounts of alkali and mineral acids are mark- edly antilytic, probably by means of a destructive influ- ence upon complement and if present in test tubes or other glassware may yield falsely positive Wassermann reac- tions, conversely larger amounts may prove hemolytic and thereby yield nonspecific or falsely negative Wassermann reactions." Similar results have been obtained by Manwaring,150 Hektoen and Ruediger,151 and Cumming.152 The pipettes used for measuring the patient's serum should be graduated in tenths and do not need to be grad- uated in hundredths, thus saving considerable expense TECHNIC OF THE WRITER'S MODIFICATION 103 where many tests are being done, as an individual pipette should be used with each patient's blood serum. At least a dozen 1 c.c. pipettes, graduated plainly in hundredths, the graduations being at such a distance apart that 0.02 c.c. can be accurately measured, should be in stock, for titration wTork and in making dilutions. In using any pipette graduated to the tip the last unit of graduation should not be used as it is very apt to be inaccurate. Great care should be exercised in using the pipettes to avoid breakage, especially of the tip, and they should not be al- lowed to remain unwashed after using but should be im- mediately cleaned, as otherwise the blood serum will dry in them and will be found very difficult to remove owing to the small caliber of the pipette. The greatest care should be taken that the pipettes used in measuring the patient's serum be cleaned thoroughly, first with hot soapy water and finally with hot water, as otherwise erroneous results may occur, owing to serum re- maining in the tubes. Brown and Kolmer found that 1 c.c. pipettes used with a positive serum and washed out for from one to three times with normal salt solution gave weakly positive results when they were used with a nega- tive serum, thus proving the importance of thorough cleansing before use. Pipettes should never be placed in a cylinder containing water or salt solution, especially when a large number of tests are being made, as enough positive and negative sera are thus carried to the fluid in the jar to produce erroneous results if the pipettes are not very care- fully cleaned after removal from the jar. The best prac- tice is to rinse each pipette out with running water imme- diately after it is used and then clean with hot water as soon as possible. Collection and Preparation of the Patient's Blood Se- rum.-When specimens of blood for the Wassermann test are to be sent through the mail or kept for more than two days before they are tested, the blood should be collected 104 THE WASSERMANN TEST with aseptic precautions and placed in sterile glass con- tainers. If the test is to be made at once or within a day or two such precautions are not necessary but it is always essential that an aseptic technic be employed in collecting the blood, in order to avoid any danger of infection to the patient. The reason for observing these precautions has been noted in the discussion of the effect of contaminating bacteria upon the result of the Wassermann test. All specimens of blood serum for the test should be kept in the ice box until used, as exposure to room temperature may cause the serum to become anticomplementary. Before taking the blood from the patient he should be questioned regarding the use of alcohol and if he states that he has used alcoholics within thirty-six hours the blood should not be taken but the patient told to return upon the following day, meanwhile abstaining from the use of alcohol in any form. This precaution is necessary owing to the fact that alcohol is capable of rendering a positive reaction negative, as shown by the writer and Nichols and already discussed. Patients undergoing treatment for syphilis frequently give a negative Wassermann reaction during the time treat- ment is being administered, so that it is better to test the blood serum after treatment has been stopped for a period of two or three weeks. Blood for the Wassermann test should not be taken during a malarial paroxysm, as some- times a positive reaction is given in such instances, nor should it be collected immediately after anesthesia or after eating. Bile stained blood sera may inhibit hemolysis and should not be tested if it is possible to secure a specimen from the patient that is not bile stained. If a positive re- action occurs with a bile stained serum it should be re- membered that it may be a false positive and the diagnosis of syphilis should only be made provisionally. Blood serum should not be over five or six days old, at the most, when tested, as older sera frequently are anti- complementary. If a serum inhibits hemolysis in both the TECHNIC OF THE WRITER'S MODIFICATION 105 antigen and control tube another specimen should be re- quested. Inactivation often removes the anticomplemen- tary properties of old sera but it is best not to employ them in the test if it is possible to avoid doing so. Collection of the Blood.-The amount of blood serum required in the writer's modification of the Wassermann test does not exceed 0.2 c.c. so that 2 c.c. of blood is amply sufficient for the test. This amount may be easily collected from a puncture with a needle or glass needle just above the finger nail upon the dorsal surface of the middle finger in the adult, allowing the blood to drop into a small, sterile and perfectly dry glass vial until from one and one-half to two cubic centimeters are collected, or the blood may be collected from a puncture in the lobe of the ear with a Wright tube. Any form of special apparatus that has been devised for collecting blood for the Wassermann test is unnecessary and if venipuncture is objected to by the pa- tient enough blood can easily be obtained as described above. In the writer's opinion, the simplest and best method of obtaining blood for the test is to remove from 2 to 5 c.c. of blood from one of the large veins in the forearm or at the bend of the elbow with a sterilized glass syringe. The arm should be cleaned, the site of the puncture of the vein brushed with iodine, and the needle, held almost par- allel with the surface of the arm, pushed through the skin and directly into the vein. The blood should be withdrawn slowly or the syringe can be dispensed with entirely and the blood allowed to flow directly into a container from the needle. The operation is practically painless and with a little practice anyone can become very expert in the pro- cedure. The advantage of the method is that more blood can be collected, thus insuring an abundant supply of the serum to be tested. After collection, the blood is placed in a suitable glass vial, the serum allowed to separate, after which it is pipetted into a sterile glass vial or tube if it is to be mailed, 106 THE WASSERMANN TEST or used at once in the test. The serum separates best if the blood be kept at room temperature for an hour and then placed in the ice box. All blood serum should be kept in the ice box until used, and when specimens are received by mail, they should be immediately placed in the ice box. A positive and negative blood serum should always be kept in stock for control sera and these should be kept in the ice box when not in use. All blood sera used for the Wassermann test should be clear in appearance and with no disagreeable odor. Milky or cloudy sera are often due to chyle but it is best, if possible, not to use such sera, although they may not cause inhibition of hemolysis. Inactivation of the Blood Serum.-All blood sera used in the Wassermann test as recommended by the writer must be inactivated by heating at 56° C. for one-half hour. This is done for the purpose of removing the native com- plement in the blood serum and also for removing any thermolabile anticomplementary substances that may have developed through age or bacterial contamination. While the inactivation of the serum also tends to make it weaker in substances that give a positive reaction with syphilitic serum, as so well shown by Noguchi,50 it is absolutely nec- essary that inactivation be employed in any modification of the Wassermann test that employs either aqueous or alcoholic extracts of tissues ivhich contain proteins and this is true of the original Wassermann test and most of its modifications. When Noguchi's acetone insoluble anti- genic extract is used the blood serum does not require in- activation and this is one of the advantages that its author claims for it. However, in practice, the amount of harm done by inactivation, in weakening the reacting power of the blood serum, is small when compared with the harm which would result in using active serum, and it is only in very early infections or in those which have received treatment that inactivation might so weaken the serum as to render it negative. In inactivating the blood serum extreme care should be TECHNIC OF THE WRITER'S MODIFICATION 107 taken that 56° C. is not exceeded, for exposure of the serum to a temperature of 60° C. will practically destroy its power of reacting with syphilitic serum. The temper- ature may be allowed to stand at 55° C. with as good results as at 56° C., but the latter figure should never be exceeded. At the same time that the sera to be tested are inac- tivated, a known syphilitic serum and a known normal serum are also inactivated, to serve as controls of the test. In addition, there is always provided a tube for the serum to be tested, in which no antigen is placed, the control tube, and this is chiefly for the purpose of deter- mining that the serum tested is not anticomplementary of itself. This control tube of each serum tested should never be omitted for if it is no dependence can be placed upon the results of the test. The Writer's Technic.-Having inactivated the sera to be tested and the positive and normal control, and being provided with the proper blood suspension, hemolytic am- boceptor, and syphilitic antigens, the test can be proceeded with. Before doing anything else, hoivever, the complement must be titrated as already described and the exact comple- ment unit determined, and this is doubled for each tube in making the test. This procedure should never be omitted under any consideration. Some authorities prefer to ti- trate the amboceptor instead of the complement, but as the complement is the substance in the test that is most apt to vary in strength, it should be titrated in preference to the amboceptor. As a matter of fact, the titration of the com- plement with a fixed dose of amboceptor, also determined by previous titration, can not result in any fallacies in the test, as the amount of complement necessary to produce hemolysis with the dose of amboceptor used, when doubled, is always sufficient to give accurate results when used in the test. 108 THE WASSERMANN TEST It is recommended that two antigens be used in testing each serum, one, the alcoholic extract of fetal syphilitic liver; the other, the alcoholic extract of normal human heart muscle, to which 0.4 per cent of cholesterin has been added. If both antigens can not be obtained, good prac- tical results can be obtained by using one, as it is seldom that a positive result is obtained with one antigen and not with the other. If two antigens are used three test tubes are needed for each serum to be tested, while if only one antigen is used, two tubes will be sufficient. The test tubes should be placed in racks having a double row, an anterior and a posterior. When three tubes are used, two should be placed in the an- terior row of the rack and one in the posterior, while if two tubes are used, one should be placed anteriorly and one posteriorly. The anterior tubes contain the patient's blood serum together with the syphilitic antigen, while the posterior contains the patient's blood serum without the antigen, this tube acting as a control of the serum tested. Besides the tubes for the patient's blood serum, there must be an equal number of tubes for a known syphilitic serum, the positive control; and for a known normal serum, the normal serum control. In making the test proceed as follows: In all of the tubes enumerated place 0.9 c.c. of normal salt solution (0.85%). In Tube 1 and Tube 2, anterior, if two antigens are used, or in Tube 1, if only one antigen is used, place 0.1 c.c. of the inactivated blood serum of the patient to be tested, and the same amount in Tube 1, posterior. The same amount of syphilitic and normal blood serum is placed in the control tubes mentioned and in the same man- ner. Thus, if two antigens are used, in Tubes 3 and 4 0.1 c.c of the known syphilitic serum is placed, and the same amount in Tube 3, posterior, while if only one anti- gen is used, 0.1 c.c. of the syphilitic serum is placed in Tube 2, anterior, and in Tube 2, posterior. There is now added to every tube used in the test kvo units of comple- TECHNIC OF THE WRITER^ MODIFICATION 109 ment and to each anterior tube one unit of the antigen used. If two antigens are used, one unit of each should be placed in the anterior tubes, alternately. In diluting the antigen with normal salt solution, the salt solution should be added to the stock antigen in a test tube, in the proportion of nine parts of salt solution to one part of antigen. The required amount of stock antigen is placed in the test-tube and the necessary amount of salt solution added to it from a pipette, the addition being made very slowly so that a markedly opalescent and cloudy mixture is obtained. Unless the salt solution is added slowly the mix- ture will not be cloudy and negative or partial reactions may be obtained in positive syphilitic sera. If the antigen solu- tion, after the addition of the salt solution, is not cloudy and opalescent in appearance, it will not be found satisfac- tory for use in the test. The importance of adding the salt solution slowly to the stock antigen and the obtaining of a cloudy opalescent solution was first emphasized by the writer and has been repeatedly confirmed by others, espe- cially by Ruediger.153 The tubes are now placed in a water-bath at 37° C. for one half hour, or in an ordinary bacteriological incubator, at the same temperature, for one hour, and at the expira- tion of this time there is added to each tube 0.1 c.c. of the 5 per cent suspension of human red blood corpuscles and two units of the amboceptor paper. The tubes are again in- cubated in the water-bath at 37° C., this time for one hour, or in the incubator for two hours, being shaken every fif- teen minutes in order to facilitate the liberation of the amboceptor serum from the filter paper. After this incu- bation the tubes are placed in an ice box for one or two hours and the results then read. Neglect of shaking the tubes, as recommended, will oftentimes result in partial positive reactions in sera that should react negatively. If desired a hemolytic and antigen control tube may be included but this is not necessary in practice as the titra- 110 THE WASSERMANN TEST tions of the complement will show if anything is wrong with the hemolytic system, and an antigen control is unnec- essary unless it has been some time since the antigen was titrated. The accompanying graphic table illustrates the writer's method of performing the complement-fixation test for syphilis. Method of Performing Complement-Fixation Test for Syphilis SERUM FOR DIAG- POSITIVE CONTROI. NEGATIVE CON- NOSIS SET TROL SET Anterior Tubes Anterior Tubes Anterior Tubes Patient's serum, Positive serum, Normal serum, 0.1 c.c. 0.1 c.c. 0.1 c.c. Complement, 2 Complement, 2 Complement, 2 units units units Antigen, 1 unit Antigen, 1 unit Antigen, 1 unit Salt solution, Salt solution, Salt solution, 0.9 c.c. 0.9 c.c. 0.9 c.c. Posterior Tubes Posterior Tubes Posterior Tubes Patient's serum, Positive serum, Normal serum, 0.1 c.c. 0.1 c.c. 0.1 e.c. Complement, 2 Complement, 2 Complement, 2 units units units Salt solution, Salt solution, Salt solution, 0.9 c.c. 0.9 c.c. 0.9 c.c. Incubate for one-half hour in water-bath at 37° C. or for one hour in incubator at same temperature. Add two units of amboceptor paper to each tube and 0.1 c.c. of a 5 per cent sus pension of human red blood corpuscles. Incubate for one hour in a water-bath al 37° C., or for 2 hours in an incubator at the same temperature. Results of the Test.-If all of the reagents used in the test are working properly the results, after the time al- lowed in the ice box, should be as follows: The anterior tube or tubes, containing the patient's blood serum, and the antigen or antigens, should show complete inhibition of hemolysis (positive reaction), if syphilis is present; while the posterior tube, containing the patient's serum without antigen, should show complete hemolysis. Of course, various degrees of inhibition will be shown in the anterior tube or tubes if the case is an early one, treat- ment has been given, or if the reaction is weak. If the patient is not syphilitic there should be complete hemolysis in the anterior tube or tubes as well as in the posterior tube. The posterior, or control tube of the patient's se- TEST UPON THE CEREBROSPINAL FLUID 111 rum, should always show complete hemolysis, and if it does not, another specimen should be requested. The anterior tube, or tubes, containing the known syph- ilitic blood serum, or the positive control, should show complete inhibition of hemolysis (positive reaction) while the posterior tube of the same serum without antigen, should show complete hemolysis. If inhibition occurs in this tube it demonstrates that the control positive serum is worthless. The anterior tube, or tubes, containing the known nor- mal serum, or the normal control, with antigen, should show complete hemolysis, as should the posterior tube of this set. If a control tube for the antigen and for the hemolytic system be used, both should show complete hemolysis. The Test Upon the Cerebrospinal Fluid The great value of the Wassermann test when per- formed with the cerebrospinal fluid in demonstrating the existence of syphilitic infection of the central nervous sys- tem renders it almost imperative that, before assuring a patient that he is free from infection, the spinal fluid be tested. Not only is this true, but the fluid should be tested in all cases early in the infection, if possible, in order to ascertain whether the nervous system is involved, and if this were more generally done, it is probable that many of the severe complications characteristic of involvement of the central nervous system could be avoided. Collection of the Cerebrospinal Fluid.-The cerebrospinal fluid is collected by lumbar puncture and when properly performed, this operation is devoid of danger. The pa- tient should be in bed and should remain in bed for twenty- four hours after the puncture. The writer is well aware of the fact that scores of punctures of the spinal canal are made in the physician's office, the patient afterward being allowed to walk home or to attend to his business, but, in PLATE II Fig. 1.-Reading of Results of Wassermann Test. The first tube shows com- plete inhibition of hemolysis and may be read four-plus or double-plus, accord- ing to the nomenclature used in the individual laboratory. Tubes 2 and 3 show three-plus and two-plus reactions, or both are included and reported as a plus reaction. Tubes 4 and 5 show single-plus and plus-minus reactions, or may both be included and reported as a plus-minus reaction. Tube 6 shows a minus or negative reaction. Fig. 2.-Wassermann Test. Positive Reaction. Tube 1 contains the pa- tient's blood serum and syphilitic antigen and shows complete inhibition of hemolysis or a positive reaction. Tube 2 is a control of the patient's serum without antigen, and shows complete hemolysis. Tube 3 is a known syphilitic control serum. Tube 4 is a control of the positive serum, without antigen. Tube 5 is a known normal serum and Tube 6 a control of the normal serum without antigen. Fig. 3.-Wassermann Test. Negative Reaction. Tube 1 contains the pa- tient's blood serum and syphilitic antigen and shows complete hemolysis, or a negative reaction. Tube 2 is a control of this serum without antigen. Tube 3 is a known positive serum and shows complete inhibition of hemolysis. Tube 4 is a control of the known positive serum without antigen. Tube 5 is a known normal serum with the antigen and Tube 6 a control of the normal serum without antigen. • ' T T " Fig.I Reading of Results of Wassermann Test j 2 5_ 5 6 Fig.z Wassermann Test Positive Reaction Fiq.3 Wassermann Test Negative Reaction Plate II. TEST UPON THE CEREBROSPINAL FLUID 113 his opinion, this practice is unjustifiable and should be dis- couraged, for while the danger of complications may be small, deaths have occurred during or shortly after the procedure when thus performed, and severe symptoms are not uncommon. The lumbar region is washed carefully with soap and water, alcohol and ether, and the region over the lumbar vertebrae brushed with iodine solution. The site of punc- ture is most easily ascertained by running the finger along the spines of the vertebrae until the so-called "soft spot" is reached, which is indicated by a distinct feeling of soft- ness to the finger and is situated between the third and fourth lumbar vertebrae. The patient may sit upon a stool with his back bent so as to bring the spinous proc- esses of the vertebrae close to the surface, or lie upon the left side on the edge of the bed. The latter posture is necessary in the case of the sick but the sitting posture is suitable for ordinary ambulant patients. The needle used in the puncture should be of flexible steel, measuring 10 cm. long, with a bore of 1 to 1.5 mm., while for a child the needle should be shorter but of the same bore. The needle should be sterilized before being used, as well as the stylet which should accompany it. The space between the third and fourth lumbar vertebra having been ascertained the puncture is made directly in the median line, which I have found much more satisfac- tory than the lateral puncture recommended by many writ- ers. The needle is held firmly and the puncture made quickly, the direction being straight forward. Enough force should be used to push the needle quickly through the skin and muscles but as soon as the spinal ligaments are reached it may be pushed forward more slowly until there is a sudden sensation of loss of resistance when the needle is in the spinal canal. The stylet is now withdrawn and the spinal fluid, if the puncture is successful, will flow from the needle. If it does not flow, the needle may be shifted very gently or the patient told to take a deep breath, or the 114 THE WASSERMANN TEST stylet may be replaced in the needle and gently pushed forward to clear the tube of any materials that have been carried into it by the influx of the spinal fluid. If none of these procedures are successful the puncture is said to be a "dry" one and there is little use in repeating it. Not over 5 c.c. of fluid should be withdrawn, as that is amply sufficient for the test and the less fluid withdrawn the slighter will be the symptoms following the puncture. The fluid often comes from the needle under considerable pressure and this should be noted upon the slip sent to the laboratory with the fluid. It should be free from blood but can be used even if blood corpuscles are present, by centri- fuging the fluid until they are deposited at the bottom of the tube and pipetting off the clear fluid. The spinal fluid is kept in the ice box until used and should not be inactivated, as is the blood serum, before it is tested. The same procedure is followed in testing spinal fluid as in testing blood serum, but different quantities of the fluid are tested, as in some infections of the central nerv- ous system, as paresis, a very small amount of the fluid will give a positive reaction, while the same amount in cere- brospinal syphilis would be negative. Therefore it has always been the writer's practice to test 0.2, 0.5, 0.7 and 1 c.c. of the spinal fluid, thus making four tubes for the antigen tubes and four for the control amounts of fluid, if only one antigen is used, and double this number if two antigens are used in the test. Precautions to be Observed in Lumbar Puncture.-Pa- tients who are very weak or debilitated should not have a lumbar puncture performed unless it is absolutely neces- sary. Those suffering from paresis and tabes bear punc- ture exceedingly well and seldom complain of pain, but nervous normal adults frequently speak of the pain of the puncture and of disagreeable symptoms following it. Headache may be severe if the fluid is withdrawn rapidly and if the bone is punctured or bruised by the needle dur- TEST UPON THE CEREBROSPINAL FLUID 115 ing the procedure pain may last for several hours after- wards. Nausea is sometimes observed, especially if the fluid is rapidly withdrawn, so that it is always best to allow the fluid to flow as slowly as possible, thus relieving pressure gradually. Nonne51 warns against lumbar punc- ture in cases of cerebral tumor, several deaths having fol- lowed the operation in this condition, and states that he has himself observed four cases of sudden death in such patients following the puncture. The patient should remain in bed for twenty-four hours after the puncture, if possible, and neglect of this precau- tion often leads to disagreeable symptoms. In a case which was observed in the writer's practice, very severe head- aches and attacks of dizziness occurred for several days after puncture because the patient insisted upon immedi- ately returning to work. Diarrhea is a symptom some- times observed, especially if the fluid has been withdrawn rapidly. The Reading of the Reaction and Nomenclature Em- ployed.-The results of the Wassermann test are variously reported from different laboratories and unless one is ac- quainted with the meaning of the various terms used, grave errors may result in the interpretation of the results of the test. In perhaps the large majority of Wassermann labora- tories the results of the test are recorded as four-plus (+ + + +), indicating complete inhibition of hemolysis, and, therefore, a strongly positive reaction; three-plus (+ + +), indicating about 75 per cent inhibition of hemolysis, or a moderately positive reaction; two-plus (++), indicating 50 per cent inhibition of hemolysis, or a weakly positive re- action; plus (+), indicating from 25 to 50 per cent inhibition of hemolysis, or a doubtful reaction; plus-minus (+-), in- dicating less than 25 per cent inhibition of hemolysis or a doubtful reaction; and minus (-), indicating complete hem- olysis, or a negative reaction. 116 THE WASSERMANN TEST In the army laboratories a more simple nomenclature has been employed for the reason that only absolute inhi- bition of hemolysis is considered a positive reaction, all other degrees of inhibition being considered doubtful. This standard has been adopted principally that the soldier may be given the benefit of any doubt that may exist re- garding syphilitic infection. Thus four designations are used in the army laboratories, as follows: Double-plus (++), indicating complete inhibition of hemolysis, and, therefore, a positive reaction; plus (+), indicating anything between complete inhibition and 50 per cent of hemolysis; plus-minus (+-), indicating anything between 50 per cent of hemolysis and complete hemolysis; minus (-), indicat- ing complete hemolysis, or a negative reaction. Comparing this nomenclature with the foregoing, the three-plus and two-plus reactions would correspond to our plus reaction, the one-plus with our plus-minus reaction, and the four-plus with our double-plus reaction. The writer believes that for use in civilian life the no- menclature of results first mentioned, i. e., four-plus, three- plus, two-plus, plus, plus-minus and negative, is the most useful; but anything below a three plus reaction, in his opinion, should be considered doubtful, unless there is very good evidence of syphilitic infection, or the patient has received antisyphilitic treatment. The Original Wassermann Technic In the original Wassermann technic an aqueous extract of syphilitic liver was employed as the antigen; the blood sera were inactivated by heating them at 55° C. for one- half hour before testing; and a sheep hemolytic system was used. The complement was fresh guinea pig serum diluted 1:10 with normal salt solution, the dose being arbitrarily fixed at 1 c.c. of the diluted serum. The blood suspension was a 5 per cent suspension of sheep erythrocytes in nor- mal salt solution, and the dose for each tube was 1 c.c. of ORIGINAL WASSERMANN TECHNIC 117 the suspension. The hemolytic amboceptor was the blood serum of a rabbit immunized with sheep's erythrocytes and this was titrated carefully before each series of tests. The antigen, as stated, was an aqueous extract of fetal syphilitic liver, prepared in the manner already described and carefully titrated before using. Two tubes were used for each of the blood sera tested; two for the positive con- trol; two for the normal control; and three tubes for con- trol of the hemolytic system, antigen and blood suspension. The tubes are arranged in double rows except the three control tubes. For each serum tested there should be an anterior and posterior tube and the same for the known positive serum and for the known normal serum. In each of these tubes there is placed 1 c.c. of the complement, di- luted 1:10. In the anterior tube of the test set of two there should be placed 0.2 c.c. of the patient's inactivated blood serum, and the same amount in the posterior tube of this set. In the anterior tube and posterior tube of the positive control set there should be placed 0.2 c.c. of the known pos- itive serum, and in the two tubes of the normal serum con- trol there should be placed the same amount of a known negative blood serum. In the anterior tubes the proper dose of antigen is placed and enough normal salt solution to bring the total amount in the tubes up to 8 c.c. Of the three control tubes, the first is the antigen control and should contain the dose of antigen used, 1 c.c. of the com- plement, and enough salt solution to equal 3 c.c.; the sec- ond is the hemolytic system control, and should contain 1 c.c. of the complement and 2 c.c. of the normal salt solu- tion; while the third is the blood suspension control, and contains 1 c.c. of the corpuscle suspension and 2 c.c. of salt solution. The tubes are shaken and incubated for one hour in the incubator and to each tube is then added two units of am- boceptor serum, as determined by the titration and 1 c.c. of corpuscle suspension. The blood suspension control does 118 THE WASSERMANN TEST not, of course, receive this suspension. The tubes are shaken and incubated until the blood serum controls (the posterior tubes) are hemolyzed, and the results read at once or after several hours in the ice box. If the patient's serum is positive, the anterior tube will show complete or almost complete inhibition of hemolysis, while the posterior tube, containing the serum without an- tigen, should show complete hemolysis; if the patient's serum is negative, the anterior and posterior tubes should both show complete hemolysis. The anterior tube of the positive control set should show complete inhibition of hemolysis, while the posterior tube should show complete hemolysis; the anterior tube of the normal serum control and also the posterior should show complete hemolysis. The antigen control tube should show complete hemolysis, as should the hemolytic system control, while the blood suspension control should show complete inhibition of hemolysis. The nomenclature used in reporting the test is that al- ready described, but in reporting results to physicians it is better to report them simply as positive, very suspicious, suspicious, doubtful or negative. Other Modifications of the Wassermann Test There have been many modifications of the original Was- sermann technic for the complement-fixation test in syph- ilis and the most important will be here considered. The Hygienic Laboratory Modification.-The method used at the Hygienic Laboratory of the United States Pub- lic Health Service is described by M. H. Neill, P. A. Sur- geon, United States Public Health Service, in Reprint 483, from the Public Health Reports. The method is used not only at the Hygienic Laboratory but also in the Venereal Disease Division of the United States Public Health Serv- ice. A description of the method follows. MODIFICATIONS OF THE WASSERMANN TEST 119 REAGENTS Sodium Chlorid Solutions.-A nine-tenths per cent solu- tion of the chemically pure salt in distilled water is made up by weight, distributed in 500 c.c. Erlenmeyer flasks and sterilized by steam. Saturated Sodium Chlorid Solution.-Is made by satu- rating a small sterile bottle of distilled water with the chemically pure salt. The bottle is to be kept at room tem- perature. Sheep's Blood Corpuscles.-These are best obtained by bleeding a sheep from the jugular vein by use of a sterile syringe, previously rinsed with saline solution, transfer- ring the blood immediately to a sterile 50 c.c. Erlenmeyer flask containing sterile glass fragments, and agitating for 15 minutes, avoiding foaming, thus defibrinating the blood. Sheep's corpuscles can also be obtained at the abattoir by catching the blood coming from the vessels of the sheep's neck directly in the defibrinating flask. In doing this due care should be taken to avoid the entrance of gross dirt into the flask. After defibrinating, decant the blood into a graduated cylinder which is either perfectly dry or has just been rinsed out with 0.9 per cent sodium chlorid solution. Note the amount of blood and divide it about equally be-, tween two or more 50 c.c. centrifuge tubes. Add 0.9 per cent sodium chlorid solution till tubes are nearly full, and mix thoroughly. The glass centrifuge tubes in place within the metal tubes and trunnion rings in place should be bal- anced against each other in pairs. Centrifuge till the cor- puscles are completely precipitated. Pipette off super- natant fluid, add fresh 0.9 per cent sodium chlorid solution, mix thoroughly, centrifuge again, pipette off the superna- tant fluid, add fresh sodium chlorid solution, mix, and again centrifuge. After the third centrifuging, pipette off the fluid above the blood cells, transfer the cells to the cylinder previously used for measuring the volume of the blood, which should be freshly rinsed with saline solution, and 120 THE WASSERMANN TEST make up to the original volume with 0.9 per cent sodium chlorid solution. Keep blood corpuscles in the ice box at a temperature not above 6° C. Cells may thus be preserved for at least 48 hours. The blood cells should not be used if the supernatant sodium chlorid solution shows a reddish coloration. If after the final washing a sterile 5 per cent dextrose solution be used to make corpuscle suspension up to the original volume of the blood, the corpuscles should show no evidence of disintegration at the end of 10 days, if kept at a temperature of not more than 6° C. Under these conditions they may be used in the test the same as fresh cells. Hemolytic Amboceptor.-This substance is produced in the serum of a rabbit by injecting this animal with the washed blood corpuscles of a sheep prepared as above. The following methods of producing hemolytic amboceptor may be used. 1. Inject rabbits intravenously with 1 c.c. of fresh sheep cell suspension on the first, fourth, seventh, and tenth days of the process. Make test bleeding on the fifteenth day. 2. Inject rabbits intravenously with 1 c.c., 1 c.c., and 2 c.c. washed sheep corpuscles on three successive days. Wait five days and repeat the injections. Make test bleed- ing five days after the last injection. When the rabbit sera have been found to contain hemo- lytic amboceptor of a suitable strength, the animals should be exsanguinated into large sterile centrifuge tubes or test tubes. When the blood has clotted, separate the clot from the sides of the tube by means of a platinum wire, or glass rod, and place the tubes in the 37° water-bath for an hour, then place in the cold box overnight. Separate the serum, add 0.3 per cent phenol, and preserve in the ice box at a temperature not higher than 6° C. Amboceptor serum should be stored two weeks before use. Complement.-Bleed not less than five full-grown guinea pigs from the heart by means of a sterile syringe, previ- MODIFICATIONS OF THE WASSERMANN TEST 121 ously rinsed with salt solution. With proper technic, from 5 to 10 c.c. may be obtained without injury to the animal. Pigs may be kept for this purpose and bled once in two weeks. Place the blood drawn from each pig in a separate centrifuge tube, and allow to clot. Separate the clot from the side of the tubes with sterile needle or pipette and place in 37° water-bath for one hour. Then place in cold box overnight, or, if serum has separated, pipette off the serum immediately. Centrifuge clear of red blood corpuscles and pool the sera in a sterile glass container. For each cubic centimeter of pooled sera add one-tenth cubic centimeter saturated sodium chlorid solution, mix well, and preserve in the cold box at a temperature not above 6° C. Guinea pig serum so preserved will retain its complement undimin- ished in hemolytic properties for two weeks. Just before use dilute with three volumes of distilled water, to restore the normal tonicity, and dilute as convenient with 0.9 per cent sodium chlorid solution. Antigen.-A suspension of ether and alcohol soluble, ac- etone insoluble lipoids is used in the test. This is prepared as follows: A fresh beef heart is freed from fat and connective tissue and the muscle ground in a meat grinder. Further com- minution is not essential but a greater yield of antigen is secured by hand grinding in a mortar with sand. Place 200 gm. of the ground meat in a bottle, add 2,000 c.c. of alcohol (either absolute or 95 per cent) and stopper tightly. Extract for at least two weeks at 37° C., thoroughly shak- ing the bottle three times daily. Filter through filter pa- per, and evaporate with the aid of a vacuum and condenser, keeping the solution at somewhat below 40° C. by means of a water bath. Agitate the flask from time to time to keep the material washed down as it dries. When evaporated to dryness, wash out the residue with 100 c.c. of ether, or as much more as may be necessary, and allow this to stand in a stoppered separatory funnel overnight. Decant the clear, 122 THE WASSERMANN TEST supernatant liquid and discard the residue. Evaporate the ethereal solution to secure a saturated solution which will reduce the volume to about 30 c.c. Add at least ten volumes of acetone, stir, and allow to stand covered in a cool place overnight. Collect the precipitate and bottle it with a lit- tle acetone. Preserve in the cold box. For use, dissolve 0.3 gm. of the solid in 1 c.c. ether and 10 c.c. of the best obtainable grade of methyl alcohol. Pre- serve this solution in the cold. When about to make a test, place the necessary amount of the methyl alcohol solution in a dry vessel and add the saline for dilution, the first 10 c.c. slowly, and the rest at convenience. The Patient's Serum.-Blood may be obtained from the arm vein. To do this sterilize a syringe and its needle by boiling, and also sterilize the skin of the front of the arm at the bend of the elbow. Rinse the syringe with saline solu- tion. After the patient has opened and closed the fist vig- orously several times, to pump the blood into the veins, place a tourniquet above the elbow just tightly enough to cut off the venous circulation. Puncture a prominent vein with the needle; draw up 5 c.c. blood. Loosen the tourni- quet and discharge the blood into a 15 c.c. centrifuge tube. The venous puncture ordinarily requires no dressing. The procedure is best carried out with the patient lying down. When the serum has clotted, separate the clot from the side of the tube, and set in a cool place, to allow the clot to con- tract. When the serum has separated, pipette off and transfer to another sterile glass container till tested. Keep in the ice box at a temperature not above 6° C. Sera should always be separated from the clot before shipment, as if this is not done more or less hemolysis will take place en route, rendering the serum unfit for testing. Serum should be shipped in a sealed glass capsule, or small, sterile, rubber-stoppered bottle. The stopper should be firmly inserted and a strip of adhesive plaster pasted over the top to guard against its being dislodged. Blood MODIFICATIONS OF THE WASSERMANN TEST 123 serum in sealed glass capsules, or sterile vials, should be well wrapped in cotton and placed in a double mailing case, as specified by the postal regulations. (See reprint from Public Health Reports No. 438.) Just before subjecting the sera to the Wassermann test they should be heated in a water-bath at 54° to 56° C. for one-half hour, but spinal fluids do not require heating. The sera should be fresh-i. e., not more than 24 hours old. Tests may be performed with sera older than this, but in that case more negative results with the sera of syphilitics are to be anticipated than if the sera were fresh. STANDARDIZATION The Wassermann reaction, properly performed, is a quantitative biochemical reaction, our knowledge of which is wholly empirical. To perform it properly, the various quantities of the elements entering into it should be meas- ured as precisely as possible. Furthermore, it is evident that the substances used are extremely complex organic materials, and the greatest possible care should be taken to handle them properly and measure them accurately. The substances first requiring attention are sheep's cor- puscles, hemolytic amboceptor (rabbit) serum, and comple- ment (guinea pig) serum. The proper adjustment of these substances in relation to each other, known as the adjust- ment of the hemolytic system, is an essential preliminary to the tests for syphilis. Sheep's Blood Corpuscles.-The sheep cell suspension previously described is added to 0.9 per cent sodium chlo- rid solution in the proportion of 5 c.c. of the suspension to 95 c.c. of the saline solution. (For details of preparation see amboceptor titration and complement titration.) This suspension is taken as an arbitrary starting point in meas- uring the amount of hemolytic amboceptor present in the rabbit serum, and the quantity of complement present in 124 THE WASSERMANN TEST the guinea pig serum to determine the proper amounts of these substances to use in the tests. It should always be remembered, however, that the red blood corpuscles of different sheep vary considerably in the ease with which they are hemolyzed by complement and amboceptor; so that the substitution of the corpuscles of one sheep for those of another may cause an actual varia- tion of as much as 100 per cent in the quantity of ambo- ceptor or complement serum necessary to cause complete hemolysis, and thus give the appearance of a sudden change in the potency of these reagents. This variability of the sheep corpuscles is taken into account and provided for, as is the variability of other reagents, by daily titration of complement just before setting up the tests for syphilis. Titration of Hemolytic Amboceptor.-The ''unit of am- boceptor" is the smallest amount of amboceptor serum which with 0.05 c.c. fresh-pooled guinea pig serum will com- pletely hemolyze 1.0 c.c. of the 5 per cent suspension of sheep cells, when exposed to a temperature of 37° C. for one hour. Select a specimen of antisheep rabbit serum at least two weeks old. Place 0.1 c.c. in a conical glass and add precisely 19.9 c.c., 0.9 per cent sodium chlorid solution (i. e, 1 in 200 dilution); mix thoroughly by blowing air through the solu- tion; then each cubic centimeter of the final dilution will contain 0.005 c.c. rabbit serum. Now place the following amounts of the final dilution in a row of test tubes: 0.1 c.c. containing 0.0005 c.c. amboceptor serum. 0.2 c.c. containing 0.001 c.c. amboceptor serum. 0.3 c.c. containing 0.0015 c.c. amboceptor serum. 0.4 c.c. containing 0.002 c.c. amboceptor serum. 0.5 c.c. containing 0.0025 c.c. amboceptor serum. 0.6 c.c. containing 0.003 c.c. amboceptor serum. Add one tube, containing no amboceptor serum, to the row and make up the volumes in all tubes to 2 c.c. with 0.9 MODIFICATIONS OF THE WASSERMANN TEST 125 per cent sodium chlorid solution. Take 1 c.c. of the pooled sera of at least 5 guinea pigs, which has been obtained within 5 hours and kept cold, the serum to be unsalted, and add 19 c.c. 0.9 per cent sodium chlorid solution. Each c.c. will therefore contain 0.05 c.c. of the guinea pig serum. Add 1 c.c. of the diluted guinea pig serum containing com- plement to all the test tubes. Next add to all the tubes a 5 per cent dilution in saline solution of the sheep cell sus- pension already described, making the total volume in each tube 4 c.c., mix thoroughly by agitating the tubes, place in the 37° water-bath for one hour, and keep at about 15° C. overnight. Note the tube containing the least amount of hemolytic amboceptor serum which shows complete hemol- ysis. By complete hemolysis is meant a cloudless red solu- tion with no undissolved corpuscles at the bottom of the tube. The amount of rabbit serum in this tube is the "unit of amboceptor." Reject, as unsuitable, those specimens of rabbit serum which fail to give complete hemolysis in amounts of 0.002 c.c. or less, with 0.05 c.c. pooled comple- ment sera. Amboceptor serum should be retitrated every six weeks. In titrating a new specimen of amboceptor se- rum set up a duplicate test, using a specimen of ambocep- tor serum of known titer. Titration of Complement.-This is to be done daily just before the syphilis tests are set up. The "unit of complement" is the smallest amount of complement serum which, with two units of amboceptor, will completely hemolyze 1 c.c. of the 5 per cent sheep cell suspension when kept at a temperature of 37° C. for one- half hour. Estimate, in round numbers, the number of cubic centi- meters of red cell suspension needed for the day's work; for example, 100 c.c. Multiply the unit of amboceptor by 200 and place that amount of amboceptor serum in a 100 c.c. glass-stoppered graduated cylinder. Add about 50 c.c. of 0.9 per cent sodium chlorid solution, taking care to wash 126 THE WASSERMANN TEST down the serum adhering to the sides of the cylinder; next add 5 c.c. of the undiluted sheep corpuscles which have been made up to the volume of the defibrinated blood. Then make up to 100 c.c. with 0.9 per cent sodium chlorid solution. Invert 50 times to mix thoroughly. Set aside for 15 minutes. Dilute some of the salted complement serum as follows: Serum 0.3 c.c. Water 0.9 c.c. 0.9 per cent solution sodium chlorid 1.8 c.c. Total 3.0 c.c. Set up seven test tubes, adding the following amounts of the above solution to them: 0.6 c.c., 0.5 c.c., 0.4 c.c., 0.3 c.c., 0.2 c.c., 0.1 c.c., 0 c.c., using a 1 c.c. pipette. Make up the volume in each tube with saline solution to 3 c.c. Use a 1 c.c. pipette to make up tenths and a 5 or 10 c.c. pipette to add the necessary 2 c.c. Add to each tube 1 c.c. of the amboceptor-corpuscle suspension, incubate in a water-bath at 37° C. for one-half hour, and read the unit at once by noting the tube containing the least amount of guinea pig serum in which the cells are completely dissolved. Titration of Antigen.-In determining the suitability and amount of a specimen of the acetone-insoluble lipoids for use as antigen, the following properties of this substance especially concern us. 1. The property of the antigen of combining with com- plement in the presence of syphilitic sera. 2. The property of the antigen, in much larger amounts, of combining with complement in the presence of normal sera. 3. The property of the antigen of hemolyzing the red blood cells. The first two properties are present to a degree in nearly all antigens, while the third occurs only occasionally and MODIFICATIONS OF THE WASSERMANN TEST 127 is reason for the rejection of the particular specimen in question. The quantitative estimation of the first is called the anti- genic titration; that of the second the anticomplementary titration. These processes may be combined as follows: Set up two parallel rows of 12 tubes and add to them, in pairs, graded amounts of methyl alcohol solution of the antigen to -be tested, leaving one pair without antigen for control-viz., 0.2, 0.16, 0.14, 0.1, 0.08, 0.06, 0.04, 0.02, 0.01, 0.006, 0.004, 0.000 c.c. In adding the antigen solution, di- lute with 0.9 per cent sodium chlorid solution. Make up to 2 c.c. with 0.9 per cent sodium chlorid solution. To each tube of one row add 0.2 c.c. of known positive syphilitic se- rum, and to each tube of the other 0.2 of known negative serum. Add to each tube two units of complement, just previously determined as already described, contained in 1 c.c. of 0.9 per cent sodium chlorid solution. This may conveniently be done as follows: Total number tubes = 24 (allowing for fluid lost in measurement, 25 c.c. complement solution will be needed). Unit of complement = 0.03 x 2 = 0.06 Then 0.06x25 = 1.50. Take salted complement serum 1.50 c.c. Water 4.5 e.e. (3 x 1.5) 6.0 c.c. 0.9 per cent sodium chlorid solution 19.0 c.c. (25-6 = 19) Total 25.0 c.c. Then each cubic centimeter of the solution contains two units of complement. Mix contents of tubes thoroughly, place in 37° water- bath for one hour. Remove and add to each tube 1 c.c. of amboceptor sheep corpuscles, used in determining the unit of complement. Incubate one-half hour and set in a cold place about 15° C. overnight. Note the least amount of antigen completely preventing hemolysis in the tubes containing positive syphilitic serum. 128 THE WASSERMANN TEST Note the largest amount of antigen not interfering with complete hemolysis in the negative serum tubes. The best antigen gives a wide margin between these readings. Now select the unit of antigen for use in the syphilis tests between these values. It should be several times the least amount of antigen completely preventing hemolysis in the tubes containing positive sera; as other positive sera may be encountered, weaker in the syphilitic reacting sub- stance than the specimen used. On the other hand, the unit should not be more than one-half the largest amount of an- tigen not interfering with complete hemolysis in the tubes containing negative serum, as other negative sera may be more anticomplementary than the one used in the test, and false positive reactions might result from the use of too much antigen. Example of selection of antigenic unit: With Negative Serum With Positive Serum AMOUNT ANTIGEN HEMOLYSIS amount antigen HEMOLYSIS 0.2 c.c. None 0.02 c.c. None 0.16 c.c. Partial 0.01 c.c. do 0.14 c.c. do 0.006 c.c. do 0.1 c.c. Complete 0.004 c.c. Partial 0 do 0 Complete Unit selected = .02 c.c. (methyl alcohol solution). When the unit has been selected, place 2 units diluted to 3 c.c. in a test tube and add 1 c.c. of the cell suspension. Incubate 1 hour at 37° C. Should any hemolysis occur, the specimen of antigen should be rejected as being hemolytic. PERFORMING THE TESTS* After the units of amboceptor and antigen have been de- termined, the unit of complement titrated, and the patient's sera heated to 54°-56° C. for one-half hour on the day of the test, the test may be set up as follows (see diagram): Set up a pair of tubes for each new serum to be tested and for the positive and negative control sera which are to be MODIFICATIONS OF THE WASSERMANN TEST 129 retested. Add one tube for the antigen control, one for the hemolytic system control, and one for the sheep cor- puscle control. The paired tubes are conveniently placed in two rows, front and back. To each of the front row tubes add one unit of antigen contained in 1 c.c. of 0.9 per cent sodium chlorid solution, and to the antigen control tube add 2 units contained in 2 c.c. To dilute the methyl alcohol solution of antigen for this purpose multiply the unit by the number of sera to be examined plus five. Place this amount of the methyl alcohol antigen solution in a conical glass and add sufficient of the diluent to make the total volume equal to the number of specimens to be exam- ined plus five. Eg. Antigen unit = 0.03. Number of sera including positive and negative controls plus 5 = 40. 0.03x40 = 1.20. Use 1.2 c.c. antigen solution and add to it 40 c.c. minus 1.2 c.c., or 38.8 c.c. 0.9 per cent sodium chlorid solution. Then, each cubic centimeter of the antigen suspension as made up will contain 0.03 c.c. of the methyl alcohol solution. Now to each pair of tubes, corresponding to the sera which are to be tested, add 0.2 c.c. of the sample of serum to the front tube and 0.4 c.c. to the back tube. Make the volume of all tubes equal 2 c.c. with 0.9 per cent sodium chlorid solution. To do this add 2 c.c. of it to the hemolytic control tube, 3 c.c. to the corpuscle control tube, 1.6 c.c. to the tubes in the back row containing sera and 0.8 c.c. to the tubes in the front row. Add to all tubes, save the corpuscle control tube, 2 units of complement contained in 1 c.c. 0.9 per cent sodium chlo- rid solution (see antigen titration). Add 1 c.c. of this mix- ture to each tube, except the corpuscle control tube. The volume contained in each tube will now be 3 c.c. Mix well, by individually agitating each tube. Incubate in a water bath at 37° C. for one hour. Add to each tube 1 c.c. of the amboceptor-sheep corpuscle suspension. Mix well and in- 130 THE WASSERMANN TEST cubate as above for one hour. Place in a cool place at about 15° C. overnight. READING AND RECORDING RESULTS The morning after performing the tests, first examine the control tubes. All the red cells in the antigen and hem- olytic system control tubes should be hemolyzed, but there should be no trace of hemolysis in the corpuscle control tube. Next examine the tubes containing known positive and negative sera. The rear tubes of both these pairs should show complete hemolysis, as should the front tube of the pair containing negative serum. The front tube of the positive pair should, however, show little or no hemol- ysis, indicating complete, or nearly complete, fixation of complement. In like manner examine all the tubes contain- ing serum to be tested. Inspect the back tubes first, if com- plete hemolysis is not present it may be concluded that the serum was anticomplementary, i. e., was capable of fixing complement in the absence of antigen, and that any fixation in the front tube is of doubtful significance. In some sera, in which hemolysis is complete in the back tube, various degrees of fixation, i. e., weakening of hemolysis, will be noted in the corresponding front tubes. If the appearances of the controls previously mentioned are satisfactory, it is permissible to conclude that these sera are positive. Record the result of the tests, as indicated by the amounts of fixation in these tubes as compared with a spec- imen showing complete fixation (no hemolysis) and one showing no fixation (complete hemolysis) reporting the re- sult as follows: 70 to 100 per cent fixation = ' ' strongly positive. ' ' 40 to 70 per cent fixation = ' ' positive. ' * 20 to 40 per cent fixation = " weakly positive." 0 to 20 per cent fixation = '1 negative. ' ' In actual practice, with the technic described, the experi- ence has been that nearly all sera give either "strongly positive" or "positive" reactions, or are frankly negative. 131 MODIFICATIONS OF THE WASSERMANN TEST BACK ROW Known positive serum Known negative serum Unknown serum* to be tested for syphilis Antigen control Hemolytic system control Sheep corpuscle control Serum .4 c.c. NaCl sol. 1.6 c.c. Complement dilu- tion 1 c.c. Amboceptor-cor- puscle suspen- sion 1 c.c. Serum .4 c.c. NaCl solution 1.6 c.c. Complement dilu- tion 1 c.c. Amboceptor-cor- puscle suspen- sion 1 c.c. Serum .4 c.c. NaCl solution 1.6 c.c. Complement dilu- tion 1 c.c. Amboceptor-cor- puscle suspen- sion 1 c.c. Antigen suspension 2 c.c. Complement dilu- tion 1 c.c. Amboceptor-cor- puscle suspen- sion 1 c.c. NaCl solution 2 c.c. Complement dilu- tion 1 c.c. Amboceptor-cor- puscle suspen- sion 1 c.c. NaCl solution 3 c.c. Amboceptor-cor- puscle suspen- sion 1 c.c. *But one tube shown in diagram. FRONT ROW Known positive serum Known negative serum Unknown serum* to be tested for syphilis Antigen control Hemolytic system control Sheep corpuscle control Antigen suspen- sion 1 c.c. Serum .2 c.c. NaCl solution .8 c.c. Complement dilu- tion 1 c.c. Amboceptor-cor- puscle suspen- sion 1 c.c. Antigen suspen- sion 1 c.c. Serum .2 c.c. NaCl solution .8 c.c. Complement dilu- 1. c.c. Amboceptor-cor- puscle suspen- sion 1 c.c. Antigen suspen- sion 1 c.c. Serum .2 c.c. NaCl solution .8 c.c. Complement dilu- tion 1 c.c. Amboceptor-cor- puscle suspen- sion 1 c.c. The squares represent the arrangement of tubes as seen by one looking down on the rack. Inside the squares appear the reagents in the order in which they are introduced together with the amounts. The preliminary phase of the incubation is carried out at 37° C. for 1 hour. Add amboceptor-cell suspension, incubate at 37° for one-half hour, and keep at about 15° C. overnight. *But one tube shown in diagram. Diagram of Complement Fixation Test for Syphilis 132 THE WASSERMANN TEST Noguchi's Modification.-In Noguchi's modification a human hemolytic system is used, the serum to be tested is not inactivated, and the acetone insoluble lipoid antigen is employed, the preparation of which has already been de- scribed. If the blood serum can not be tested at once, in- activation should be employed but better results are ob- tained with active serum. Capillary pipettes are used in measuring the patient's serum, one drop of the active se- rum being employed, or practically 0.02 c.c. and four drops of the inactivated serum, or practically 0.08 c.c. Noguchi thus describes the technic of the test:52 A rack containing two rows of holes should be used, an anterior and a posterior row. "For each test two tubes are re- quired, one in the front row and its control in the rear row. There will also be two pairs of tubes to serve as positive and negative controls. ' ' Put into each of two small test tubes front and rear one drop (0.02 c.c.) of the serum to be tested from a capillary pipette. When using inactivated serum put 4 drops (0.08 c.c.) into each tube. Use 0.2 c.c. of cerebrospinal fluid not inactivated. Add to each tube 0.1 c.c. of 40 per cent fresh guinea pig serum made by adding 1 part of complement to 1^2 parts of 0.9 per cent salt solution. To the front tube add the unit of antigen determined by titration. Then to both tubes add 1 c.c. of the one per cent suspension of washed human corpuscles (the first procedure). In case of using the ten per cent corpuscle suspension add 0.9 c.c. of salt solution and two units of amboceptor (the second procedure). Shake the tubes thoroughly from time to time to distribute the reagents throughout the mixture. "With every series of tests it is necessary to carry two sets of controls and for this purpose four additional tubes are necessary. To each of the first pair of these, one in the front and one in the rear row, one capillary drop of a syph- ilitic serum known to give a positive reaction is added. This will serve as a positive control. To the second pair MODIFICATIONS OF THE WASSERMANN TEST 133 one drop of normal serum known to give a negative re- action should be added. This tube will serve as a negative control. Now put into each tube complement and into the tubes of the front row antigen, adding finally 1 c.c. of the one per cent corpuscle suspension (first procedure), or 0.9 c.c. of salt solution and 2 units of amboceptor (second pro- cedure) to each tube.- "Place the rack holding these pairs of tubes in a water- bath, thermostat, or warm place not over 37° C. Allow an hour from the time the mixture is made for the antibody to combine with the antigen and for complement to be fixed. If a water-bath is used 30 minutes is a sufficient length of time. The contents of the tubes in the first procedure (first method) are as follows: "Rear. Test serum plus complement (2 units) plus 1 c.c. 1 per cent corpuscle suspension. "Front. Test serum plus complement (2 units) plus antigen plus 1 per cent corpuscle suspension. 4 ' The contents of the tubes in the second procedure are as follows: "Rear. Test serum plus 2 units complement plus 2 units amboceptor paper. "Front. Test serum plus complement (2 units) plus antigen plus 2 units amboceptor. "First incubation at 37° C. for 1 hour, then add to each tube of the first procedure a slip bearing two units of am- boceptor, as follows: "Rear. Above, plus amboceptor (2 units). "Front. Above, plus amboceptor (2 units). "To each tube of the second procedure 0.1 c.c. of the ten per cent corpuscle suspension, as follows: "Rear. Above, plus 10 per cent corpuscle suspension (.1 c.c.). "Front. Above, plus 10 per cent corpuscle suspension (0.1 c.c.). "Allow two hours in the thermostat or one hour in the water-bath. After final incubation the tubes are kept at 134 THE WASSERMANN TEST room temperature for a few hours before the results are recorded." It will be noted that Noguchi gives two procedures or methods of making the test but the first procedure is to be preferred, in the opinion of the writer. Table XII illustrates the method of performing Noguchi's modification of the Wassermann test by the first proce- dure recommended: Table XXI Noguchi's Method of Performing the Wassermann Test with His Modification SET FOR DIAGNOSIS POSITIVE CONTROL SET NEGATIVE CONTROL SET Posterior Bow Posterior Bow Posterior Bow Unknown serum, 1 drop (active) Complement, 2 units 1 c.c. 1% corpuscle suspension Positive serum, 1 drop Complement, 2 units 1 c.c. 1% corpuscle suspension Normal serum, 1 drop Complement, 2 units 1 c.c. 1% corpuscle suspension Anterior Bow Anterior Bow Anterior Bow Unknown serum, 1 drop (active) Complement, 2 units 1 c.c. 1% corpuscle suspension Antigen, 1 unit Positive serum, 1 drop Complement, 2 units 1 c.c. 1% corpuscle suspension Antigen. 1 unit Normal serum, 1 drop Complement, 2 units 1 c.c. 1% corpuscle suspension Antigen, 1 unit Addition of 2 units of antihuman am- boceptor to all tubes. Incubation at 37° C. for two hours longer, then at room temperature. One hour in water-bath sufficient and one-half hour for first incubation. Incubation at 37° C. for one hour. Hecht-Weinberg's Modification.53-In this modification the natural amboceptor present in human blood serum to sheep's corpuscles is used for the amboceptor, and for complement, that present in the patient's blood serum is utilized. In other words, the serum to be tested furnishes both amboceptor and complement. Therefore, the blood serum must not be inactivated before use and should be used as soon as possible after the specimen is taken. The modification is claimed to give more delicate results than the usual Wassermann test but the writer does not believe that it should be used generally, owing to the errors that may occur through false positive reactions occurring with the usual antigens employed and because the amount of natural amboceptor to sheep erythrocytes is such a vari- MODIFICATIONS OF THE WASSERMANN TEST 135 able quantity in human blood, being excessive in some in- stances and practically absent in others. The acetone in- soluble antigen of Noguchi should be used with this modi- fication. Gradwohl's Modification of the Hecht-Weinberg Modifi- cation.-Gradwohl54 has greatly improved the technic of the Hecht-Weinberg modification of the Wassermann test by ascertaining the hemolytic strength of each serum to be tested to sheep corpuscles; i. e., the hemolytic index of each serum to sheep corpuscles is worked out before the serum is finally tested. Gradwohl55 thus describes his technic: "Place in a rack fourteen small test tubes. The first ten of these tubes are used to determine the hemolytic in- dex of the suspected blood. By this I mean the exact amount of hemolytic amboceptor present in the given blood serum. The last four tubes are used in the actual test. Add 0.1 c.c. of fresh, unheated, patient's blood serum to each of the first ten tubes. Then add decreasing amounts of normal salt solution to these tubes, beginning with 1 c.c., then 0.9, 0.8, 0.7, 0.6, 0.5, 0.4, 0.3, 0.2 and 0.1 c.c. to the succeeding nine tubes. Next add increasing amounts of fresh 5 per cent suspension of sheep's corpuscles, start- ing with 0.1 c.c. and ending with 1 c.c. Place the rack in the water-bath for one-half hour. The tube which last shows complete hemolysis constitutes the 'hemolytic in- dex;' if it is tube 4, the index is 4, because this tube had received 0.4 c.c. of sheep's corpuscles and therefore we have obtained an idea as to how much sheep's blood is to be added to the last four tubes. The first three of these tubes (11, 12, and 13) constitute the tubes for the actual test, while the last tube in the rack (14) serves as the se- rum control tube. Tubes 11, 12, and 13 receive, therefore, the patient's serum, the proper amount of sheep's corpus- cles, dependent on hemolytic index, rising strengths of an- tigen, but no complement and no amboceptor. Tube 14 re- ceives only sheep corpuscles, but no antigen." 136 THE WASSERMANN TEST "Ill my technic I use 0.1 c.c. of a diluted antigen, deter- mined by titration, in tube 11, 0.15 c.c. antigen in tube 12, and 0.2 c.c. in tube 13. In order to equalize the volume of fluid in these tubes, I add 0.2 c.c. normal saline to tube 11, 0.15 c.c. to tube 12, 0.1 c.c. to tube 13, and 0.3 c.c. to tube 14. The tubes are then agitated and placed in the water-bath for half an hour. These last four tubes are filled at the time I make the additions to the first ten and are left with them in the water-bath for one-half hour for fixation of complement, the rack is then taken out and the hemolytic index computed. If the index is low, say from 1: 4, I add 0.1 c.c. of sheep's blood to the last four tubes. If the in- dex is between 5 and 7, I use 0.15 c.c. sheep's blood to the last four tubes; if between 8 and 10, I add 0.2 c.c.; if the index is over 10, I rack up ten more tubes and repeat the titration of the natural complement and amboceptor, then I estimate that; if between 11 and 15, I use 0.25 c.c.; if between 15 and 18, I use 0.3 c.c.; and if between 18 and 20, I use 0.35 c.c. If the patient's serum has an index be- low 2, I regard the reaction of doubtful value. If it is above 2, I regard it as absolute. The reaction is read off exactly as in the Wassermann, that is, inhibition or non- inhibition of hemolysis. If the amount of complement or amboceptor is very low, we can add the proper amount of guinea pig's serum or rabbit's immune serum, ascertained by preliminary titration. ' ' Gradwohl claims that this modification, in his hands, is superior to the Wassermann reaction, giving 15 to 20 per cent more positive reactions than the latter, but the writer believes that it is so cumbersome that it could not be used in laboratories where scores and, sometimes, hundreds of tests are being made each day, as in army practice, and that the fact that false positives have been repeatedly re- ported by different observers who have used the test mil- itates greatly against its value. It is an excellent control MODIFICATIONS OF THE WASSERMANN TEST 137 of the Wassermann test but should not replace it in gen- eral laboratory practice. Kolmer's Modifications.-Four different methods for performing the Wassermann reaction are recommended by Kolmer.56 The first method is exactly similar in technic to the original Wassermann, except that an alcoholic extract of syphilitic liver is used instead of an aqueous extract. In the second method the technic is the same except that three different antigens are used with each blood serum tested and the amounts of each reagent are reduced one- half. Kolmer strongly recommends this method as being "simple, accurate, and reliable." The three antigens used are: 1. A cholesterinized alcoholic extract of human heart; 2, alcoholic extract of syphilitic liver; 3, acetone insoluble lipoids. It will thus be noted that four test tubes are re- quired for each serum tested, one as control, and three for the antigen tubes. As regards the results, Kolmer says: "With strongly positive serums there is complete inhi- bition of hemolysis with all three antigens. * * * In from 15 to 20 per cent of cases the cholesterinized extract shows a 50 per cent or more inhibition of hemolysis, whereas with the other two antigens the reactions are negative. In our experience the majority of such serums were taken from patients giving a frank history of syphilis of many years' standing, and from known cases undergoing treatment, fur- ther therapy being indicated until the reaction finally be- comes negative when cholesterinized extracts are used. In a small proportion of cases a feebly positive reaction of 25 per cent or even less of inhibition of hemolysis may be found with the cholesterinized extract alone. Many of these reactions occur with serums of treated cases of syphilis; on the other hand a similar reaction may occur with about 5 per cent of normal serums, so that if the history and clin- ical conditions are clearly negative, a slight degree of in- hibition of hemolysis (5 to 10 per cent) with the choles 138 THE WASSERMANN TEST terinized extract and marked hemolysis with the other two antigens may be interpreted as a negative reaction." The writer has quoted Kolmer regarding his results with cholesterinized extracts for antigens because his experi- ence has been exactly similar and for this reason the writer firmly believes in the value of such extracts in the diagno- sis of syphilis with the Wassermann test and in the control of the treatment of the disease. Properly titrated and con- trolled, and with experience gained by using them, so that a proper interpretation may be placed upon the results ob- tained, cholesterinized antigenic extracts are perfectly safe and furnish a much greater amount of information regard- ing the existence of a syphilitic infection than do other antigenic extracts. Kolmer's third method was devised in order to measure the amount of syphilitic antibody in the patient's blood se- rum. He employs only one antigen, either an alcoholic extract of syphilitic liver or the acetone insoluble lipoid antigen. The technic is about the same as in the original Wassermann but eight tubes are used for each patient's serum, increasing amounts of the serum, diluted 1:10 being placed in them, together with the unit of antigen deter- mined by titration, only one antigen being employed. The fourth method is somewhat similar, but in this the amount of syphilitic antibody in the patient's blood serum is meas- ured by the number of hemolytic doses of complement fixed by a constant amount of antigen. As routine methods, where many tests are being per- formed, the third and fourth methods of Kolmer are not practical, as they consume too much time and material. The second method is an excellent one, but it appears to the writer that practically the same results are obtained by using two antigens, instead of three, as recommended in his modification. The writer also believes, as already stated, that a human hemolytic system is preferable to the sheep system, which is used in all of Kolmer's methods, owing to the possibility of the natural antisheep ambocep- MODIFICATIONS OF THE WASSERMANN TEST 139 tor present in nearly all human blood serum interfering with the delicacy of the test. Despite all the claims that have been made that the pres- ence of antisheep amboceptor in human serum is a negligi- ble factor in the Wassermann test, the fact remains that in some blood sera there is so much of this amboceptor present that no dependence can be placed upon the results of a Wassermann test with them unless it is first removed, and it has been proved that, in some cases, this ambocep- tor, when combined with the amboceptor added, will ren- der a reaction negative that, with the human hemolytic sys- tem, would be positive. This factor would frequently be of very great importance in any method of performing the test in which titrations of the amount of syphilitic antibody in the patient's serum were used. In all complement-fixation tests in which quantitative results are desired an antisheep hemolytic system should never be employed, if human blood sera are to be tested, and it is now the consensus of opinion of the best author- ities that, when an antisheep hemolytic system is employed, the natural antisheep hemolysin so frequently present in human serum should be removed before such sera are tested. More recently Noguchi154 has described a modification of the test in which the complement present in fresh blood se- rum from the individual to be tested is utilized for com- plement. The blood serum to be tested is not inactivated, seven times the usual amount of antihuman amboceptor is used, owing to the small amount of complement in human serum, and the acetone-insoluble antigen of Noguchi is in- variably employed, as alcoholic extracts, especially when fortified with cholesterin, give false results. This method has not proved satisfactory, in the experience of most se- rologists, as known syphilitic sera have been found to give negative results with it, in a considerable proportion of cases. 140 THE WASSERMANN TEST Somewhat similar modifications have been devised by Thompson155 and by Bartlett and O'Shansky,156 but neither have become popular, although, when properly performed, the methods give accurate results. In an investigation of the various methods of comple- ment fixation for syphilis, in which the complement pres- ent in human blood serum is utilized instead of guinea pig blood serum, Kolmer and Flick157 obtained the following results: From 2 to 10 per cent of human sera do not contain suffi- cient complement or natural hemolysis to secure definite results. A higher percentage of positive results is obtained with tests employing human complement, the majority be- ing correct, and are due to using the active (unheated) blood serum of the patient. False positive reactions are observed in nonsyphilitic individuals, especially if crude alcoholic extracts are used as antigen. The use of human complement, when the patient's serum is inactivated, results in a smaller proportion of positive results in known sypjhilitics, and false negative results oc- cur in such instances more frequently when human comple- ment is employed. As summed up by Kolmer and Flick the advantages and disadvantages of employing human complement instead of guinea pig complement are as follows: Advantages.-1. Greater delicacy of the test if patient's serum is not heated. 2. Guinea pigs are not required. Disadvantages.-1. False positive reactions. 2. False negative reactions. 3. The blood sera to be tested must be freshly collected. 4. Unsatisfactory results are obtained with spinal fluids and with inactivated sera. 5. Comple- ment, or natural hemolysin, or both, are insufficient in some sera to obtain results. 6. Isoagglutinins interfere with the test in some cases. MODIFICATIONS OF THE WASSERMANN TEST 141 As a result of their studies they state that no tests em- ploying human complement should be adopted in devising a standardized Wassermann test. Of the methods that have been devised they obtained the best results with that of Thompson, which employs the complement present in each patient's serum and an antihuman hemolysin, and that of Bartlett and O'Shansky, in which the complement and hemolysin present in each serum tested is utilized. Among the other modifications of the Wassermann test that have been recommended, from time to time, may be mentioned that of Tschernogubow57 using the natural com- plement and amboceptor against guinea pig corpuscles present in human serum; of Detre and Brezovsky,58 in which an antihorse hemolytic system is employed with rab- bit complement; of Margarette Stern,59 who used fresh, ac- tive blood serum and the complement present in the blood serum tested; of Bauer,60 who makes use of the antisheep hemolysin present in human blood serum; of Browning and McKenzie,61 who use an antiox hemolytic system and a quantitative technic, and of Thompson.62 There are many other modifications noted in the literature but none of them appear to be as reliable as the technic recommended by the writer or the modifications that have been more fully dis- cussed in the preceding pages. The Effect of Low Temperatures upon the Results of the Wassermann Test.-Quite recently several observers have called attention to the effect of incubating the com- plement, antigen, and patient's blood serum at ice box tem- perature, claiming that more delicate results are obtained than when this incubation is made in the water-bath at 37° C. Zinsser63 states that in his laboratory two series of tests were run, one with the first incubation at 37.5° C. for 30 minutes, the other with the first incubation in the ice box at 8 to 10° C., for three hours, the same blood sera being used in each series. The results showed 15 more positive reactions in the series incubated at ice box tern- 142 THE WASSERMANN TEST perature, than in that incubated in the water-bath, and of these cases 7 were unquestionably syphilitic, two were treated syphilitics, and four were probably syphilitic. In a recent valuable contribution, Smith and MacNeal64 compared the results obtained with different antigens and different temperatures of incubation in the Wassermann test. They experimented with three antigens as follows: 1. A cholesterinized alcoholic extract of beef heart. 2. A plain alcoholic extract of beef heart. 3. An acetone insolu- ble extract of beef heart. Two series of tests were made, with the same blood sera, one in which the first incubation, with each antigen, was at 37° C. and one in which the first incubation was for 4 hours in the ice box at 8° C., with each antigen. Over 500 different sera were tested and their con- clusions are important. They found that the cholesterin- ized antigen, with incubation at 8° C. for 4 hours proved a more sensitive test for syphilis than any of the other methods tried but that it was apt to give nonspecific com- plement fixation, and, therefore, if used alone it could not be depended upon in diagnosis. The simple extract antigen (alcoholic extract of beef heart), with the first incubation for 4 hours at 8° C. in the ice box, gave more sensitive re- sults than the cholesterinized extract at 37° C. and in these experiments did not give any false positive results. The acetone insoluble extract was found less sensitive either at 37° C. or 8° C. than the cholesterinized extract and also less sensitive than the simple alcoholic extract at 8° C. It was more sensitive than the simple extract at 37° C., and gave no false positive results in this series of tests. From their results it is evident that the cholesterinized ex- tract at 8° C. gave the highest percentage of positive re- sults, but sometimes gave a false positive reaction, while the simple alcoholic extract gave the best practical results, when the first incubation was made at 8° C., as it gave better results than the cholesterinized extract at 37° C., and did not give any false reactions. MODIFICATIONS OF THE WASSERMANN TEST 143 The subject of the ice box temperature for the first in- cubation in the Wassermann test deserves further study, and, if these observations are confirmed, this temperature should be adopted in practice. If it is found reliable it will probably be best to use simple alcoholic extracts for antigen instead of cholesterinized extracts, but more work should be done upon the subject before the method is gen- erally adopted. Since the above was written a great deal of experimental work has been done upon the ice box incubation for com- plement fixation in the diagnosis of syphilis and interesting and important results have been obtained by several ob- servers. Ruediger158 found that incubation at 1° C., for 10 hours gave the best results and that alcoholic extracts of beef or human heart gave better results, when used as anti- gen, than acetone insoluble fractions of either. The re- searches of Jacobstahl, McNeill, Altmann, Leredde and Rubenstein, Ottenberg, and others, have shown that a slightly higher percentage of positive results are obtained in syphilitic sera with ice box incubation and that the best results are obtained with simple alcoholic extracts of tis- sue, as beef and human heart, than with any other form of antigen. Cholesterinized extracts should never be used, in the experience of all who have worked with ice box incubation, owing to the danger of obtaining false positive results. From a review of the literature and personal informa- tion, the writer believes that the general consensus of opinion regarding this matter is that a slightly greater number of positive cases are detected when ice box incuba- tion is used, but that its greatest value is obtained in test- ing treated cases of syphilis. There is a general agreement that cholesterin fortified antigens should not be used when ice box incubation is employed. The technic of the test, when ice box incubation is used, is the same as that recommended by the writer, with the 144 THE WASSERMANN TEST exception that after adding the patient's serum, antigen and the complement, the tubes should be placed in an ice box and kept at a temperature between 4° and 10° C. for a period of from 4 to 10 hours, before adding the amboceptor and blood cell suspension. The antigen should be an alco- holic extract of human heart muscle. The Titration of the Strength of the Wassermann Reac- tion.-In using the Wassermann test as a control of the efficiency of treatment, it is often advisable to test the strength of the reaction, from time to time, especially if sera show a double-plus (four-plus reaction) repeatedly, although intensive treatment is being administered. In such cases, the physician can not be sure that the treat- ment is having any effect unless the blood serum is titrated as to its fixing or absorbing power of complement in the presence of suitable antigen. This is done by diluting the blood serum in the following manner, as recommended by Vedder:65 Six tubes are used in addition to the usual num- ber used in the Wassermann test as recommended by the writer. In the first tube is placed 1.9 c.c. of normal salt solution, and in each of the other tubes 1 c.c. of normal salt solution. In the first tube is placed 0.1 c.c. of the serum to be titrated and the tube is thoroughly shaken. One cubic centimeter is taken from this tube and placet; in the second tube, which is shaken, and the same amount is carried to the third tube, and the operation repeated until the sixth tube is reached, from which, after it is shaken, 1 c.c. is discarded. After the dilutions are thus made each tube will contain 1 c.c. of fluid and the dilutions of the blood serum will be as follows: Tube 1, 0.05 c.c.; Tube 2, 0.025 c.c.; Tube 3, 0.012 c.c.; Tube 4, 0.006 c.c.; Tube 5, 0.003 c.c.; and Tube 6, 0.0015 c.c. The test is then made in the usual manner with all of the dilutions and the results recorded as in Table XXII. This method of titrating the blood serum of patients under treatment is recommended in all cases in which per- 145 MODIFICATIONS OF THE WASSERMANN TEST Table XXII Results of Titration of Blood Serum of DATE OF TEST DILUTIONS 0.1 C.C. 0.05 0.025 0.012 0.006 0.003 0.0015 Previous Result. + + with 0.05 c.c. + + + + - - - - - sistent double-plus (four-plus) reactions are reported, as it is the only way in which the physician can be sure that the particular method of treatment being pursued is prov- ing efficient. In many instances the titration will demon- strate that the smaller amounts of serum become progres- sively negative, thus proving the efficiency of the treatment, although the usual amount of serum tested, i. e., 0.1 c.c., may give a positive reaction. In making quantitative complement-fixation tests upon human blood sera an antisheep hemolytic blood system should never be used, as the natural antisheep hemolysin present in many human blood sera will cause erroneous results, especially if the sera tested contain only small amounts of substances reacting to the antigen employed in the complement-fixation test for syphilis, as is very fre- quently the case in patients who have received treatment. A human hemolytic system should always be employed if the most accurate results are desired in quantitative tests of the serum of syphilitics as regards the strength of the complement-fixation reaction. CHAPTER VI COMPLEMENT FIXATION IN SYPHILIS WITH ANTIGENS PREPARED FROM PURE CULTURES OF TREPONEMA PALLIDUM The success attending Noguchi's efforts to obtain Tre- ponema pallidum in pure culture lead to the hope that an- tigens prepared from such cultures would be effective in the Wassermann test and that in this manner a truly spe- cific antigen would be obtained and the reaction become a true antigen-antibody reaction. It is obvious that if such a specific antigen could be obtained the value of the Wassermann test, great as it is at present, would become still greater, as the positive reactions frequently observed in yaws, tubercular leprosy, and with less frequency, in malaria and trypanosomiasis, would be eliminated. The subject has already been briefly discussed but it is thought of enough importance to consider somewhat more exten- sively in this chapter. Schereschewsky, in 1909,66 was the first to attempt to secure an antigen from cultures of Treponema pallidum. Unfortunately, his culture was an impure one and for this reason his results were not conclusive. He was able to obtain complement fixation with syphilitic serum, with his antigen, but also obtained positive results with some of the control antigens with which he worked. It is now gen- erally admitted that he was not working with cultures of Treponema pallidum at all but with Treponema microden- tium, and, as will be shown later, his positive results were in the nature of a group reaction, as the writer and Nichols have shown that this organism will react positively with syphilitic serum, if used as an antigen. In addition, 146 COMPLEMENT FIXATION IN SYPHILIS 147 Schereschewsky's cultures were contaminated with bac- teria and it is possible that some of his positive results were due to this fact, for, as already noted, certain bacteria, when growing in normal blood serum, produce substances which give a positive reaction with the Wassermann test, and it is not improbable that bacteria growing in the horse serum medium, which he used for his cultures, may have produced lipoidal substances which were capable of giving positive results. However, his work was very suggestive and, without doubt, stimulated research along these lines. In 1912, Noguchi,67 who had succeeded in isolating and growing Treponema pallidum in pure culture, published the results of his work upon complement fixation in syph- ilis with an antigen prepared from such cultures. His an- tigen was an aqueous extract of Treponema pallidum, and he also worked with an emulsion of the organisms obtained from the testicles of infected rabbits. His researches dem- onstrated beyond doubt that, in certain cases of syphilitic infection, it is possible to obtain complement fixation with antigens prepared from a pure culture of Treponema palli- dum and that, in such instances, the reaction is indeed a true antigen-antibody reaction. Noguchi found that, in cases of syphilis that had been treated or in which the disease was latent, he often ob- tained a positive reaction with the culture antigen when the ordinary antigens used in the Wassermann test gave a negative result, and that the culture antigen very rarely gave a positive result in primary and secondary cases of the disease. Therefore, he concludes that the reaction with the pallidum antigen may be regarded as an index of the resistance of the patient to the infection, stating "We have in the Wassermann reaction a fair measure of the ac- tivity of the infecting agent, and now we will have in the pallida fixation reaction a gauge for the defensive activity of the infected host." Noguchi's conclusions regarding his work with the pal- lidum antigen are as follows: 148 THE WASSERMANN TEST 1. The Wassermann reaction is caused by the lipotropic substances in sera but not by the antibodies which combine specifically with the pallida antigen. 2. The fixation produced by the culture pallida antigen with certain syphilitic sera is caused by the specific anti- bodies contained in the latter and may constitute a specific diagnostic method for syphilis. 3. The fixation caused by testicular extracts behaves like the culture pallida antigen in the majority of cases, but when the sera (syphilitic or leprous) contains abundant lipotropic substances, they may give a Wassermann re- action as well, which is not the case with the culture pallida antigen. 4. In the serum of rabbits with active syphilitic orchitis there is no indication of the presence of a sufficient amount of antibodies for the pallida antigen, although it gives a strong Wassermann reaction. In 1912, the writer and Major Nichols published the results of some work upon specific antigens in complement fixation with syphilitic serum,68 which differed somewhat from those of Noguchi, chiefly because of the nature of the antigen employed. In his work Noguchi used aqueous extracts of Treponema pallidum, while in our work alco- holic extracts were employed. Noguchi stated that it was impossible to extract the antigenic properties of the or- ganism by alcohol and that only aqueous extracts contained them, but our results showed that alcoholic extracts were efficient, although they were not as efficient as aqueous ex- tracts, and our work has since been confirmed by other au- thorities. It is true that alcoholic extracts owe a part of their activity to other substances than the specific com- plement-fixing substance of Treponema pallidum which is best extracted in water, but it is also true that the alco- holic extracts do contain this specific substance, as the re- sults obtained can be explained in no other way, and dif- fer so greatly from those obtained with ordinary alcoholic antigens used in the Wassermann test. COMPLEMENT FIXATION IN SYPHILIS 149 In our experiments we tested 51 syphilitic sera and 54 nonsyphilitic sera. Of the latter, 38 were from patients suffering from diseases other than syphilis and 16 were from normal individuals. Eleven rabbit sera were also tested. The sera were all inactivated by heating them at 55-56° C. for half an hour before testing. The stock antigen used was an alcoholic extract of fetal syphilitic liver and the culture antigens were prepared from cultures kindly sent us by Dr. Noguchi. Alcoholic extracts were made of the cultures, filtered, the filtrate evaporated to one-third its original volume and titrated in a dilution of 1-10 with normal salt solution. Antigens were thus prepared from cultures of Treponema pallidum, Treponema pertenue, and Treponema microdentium. The test was performed in the same manner as that recom- mended in this work by the writer for the Wassermann test. As a control of the culture antigens used an alcoholic ex- tract of the medium employed in growing the treponemas was prepared in the same manner as the culture antigens. Results with Normal Sera.-The blood serum of 16 nor- mal individuals was tested with the stock antigen and with pallidum, pertenue, and microdentium antigen, and all gave a negative result. Not a trace of a reaction occurred in any of these tests so that it is evident that none of the an- tigens employed reacted with normal blood serum. Results with Serum from Other Diseases than Syph- ilis.-The blood serum of 38 patients suffering from dis- eases other than syphilis was tested with the same an- tigens and was negative in every instance, except that a plus reaction was obtained with the microdentium antigen in a case diagnosed as arthritis. Syphilis could not be eliminated in this case but the stock syphilitic antigen gave a negative result with the serum. Results with Serum from Syphilitic Patients.-The blood sera of 51 patients suffering from syphilis were tested and Table XXIII shows the results obtained. In recording 150 THE WASSERMANN TEST these tests the sign, + +, means absolute inhibition of hem- olysis; the sign, +, at least 50 per cent of inhibition; and the sign, +-, weaker degrees of inhibition. The sign, indicates total lack of inhibition, while the sign 0 indicates that no test was made. Table XXIII Complement Fixation in Syphilis with Stock Antigen and Specific Antigens in Forty Infections* SERUM NO. STOCK ANTIGEN T. PAL- LIDUM ANTIGEN T. PER- TENUE ANTIGEN T. MICRO- DENTIUM ANTIGEN CONTROL ANTIGEN REMARKS 1 + + + + 0 o - Secondary, early 2 + + + +- 0 - Primary 3 + + + - + - 0 - Secondary 4 + + - 0 0 0 Latent, 18 years 5 + + + + + Latent, 4 years 6 + + + - - - Latent, 3 years 7 + + - 0 - Tertiary, 10 years 8 + + - 0 0 - Secondary, 9 months 9 + + 0 0 0 Secondary, early 10 + 0 0 0 - Primary 11 + + + 0 0 - Primary, 2 weeks 12 + + 0 0 0 Latent, 2 years 13 + - 0 + - 0 Primary 14 + - 0 + - Primary 15 + + 0 + Secondary, early 16 + + - - Secondary, 1 year 17 + - 0 + - Tertiary 18 + + + - + - Secondary 19 + + + 0 + - Tertiary 20 + + - + - +- - Latent, duration un- known 21 + - +- 0 0 - Latent, 3 years 22 + - + 0 0 - Primary, 3 weeks 23 + - + - 0 0 0 Secondary, early 24 + - - - - - Primary, 2 weeks 25 + - - - 0 - Epilepsy 26 + - + - 0 + - - Arthritis, syphilitic 27 + - + - 0 + - 0 Latent, 2 years 28 + - + - 0 + - - Latent, 2 years 29 + - + 0 + - Latent, 1 year Secondary, 1 year 30 + - +- - + - - 31 + - + + + + + - Latent, duration un- known 32 + - +- 0 + - Latent, 3 years Secondary, 3 months 33 + - + - + - - - 34 + - + + - + - - Latent, 4 years 35 + - + - - + - - Latent, 6 years 36 + - +- + - - - Secondary, early 37 - + - - + - - Latent, 2 years 38 - + + + - - Latent, 1% years 39 - + - - - - Latent, 6 years 40 - + - - 0 0 Primary, 2 weeks *Each serum tested was properly controlled as is usual in the Wassermann test. COMPLEMENT FIXATION IN SYPHILIS 151 It will be noted that no case reacted with the control culture medium antigen and that of the 36 cases giving any reaction at all with the stock extract antigen, no less than twenty-seven, or 75 per cent reacted also with the pallidum antigen, but the reactions were, as a rule, much weaker with the pallidum antigen than with the stock an- tigen. However, there were instances in which the pal- lidum antigen gave a stronger reaction than the stock an- tigen, as illustrated in serums 11 and 19, while in serum 38 the stock antigen gave a negative result and the pal- lidum antigen a plus reaction. As regards the results given with the antigen prepared from the pure cultures of Treponema pallidum, 17 of the sera gave the same result as with the stock antigen, four gave weaker reactions, six gave stronger reactions, and four gave positive results in sera in which the stock an- tigen gave a negative result. However, the pallidum an- tigen gave seven negative reactions in twenty sera that gave absolute inhibition of hemolysis with the stock an- tigen, (not shown in table) thus proving that the pallidum antigen is untrustworthy in the diagnosis of syphilis. Of the various syphilitic sera tested, fifteen were tested with all three antigens, and the results are valuable be- cause they indicate the group nature of the complement fixation reaction obtained with these alcoholic extracts of Treponema pallidum, Treponema pertenue, and Treponema micro dentiiim. Table XXIV gives the results obtained: In four of the cases, the results were the same with all of the culture antigens but in the remaining cases they varied, the pallidum antigen giving results more like those of the stock antigen, while the pertenue and microdentium antigens gave quite similar results, although those of the microdentium approached nearer to those of the pallidum than did the pertenue. Unlike Noguchi, we obtained positive results not only in latent syphilitic infections but also in primary and second- 152 THE WASSERMANN TEST Table XXIV Complement Fixation in Syphilitic Serum with Culture Antigens SEBUM NO. STOCK ANTIGEN T. PAL- LIDUM ANTIGEN T. PER- TENUE ANTIGEN T. MICRO- DENTIUM ANTIGEN CONTROL ANTIGEN REMARKS 1 + + + + + - Latent, 4 years 2 + + - - - - Latent, 3 years Q + + - + - + - - Latent, duration? 4 + - - - - Primary, 2 weeks 5 + + - - - Secondary, 1 year 6 + + + - + - Secondary, duration? 7 + - + - - + - - Latent, duration? 8 + - + + + + + - Secondary, 3 months 9 + - + - + - - - Latent, 4 years 10 + - + + - + - - Latent, 6 years 11 + - + - - + - - Latent, 2 years 12 - + - - + - - Latent, 1% years 13 - + + + - - Latent, 6 years 14 - + - - - - Primary, 2 weeks 15 - + - - 0 0 Secondary, 1 year ary infections with the culture antigens. It should be re- membered, in this connection, that Noguchi worked with an aqueous extract of Treponema pallidum while we used alcoholic extracts, and thus the variation may be explained, although the writer does not agree with Noguchi that it is impossible to extract from this organism specific comple- ment-fixing substances with alcohol. Our results with these alcoholic extracts of treponemas are too much at variance with those obtained with the ordinary alcoholic extract of fetal syphilitic liver to make it at all probable that the entire complement-fixing or absorbing power of these cul- ture antigens was entirely nonspecific in character, and our experiments with rabbit blood serum, which gave identical results with those of Noguchi, are evidence of this fact. Results With Rabbit Blood Serum.-We tested the blood serum of two rabbits presenting acute symptoms of syphilitic infection; of three that had recovered from a syphilitic orchitis; of two that had recovered from yaws; and of four normal rabbits, and the results are given in Table XXV. The important fact shown by these experiments is that COMPLEMENT FIXATION IN SYPHILIS 153 Table XXV Complement Fixation with Specific Treponema Antigens in Rabbit Sera RABBIT NO. STOCK ANTIGEN T. PAL- LIDUM ANTIGEN T. PER- TENUE ANTIGEN T. MICRO- DENTIUM ANTIGEN CONTROL ANTIGEN REMARKS 1 + + - - - Active lesions 2 + + - - - - Active lesions 3 - - - - 0 Recovered from syphilis 4 - - - - 0 Recovered from syphilis 5 - - - - - Recovered from syphilis 6 - - 0 - Recovered from yaws 7 - - - - Recovered from yaws 8 - - - 0 - Normal rabbit 9 - - - 0 - Normal rabbit 10 - - - 0 - Normal rabbit 11 - - 0 0 Normal rabbit no reaction with the culture antigens occurred in any of the rabbit sera, although the two animals showing active le- sions of syphilis gave absolute inhibition of hemolysis with the stock antigen. These results are in entire agree- ment with those of Noguchi, who found that his pallida an- tigen did not react with rabbit serum even when the ani- mals presented active lesions of the infection. On the other hand, antigens prepared by extracting with alcohol the testicles of rabbits presenting a syphilitic orchitis will give positive results with syphilitic sera and also with the blood of rabbits injected with syphilis, thus proving that the alcoholic extract of cultures of Treponema pallidum is an entirely different antigen from that obtained from the infected testicle of rabbits and that the reactions ob- tained with such an extract are, at least, partly specific in character. In 1911, Nichols published some experiments of ours upon complement fixation with alcoholic antigens made from the testicles of rabbits suffering from experi- mental syphilis and yaws, and we found that while such antigens gave positive results with certain syphilitic sera, they also reacted with certain normal sera, as well as sera from other diseases, and that an antigen prepared from the testicles of rabbits inoculated with syphilis give posi- tive results with the sera of rabbits inoculated with yaws 154 THE WASSERMANN TEST and vice versa. In addition, we found that alcoholic an- tigens prepared from the testicles of normal rabbits also gave positive results, so that we did not consider that this type of antigen was at all specific in character. From our experiments with antigens prepared from pure cultures of the treponemas named we reached the following conclusions: 1. Alcoholic extracts of pure cultures of Treponema pal- lidum can not be depended upon in the diagnosis of syph- ilitic infection by the complement-fixation test, for many cases of undoubted syphilis, giving absolute inhibition of hemolysis with the stock antigen, an alcoholic extract of fetal syphilitic liver, gave a negative reaction with the culture antigen, and the reactions obtained with the pal- lidum antigen were generally weaker and less conclusive than with the stock antigen. 2. The results appear to prove that there are specific group reactions, in the way of complement fixation, or absorption, between Treponema pallidum, Treponema per- tenue and Treponema microdentium, as antigens prepared from pure cultures of these organisms may all give comple- ment fixation in the presence of syphilitic serum. Later experiments69 with antigens prepared from pure cultures of Spirochata refringens, Spirillum cholera, and Bacillus typhosus demonstrated that, while Spirochata re- fringens sometimes reacted positively with the sera tested, the antigens prepared from the cultures of Spirillum chol- era and Bacillus typhosus never give a positive reaction, thus proving that the theory that positive reactions with these pure culture antigens might be due to lipoidal sub- stances produced in the media itself by the growth of the organisms is practically untenable, and that there are group reactions among this class of organisms and that alcoholic extracts of these organisms contain some specific substance common to all, to a greater or lesser extent. Since our work a considerable number of investigators have studied the immunological phenomena of cultivated COMPLEMENT FIXATION IN SYPHILIS 155 treponemas and spirochetes, and our results have been confirmed. Thus, Kolmer, Williams, and Laubaugh,70 working with aqueous and alcoholic extracts of Treponema pallidum obtained positive reactions in secondary, tertiary, and congenital syphilis. They found that the aqueous ex- tracts gave better results than the alcoholic but that, with both, the reactions were generally weaker than those ob- tained with the ordinary antigenic extracts used in the Wassermann test. They also obtained weak reactions with control antigens made from uninoculated media and other organisms. Zinsser, Hopkins, and McBurney71 in an ex- tended series of studies upon Treponema pallidum and syphilis obtained specific agglutination and complement fixation with culture treponemas but conclude that their work lends no support to the belief that either agglutina- tion or complement fixation with cultures of Treponema pallidum will prove of any practical value in the diagnosis of syphilis. Finally, the work of Noguchi and Akatsu72 proves that group reactions occur between certain tre- ponemas and that specific immune bodies capable of pro- ducing agglutination, complement fixation, and opsoniza- tion could be produced in the blood serum of rabbits by injection of these organisms. They worked with Tre- ponema pallidum, Treponema calligyrum, Spirochaeta re- fringens, Treponema microdentium, and Treponema mu- cosum, and found that the pallidum produced higher de- grees of agglutination and complement fixation, than the others, but that group reactions with complement fixation occurred between Treponema pallidum and Treponema cal- ligyrum, between calligyrum and Spirochaeta refringens, and between Treponema microdentium and Treponema muco- sum, while there was slight group reaction between the lat- ter and Treponema pallidum. The present status of this subject is, that while specific complement fixation, at least within the limits of a group, is possible with either aqueous or alcoholic extracts of pure cultures of Treponema pallidum, and that aqueous extracts 156 THE WASSERMANN TEST give the best results there is no evidence sufficient to prove that such antigens are of real practical value in the diagnosis of syphilis by the Wassermann test. The re- actions obtained with them are generally weak and doubt- ful and they are so often negative where the ordinary anti- genic extracts are positive that their use in diagnosis is not justifiable. Further research may result in giving us a specific antigen for the Wassermann test but at present the outlook is very doubtful that such an antigen can be evolved. CHAPTER VII THE RESULTS OF THE WASSERMANN TEST IN THE VARIOUS STAGES OF SYPHILIS. TIME OF APPEARANCE OF THE REACTION. THE SPECIFICITY OF THE WAS- SERMANN TEST The results of the Wassermann test vary with the dif- ferent stages of syphilitic infection, the highest percentage of positive results being obtained in the secondary stage of the disease and the lowest in latent infections. The results also vary with the methods employed by different observers, although, in the hands of experienced serol- ogists, the results obtained with one method are practically the same as those with another, and even in the hands of the inexperienced the results often are almost as good as when the test is done by the experienced investigator. It is one of the most fortunate things connected with this test, that it has given such accurate results with so many different modifications of the original test, and in the hands of such a great number of workers, many of whom were, and are, far from being experts in serology. As Vedder well says :73 "One is impressed by the fact that the Wassermann re- action must be a test of most surprising merit to have sur- vived all the clumsy technic that has been perpetrated in its name." The figures as to the results of this test in the various stages of syphilis which are here given, aside from those given by authorities quoted, are based upon more than fifty thousand Wassermann tests personally made by the writer, using the method which has been recommended in Chapter V. A great many thousands of these tests were made upon conditions due to other causes than syphilis 157 158 THE WASSERMANN TEST while a still greater number were reexaminations made upon the same case, sometimes as many as twenty or more tests having been made upon a single patient, in the course of months or years. In the army it has been customary to control the treatment of syphilis by repeated Wassermann tests and this practice has not only enabled army surgeons to have scientific control of the treatment of their syphi- litic patients but has also enabled us to have a very ac- curate check upon the efficiency and reliability of the modi- fication of the Wassermann test which is recommended in this work. The writer has tried to keep himself informed as to its success in the hands of other laboratory workers and it is gratifying to be able to report that the modification has proved satisfactory in every laboratory in the army in which it has been adopted, and this includes nearly all, and that it has been recommended as a standard method for all army laboratories by Lieut. Col. Vedder,74 who has devised special apparatus which makes such standardiza- tion possible. Before considering the results of the Wassermann test in each stage of syphilis, it is well to call attention to the fact that many of the statistics which are to be found in the literature are based upon observations made during the first few years after the test was discovered and are, there- fore, not representative of the results that are really being obtained today. A great amount of research has been ex- pended upon the various reagents used in the Wassermann test and a great many factors have been discovered that influence the results of the test, in one way or another, so that at the present time much better results are being obtained with the test than formerly and many inaccura- cies in technic have been corrected. This fact must be borne in mind in considering the percentages of positive results recorded by the earlier observers in the various stages of syphilis and especially the numerous reports of positive results in diseases other than syphilis. In fact, the results now being obtained with this test are hardly at RESULTS IN VARIOUS STAGES OF SYPHILIS 159 all comparable with those which were obtained with it dur- ing the first year or two after Wassermann's publication of his technic. As an illustration of the results obtained with the au- thor's method in the various stages of syphilis Table XXVI is inserted, giving the results of the test in 5,600 syphilitic infections, personally tested by him: Table XXVI The Results of the Wassermann Test in 5,600 Cases of Syphilis STAGE TOTAL CASES POSITIVE % NEGATIVE % Primary 1080 970 89.8 110 10.1 Secondary 2217 2132 96.1 85 4.9 Tertiary 728 633 87.4 95 13.5 Latent 1525 1039 68.1 486 31.8 Congenital 28 25 82.2 3 10.7 Parasyphilis 22 15 68.1 7 31.8 TOTALS 5600 4814 85.9 786 14 In the above cases no test was considered positive un- less there was complete inhibition of hemolysis (a four- plus reaction) except in the very early primary cases when a three-plus reaction was considered positive in a few in- stances. By insisting upon absolution inhibition of hem- olysis as indicating a positive reaction in these cases it is undoubtedly true that some cases of infection, which gave reactions almost absolute, have been erroneously consid- ered as negative, but in most of the army laboratories we have not felt justified in basing a positive diagnosis of syphilis upon anything less than absolute inhibition of hemolysis. In considering this table it should also be re- membered that in at least 95 per cent of the cases tabulated only one Wassermann test was performed and that had repeated tests been made it is undoubtedly true that a higher percentage of positive results would have been ob- tained in all of the stages of syphilis tested. Our results are in practical agreement with those of most recent observers, so that it may be stated that about 86 per cent of infections with syphilis will give a positive 160 THE WASSERMANN TEST reaction with the Wassermann test, if they are distributed among the various stages as in the table given; that 10 per cent of primary cases will give a negative result; about 5 per cent of secondary cases; about 13 per cent in the tertiary stage; and about 30 per cent in the latent stage of the disease. With good technic and efficient antigenic extracts one should obtain 90 per cent of positive results in the primary stage of syphilis; 95 per cent of positive re- sults in the secondary stage; 87 per cent of positive results in the tertiary stage; and 70 per cent of positive results in the latent stages. Results in the Primary Stage of Syphilis.-It is now a generally accepted fact that the earlier a syphilitic infec- tion is diagnosed the better will be the results of any method of treatment that may be instituted. Thus the early diagnosis of the disease becomes of paramount im- portance and with the dark-field apparatus and the Was- sermann test we are supplied with agencies that, if prop- erly used, should enable us to make the diagnosis during the early portion of the primary stage of the disease in the vast majority of infections. If possible every patient exhibiting a sore upon the penis should be given the benefit of an examination of the serum from the sore by the dark-field apparatus. If this were done, the Treponema pallidum would be found in most instances of true syphilitic infection, and thus the dis- ease would be diagnosed at a very early stage. However, the dark-field apparatus is often unobtainable and the ex- amination can not be made, so that the question arises as to the diagnostic value of the Wassermann test in the primary stage of syphilis. If one were to believe some writers it would be concluded that during this stage of syphilis the Wassermann test is of very little value, as so many cases give a negative reaction, but this has not been the writer's experience with his modification of the test, for, as already mentioned, practically 90 per cent of the primary cases he has tested have given a positive result before the de- RESULTS IN VARIOUS STAGES OF SYPHILIS 161 velopment of secondary symptoms. Therefore, if a dark- field examination can not be made, or is negative, it is recommended that every patient presenting a suspicious lesion upon the penis be given a Wassermann test, for if modern research upon syphilis has proved anything, it is that the morphology of a sore upon the penis is of much less value than formerly supposed, and that the typical Hunterian chancre is almost as much the exception as the rule. Mixed infections of chancre and chancroid, with the resultant confusion in clinical pictures, have been quite common in the writer's experience, and with the admis- sions by the best authorities upon syphilis that it is often impossible to diagnose syphilis from the appearance of the initial lesion, the use of the Wassermann test and the dark- field apparatus in every suspicious case is imperative. From experience and experimental observations it is certain that a syphilitic infection must exist for a certain length of time in man before the substance or substances which give rise to a positive Wassermann reaction appear in the blood. Just how long this is will vary with the viru- lence of the infection and the resisting power of the pa- tient, but it is beyond controversy that sometimes the test proves positive a very few days after the appearance of the initial lesion, although this occurs but rarely. The writer has observed and reported75 one case in which the reaction occurred only five days after the appearance of the chancre and two cases in which it occurred after eight days of the appearance of the initial lesion. As will appear in Table XXIX, no less than 27 of 77 cases, or 36 per cent of primary cases tested by the writer, gave a positive reaction during the first week after the chancre appeared and al- most 60 per cent during the second week after the appear- ance of the chancre. It is thus evident, that even very early in the primary stage of syphilis a positive Wasser- mann reaction may occur, and thus the test will prove of value in diagnosis. As would be expected, the statistics given by different au- 162 THE WASSERMANN TEST thorities regarding the percentage of positive results in the primary stage of syphilis vary greatly, especially those published shortly after Wassermann's original paper upon the subject. Kolmer78 states that in general, in primary syphilis positive results will be found in from 80 to 90 per cent of the cases tested, and the writer agrees with this statement, as it has been his experience that when properly performed the Wassermann test should give at least this number of positive results. Table XXVII gives the results obtained by different observers in the primary stage of syphilis: Table XXVII Results of the Complement-fixation Test for Syphilis in the Primary Stage of the Disease NAME OF OBSERVER NO. OF CASES POSITIVE % POSITIVE Arning 48 25 60 Bering 56 47 84 Boas 50 30 60 Brack-Stern 27 13 48.5 Fischer-Meier 8 6 75 Grosser 20 19 95 Hancken 17 15 88 Craig 1080 970 89.8 Kaplan 138 125 90 Noguchi 70 65 92.8 Swift 16 13 81 Vedder 242 178 73.5 From this table it is evident that the percentage of posi- tive results with the Wassermann test in the primary stage of syphilis has varied greatly in the hands of different ob- servers, the percentage being as low as 48 per cent and as high as 95 per cent. However, as already stated, from 80 to 90 per cent of the usual run of primary infections com- ing to the physician should give a positive reaction. Of course, if tests were made upon a very large number of very early infections in a series reported, the percentage would be much lower than if cases were tested from two to three weeks after the appearance of the initial lesion, but the writer's experience has been that at least 80 per cent of cases of primary syphilis should give a positive RESULTS IN VARIOUS STAGES OF SYPHILIS 163 Wassermann test and it is believed that, in the army, the initial lesion is reported to the surgeon at least as soon as it is, in the vast majority of cases, in civil life. It is true that the percentage of positive results with this test varies more in the primary stage of syphilis than in other stages, with the possible exception of the latent stage, due to many reasons, among the most important being that the test is made within a very few days of the ap- pearance of the initial lesion or early treatment often ob- scures the reaction at this stage of the disease. The writer has often observed negative cases become positive with the cessation of treatment, during the primary stage, and hundreds of instances in which, in untreated cases, a negative reaction became positive after an interval of days or weeks, and before the occurrence of secondary symptoms. It should be remembered, however, that a cer- tain proportion of negative results must be expected in the primary stage of syphilis and that a negative result is no proof that a suspicious lesion is not syphilitic. Time of Appearance of the Reaction The date after infection with syphilis that one may ex- pect the Wassermann reaction to be positive is a point of great practical interest, for the earlier treatment is com- menced in syphilis the better are the therapeutic results, and unless the Wassermann test can assist in diagnosis early in the disease, it loses much of its value as a diag- nostic agent. While the ideal method of diagnosing syph- ilis in the primary stage is by the demonstration of Tre- ponema pallidum with the dark-field apparatus, this meas- ure is not infrequently unsuccessful, especially if local treatment has been given the initial lesion, and, in addi- tion, comparatively few practitioners possess the appara- tus. The demonstration of the treponema in stained prep- arations or by the India ink method should always be at- tempted but these methods are also often unsuccessful, so 164 THE WASSERMANN TEST that one must depend upon the Wassermann test fre- quently for an early diagnosis. Contrary to the opinion of some observers, the writer believes that this test is a most valuable aid to the diag- nosis of syphilis even thus early in the disease, for, in his experience, nearly 90 per cent of primary cases tested by him have shown a positive result before the appearance of secondary symptoms, and while a single test performed at a very early date after the appearance of the chancre will be negative in a very large proportion of cases, repeated tests will prove that the reaction becomes positive before the appearance of secondary symptoms in most patients. In Table XXVIII Vedder77 shows the results of the ex- amination of cases of syphilis in the primary stage by both the dark-field apparatus and the Wassermann test, and it is of interest as demonstrating that both methods fail, at times, to aid in the diagnosis of the disease. Table XXVIII Result of Examination of Primary Cases of Syphilis by Dark-Field and THE WASSERMANN TEST* DURATION OF DARK- FIELD WASSER- MANN DURATION OF SORE DARK- FIELD WASSER- MANN SORE 2 days + - 1 month - + + 3 days + - 7 days + + + 7 days + - 9 days + + 7 days + - 14 days + + 10 days + - 15 days •1- + 2 weeks + - 1 month + + + 2 weeks + - 1 month + + 3 weeks + + - 1 month + + + 3 weeks + - 1 month + + + 3 weeks + - 1 month + + + 1 month + - 1 month + + + 1 week - + + 1% months + + + 1 week - + 2 months + + + 2 weeks - + + 1 month - - 3 weeks - + 1 month - - 1 month - + + *The sign 4- 4- equals absolute inhibition of hemolysis, the four-plus reaction of many laboratories. Table XXVIII is conclusive evidence of the value of the Wassermann test as an aid in the diagnosis of syphilis in RESULTS IN VARIOUS STAGES OF SYPHILIS 165 the primary stage. It will be noted that in no less than three cases the Wassermann test was positive one week after the appearance of the chancre and that in two of these cases the dark-field examination was negative. The table also demonstrates that in every case of primary syphilis both a dark-field examination and a Wassermann test should be made, if possible. Table XXIX, compiled from personal tests, gives the date of appearance of the positive reaction in periods of weeks after the appearance of the initial lesion. Six hun- dred cases of primary syphilis are here considered and the Wassermann tests were performed by the writer in the ordinary course of routine examinations in the army. A great many more cases of primary infection were really tested but are not included for the reason that we could not be sure of the data regarding the date of appearance of the initial lesion. Table XXIX Date of Appearance of Wassermann Reaction in Weeks in 600 Cases of Primary Syphilis WEEK AFTER APPEARANCE OF CHANCRE TOTAL CASES POSITIVE % NEGATIVE % First week 77 27 36.3 50 64.9 Second week 155 92 59.3 63 40.3 Third week 158 109 68.9 49 31. Fourth week 167 129 77.2 38 22.7 Fifth week 43 35 81.3 8 18.6 From this table it is evident that, in the writer's ex- perience, 36 per cent of cases of primary syphilis have given a positive Wassermann reaction by the end of the first week after the appearance of the initial lesion; almost 60 per cent by the end of the second week; almost 70 per cent by the end of the third week; over 77 per cent by the end of the fourth week; and over 80 per cent by the end of the fifth week after the appearance of the chancre. It is also evident that this test is of very distinct value in the diagnosis of syphilis during the primary stage of 166 THE WASSERMANN TEST the disease, and should always be resorted to if a dark- field examination is impossible or gives a negative result. A single negative test, at this stage especially, should not be considered as proving that a suspected lesion is non- syphilitic, but repeated tests should be made before a pa- tient is assured that the condition present is not due to syphilis. In the army, owing to the situation of the va- rious Corps Area Laboratories, the Wassermann test is al- ways available and suspected cases can be examined at weekly intervals, a practice which enables the army sur- geon to establish the diagnosis very early in the disease, but in civil practice this is generally impossible, owing to the cost of the test. However, whenever possible, repeated Wassermann tests should be made upon patients suspected of presenting the primary lesion of syphilis. As regards the question as to how early the positive Wassermann may appear in the blood after the primary lesion appears, there are a considerable number of obser- vations upon record showing that positive reactions have been obtained as early as the third and fourth days after the appearance of the chancre, and a few in which a posi- tive reaction has been reported even before the appearance of the initial lesion. In view of our knowledge of the na- ture of the syphilitic poison it may be stated that one must regard with suspicion the reports of a positive Wasser- mann before a primary lesion was observed, for it is prac- tically impossible to rule out previous syphilitic disease in most patients and unless this can be done or it can be definitely proved that the lesion supposed to be primary in such cases is really so, by the demonstration of Tre- ponema pallidum, the evidence can not be considered as con- clusive. However, it may be accepted as proved beyond a doubt that the Wassermann test is not so very infrequently posi- tive by the end of the first week after the appearance of the chancre, and the writer has observed two cases in which it was positive five days after the sore was first noticed, RESULTS IN VARIOUS STAGES OF SYPHILIS 167 and in both cases Treponema pallidum was demonstrated in serum from the lesions. Well authenticated cases are on record in which the test became positive two or three days after the initial lesion was noted, so that the possi- bility of a positive Wassermann test before the appearance of such a lesion can not be denied, although it must be ex- ceedingly rare. The results obtained by Klauder159 in the early diagnosis of syphilis by means of the Wassermann test and the dark- field examination are interesting and valuable, as showing the relative value of the two methods and the time of ap- pearance of complement fixation. He tested 115 cases and obtained the results shown in the accompanying tables. Dark-Field Examination of 115 Chancres DURATION OF CHANCRE NO. OF CASES POSITIVE NEGATIVE PER CENT POSITIVE IN DAYS 1-10 33 31 2 93.9 10-20 17 9 8 52.9 20-30 20 10 10 50.0 30-40 5 3 2 60.0 Over 40 40 12 28 30.0 Totals 115 65 50 57.3 Results of Wassermann Test in Same 115 Infections DURATION OF CHANCRE NO. OF CASES POSITIVE NEGATIVE PER CENT POSITIVE IN DAYS 1-10 33 12 21 36.0 10-20 17 11 6 64.7 20-30 20 14 6 70.0 30-40 5 5 0 100.0 Over 40 40 40 0 100.0 Totals 115 82 33 74.1 Klauder's results, as regards the time of the appearance of complement fixation in syphilis, are practically identical with the writer's, while the fact is evident that the earlier the dark-field examination is made the greater will be the positive findings and that very early in the primary stage such an examination is more valuable than is the Wasser- mann reaction. However, it is also evident that the latter 168 THE WASSERMANN TEST test is really more valuable in the primary stage of the disease if the results obtained after ten days from the date of appearance of the chancre are considered. It is now generally conceded that when the primary le- sion of syphilis appears the disease has already become a generalized infection, so that it is not difficult to under- stand why a positive Wassermann reaction may occur very soon after the chancre appears, and the variation that oc- curs as to the time of the appearance of the positive re- action in different patients must depend upon the virulence of the treponema and the reaction of the tissues of the patient to the organism. The writer has observed many cases in which repeated Wassermann tests always proved negative, though there was no question of the presence of syphilis and marked lesions were present at the time that the tests were made, so that in a certain percentage of cases of syphilis it is probable that a positive reaction will never be obtained. The writer has also observed two pa- tients in whom the primary lesion of syphilis was present, as demonstrated by the finding of Treponema pallidum by the dark-field, and whose blood serum gave a positive re- action prior to the development of secondary symptoms, but in whom no secondary symptoms ever developed, although they were under observation for nearly one year, and in whom the Wassermann reaction became negative, although no specific treatment was ever administered. In his opin- ion, these cases demonstrate that syphilis is sometimes spontaneously cured, and it is not improbable that this occurs much more frequently than is generally supposed. In view of the reports of positive Wassermann reactions occurring before the initial lesion appears, the recent work of Woods and Morris78 upon complement fixation in trypanosomiasis is very suggestive. They found that with an antigen prepared from the spleen of a rat heavily in- fected with Trypanosoma equiperdum, complement fixa- tion usually appeared in an infected animal's blood serum RESULTS IN VARIOUS STAGES OF SYPHILIS 169 within eight days after inoculation, usually at the time that the trypanosomes appeared in the blood of the infected animal, but sometimes even before this occurred. In every instance, however, complement fixation preceded the ap- pearance of symptoms of the infection. Results in the Secondary Stage of Syphilis.-It is in the secondary stage of syphilis that the highest percentage of positive results are obtained with the Wassermann re- action. The per cent of positive reactions reported by va- rious observers varies somewhat, but it may be stated that in this stage of the disease practically 95 per cent of the patients should give a positive result, provided, of course, that treatment has not been administered, but even in such cases one would expect from 90 to 95 per cent of positive results. Kolmer79 states that in untreated secondary syph- ilis he has never obtained a negative result and Boas80 in 437 untreated secondary cases obtained 100 per cent of positive reactions. The writer, in 2,217 cases of secondary syphilis, many of whom had received some specific treat- ment, obtained 2,132 positive results, or 96.1 per cent, and Vedder,81 in 310 secondary cases of similar nature, obtained 91 per cent of positive results. Table XXX gives the results obtained in the secondary stage of the disease by various observers. From this table it will be noted that the percentage of positive results in the secondary stage of syphilis varies, according to different observers, from 79 per cent to 100 per cent, but it should be remembered that many of the cases tested by one observer may have had specific treat- ment, while those tested by another had had no treatment whatever, so that the low percentages reported can not be considered as giving any indication of the real value of the test in the absence of data regarding specific treatment. As already stated, the percentage of positive results in untreated cases of secondary syphilis should be something 170 THE WASSERMANN TEST Table XXX The Results of the Complement-fixation Test for Syphilis in the Secondary Stage of the Disease OBSERVER NO. OF CASES POSITIVE % POSITIVE Arning 107 99 93 Bering 113 111 98 Bruck-Stern 115 101 87.1 Blumenthal-Roscher 131 130 99 Craig 2217 2132 96.1 Hoene 376 260 79.1 Lesser 204 186 91 Lederman 110 108 98 Merz 377 366 97 Noguchi 197 190 96 Schonnefeld 112 112 100 Swift 76 70 92 Vedder 310 285 91.9 over 95 per cent, and unless as high a percentage as this is obtained the complement-fixation system used should be carefully studied for the occurrence of possible errors, as the Wassermann test, as at present performed, should not be given a lower percentage of positive results in the secondary stage of syphilis than that mentioned. A negative reaction in this stage of syphilis, when there is a clear history of infection and lesions are present, is of no value whatever in excluding the disease; and as nega- tive reactions may occur, even when the test is repeated many times during this stage of the disease, the diagnosis of syphilis should be made, in such instances, without ref- erence to the result of the test. It should also be remem- bered that the severity of the lesions present has nothing to do with the type of reaction obtained, for cases with very mild symptoms will frequently react as strongly as those showing severe lesions, and not infrequently a pa- tient presenting the most pronounced secondary eruption and mucous patches will give a weak Wassermann reaction, which, unless repeated, might be interpreted as doubtful. However, in these cases repeated Wassermann tests almost invariably result in the appearance of a strong positive re- action. RESULTS IN VARIOUS STAGES OF SYPHILIS 171 Results in the Tertiary Stage of the Disease.-Perhaps in no other stage of syphilis has the Wassermann test proved as valuable as in the tertiary stage, and especially in those cases in which there is involvement of the central nervous system. The percentage of positive results ob- tained will vary, of course, with the amount of treatment that has been given the patient, for in a large majority of the cases of tertiary syphilis, observed in practice, a considerable amount of specific treatment has been admin- istered prior to the time that the blood was collected for a Wassermann test. If cases are tested that have received no specific treatment and in which active tertiary lesions are present, the percentage of positive results should be above 95 per cent, but in the usual run of cases examined the Wassermann test, if properly applied, should give posi- tive results in from 80 to 85 per cent of cases examined. The writer, in 728 cases of tertiary syphilis obtained 633 positive results, or 87.4 per cent, using the method recom- mended in this work, and Vedder,82 using the same method, obtained 227 positive reactions in 263 cases tested, or 86.3 per cent. Table XXXI gives the results reported by various ob- servers of the complement-fixation test for syphilis dur- ing the tertiary stage of the disease. Table XXXI Results of the Complement-fixation Test for Syphilis in the Tertiary Stage of the Disease OBSERVER NO. OE CASES TESTED POSITIVE % POSITIVE Arning 30 27 90 Bering 45 37 82 Bruck-Stern 47 27 57 Fleishman 41 40 98 Hohne 33 21 63 Craig 728 633 87.4 Lederman 78 75 96 Lesser 131 119 90 Merz 158 127 80.3 Noguchi 177 159 89.9 Swift 45 37 80 Vedder 263 227 86.3 172 THE WASSERMANN TEST From this table it is evident that the results of the Was- sermann test in the tertiary stage of syphilis varied from 57 per cent to 96 per cent, according to various observers. This great variation in the results may probably be ex- plained by certain of the observers including under the tertiary stage cases tested, 1, cases without definite symp- toms; 2, cases that had received specific treatment within a recent date or were receiving treatment; and 3, cases in which tertiary symptoms had never appeared. It is evi- dent that, until statistics are based upon similar cases a great deal of variation must be expected in the results of this test in the hands of different observers. In syphilitic disease of the blood vessels, as aortitis, aortic aneurism, and aortic insufficiency, the Wassermann test has proved of the very greatest diagnostic value and in syphilitic disease of the central nervous system the test is invaluable. This subject will be discussed more fully in Chapter X. Results in the Latent Stages of the Disease.-The Was- sermann test in the latent stages of syphilis has been of the greatest use in indicating that treatment was neces- sary in patients thought to be cured of the infection be- cause of the absence of symptoms for long periods of time. There is probably no disease occurring commonly in the practice of the physician which so often exhibits intervals during which clinical symptoms are absent as does syph- ilitic infection and often a positive Wassermann reaction is the only tangible evidence of the disease. The patho- logical investigations of careful workers, as Warthin, have shown that the human body may be extensively invaded by Treponema pallidum without clinical symptoms being evi- dent and with a positive Wassermann reaction as the only evidence of such infection, and one of the most valuable of assets of the Wassermann test is its capacity for detecting such infections. Of course, the test is frequently negative RESULTS IN VARIOUS STAGES OF SYPHILIS 173 in this class of patients, as would be expected, for, as the writer has shown, a certain amount of the substance or substances causing complement fixation in syphilis must be present in order to accomplish this result, and in the latent cases of syphilis a sufficient amount is not present in a considerable percentage of cases, as shown by the re- sults of the test. The writer has tested 1,525 cases of latent syphilis, with a positive reaction in 1,039, or 68 per cent, using the method of performing the test recommended in this work. Ved- der,83 using the same method, tested 114 cases with a posi- tive result in 92, or 80.7 per cent. Boas,84 in 363 cases of early latent syphilis obtained positive results in only 40 per cent of the cases, while in older cases he obtained posi- tive results in but 22 per cent of treated cases but in cases treated carelessly the positive results equaled 74 per cent. Table XXXII gives the results of the Wassermann test obtained by the observers mentioned in the latent stage of syphilitic infection. Table XXXII Result of the Complement-fixation Test foe Syphilis in the Latent Stages of the Disease OBSERVER NO. OF CASES TESTED POSITIVE % POSITIVE Bering 147 75 48 Grosser 35 12 33.3 Craig 1525 1039 68.1 Fox 54 25 46 Lederman 78 36 46 Noguchi 265 206 77.7 Swift 39 25 64 Vedder 114 92 80.7 As would be expected, there is a great variation in the number of positive reactions obtained with the test in the latent stages of syphilis, as shown in this table. When it is remembered that in most of the cases tested specific treatment had been given for a longer or shorter period of time; that there were no symptoms of the disease pres- ent at the time the blood was tested; and that in some the 174 THE WASSERMANN TEST only history of a syphilitic infection was a doubtful one; the low percentage of positive results sometimes obtained in this class of cases can be understood. In the writer's cases there was always a clear history of infection but in the majority of the patients some specific treatment had been taken, so that the percentage of positive results ob- tained ; i. e., 68 per cent, may be considered the average percentage obtained in latent infections in such cases. The value of the Wassermann test in the diagnosis of syphilis in the latent periods of the disease can not be overestimated either from the sociological or therapeutic standpoint. The sociological aspect of latent syphilitic in- fection will be treated of more fully in the discussion of the subject of Wassermann surveys, but it can not be too strongly emphasized that there is a very large amount of syphilitic infection present in all countries that is never recognized and it has only been within very recent years that, because of researches upon various classes of the population by means of the Wassermann test, it has been realized to what a great extent this insidious disease has invaded society. From the therapeutic standpoint, the Wassermann test is most valuable in the latent stage of syphilitic infection because a positive reaction is the only index which we pos- sess as to the existence of the infection and the necessity for specific treatment. The test is often positive even after specific treatment has been administered for months and years, and the vast majority of patients treated with mer- cury, before the days of salvarsan, and supposed to have been cured, will give a positive reaction with the Wasser- mann test. Some authorities have gone so far as to state that they do not believe that any syphilitic infection was ever cured by mercury alone, because of the revelations of the Wassermann test in latent infections, but the writer believes this to be too strong a statement, although it must be admitted that the cases proving the contrary are com- paratively few in number. RESULTS IN VARIOUS STAGES OF SYPHILIS 175 Results in Congenital Syphilis.-The results of the Was- sermann test in congenital syphilis vary with the character of the case tested. In children showing the lesions of syphilis at birth the test gives 100 per cent of positive results, in the writer's experience, but in children exhib- iting the lesions of late congenital syphilis the percentage of positive results is lower, averaging from 80 to 85 per cent. In still older individuals in whom the disease is supposed to be congenital the percentage of positive re- sults will not exceed 70 to 75 per cent. Table XXXIII gives the results obtained by several ob- servers with the Wassermann test in congenital syphilis: Table XXXIII Results of the Complement-fixation Test for Syphilis in Congenital Syphilis OBSERVER NO. OF CASES TESTED POSITIVE % POSITIVE Arning 5 3 60 Craig 28 25 82.2 Jesionek-Meirosky 18 16 88.8 Noguchi 17 17 100 Vedder 128 123 96 The variation obtained by different observers in the re- sults of the Wassermann test in congenital syphilis are well shown in the table and depend upon the causes already mentioned. The recent work of Veeder85 upon the subject is of interest. He studied 100 syphilitic families with the following results: In the 100 families, 331 pregnancies occurred resulting as follows: abortions, 100, or 30.2 per cent; stillbirths, 31, or 9.3 per cent; living births, 200, or 60.5 per cent. Of the 200 living births, 39 of the children had died at the time of the investigation and of the 161 living children, 12 died during the course of the investiga- tion. Of the 161 tested, 107 had both clinical symptoms of syphilis and positive Wassermann reactions; 5 were clinically positive but gave negative Wassermann tests; 16 were clinically negative but gave positive reactions, leaving 33 children who were both clinically negative and 176 THE WASSERMANN TEST who gave a negative Wassermann test. In this series of cases of congenital syphilis in young children born of syphilitic parents Veeder obtained 96 per cent of positive results, and his work well illustrates the value of the test in the diagnosis of congenital infection. The Wassermann test has changed our conception of Colles' and Prof eta's laws and has explained these appar- ent contradictions in the etiology of syphilis. Colles' law, that an apparently healthy mother of a syphilitic child may suckle the child with impunity, even though it pre- sents the most infectious lesions of the disease, is ex- plained by the fact that the mother, although showing none of the lesions of syphilis is really in the latent phase of the disease, as demonstrated by the positive Wassermann reaction that is almost invariably obtained with the blood of such mothers. Prof eta's law, that the child bom of a syphilitic mother, but presenting no evidence of the dis- ease, may suckle its mother with impunity, has also been shown by the Wassermann test to be due to latent syphilis in the child, as in these children the test is generally pos- itive although no symptoms of syphilis are present and none may appear for long periods of time. In other words, children born of syphilitic parents possess no immunity to syphilis as was long believed, nor does a mother who with impunity suckles her syphilitic child, possess an im- munity, but is herself syphilitic. Just what part late hereditary syphilis plays in the nu- merous instances in which a positive Wassermann reaction is found in patients who deny any syphilitic infection and who may, or may not, present evidences of the disease, is still a mooted question. In the writer's experience it has not infrequently occurred that patients who are absolutely truthful and who have given a positive Wassermann re- action, have denied even any exposure to infection or any lesions that could be considered syphilitic, and there has been no reason to doubt the truth of their statements. In RESULTS IN VARIOUS STAGES OF SYPHILIS 177 such cases, especially if there are no lesions of the dis- ease present, one should be very careful to eliminate any condition that has been proved to give a positive Wasser- mann reaction, but after this has been done the explanation of the reaction is often very difficult and even impossible. That hereditary syphilis, latent in character, is the cause of some of these reactions can not be doubted but the writer is loth to believe that most of them are due to such a cause. It is more probable that an accidental infection may have occurred in infancy or childhood, the initial le- sion of which went unrecognized, than that the positive test is due to a hereditary infection that has never caused symptoms or lesions of syphilis. The Specificity of the Wassermann Reaction While it must be admitted that the Wassermann reac- tion is not absolutely specific for syphilis, for positive results have sometimes been observed with other diseases, the fact remains that a positive reaction with this test indicates syphilis in so immense a proportion of individ- uals giving it that the practical value of the reaction in the diagnosis of syphilis is hardly at all decreased by the comparatively few instances in which such a result is ob- tained in other conditions. If one reads the works of writers published shortly after the discovery of the Wassermann test, the attention is at once called to the large number of conditions other than syphilis in which positive reactions are reported; but if these writings be compared with those published by in- vestigators during the past five years, it is evident that most of the so-called positive reactions must have been due to poor technic, for as our knowledge of the technic has increased, these nonspecific reactions have decreased in number, until today a positive reaction with the test is reported in but few conditions that are not due to syphilis. Any serologist who reports any considerable number of 178 THE WASSERMANN TEST positive reactions with the Wassermann test in diseases other than syphilis proclaims that his technic is poor, for, if properly performed, this test will give positive results in very few nonsyphilitic conditions. The earlier writers upon the subject reported positive results in such diseases as scarlet fever, septicemia, diph- theria, carcinoma, and most of the acute infections; but it is now well known that these conditions do not give a positive reaction if the test is properly performed. There is still an opinion prevalent among the profession that the test is frequently positive in scarlet fever, but the work of recent investigators proves conclusively that the test is not positive in this disease but that when a positive reac- tion does occur the child is syphilitic or syphilis can not be excluded. Hecht, Lateiner, and Wilenko86 tested 105 cases of scarlet fever and obtained but one positive result, and Kolmer87 in 250 cases which he tested with both the original Wassermann method and the Noguchi modifica- tion obtained positive reactions in only 5 cases, or 2 per cent, and in these cases syphilis could not be excluded. Browning and McKenzie88 examined 37 cases of scarlet fever and did not obtain a positive reaction in one of them. These results prove that the Wassermann test is not pos- itive in scarlet fever and that the large proportion of pos- itive results reported in the early literature of the test must have been due to improper technic. Another disease in which a considerable proportion of positive reactions have been reported is tuberculosis. Some authorities have reported as high as 30 to 40 per cent of positive results in this disease but such reports are absolutely unreliable and prove that the method of per- forming the test must have been erroneous. The writer has tested hundreds of cases of tuberculosis and while a few have shown a positive reaction, syphilis could not be excluded in any case, and the majority of the patients ad- mitted infection. There are many instances on record of syphilis of the lung existing along with tuberculosis of RESULTS IN VARIOUS STAGES OF SYPHILIS 179 the same organ and many in which syphilis of the lung was diagnosed as tuberculosis, and it is these cases that have caused the impression that tuberculosis frequently gives a positive reaction with this test, together with the large number of latent syphilitic infections which are found in patients suffering from other diseases. There is also con- siderable evidence to prove that tuberculosis renders an individual much less resistant to infection with Treponema pallidum and this probably has much to do with the prev- alence of the disease in the tubercular. However, that tuberculosis is a disease which causes a positive Wasser- mann reaction can not be longer maintained in view of the negative results of the best observers as reported in the most recent literature. There are, however, a few diseases in which a positive Wassermann reaction is sometimes obtained and where the most careful examination fails to disclose any evi- dences of syphilitic infection and the most careful in- quiries any history of the disease. These diseases are yaws or frambesia; leprosy, especially the tubercular form; some cases of relapsing fever; some malarial infections, during the febrile stage; and some instances of experimental try- panosomiasis in animals. Of these, yaws is the most im- portant because some of the lesions of this disease re- semble those of syphilis and in the lesions the causative organism, Treponema pertenue, may be found either by a dark-field examination of the secretions from the lesion or by staining methods. As this treponema is almost indistinguishable morphologically from Treponema pal- lidum, and the blood often gives a positive Wassermann reaction, some confusion in diagnosis may result, but a careful study of the case should enable the diagnosis of yaws to be made from the clinical history and symptoms alone. The writer has observed positive reactions in five cases of tertian malarial infection, during the febrile stage of the paroxysm, the blood becoming negative during the 180 THE WASSERMANN TEST afebrile periods. Thompson89 states that he has observed positive results several times in malarial disease during the stage of fever which became negative with the disap- pearance of the -fever and plasmodia, and our results have been confirmed by many observers. On the other hand, the occurrence of a positive reaction in malaria, even dur- ing the paroxysm of fever, is comparatively rare, for the writer has examined many cases in which the test was neg- ative and other investigators report similar results. How- ever, the microscopical examination of the blood should suffice to diagnose malarial infections, so that the occur- rence of a positive Wassermann reaction should cause no trouble in the diagnosis of the disease. If the reaction persists after the disappearance of the fever and plas- modia the diagnosis of syphilis should be made, for all investigators report that the positive reaction in malaria is only temporary in character. In the tubercular forms of leprosy numerous positive reactions have been reported but it is impossible to say just how many of these were true false reactions and how many were due to a complicating syphilitic infection. If it be remembered that most of the cases of leprosy that have been tested occurred in native races which were thor- oughly syphilized the difficulty of ruling out that infection can be appreciated. The writer believes that a positive re- action does occur in tubercular leprosy in certain stages of the disease but he also believes that the majority of the positive reactions reported in leprosy have been due to a concurrent syphilitic infection. In the anesthetic form of leprosy the Wassermann test is generally negative, so far as reports indicate. As tubercular leprosy can be easily diagnosed by the finding of the bacillus in the le- sions a positive Wassermann reaction should occasion no confusion in the diagnosis. Cases of relapsing fever have been found to give a posi- tive reaction and also some experimental infections with trypanosomiasis but it is yet unsettled whether or not the RESULTS IN VARIOUS STAGES OF SYPHILIS 181 blood of patients suffering from infection with Trypano- soma gambiensi or other human trypanosomes gives a pos- itive reaction. Positive reactions have been reported in diabetes mel- litus, in cases suffering from acidosis, and in pellagra, but so far as the latter disease is concerned it is now definitely proved that the Wassermann test gives only negative results. Warthin and others have observed positive results in cases of generalized carcinomatosis and sarcomatosis, which could not be explained by a complicating syphilitic infection. It is very probable that most of the positive results that have been observed in relapsing fever were really due to syphilitic infection, although, in view of the researches of numerous investigators who have shown that positive re- sults with this test may be obtained with antigens prepared from several species of spirochsetes, it might be expected that such results would be often obtained in the relapsing fevers, which are caused by parasites belonging to this genus. As a matter of fact, the subject demands much more study before it can be accepted that the relapsing fevers give a positive Wassermann reaction. As regards the number of positive reactions occurring in diseases other than syphilis it may be stated that, with a properly performed and controlled method of making the test, it will be so small as to be of practically no im- portance from a diagnostic standpoint. The writer has ob- tained 12 positive reactions among 4,000 individuals suffer- ing from diseases other than syphilis or 0.3 of 1 per cent. Of these patients five were suffering from tertian malarial infection and gave a positive result during the febrile stage, which became negative when the temperature re- turned to normal; three were diagnosed as tuberculosis; three as pityriasis rosea; and in one the diagnosis was undetermined. In the pityriasis rosea cases the reactions 182 THE WASSERMANN TEST were almost double-plus (four-plus), but would not have been reported as positive without a qualifying statement, while in two of the tuberculosis cases a history of syphi- litic infection was afterwards obtained, and the disease could not be excluded in the remaining case. But, even admitting that all of these cases gave a false positive re- action, the real value of the test is but very little affected as a diagnostic measure. Vedder90 has obtained results almost identical with those of the writer, using the writ- er's modification of the Wassermann test, as he obtained only four positive reactions among 1,049 individuals suf- fering from diseases other than syphilis. When it is remembered how difficult it is to absolutely exclude a syphilitic infection in any individual and the comparatively few instances in which the Wassermann test gives a positive result in diseases other than syphilis, one should accept with great caution the reports of positive reactions in such diseases, for the experience of those who have worked with this test for years and in hundreds and thousands of cases has been that a positive reaction, in the overwhelming majority of instances, demonstrates the presence of syphilis. There are a very few conditions in which a positive reaction is obtained with this test and in which syphilis can be eliminated, but as most of them can be easily diagnosed by laboratory and clinical methods, there should be no confusion caused by the result of the Wassermann test, but even in these cases a very careful inquiry and examination should be made before a possible complication of syphilis is excluded. The really remarkable results of the complement-fixation test for syphilis, as regards accuracy, are shown by the nu- merous postmortem investigations of different authorities, in which patients who gave a positive test during life, al- though no clinical evidences of syphilis could be discovered, were found to show syphilitic lesions of the organs at autopsy, or the Treponema pallidum was demonstrated in RESULTS IN VARIOUS STAGES OF SYPHILIS 183 the heart, blood vessels, or elsewhere upon histological ex- amination. The work of Warthin is especially notable in this direction and the recent report of Turnbull,160 for the Medical Research Committee of England, proves beyond controversy the paramount value of this test in the diag- nosis of syphilis. He made a careful postmortem examina- tion of 121 individuals who had been tested before death and concluded that there was no proof of the inaccuracy of the Wassermann test in the whole 121 cases and he con- cludes his report by the statement that "no other conclu- sion would appear to be possible than that the antemortem Wassermann test is a diagnostic weapon of astonishing precision." In surgical practice weak positive reactions are some- times noted in blood sera collected shortly after ether and chloroform anesthesia, so that the test should not be made upon patients who have been anesthetized until at least two days have elapsed since the anesthetic was admin- istered. In conclusion, it may be stated that the substance or substances that cause a positive Wassermann reaction ap- pear to be practically peculiar to the blood serum of pa- tients suffering from syphilitic infection, and while it can not be claimed that the test, when positive, is absolutely specific of syphilis, from a practical standpoint it is doubt- ful if a more specific test is employed in medicine, the margin of error appearing to be less than five-tenths of one per cent. CHAPTER VIII THE INTERPRETATION OF THE RESULTS OF THE WASSERMANN TEST. THE WASSERMANN TEST AS AN INDEX OF THE PREVA- LENCE OF SYPHILIS IN COMMUNITIES The writer has often been impressed with the evident misconceptions of many practitioners of medicine regard- ing the nature of the Wassermann reaction, the percentage of positive results that may be expected in the various stages of the disease, but, especially the exact significance that may be attached to a positive or negative result. In his experience, many patients have been told by their med- ical attendant that they were free from syphilis, the state- ment being based upon a single negative result with this test, while others have been told that they were suffering from the disease upon the strength of a plus or plus-minus reaction, in the absence of either a history of infection or of any symptoms of the disease. Such interpretations of the results of the Wassermann test are unwarranted and have brought the test into dis- repute in certain localities. In addition, innocent individ- uals have suffered great mental anguish from being un- justly stigmatized as afflicted with syphilis, and others have been infected by those who have been told that they were free from the disease by some careless or ignorant practitioner. Even today, over ten years since the Was- sermann test has been in general use by the profession, there is a surprising amount of ignorance regarding the interpretation of the various grades of the reaction, as re- ported by different laboratories, and much of the fault lies with the laboratory reports, for the terms strongly positive, positive, and weakly positive, so commonly used, 184 INTERPRETATION OF RESULTS 185 are misleading, unless they are carefully explained in ref- erence to each case. In fact, the writer has always believed that only three terms should be used in reporting the re- sults of the Wassermann test; i. e., positive, doubtful, and negative, leaving it to the clinician to decide how much weight should be attached to any grade of the reaction which is reported as doubtful, and which, in the army, would mean any degree of the reaction less than complete inhibition of hemolysis. While there is no doubt that many cases of syphilis react weakly consistently, especially in the early and late latent stages, never presenting complete in- hibition of hemolysis, the writer believes that all cases show- ing less than complete inhibition should be reported as doubtful, and the interpretation of the reaction left with the clinician. Of course, if there is a clear history of infection and typical clinical signs are present, the doubtful reaction would at once, in the opinion of the clinician, be considered as positive and as supporting the diagnosis of syphilis, but if, on the other hand, there was no history of infection and no symptoms are present, a doubtful Wassermann reac- tion is not sufficient proof upon which to base a diagnosis of syphilitic disease. The statements here made regarding the interpretation of the results of the Wassermann test are based upon over 50,000 tests personally performed by the writer, the major- ity of them being reexaminations made as a control of the treatment of the disease, the remainder being upon patients presenting the symptoms common to the various stages of syphilis. The percentage of positive and negative results obtained in these cases has already been given (Chapter VII), as well as the terminology used in reporting the re- sults, but it is well here to repeat the terminology used in the army laboratories. Four degrees of the reaction are noted in reports from the army laboratories. A positive reaction is reported as double-plus (+ +) and means that there was absolute inhibi- tion of hemolysis. A doubtful reaction is reported as 186 THE WASSERMANN TEST plus (+) or plus-minus (+-), the former term indicating that there was over 50 per cent inhibition of hemolysis, the latter that there was less than 50 per cent inhibition of hem- olysis. A negative reaction is reported as minus (-). In most civilian laboratories the results of the Wassermann test are reported as four-plus (+ + + +), three-phis (+ + +), two-plus (++),plus (+),plus-minus (+-), and negative (-). The four-plus reaction corresponds to the army double- plus; the three-plus and two-plus to the army plus; and the phis and plus-minus to the army plus-mimis. These distinctions between the terminology used should be borne in mind in the following discussion of the inter- pretation of the results of the test. . The writer has already discussed the specificity of the Wassermann reaction (Chapter VII) and there remains the question as to the significance of the positive reaction. Some authors have denied that a positive reaction with this test indicates the presence of syphilis, claiming that after the disease disappears the reaction still persists, and that the positive result is simply an indication that the patient has had syphilis sometime in the past. In the light of mod- ern research upon the disease this theory can no longer be considered as worthy of credence, for it is the unanimous experience of those who have had the most to do with the treatment of syphilis, as controlled by the Wassermann test, that with the disappearance of the treponema and of symptoms the reaction also disappears and that, in cases that are really cured, the reaction remains permanently absent. That this is true has been proved experimentally upon infected animals, and the disappearance of the reac- tion in these animals after the administration of salvarsan which has caused the disappearance of the treponemas, is conclusive evidence that the positive reaction does not per- sist after the cure of the syphilitic infection. There are still others who maintain that, in the absence of symptoms, a positive Wassermann reaction does not in- dicate syphilis. To those who still hold to this opinion, the INTERPRETATION OF RESULTS 187 writer has only to say that all experience has shown that, if the few conditions in which a positive result sometimes occurs can be eliminated, a positive reaction means the presence somewhere in the body of living treponemas and that the question of history or symptoms should have no influence in deciding the advisability of treatment. It is just this class of cases that later furnish the great bulk of our patients suffering from paresis, tabes, aortitis, aneu- rism, and syphilitic disease of the viscera, the latter often only discovered at autopsy. In cases in which there is good reason to question the possibility of syphilis the pa- tient's blood serum should be titrated, for, as Vedder91 well says: "One-tenth of a cubic centimeter of some sera is re- quired to give a double-plus reaction, while one-thousandth of a cubic centimeter of another serum may be sufficient to give a double-plus reaction. There would be little hes- itation in making a diagnosis of syphilis in the latter case whether clinical symptoms were present or not. There might be doubt in the former case." As the interpretation of the results of the test varies in the different stages of the disease, each stage will be con- sidered separately. The Interpretation of the Results of the Wassermann Test in the Primary Stage of Syphilis.-It is in this stage of the disease that the weaker grades of reaction are of most value from a diagnostic standpoint. As would be ex- pected, the earlier in the primary stage the blood is tested, the larger will be the percentage of negative results and also the larger the percentage of doubtful reactions. As a certain amount of complement-binding substance must be present to give a positive result, the longer the infection has lasted the greater the chance that this amount is pres- ent and that the test will be positive. This fact has al- ready been shown in the table giving the date of the ap- pearance of the positive Wassermann reaction in 600 cases of syphilis, but, with a clear history of infection and a 188 THE WASSERMANN TEST suspicious lesion, it is proper to consider as positive, in this stage of the disease, reactions that would be always doubtful at other stages. A plus-minus reaction (plus of most laboratories) should be considered as a negative reaction, in most in- stances, but if such a reaction is obtained within a week or two after the appearance of the initial lesion it may be considered as of some confirmatory value, but never as diagnostic of syphilis. A plus reaction (+), the double-plus (++), and three-plus (+ + +) reaction of other laboratories, when obtained in the primary stage of syphilis, may be interpreted in practice as a positive reaction, provided there is a history of ex- posure and a suspicious lesion is present. The writer often obtains this type of reaction in cases in which the initial lesion has lasted for from one to three or four weeks, and he has never seen a plus reaction in such cases become negative, except as the result of treatment, and if un- treated, the patients invariably developed secondary symp- toms. In the absence of a history of infection or the initial lesion a plus reaction should not be considered as diag- nostic of syphilis. A double-plus (+ +) reaction, the four-plus (+ + + +) re- action of some laboratories, is always diagnostic of syph- ilis in this, as in every other stage of the disease, provided the other conditions that sometimes give such a reaction, can be eliminated. This type of reaction, which means ab- solute inhibition of hemolysis, in the writer's experience has occurred in 14 per cent of the cases of primary syph- ilis tested during the first week after the appearance of the initial lesion; in 22 per cent tested during the second week; in 41 per cent tested during the third week; in 53 per cent tested during the fourth week; and in 61 per cent of cases tested during the fifth week after the appearance of the initial lesion. Reactions so strong as to almost equal complete inhibition of hemolysis occur in a large number of INTERPRETATION OF RESULTS 189 cases of primary syphilis and these should be reported as double-plus or four-plus reactions. In these, the super- natant salt solution, after the undissolved corpuscles have settled to the bottom of the test tube, shows a slight red- dish yellow tinge, demonstrating that a small amount of hemoglobin has been dissolved from the erythrocytes, but practically such reactions are positive and should be re- ported as such. A negative reaction, in the primary stage of syphilis, is of no value whatever in eliminating syphilis. Even though it persists until the initial lesion disappears and during the latent stage before the occurrence of secondary symp- toms, it can not be relied upon as evidence that syphilis does not exist. As the writer has already shown, at least 10 per cent of patients suffering from primary syphilis will give a negative Wassermann reaction, so that, in the presence of a suspicious lesion, the negative reaction is of no value, and the patient should never be assured, upon the strength of such a reaction, that his symptoms are not due to syphilis. As already emphasized, every suspicious lesion follow- ing exposure, whether situated upon the penis or else- where, should, if possible, be examined with the dark-field apparatus for Treponema pallidum. This should never be neglected and a Wassermann test made in preference, for the dark-field examination is the diagnostic method of choice during this stage of the disease, but where such an examination can not be made, or w'here it gives negative results, a Wassermann test should always be performed. The writer has observed many instances in which it was impossible to demonstrate Treponema pallidum, either by the dark-field apparatus or in stained preparations made from the suspicious lesion, but in which the Wassermann test was positive and thus the diagnosis was established. In extraurethral chancres the test has been of the greatest value in enabling us to make an early diagnosis and in the latent period, after the healing of the initial lesion and be- 190 THE WASSERMANN TEST fore the appearance of secondary symptoms, the Wasser- mann test is the only method available for making a diag- nosis. It is in this latter class of cases, where there is only a history of a healed lesion to assist one, that the test is of the very greatest clinical value in diagnosis and every such case should be tested repeatedly, if possible, as treat- ment applied before the appearance of secondary symp- toms is much more efficient than when it is administered afterward. The Interpretation of the Results of the Wassermann Test in the Secondary Stage of Syphilis.-In the secondary stage of syphilis a positive result is obtained in from 95 to 96 per cent of all cases, so that in this stage of syphilis the Wassermann test may be stated to be almost always confirmatory of the clinical findings. However, from 4 to 5 per cent of secondary cases, in the writer's experience, will give a negative reaction, even when repeatedly tested, and when marked secondary lesions are present, so that the mere absence of a positive reaction is no proof that the patient is not suffering from the disease. A double-plus (four-plus) reaction in patients present- ing the lesions of secondary syphilis or lesions that are suspicious of this stage of the disease, is conclusive evi- dence of the presence of syphilis under the limitations al- ready mentioned in discussing this same type of reaction in the primary stage; i. e., if the conditions which some- times give a positive reaction can be eliminated in the particular case examined. A plus (two- or three-plus) reaction is not of as much value in the secondary stage of syphilis as in the primary, but even in this stage this type of reaction occurs in about 10 per cent of cases, in the writer's experience, and when accompanied by suspicious symptoms, should be inter- preted as a positive reaction. In the majority of such in- stances the reaction is equal to the three-plus reaction reported by most laboratories, inhibition of hemolysis be- ing almost complete, but even if the reaction is less than INTERPRETATION OF RESULTS 191 this it should be regarded as confirmatory of the presence of the disease in patients presenting clinical symptoms or a clear history of a primary lesion. A plus-minus {plus or plus-minus) reaction in the sec- ondary stage of syphilis is valueless from a diagnostic standpoint. In these cases the diagnosis will have to rest upon the clinical symptoms present or upon the history of the case. On the other hand, the presence of this type of reaction should not lead to any hope that the disease is not present, as some secondary infections never give more than a plus-minus reaction. A negative reaction when suspicious symptoms of sec- ondary syphilis are present is of greater value than at any other stage of the disease, but it should be remembered that practically five per cent of cases of secondary syph- ilis give a negative reaction and that many of these cases remain negative for weeks and months. It is only when the negative reaction persists for several months and no symptoms of syphilis are present during this time that it may, with justice, be regarded as conclusive evidence of the absence of syphilis. On the other hand, if the case examined gives no history of a primary lesion and the symptoms present are only slightly suspicious, a negative reaction which persists for two or three months may be regarded as proving the absence of syphilis, provided treatment has not been administered in the meanwhile. The Interpretation of the Results of the Wassermann Test in the Tertiary Stage of Syphilis.-In the tertiary stage of syphilis, when definite tertiary lesions are present, the Wassermann test gives as high a percentage of posi- tive results as in the secondary stage, but there are many cases tested in this stage in which the symptoms are atyp- ical or only slightly suspicious, or in which the nervous system alone is involved, which give weak or negative re- actions, so that the total percentage of positive results in this stage of the disease is slightly below that encountered in the secondary stage. In many of these cases, also, spe- 192 THE WASSERMANN TEST cific treatment has been administered and although not successful in preventing the development of lesions, it may have reduced markedly the strength of the Wassermann reaction. A double-plus (four-plus') reaction with the test in the tertiary stage of syphilis is diagnostic of the disease under the limitations already noted in discussing the interpreta- tion of this type of reaction in the primary and secondary stages. A plus (two-plus or three-plus) occurs more frequently in this stage of syphilis than in the secondary stage and is, therefore, of greater diagnostic value. In patients pre- senting suspicious lesions of this stage, especially in those in which there is apparent involvement of the central nerv- ous system, the writer believes that this type of reaction should be given the same diagnostic value as the double- plus (four-plus) reaction, with the same limitations. A plus-minus (two-plus and plus-minus) reaction, unless permanent, or treatment has been administered, is of no value in the diagnosis of syphilis during this stage of the disease. In those instances in which there is no history of infection and the symptoms present are merely slightly suspicious, this type of reaction should be considered as negative, provided it persists over two or three months, but in the presence of a clear history of infection, especially if treatment has been administered, a plus-minus reaction should not lead one to consider the case as nonsyphilitic. A negative reaction in patients suspected of tertiary syphilis is of no value in excluding the disease unless it persists for several months and symptoms have not de- veloped meanwhile. In the absence of a history of infec- tion and of symptoms which are typical, a negative re- action may be interpreted as excluding syphilis, provided no specific treatment has been administered and it remains negative for a period of three months. In all suspected tertiary cases, when the Wassermann test is negative, a test should be made upon the cerebrospinal fluid. INTERPRETATION OF RESULTS 193 The Interpretation of the Results of the Wassermann Test in the Latent Stages of Syphilis.-The results of the Wassermann test differ in the early and late latent cases but the difference is not sufficient to justify our classify- ing the cases into early and late cases, so far as the inter- pretation of the results is concerned. In the early latent cases, that is, the cases tested during the latent period be- tween the healing of the initial lesion and the occurrence of secondary symptoms, a higher percentage of positive results is obtained than in latent periods occurring later in the disease, and the lowest percentage of positive re- sults is obtained in latent cases in which the period of latency occurs after the development of secondary or ter- tiary lesions. The percentage of positive results obtained by the writer in latent infection, including both early and late cases, has been 68 per cent, so that in this class of infections one must expect about 30 per cent of negative results. One of the most important reasons for the lower percentage of positive results in latent cases is that the vast majority of these patients have received more or less specific treatment and in most of the cases the test was made in order to determine the result of this treatment. Therefore, it follows that the plus reaction will have more diagnostic value in these treated cases than in other stages of syphilis when treatment has not been given, and that a negative reaction will have less value in excluding the disease. A double-plus (four-plus) reaction occurs less frequently in latent cases than in other stages of syphilis, but when it is obtained it is absolutely diagnostic of the disease under the limitations already mentioned. A plus (two-plus or three-plus) reaction, in latent cases, where there is a clear history of infection, should be given the same interpretation as the double-plus (four-plus) re- action, especially if specific treatment has been admin- istered. It is the writer's belief that this type of reaction 194 THE WASSERMANN TEST in latent cases, under the conditions named, is diagnostic of syphilitic infection. A plus-minus (plus or plus-minus) reaction in latent in- fections, when there is a clear history of infection, symp- toms have been present in the past, or specific treatment has been given, should arouse the suspicion that the disease is still present, but the test should be repeated two or three times, with a similar result, before a diagnosis of syphilis should be entertained. In such cases a plus-minus reaction usually means insufficient treatment, and indicates that further treatment is required. In the absence of a his- tory of infection, or of previous symptoms, a diagnosis of syphilis should never be made upon a plus-minus reaction. A negative reaction in cases suspected of being in the latent stage of syphilis has no value as excluding the dis- ease unless it is consistently negative over several months, in the absence of specific treatment and of symptoms of the disease. A negative result is obtained in at least 30 per cent of latent syphilitics so that it has little value in excluding the disease. Interpretation of the Results of the Wassermann Test in Syphilitic Disease of the Central Nervous System.-The in- terpretation of the results of the Wassermann test in syph- ilitic disease of the central nervous system is considered in Chapter X of this work. Interpretation of the Results of the Wassermann Test after Specific Treatment.-The interpretation of the results obtained in patients who have received specific treatment is considered in Chapter IX of this work. General Summary.-From the writer's experience he be- lieves that the following general conclusions regarding the interpretation of the results of the Wassermann test are justified: 1. If the diseases, other than syphilis, that sometimes give a positive result with the Wassermann test, can be excluded, a double-plus (four-plus of many writers) re- action is absolutely diagnostic of syphilis. It is believed INTERPRETATION OF RESULTS 195 that, under such conditions, this type of reaction is specific of the disease, whether symptoms are present or not, or whether there is or is not a history of infection. 2. Under the same conditions, a plus reaction (three- plus or two-plus of many writers) may, in primary, ter- tiary, and latent infections be regarded as diagnostic, pro- vided there is a clear history of infection, or suspicious clinical symptoms are present. In secondary cases this type of reaction is not diagnostic unless typical secondary symptoms are present. In the absence of either history or symptoms this type of reaction should not be regarded as diagnostic of syphilis. 3. A diagnosis of syphilis should never be made upon a plus-minus reaction alone. Many perfectly normal individ- uals give this type of reaction at times in their blood serum. In latent infections a plus-minus reaction, if persistent, should be followed by further specific treatment. 4. A single negative reaction, where there is no history of infection and where symptoms are not present, is of con- siderable value as a corroborative sign that syphilis is not present, but where there is any suspicion that the disease may be present it has very little value in excluding syphilis. The experiments already mentioned upon the variation in the complement-binding power of the blood serum of known syphilitics illustrate the truth of this statement, and it is only when a negative reaction is repeatedly obtained over a long period of time that it can be considered as good evidence of the absence of the disease, and even then the spinal fluid should be tested and, if possible, a provocative Wassermann test should be made upon the blood. The same remarks apply to cases that have been given specific treatment, a negative reaction being of no value as indi- cating a cure of the infection, unless the reaction remains persistently negative over a period of at least one year. In such cases a provocative "Wassermann test should be made, and the spinal fluid should be tested. A luetin test 196 THE WASSERMANN TEST should also be made in certain cases before the patient is discharged as cured. In cases where the so-called "delayed negative reaction" occurs, a careful investigation should be made as to a pos- sible history of infection or the presence of suspicious symptoms. This delayed negative reaction, in which com- plete hemolysis does not occur until near the end of the incubation period for hemolysis employed in the test, is ob- served in many treated cases and in the early primary stage of the disease, and should undoubtedly receive much greater consideration than has heretofore been accorded it. In a recent contribution, McConnell161 has called attention to the diagnostic importance of the delayed negative reaction, having found it to occur in 18 instances of 1793 individual tests, hemolysis not occurring until after 30 minutes had elapsed after adding the amboceptor and cell suspension. He found, on investigation, that the cases in which it oc- curred could be placed in three classes: those giving a frank history of infection with little treatment; those deny- ing infection but having loose sexual relations; and those with no history or clinical symptoms. Of the 18 cases in- vestigated by him, eight belonged to the first class; six to the second class, all admitting having had gonorrhea; while only four could be placed in the third class. He urges the most careful investigation of all cases giving a delayed negative reaction. The writer believes that it would be well to make a read- ing of the tests at the end of 30 minutes and another at the end of 45 minutes after adding the amboceptor and cell sus- pension, and those sera showing little hemolysis at the end of that time but complete hemolysis at the end of the hour should be noted and carefully checked with the history and clinical condition of the patient, and the data considered in the interpretation of the result of the test. The interpretation of the results of the Wassermann test must rest very largely with the clinician, for the clin- INDEX OF THE PREVALENCE OF SYPHILIS 197 ical picture present is often more decisive than is the re- sult of the test, and it is the clinician's place to reconcile the result of the test with the clinical picture, rather than the serologist's. The latter should simply report what actually occurs when the patient's blood serum is tested, without any reference whatever to the clinical history or the symptoms which may be present, and it remains with the clinician to interpret the report in the light of his patient's condition. The writer believes that much harm has been done by the serologist allowing the history of a case, or the symptoms present, to influence his reading of the Wassermann test, and it is always better to have the test read without the serologist knowing anything what- ever about the history or symptoms of the patients from whom the specimens were collected. The Wassermann Test as an Index of the Prevalence of Syphilis Since the discovery of the Wassermann test it has been largely used in determining the prevalence of syphilis in communities, hospitals, and among various classes of peo- ple, and has been of the greatest service in demonstrating how widespread this disease really is, and the compara- tively large number of people who suffer from it in nearly all countries and among nearly all classes of society. In fact, it may be stated, that before the days of the Wasser- mann reaction, although it was generally recognized that syphilis was a very prevalent disease in many countries, we had a very poor conception of its real prevalence, espe- cially in the United States, so that from a sociological standpoint, the test has been of such value as to almost revolutionize our ideas regarding the subject. It is now well known, and has been demonstrated time and again by Wassermann surveys of the class mentioned, that between 80 and 90 per cent of prostitutes who have pursued their calling for as long as five years or more, 198 THE WASSERMANN TEST are syphilitic, although it should be remembered, in con- sidering this high percentage, that only a comparatively small percentage of the women were actually capable of transmitting the infection at the time the tests were made, so far as clinical evidence was obtainable. However, in a disease like syphilis that has its periods of infectivity and quiescence, this high percentage of syphilitics among pros- titutes furnishes evidence enough of the danger of illicit sexual intercourse, and one often wonders why the disease is not much more common than it appears to be in the usual professional experience. As a matter of fact, syph- ilis is much more common, as shown by the results of the Wassermann test, than usually supposed and, in this coun- try as well as in others, is pretty well distributed through all classes. As a result of the application of the Wassermann test in Wassermann surveys, as they are called, a great mass of evidence has been accumulated regarding the preva- lence of syphilis in different communities and among dif- ferent classes of the population. Thus, Southard92 per- formed 6,000 Wassermann tests in the Harvard Neuro- pathologic Laboratory and found 23 per cent positive; Hammond,93 who applied the test to both patients and others at the New Jersey State Hospital, found 6.3 per cent positive; Thompson94 made 1,000 tests at the Arkansas State Hospital for Nervous Diseases and found 33 per cent positive; Rosenberger95 made 5,106 Wassermann tests upon the patients at the Philadelphia General Hospital, during 1916, and found 27.4 per cent positive; and John- son,96 who examined by means of the Wassermann test 224 children showing anemia and malnutrition, found that over 33 per cent gave a positive Wassermann reaction. There is no better place to study the prevalence of syph- ilis than among the personnel of the army, for recruits are drawn from all classes of society for the service and from all parts of the United States. The most important INDEX OF THE PREVALENCE OF SYPHILIS 199 and conclusive contribution to our knowledge of the preva- lence of syphilis in the army, and, consequently, in civil life, has been made by Lieut. Colonel Vedder, who pub- lished his results in 1915,97 and from whose work I shall quote largely in considering this subject. In his work, Vedder considered that all cases giving a double-plus (four-plus) reaction were without doubt syph- ilitic, while those giving a plus (three-plus reaction) were probably syphilitic. Summarized, his results are presented in Table XXXIV: Table XXXIV Occurrence of Syphilis in Army, from Vedder CLASSES TOTAL TESTED KNOWN CASES % WAS- SER- MANN % UNDOUBT- ED SYPH- ILITICS % WAS- SER- MANN % ESTIMATED TOTAL NUMBER OF PROBABLE SYPH- ILITICS % Recruits 1,019 0 7.75 7.75 9.02 16.77 Cadets 621 0 2.57 2.57 2.89 5.46 White enlisted 1,577 3.44 4.77 8.21 7.87 16.08 Colored enlisted 1,472 1.08 21.80 22.21 13.11 36.00 Porto Rico Regt. 531 13.55 28.58 42.37 13.55 55.93 Military convicts 1,145 6.48 9.50 15.98 5.67 21.65 Insane soldiers 567 3.51 8.29 11.80 7.41 19.21 Tuberculous soldiers 229 7.56 15.72 23.28 15.72 39.00 Soldiers home 1,171 11.62 13.40 25.02 9.73 34.75 There may be some question as to whether the cases giving a plus reaction examined by Vedder should be called probably positive, but, even if these cases are ex- cluded, the figures given in the table are very striking and show that syphilis is a far more prevalent disease than is generally supposed by the average practitioner. Among the recruits examined by Vedder, numbering 1,019, a double-plus (four-plus) reaction was obtained in 7.75 per cent, or practically 8 per cent. These men had passed a most thorough physical examination for admis- sion to the army and presumably presented no evidences of syphilis when accepted. If to this number be added 200 THE WASSERMANN TEST the 9.02 per cent of cases giving a plus (three-plus) re- action, the total percentage of syphilitics would amount to 16.77 per cent. These recruits were distributed through the areas of the United States covered by the recruiting stations at Fort Slocum, New York, and Columbus Bar- racks, Ohio, and represented numerous occupations before enlistment. Vedder concludes that his observations, to- gether with the known figures for rejection for syphilis of applicants for admission, show that practically "among young men in civil life between the ages of 20 and 30, and of the general class belonging to the occupations mentioned, the percentage of syphilis may be estimated at at least 16.77 per cent, and there is good reason for believing that it is fully 20 per cent." The writer's own observations upon the percentage of positive Wassermann reactions obtained in recruits for the army, numbering several thousands of cases, is prac- tically the same as Vedder's, as regards the occurrence of double-plus reactions, the percentage obtained by the writer being 8.64 per cent instead of 7.75. It is not be- lieved that the cases giving a plus reaction should be con- sidered as positive unless there is a history of infection or of suspicious symptoms in the past; but in the writer's experience such a history may be obtained upon careful questioning in enough of the cases to raise the percentage of syphilis among recruits of the army to 10 per cent when they are admitted to the service. If to these are added the percentage of rejections for syphilis, based upon actual symptoms of the disease, the writer believes that at least 15 per cent of all applicants for enlistment for the army, prior to the present war, were infected with syphilis, and this means that practically 15 per cent of the average male civilian population of this country between the ages of 20 and 30 are infected. This appears to be an enormous percentage but the figures are borne out in Wassermann surveys upon a similar class of men, as regards age, in INDEX OF THE PREVALENCE OF SYPHILIS 201 other countries, and in surveys made in our large hospitals upon all classes of cases admitted to them. Vedder's results upon the class of young men who have entered West Point are important, for these young men come from the average middle class family of this coun- try, have been well educated, and must be above the aver- age in intelligence in order to pass the rigid examinations for entrance to the military academy. In this class of young men Vedder found 2.57 per cent of double-plus (four-plus) reactions among the 621 cadets tested, and 2.89 per cent of plus reactions. If these be added together, there is a total percentage of over 5 per cent which he con- siders as probably syphilitic. The writer believes that probably three per cent would be about the amount of syphilis in this class of cases, and, applied to a similar class of young men in civilian life, (the class attending our colleges) it would mean that about this percentage of col- lege men are syphilitic before entering college. As regards the amount of syphilis in white and colored enlisted men in the army, Vedder found that among the white enlisted men 8.21 per cent gave a double-plus (four- plue) reaction while among the colored enlisted men no less than 22 per cent gave this type of reaction. His fig- ures as regards the greater percentage of syphilitic infec- tion among the colored men of the army agree with those of many investigators who have made Wassermann surveys upon the colored race in this country. Levin,162 for instance, examined 7387 white drafted sol- diers and found that 10.5 per cent gave a double-plus (four- plus) reaction, while of 3494 colored drafted soldiers, 18.3 per cent gave a double-plus reaction. As these tests were made upon drafted men the results give a fair survey of the prevalence of syphilis in civil life in both the white and colored races. If strong plus reactions are considered as probably positive, Levin states that the percentage of whites giving a positive result would have been 13.8 per 202 THE WASSERMANN TEST cent and for the colored 24.1 per cent. The combined sta- tistics of eleven noted dermatologists who have had a large experience with syphilis in the negro show that from 45 to 50 per cent are syphilitic. Vedder's results as to the prevalence of syphilis among the patients at the army hospital for tuberculosis, at Fort Bayard, N. M., are of interest, as they demonstrate that a much larger proportion of tuberculous patients give a positive reaction than patients suffering from other dis- eases. He found that no less than 23.28 per cent of the 229 patients examined gave a positive reaction and he states that his results are almost identical with those of Letulle, Bergeron, and Lepine in Paris, who found that one out of every five tuberculous patients was also syph- ilitic. These results explain fully the large number of positive reactions which some writers have claimed are due to tuberculosis, but the most careful investigations have shown that tuberculosis per se does not give a positive reaction with the Wassermann test, so that those patients who do react positively are undoubtedly suffering from a coincident syphilitic infection. The writer, while working upon his method of comple- ment fixation in the diagnosis of tuberculosis, had the opportunity of testing 166 patients in the General Hospital at Fort Bayard and found that nearly 10 per cent of them gave a positive reaction with the Wassermann test, while about 6 per cent were known to be syphilitic, giving a total percentage of 16 per cent of syphilitics in the number ex- amined. While these figures are not as large as those recorded by Vedder they amply demonstrate the fact that syphilis and tuberculosis are more often associated than are other diseases, and that one may expect a large per- centage of syphilitic infection among the tuberculous. Irvine,98 as the result of various Wassermann surveys, concludes that from 10 to 15 per cent of the population of the State of Minnesota are syphilitic, or from 200,000 to INDEX OF THE PREVALENCE OF SYPHILIS 203 300,000 people. He states that Wassermann surveys made in the asylum at Fergus Falls showed that 30 per cent of the inmates were syphilitic and that the average of many similar surveys is 20 per cent. Peterson" found that among 477 obstetric patients, no less than 5.8 per cent were syphilitic, and of 606 gynecologic patients the test was positive in 8.2 per cent. Regarding the value of the Wassermann test in obstetric and gynecologic prac- tice he says: "In many of the cases, both where there was a history of infection and when this was lacking, physical examination in the gynecologic clinic failed to arose sus- picion of the presence of syphilis. The point to be empha- sized is that the same careful physical examinations have been made in this particular clinic for years with only an occasional patient suspected of being luetic. When, however, such examinations are reinforced by routine Was- sermann examinations, nearly 10 per cent of the patients are found syphilitic." Ladd100 found that 25.5 per cent of patients applying for treatment at the dispensaries of the Washington, D. C., Casualty and Asylum Hospitals gave a double-plus (four-plus) Wassermann reaction, and that of these 14.6 per cent were colored, and 10.9 per cent white patients. An entire volume could be written covering the statis- tics of the almost innumerable Wassermann surveys that have been made in different localities but the illustrations here given demonstrate the value of the test in the study of the prevalence of syphilis. In fact, no longer is any data of scientific value regarding this subject unless it is based upon the results of carefully performed Wassermann re- actions, for the day has passed when clinical evidence alone is to be considered as sufficient upon which to base figures regarding the prevalence of syphilis in a community or country. The cases of syphilitic infection that are appar- ently without clinical symptoms far outnumber those that present typical symptoms and the only way that we at pres- ent possess of making a diagnosis in these latent infections 204 THE WASSERMANN TEST is by means of the Wassermann test. Aside from the value of this test in the diagnosis of the disease and in the scien- tific control of its treatment, perhaps the greatest value of the test is in the demonstration of the extent to which this insidious disease has permeated society, an extent al- together unsuspected before the use of the reaction in Was- sermann surveys. The accuracy of the results of the Wassermann test as showing the prevalence of syphilis, and that the figures so obtained are not an exaggeration of the actual conditions present, are proved by the anatomico-pathologic studies of Warthin,163 who found evidences of syphilitic infection in 300 of 750 autopsies, or 40 per cent, and in 75 of these cases there were active lesions, either macro- or microscop- ical, present, although there were no clinical symptoms pointing to syphilis observed before death. This very care- ful investigator places the incidence of syphilitic infection in this country at close to 30 per cent and considers that one-tenth of all deaths occurring in the United States are due to this disease. CHAPTERIX THE EFFECT OF TREATMENT UPON THE WAS- SERMANN REACTION. THE WASSERMANN TEST AS A CONTROL OF THE TREAT- MENT OF SYPHILIS. THE PRO- VOCATIVE WASSERMANN REACTION The treatment of syphilis, either with mercury or with salvarsan or neosalvarsan often has a marked effect upon the results of the Wassermann test, rendering strongly positive reactions negative or greatly lessening the strength of the reaction in those cases in which a positive result is not rendered negative. This fact has rendered the use of the test invaluable as a guide to the efficiency of the treat- ment of the disease with various drugs and has enabled the profession to discard many drugs vaunted as specifics for the disease but which have been proved by the Was- sermann test to be absolutely worthless. One of the most striking examples of this fact is the instance of sodium cacodylate, which at one time was announced as almost the equal of salvarsan in the treatment of syphilis, but which was soon proved by the Wassermann test to be practically worthless, so far as curing the infection, and which is not used at the present time in the treatment of the disease except by those ignorant of the work which has been done proving its worthlessness or those willfully blind to the re- sults of that work. The Wassermann reaction is generally conceded to be the most reliable and delicate of all of the symptoms of syphilis and for this reason the effect of treatment upon the reaction is of the utmost practical importance. A posi- 205 206 THE WASSERMANN TEST tive reaction, as has been noted, means the presence in the body of living treponemas, and as long as the reaction remains positive, just so long is the patient a victim of syphilitic infection, regardless of the occurrence of other symptoms of the disease. The effect of treatment with various drugs can be easily followed in the case of syph- ilis by frequently repeated Wassermann tests, and in this way one can secure a reliable picture of the results of any particular form of treatment. In view of the importance of this subject the following chapter contains the personal observations of the writer upon the effect upon the Was- sermann reaction of treatment with salvarsan and mer- cury, which have been previously published,101 together with remarks upon the test as a control of the treatment of syphilis, and upon the provocative Wassermann reaction. The Effect of Treatment upon the Wassermann Reaction Prior to 1911 there was little in the literature upon the effect of treatment on the Wassermann reaction but since that time a great deal of data has been published show- ing that the reaction is markedly influenced by treatment with certain drugs, and that positive reactions often be- come negative and remain so for considerable periods of time after the cessation of treatment. Unfortunately, in- vestigation has also shown that the early treatment with mercury and salvarsan was insufficient and that most of the cases so treated relapsed within a year after treat- ment was stopped, so that at the present time no intelligent practitioner believes that a few injections of mercury or of salvarsan or neosalvarsan will cure syphilis. The Effect upon the Reaction of Treatment with Salvar- san.-The data given in regard to the effect upon the Was- sermann test of treatment with salvarsan was collected by the writer between 1910 and 1913 and is based upon a com- paratively few injections of the drug, as Ehrlich's idea THE EFFECT OF TREATMENT 207 that one or two injections of this drug would sterilize the body, so far as Treponema pallidum was concerned, was still the belief and hope of the profession. However, the facts deduced from these experiments and observations have since been fully confirmed and the data demonstrate the effect upon a positive Wassermann reaction of treat- ment with this drug. Among those who had contributed valuable data regard- ing the effect of salvarsan upon the Wassermann reaction prior to the writer's publications may be mentioned the following: Nichols and Fordyce102 treated 11 patients with salvarsan with the following results upon the Wassermann reaction: In 8 the reaction became negative, 1 in 19 days; 2 in 21 days; 1 in 23 days; 1 in 25 days; 1 in 39 days; 1 in 3 months; and 1 in 3 months and 11 days. Michaelis103 treated 110 patients and states that the reaction may become negative within a period from 2 to 10 weeks after the injection. Fordyce104 states that the results of the Wassermann test after salvarsan have varied within wide limits, in his expe- rience, but in general a change occurred in from 4 to 5 weeks. The earliest change from a positive to a negative reaction observed by him was in 6 days after the injec- tion. He states that "several cases showed alternating negative and positive phases." Noguchi105 analyzed very carefully 102 cases treated with salvarsan, more than half of them being under observation for over three months while the remainder had been injected four weeks pre- viously. He made quantitative examinations of their blood serum for complement-binding power, the tests being made before the injection of salvarsan, and at intervals of one day, three days, one week, two weeks, three weeks, four weeks, six weeks, eight weeks, etc., after the injection. He found that 30 cases became negative, 24 reduced to less than 1 antibody unit, while the remaining 48 cases still contain more than 1 antibody unit, giving strong positive reactions. In 40 per cent of the primary cases, in 37 per 208 THE WASSERMANN TEST cent of the secondary cases, in 35 per cent of tertiary, in 33 per cent of latent, in 14 per cent of hereditary, and in 50 per cent of incipient tabes the reaction became nega- tive. The average of the negative reactions equalled 33.7 per cent of the total 102 cases. Of the negative cases re- ported by Noguchi, 34 in number, 3 became negative in two weeks, 10 in three weeks, 11 in four weeks, 5 in five weeks, 4 in six weeks, and 1 in seven weeks. Animal Experiments.-The effect upon the Wassermann reaction of treatment with salvarsan can be shown in rab- bits experimentally infected with syphilis. Through the kindness of Major H. J. Nichols, the writer had the oppor- tunity of repeatedly testing the blood serum of rabbits ex- perimentally infected with syphilis and yaws, and Tables XXXV and XXXVI illustrate the gradual increase in the strength of the Wassermann test in these animals after in- fection, and the disappearance of the positive reaction after Table XXXV Results of Wassermann Test before and after Treatment with Salvarsan in Rabbit Infected with Syphilis* AMOUNT OF BLOOD SERUM, C.C. DATE OF TESTS Jan. 23 Jan.27 Jan. 30 Feb. 6 Feb. 8 Feb. 10 Feb.13 Feb.18 0.05 - + 4- + + - + 0 * 0.10 - + + + + + + + - + + - - 0.15 0 + + + + + + + - + + 0 - 0.20 + + 0 + + + + + - + + - - •Treponema pallidum found in testicular lesion Jan. 19. 0.02 gram salvarsan per kilo administered intravenously Feb. 6. Table XXXVI Results of Wassermann Test before and after Treatment with Salvarsan in Rabbit Infected with Yaws* AMOUNT OF BLOOD SERUM, C.C. DATE OF TESTS Jan.23 Jan. 30 Feb. 2 Feb. 6 Feb. 8 Feb. 10 Feb. 13 Feb. 17 0.05 - - + + + + - - - - 0.10 - + - + + + + - + - -■ - 0.15 - + + + + + + - + - - 0.20 - + + + + + + + - + + - - *Treponema pertenue found in lesion Jan. 21. 0.02 gram salvarsan per kilo adminis- tered intravenously Feb. 6. THE EFFECT OF TREATMENT 209 treatment with salvarsan. The sign (++) indicates com- plete inhibition of hemolysis. In the rabbit infected with syphilis the treponema was found in the lesion on January 19, but the Wassermann test did not become positive, even with 0.2 c.c. of the serum, until February 6. One week after the injection of salvar- san the reaction had become negative and the table shows that it had begun to diminish in strength two days after the injection. The results of treatment with salvarsan upon the rab- bit infected with yaws were quite similar, the reaction be- coming negative one week after the intravenous injection of the drug. The following summary of the writer's observation upon the effect of treatment with salvarsan on the Wassermann reaction embraces 500 cases which were observed and tested for at least four months after treatment. Of these 500 cases only 288 became negative, or 57.6 per cent, but these results were due to the prevalence, at that time, of giv- ing but one or two intravenous injections of salvarsan in- stead of several, the intramuscular method of using the drug giving altogether the best results as regards the disappear- ance of the positive reaction. The real effect of treatment with salvarsan upon the Wassermann reaction can only be accurately understood if several factors are taken into consideration, the most important being the stage of syphilis during which the drug is administered; the intensity of the positive Was- sermann reaction before treatment; the method of admin- istration of the drug and the dose administered; as well as the amount and kind of previous mercurial treatment. The Stage of Syphilis in Relation to the Disappearance of the Reaction.-The results of the writer have shown that regardless of the method of administration or the dose, the stage of the disease has much to do with the effect of treatment with salvarsan upon the Wassermann reaction. This is shown in Table XXXVII. 210 THE WASSERMANN TEST Table XXXVII The Results of Treatment with Salvarsan upon the Wassermann Test in the Various Stages of Syphilis STAGE OF DISEASE TOTAL CASES BECAME NEGATIVE REMAINED POSITIVE NUMBER % NUMBER % . Primary 95 63 66.3 32 33.6 Secondary 285 169 59.3 116 40.7 Tertiary 52 17 32.6 35 67.3 Latent 68 39 57.3 29 42.6 TOTALS 500 288 57.6 212 42.4 An analysis of this table is interesting. In 95 primary cases 63 became negative, or 66.3 per cent; in 285 second- ary cases, 169, or 59.3 per cent became negative; in 52 tertiary cases 17 or 32.6 per cent became negative; and in 68 latent cases, including both early and late latent infec- tions, 35, or 57.3 per cent became negative. From the above it is evident that the effect upon the Wassermann reaction of treatment with salvarsan is most marked during the primary stage of the disease, as would be expected and least marked during the tertiary stage, and that the longer the infection has lasted the less is the effect upon the. reaction of the treatment. Time of Disappearance of the Reaction in Relation to the Stage of Syphilis The rapidity with which the positive Wassermann re- action disappears after treatment with salvarsan depends largely upon the stage of the disease as is shown in Table XXXVIII. Table XXXVIII The Relation of the Stage of Syphilis to the Time of Disappearance of THE WASSERMANN REACTION, AFTER TREATMENT WITH SALVARSAN STAGE OF DISEASE TOTAL CASES 1 WEEK 2 WEEKS 3 WEEKS 4 WEEKS 5 WEEKS 6 WEEKS 7 WEEKS 8 WEEKS Primary 63 6 25 20 8 2 2 0 0 Secondary 169 13 44 39 33 17 13 5 5 Tertiary 17 3 6 2 5 0 1 0 0 Latent 39 2 9 7 10 6 5 0 0 TOTALS 288 24 84 68 56 25 21 5 5 THE EFFECT OF TREATMENT 211 Regarding the time of disappearance of the reaction after treatment with salvarsan it will be noted that the best results were obtained in the primary stage of the dis- ease, no less than 51 of the 63 primary cases that became negative becoming so within three weeks and the longest time after treatment in which the reaction became negative in this stage was six weeks. In the secondary stage the reaction disappeared generally during the second, third, or fourth weeks, the longest period being eight weeks. In the tertiary stage the results appear more favorable than in the primary, 1 of the 17 cases becoming negative within three weeks but the number of cases is too small to base such an assertion upon, and the writer believes that the evidence is sufficient to prove that the best results are obtained in the primary stage of syphilis. As will be noted, none of the reactions, in any stage of the disease, became negative after eight weeks, so that it would appear that a positive reaction which remains so after a period of eight weeks has elapsed since the last salvarsan treatment, should always be an indication for further treatment. If the drug is given intravenously, as it generally is today, the period may be shortened to less than this, but it is believed that eight weeks is a practical and safe period to wait in order to ascertain if the Was- sermann reaction will become negative. The Relation of the Intensity of the Reaction to Its Dis- appearance after Salvarsan.-The stronger the Wasser- mann reaction the less is it affected by treatment with salvarsan, as is shown in Table XXXIX. In this table and those that follow, it should be remembered that a double-plus (++) reaction means absolute inhibition of hemolysis, or the four-plus reaction of most writers; a plus (+) reaction at least 50 per cent of inhibition, or the three- and double-plus reaction of some laboratories; and a plus- minus, less than 50 per cent of inhibition. An analysis of this table shows that of 295 cases in which the Wassermann reaction was double-plus (four-plus), 212 THE WASSERMANN TEST Table xxxix The Relation of the Intensity of the Wassermann Reaction to its Disappearance after Treatment with Salvarsan CHARACTER OF REACTION TOTAL CASES BECAME NEGATIVE REMAINED POSITIVE NUMBER % NUMBER % + 4- 295 145 41.9 150 50.8 + 169 110 65.0 59 34.9 + - 36 33 91.6 3 8.3 500 288 57.6 212 42.4 41.9 per cent became negative; of 169 giving a plus (three or two plus) reaction, 65 per cent became negative, and of 36 cases showing a plus-minus reaction, 91 per cent be- came negative. As regards the latter cases it may be stated that they were either very early primary cases or cases that had previously received specific treatment to a greater or lesser extent. It will thus be seen that the poorest results as regards the effect of treatment upon the Wassermann test were obtained in cases showing complete inhibition of hemolysis and the results gradually grew better with the loss in the strength of the reaction. The Relation of the Stage of Syphilis and the Intensity of the Wassermann Reaction to its Disappearance after Treatment with Salvarsan.-The percentage of negative results obtained by treatment with salvarsan in each stage of syphilis is considerably influenced by the intensity of the reaction. Table XL illustrates the effect of the inten- sity of the reaction upon its disappearance in each stage of the disease. From this table it is evident that the effect of treatment with salvarsan upon the Wassermann test, in every stage of syphilis, varies with the intensity of the reaction, being most pronounced in the cases showing the weaker reactions and least in those giving complete inhibition of hemolysis. From these data it is evident that the earlier treatment is begun in syphilis the better will be the results, for the cases of tertiary syphilis, for instance, which gave a double- plus reaction, gave a very low percentage of negative re- THE EFFECT OF TREATMENT 213 Table XL The Relation of the Stage of Syphilis and the Intensity of the Wasser mann Reaction to Its Disappearance after Treatment with Salvarsan STAGE OF SYPHILIS CHARACTER OF REACTION TOTAL CASES BECAME NEGATIVE REMAINED POSITIVE NUMBER % NUMBER % Primary + + 44 24 54.5 20 45.4 (95 cases) + 41 31 75.6 10 24.3 4- - 10 8 80.0 2 20.0 Secondary + + 185 98 52.9 87 47.0 (285 cases) + 86 58 67.4 28 32.5 + - 14 13 92.8 1 7.1 Tertiary + + 33 5 15.1 28 84.8 (52 cases) + 16 9 56.2 7 43.7 + - 3 3 100.0 0 0.0 Latent + + 33 18 54.5 15 45.4 (68 cases) + 26 12 46.1 14 53.8 + - 9 9 100.0 0 0.0 suits after treatment; i. e., only 15 per cent and even in the cases giving a plus reaction the percentage of negative reactions was only a little over 50 per cent. It will be ob- served that in the latent stage a higher percentage of cases showing a double-plus reaction became negative than of those showing a plus reaction, but this was due to previous mercurial treatment in a larger proportion of the cases giving a double-plus reaction. From these data it is also evident that the intensity of the Wassermann reaction in the various stages of syphilis is of considerable prognostic importance. While in all of the stages of syphilis, with the exception of the latent stage, the results were better in cases giving a plus reaction than in those giving a double-plus, the difference is most marked in the tertiary stage of the disease and least in the pri- mary. As regards the double-plus cases, the best results were obtained in the primary and latent stages of syphilis and the poorest results in the tertiary stage. The good re- sults in the latent stage may be explained by the fact that many of these cases were early latent cases, for the late la- tent cases are much more resistant to treatment. Of the cases giving a plus reaction the best results were obtained in 214 THE WASSERMANN TEST the primary stage, the poorest in the latent stage, while the results in the secondary stage were better than those in the tertiary. The prognosis, as regards the disappear- ance of the reaction, is most unfavorable in the tertiary cases giving a double-plus (four-plus) reaction. The Relation of the Method of Administration of Salvar- san to the Disappearance of the Wassermann Reaction in Treated Cases.-The method of administration of salvar- san has a very marked influence upon the effect of the drug on the Wassermann reaction. Of the 500 cases studied, 209 were treated by the intramuscular injection of the alka- line solution, 249 by the intravenous injection of the drug, and 42 by combined intramuscular and intravenous injec- tions. Table XLI illustrates the effect of the method of administration upon the Wassermann reaction. Table XLI The Methods of Administration of Salvarsan in Relation to the Disappearance of the Wassermann Reaction in Treated Cases METHOD OF ADMINISTRATION TOTAL CASES BECAME NEGATIVE REMAINED POSITIVE NUMBER % NUMBER % Intramuscular 209 159 76.0 50 24 Intravenous 249 101 44.5 148 55.5 Combined 42 28 66.6 14 33.3 TOTALS 500 288 57.6 212 42.4 It will be observed, from a study of the table, that the effect of intramuscular injections of salvarsan upon the reaction is much more pronounced than other methods of administration. Unfortunately, this method of adminis- tration has been almost abandoned by the profession and in the army it has been entirely replaced by the intravenous method, owing to the pain and complications that followed the injections and the time lost in hospital. However, there is no question that the intramuscular method is in- finitely more efficient in treatment than the intravenous, as shown by the results upon the Wassermann test, for of the 209 cases so treated, most of them receiving but one injec- tion of 0.6 gram of salvarsan, 159, or 76 per cent became THE EFFECT OF TREATMENT 215 negative, and, as will be noted later, in discussing the test as a control of treatment, fewer cases treated by the intra- muscular method relapsed than by the intravenous. The figures given for the intravenous method are below what they should be for the reason that the majority of the cases studied only received one intravenous injection. Further observations upon this method of administration and its effect upon the Wassermann reaction have shown that with from three to five intravenous injections the re- sults are as good as those obtained with the intramuscular injections, but the data here given are sufficient to show that the Wassermann reaction can be rendered negative by only one intravenous injection in at least 40 per cent of the cases. The effect upon the reaction of the combined intra- muscular and intravenous method of using salvarsan was determined in 42 cases, of which 66 per cent became nega- tive, thus indicating that the results wrere not as good with the combined method as with the intramuscular method alone. However, this result was due to the fact that a num- ber of the intramuscular cases had two injections of salvar- san and some three, and it is considered that one intra- muscular injection of the drug is equal to at least three in- travenous injections so far as the effect upon the Wasser- mann test is concerned. As would be expected the effect of treatment with sal- varsan upon the reaction increases with the number of doses of the drug that are administered, regardless of the method of administration. Of 200 cases receiving one intramuscular injection of salvarsan of 0.6 gm., 152, or 76 per cent became negative, while of 9 cases receiving two intramuscular injections, 7, or 77.7 per cent became negative. There is little difference between these two groups of cases so far as the apparent effect upon the Wassermann test is concerned, but the number of cases receiving two intramuscular injections is too small to allow of our basing any accurate statistics upon, as a larger 216 THE WASSERMANN TEST number of cases would undoubtedly give a higher per- centage in this class of cases. Of the cases treated by the intravenous method, 177 were given one intravenous injection of 0.6 gm. of salvar- san, and 72, or 40.6 per cent became negative; 52 were given two intravenous injections of the same dose, of which 22, or 42.3 per cent became negative; while 10 cases were given three intravenous injections, of which 7, or 70 per cent became negative. There was but little difference in the percentage of negative results obtained in those given one and two intravenous injections, and it will be noted that two intravenous injections did not have as much effect upon the Wassermann reaction, as one intramuscular in- jection. The percentage of negative results obtained with three intravenous injections, however, approaches closely to that obtained with one intramuscular injection, and jus- tifies the assertion that one intramuscular injection is equal, in its effect upon the Wassermann reaction, to three intra- venous injections of salvarsan. The time of the disappearance of a positive reaction after treatment with salvarsan varies with the method of administration as shown in Table XLII. Table XLII The Relation of the Method of Administration of Salvarsan to the Time of Disappearance of the Wassermann Reaction in Cases Treated with Salvarsan METHOD OF ADMINISTRATION TOTAL CASES 1 WEEK 2 WEEKS 3 WEEKS 4 WEEKS 5 WEEKS 6 WEEKS 7 WEEKS 8 WEEKS Intramuscular 159 9 36 39 31 18 18 4 4 Intravenous 101 13 44 24 16 4 0 0 0 Combined 28 2 4 5 9 3 3 1 1 From the table it will be noted that the positive re- action disappeared more slowly after intramuscular injec- tions of the drug than after the intravenous injections. After intravenous injections there were only four cases that became negative during the fifth week after injec- tion, and none after that period, while no less than 26 cases THE EFFECT OF TREATMENT 217 treated by the intramuscular method became negative after the fifth week. The cases treated by the combined method were also slower in becoming negative than those treated by the intravenous method alone. If the Wassermann reaction is due to the presence in the tissues of living treponemas, as is now generally ac- cepted, one would expect that a positive reaction would disappear more rapidly after an intravenous injection of salvarsan than after an intramuscular injection. The drug, when administered intravenously, is quickly brought into contact with large numbers of the organisms and they are killed, while those not immediately affected and which are accessible to the drug perish later, but as the drug is elim- inated within a few days it is obvious that what action it has occurs comparatively quickly and if the reaction is depend- ent upon the treponemas and they are destroyed rapidly, the reaction also should disappear within a comparatively short period of time. On the other hand, when salvar- san is administered intramuscularly it is absorbed very slowly and exerts its destructive action upon the trepo- nemas for days and weeks, a smaller number are killed in a given period of time, and therefore, the reaction persists for a longer period than after intravenous administration. Whether the positive Wassermann reaction is due to sub- stances liberated from the treponema during life; pro- duced by the tissues in answer to the irritation caused by the treponema multiplying within them; or to substances liberated from the organism after its destruction or to the reaction of the tissues to such substances, is immaterial, so far as the disappearance of the reaction after the admin- istration of salvarsan is concerned, for, however produced, it is evident that the reaction theoretically should dis- appear more rapidly after intravenous injection than after intramuscular, and this is what actually occurs in practice. The Effect upon the Reaction of Treatment with Mer- cury.-The Wassermann test demonstrated the true value of mercury in the treatment of syphilis by the fact that 218 THE WASSERMANN TEST this drug is capable of causing the disappearance of a positive reaction, but it has also demonstrated the fact that it is of comparatively small value when compared with salvarsan. For one case that becomes negative after treat- ment with' mercury there are hundreds that become nega- tive after treatment with salvarsan, and although the majority of both classes of cases relapse in time and again present a positive Wassermann reaction, the relapses are much less frequent among patients treated with salvarsan than with mercury. The relative efficiency of salvarsan and mercury in caus- ing a disappearance of the Wassermann reaction is well illustrated in patients who had previously been treated with mercury before receiving salvarsan. Of 90 patients who were treated with mercury by the mouth for nine months or more before the administration of salvarsan and who gave a positive Wassermann reaction, 68, or 70.5 per cent became negative within eight weeks after the ad- ministration of the drug. The intensity of the reaction at the time of receiving salvarsan and the length of time the patients had been taking mercury is given in Table XLIIL Table XLIII The Wassermann Test in Patients Who Had Taken Mercury for Various Periods of Time METHOD OF TREATMENT TIME OF TREATMENT NUMBER OF CASES CHARACTER OF THE REACTION + + + + - Mercury by mouth 9 months 17 8 7 2 ll 11 11 1 year 26 16 8 2 11 11 11 2 years 17 7 9 1 ll ll ll 3 years 8 3 4 1 TOTALS 68 34 28 6 The table shows that 26 of these patients had taken mer- cury by mouth for a period of one year and that 16 of them still gave a double-plus (four-plus) reaction; that 17 had received the same treatment for two years, of whom 7 still gave a double-plus reaction; and that 8 had received THE EFFECT OF TREATMENT 219 treatment for three years, of whom 3 still gave a double- plus reaction. Of the 68 cases not one had become nega- tive as the result of mercury administered by the mouth, but after the administration of salvarsan every one became negative within 8 weeks. This is certainly very decisive proof of the greater specific action of salvarsan and could only be ascertained by the effect of the drugs in question upon the Wassermann reaction. Objection may be raised to the results recorded above by calling attention to the fact that the administration of mercury by the mouth is now well recognized to be the very poorest of all ways of giving this drug, although it is not so long ago that three years ' treatment by mouth with mecury was considered by the best authorities upon the disease as adequate for the cure of syphilis. In answer to this objection Table XLIV is given, covering 18 patients treated with hypodermic injections of mercury, a plan of treatment generally acknowledged to be the most efficient of the many ways of administering the drug. For pur- poses of comparison the effect upon the Wasserman of treatment with salvarsan is also included. Table XLIV The Comparative Results of Treatment with Mercury and Salvarsan UPON THE WASSERMANN REACTION NUMBER OF CASES Mercurial Treatment Salvarsan Treatment INJECTIONS OF GREY OIL CHARACTER OF REACTION DOSE 0.6 GM. CHARACTER OF REACTION 2 7 + + Intramuscular - 1 8 + II Cl - 1 9 + + Cl i c - 1 11 + + ll IC - 3 15 + + Intravenous - 1 15 + Intramuscular - 4 18 + + C C 11 - 2 19 + + Intravenous - 1 20 + + I1 cc - 1 25 + Intramuscular - 1 30 + C C Cl - From a consideration of this table it is evident that cases having had as many as 18 to 20 injections of grey oil 220 THE WASSERMAN N TEST still gave a double-plus Wassermann reaction and that cases having as high as 25 and 30 injections still gave a plus (three-plus) reaction. After the administration of salvar- san all became negative within 8 weeks, and in 12 of the cases only one intramuscular injection of 0.6 gm. of sal- varsan was administered. While there is reason to believe that previous treatment with mercury may have had some- thing to do with the good effects produced by salvarsan upon the reaction, it must be admitted that the results ob- tained in these cases demonstrate beyond all question the superior specific value of salvarsan upon Treponema pal- lidum. Patients who have been previously well treated with mer- cury show a higher percentage of negative results after treatment with salvarsan than those who have received no mercurial treatment. Thus of 75 patients who had re- ceived mercurial treatment before receiving salvarsan, 84 per cent became negative, while of 110 patients who had received no treatment before the administration of salvar- san, only 74.5 per cent became negative. In order to illustrate the persistence of the positive Wassermann reaction after treatment with mercurials Ta- ble XLV has been prepared giving the method of treat- ment, the length of time the patient was under treatment, the number of cases observed, and the intensity of the reaction. The table illustrates very well the effect of treatment with mercury upon the Wassermann reaction, as it occurs in the usual routine of serological work in a Wassermann laboratory. It shows that continued treatment with the drug for long periods of time has but little effect upon the reaction in the majority of instances and that interrupted treatment continued for many years does not produce a negative reaction in many cases. It should not be thought, however, from the data given here, that the Wassermann reaction never becomes negative after treatment with mer- cury, for in a considerable proportion of cases it does, and THE EFFECT OF TREATMENT 221 Table XLV Illustrating the Persistence of the Wassermann Reaction after Treatment with Mercury METHOD OF TREATMENT LENGTH OF TREATMENT NUMBER OF CASES INTENSITY OF THE REACTION + + + + - By mouth 1 month 2 2 H C C 2 months 3 1 2 Ct 11 3 " 4 2 2 c t tt 5 " 6 3 3 t t t t 6 " 10 7 1 2 11 11 7 " 4 3 1 t t I t 8 " 5 4 1 Ct tt 9 " 10 7 1 2 Ct t t 10 " 3 2 1 Ct t t 11 " 3 3 11 tt 1 year 24 13 6 5 Ct Ct 14 months 2 2 Ct tt 15 " 2 1 1 Ct Ct 16 " 4 3 1 Ct tt 17 " 1 1 Ct tt 18 " 7 1 4 2 Ct Ct 19 " 4 3 1 11 11 20 11 2 1 1 Ct Ct 2 years 15 8 5 2 C C Ct 2.5 11 3 2 1 Ct t t 3 " 5 1 2 2 Ct t t 4 " 1 1 By mouth* 7 " 1 1 By mouth* 10 " 1 1 By mouth* 12 " 1 1 Inunctions 2 months 2 2 t c 5 " 1 1 11 6 " 2 2 11 1 year 1 1 *Interrupted treatment during this time. + + indicates complete inhibition of hemolysis. the writer has observed many cases in which a negative Wassermann reaction was obtained after from one to two years proper treatment with this drug but it has been his experience that most cases becoming negative after mer- curial treatment relapse in the course of months or a year or two. While he would not go so far as to state that the Wassermann test never becomes permanently negative after treatment with mercury, it has been his experience that it is only in very rare instances that a permanently negative result is obtained from treatment with this drug alone, no matter how it is administered or in what dosage. In fact, it has been proved in rabbits experimen- 222 THE WASSERMANN TEST tally infected with syphilis, that before the animals can be rendered sterile as regards Treponema pallidum so much mercury has to be administered as to cause either the death of the animal or very serious pathological lesions due to the drug. In the writer's rather large experience he has never personally followed a case treated with mercury alone, in which the Wassermann test was positive before treatment, that became permanently negative, although many cases have lost the positive reaction for a while but have invariably become positive again upon the cessation of treatment. However, that mercury does cure syphilis in some instances, is proved by the fact that individuals are encountered who undoubtedly had the disease but who have been without symptoms for many years and who give a negative Wassermann reaction whenever tested. The very careful investigation of Nelson and Haines164 upon the relative efficiency of salvarsan and neosalvarsan, as well as mercury, in the treatment of syphilis, amply con- firms the results obtained by the writer. They observed a large number of cases for months and their results are really the most valuable that have been recorded, so far as the writer is aware. They found that four intravenous injections of salvarsan, combined with intensive mercurial treatment gave nearly twice as many negative complement- fixation reactions as did five intravenous injections of neo- salvarsan combined with intensive mercurial treatment, 64 per cent of those treated with salvarsan becoming negative, while only 33.3 per cent of those treated with neosalvarsan presented a negative result at any time. As regards mer- curial treatment, Nelson and Anderson165 found that they did not obtain a single permanent negative reaction in 50 cases treated by injections of salicylate of mercury, al- though most of the patients received over ten injections and many of them twenty or more. Complement-fixation tests were made at monthly intervals in these cases for several months and the results prove the slight effect that WASSERMANN REACTION AS CONTROL OF TREATMENT 223 mercurial treatment has upon this reaction and the futility of relying upon mercury alone in the treatment of syphilis. If the patient's blood serum be carefully titrated after the administration of either salvarsan or mercury, it will be found that in almost every instance the Wassermann reaction is influenced to some extent, although it may never become negative. Thus, many cases showing a double-plus reaction, or absolute inhibition of hemolysis, with quanti- ties of blood serum as small as 0.02 c.c. will become nega- tive for this amount of serum, although giving a positive reaction with the usual quantity of serum used in the rou- tine tests; i. e., 0.1 c.c. By titrating the inhibitory strength of each patient's serum it will be found that, although the Wassermann test may apparently be uninfluenced when the diagnostic amount of serum is used, that is, 0.1 c.c., some diminution in the strength of the reaction will be observed in amounts less than 0.1 c.c. In using the Was- sermann test as a control of treatment the titration of the patient's serum is of great importance, as will be noted. Not only will treatment with salvarsan or mercury mark- edly influence the strength of the reaction but if treatment is commenced early in the primary stage of the disease it will sometimes prevent the development of a positive re- action, although clinical symptoms may occur. The writer has observed several such instances, in which secondary symptoms developed in spite of treatment, but in which a positive Wassermann reaction was never obtained. The Wassermann Reaction as a Control of Treatment Aside from its value as a diagnostic agent, the great- est value of the Wassermann test for syphilis is found in its use as a guide to the treatment of the disease. It will be evident, from what has been said of the effect of treat- ment upon the positive reaction that, if scientifically used, the efficacy of any form of treatment advocated for syph- ilis may be absolutely determined by employing the test as a control of that treatment. In the army the test has 224 THE WASSERMANN TEST been of the greatest practical value in ascertaining the rel- ative efficiency of the various methods of treating syph- ilis and it has been very largely used for this purpose. As an illustration of the extent of the use of the reaction in the army in the control of the treatment of syphilis, the writer may say that he has personally made almost twenty thousand Wassermann tests upon treated patients, all of these tests being for the purpose of controlling the treat- ment of the disease, and his experience has been that of all other medical officers who have had charge of the army laboratories where the Wassermann test is performed. By means of the Wassermann test, repeated at frequent intervals, one is able to diagnose a relapse of syphilitic infection long before the appearance of clinical symptoms, and in this way enabled to treat the infection before it has resulted in such symptoms, thus saving the patient much anxiety and discomfort, and obviously at a stage when it is much more amenable to treatment than after gross clinical symptoms of relapse have occurred. The writer believes that the first symptom of a relapse of a syphilitic infection, in the vast majority of cases, is the recurrence of a posi- tive Wassermann reaction, and that this recurrence often happens weeks and even months before marked clinical symptoms are noted. In other words, the positive Wasser- mann reaction is the most delicate symptom of relapse in syphilis that is known, a symptom that occurs long before definite clinical symptoms, and thus it is of the greatest value as an indication of the need of further treatment in cases where it has once disappeared and then recurs. For this reason all cases of syphilis that have been given specific treatment should be tested at frequent intervals in order to ascertain the effect of the treatment upon the reaction, and after a negative reaction is once obtained a test should be made at least every three months for a period of at least one year, when, if the reaction is still negative, a provoca- tive Wassermann test, a Wassermann test upon the spinal WASSERMANN REACTION AS CONTROL OF TREATMENT 225 fluid, and a luetin test should be made, before the patient is pronounced cured. The writer has already mentioned the great importance of titrating the complement-binding power of the syphilitic patient's blood serum in judging of the effect of any treat- ment upon the Wassermann reaction and in using the test as a control of treatment, and the method of making such titrations. The titration of the patient 's blood serum should be done whenever it is possible, but especially in those cases that show a continued double-plus (four-plus) reaction after specific treatment has been administered for some time. In such cases, also, the spinal fluid should be tested if the Wassermann on the blood shows no diminution when the serum is titrated, for it is in this class of cases that early involvement of the central nervous system is most apt to have occurred. The data derived from the use of the Wassermann test as a control of treatment has enabled us to use both mer- cury and salvarsan intelligently and has added greatly to our knowledge of the action of other drugs that have been advocated from time to time in the treatment of the disease. It has demonstrated beyond all doubt that the absence of clinical symptoms is no proof of cure in this disease, even though symptoms may be absent for months or years, and that a single negative Wassermann reaction is of absolutely no value in determining either the question of the existence of syphilis or of its cure. Perhaps the most striking proof of the value of the Was- sermann test in the control of the treatment of syphilis is the results that have been obtained by numerous workers upon the occurrence of relapse after treatment with sal- varsan, as shown by the Wassermann test. In 1913, the writer published108 certain observations upon relapse in syphilis after treatment with salvarsan and has continued collecting data upon this subject which have confirmed the observations published. They were undertaken for the purpose of determining the relation of relapse to the dose 226 THE WASSERMANN TEST and method of administering the drug but serve equally well to illustrate the value of the Wassermann test in con- trolling the treatment of syphilis and in ascertaining the therapeutic results following treatment with a specific drug. Used in this manner the Wassermann test has fur- nished the following information regarding the occurrence of relapses after treatment with salvarsan: Four hundred cases were selected for study, and were observed for a period of eight months or more. Of the 400 cases, 202, or 50.5 per cent remained positive after the administration of the drug, while 198, or 49.5 per cent be- came negative to the Wassermann test after salvarsan was given. Of the 198 cases that became negative, 172, or 86.8 per cent have relapsed; i. e., the Wassermann test has again become positive. The Stage of Syphilis in Relation to Relapse.-The re- lation of the stage of syphilis to relapse following treat- ment with salvarsan is shown in Table XLVI. Table XLVI The Stage of Syphilis in Relation to Relapse after Treatment with Salvarsan STAGE OF DISEASE NUMBER OF CASES NUMBER OF RELAPSES % OF RELAPSES Primary 35 31 88.5 Secondary 115 97 84.3 Tertiary 23 16 69.5 Latent 25 22 88.8 A consideration of the table shows that 88.5 per cent of the primary cases relapsed, as shown by the Wassermann test; 84.3 per cent of the secondary cases; 88.8 per cent of the latent; and only 69.5 of the tertiary cases. The latter had all received considerable treatment with mercury prior to the administration of salvarsan, so that the compara- tively low percentage of relapses in this stage undoubtedly was due to this fact, for the writer believes that in un- treated cases the tertiary stage would show as large, and probably a larger, percentage of relapses than either the WASSERMANN REACTION AS CONTROL OF TREATMENT 227 primary or secondary. The latent stage showed the largest percentage of relapses and this was expected, as it is this class of cases that most often prove resistant to treatment, so far as the disappearance of the Wassermann test is concerned, but the difference is only slight, as shown by the table. The Relation of the Intensity of the Wassermann Reac- tion to Relapse.-The cases giving the strongest Wasser- mann reaction furnished the largest percentage of re- lapses, as is shown in Table XLVII. Here, as in all tables in this work, the sign, + +, indicates complete inhibition of hemolysis, the four-plus reaction of many authorities: Table XLVII The Relation of the Intensity of the Wassermann Reaction to Relapses after Treatment with Salvarsan CHARACTER OF REACTION NUMBER OF CASES NUMBER OF RELAPSES % OF RELAPSES + + 110 104 94.5 + 88 68 77.2 198 172 86.8 This table demonstrates that of the 110 cases giving a double-plus reaction, no less than 104, or 94.5 per cent relapsed, while of the 88 cases giving a plus (three-plus or two-plus) reaction, only 68 or 77 per cent relapsed. Therefore it is evident that the stronger the Wassermann reaction the greater the number of relapses following treatment with salvarsan. Time of Relapse after Treatment with Salvarsan.-By controlling the treatment of syphilis with salvarsan it has been possible for us to obtain accurate data regarding the time of relapse after treatment with salvarsan; i. e., the period elapsing from the time the Wassermann test be- came negative to the time in which the test again became positive. The data accumulated show that, in the major- ity of instances, a Wassermann relapse, as it may be called, occurs within eight months after cessation of treatment. 228 THE WASSERMANN TEST Thus, of the 172 cases that relapsed after treatment with this drug, no less than 159 cases relapsed within eight months after treatment. Table XLVIII gives the time of relapse after treatment observed in these cases. Table XLVIII The Time of Relapse after Treatment with Salvarsan as Shown ry the Wassermann Test TIME OF RELAPSE NUMBER OF CASES TIME OF RELAPSE NUMBER OF CASES 1 month 8 5 months 17 5 weeks 7 6 " 12 6 c c 5 7 " 5 7 C ( 6 8 " 5 2 months 38 9 " 2 9 weeks 1 10 " 2 10 < c 7 11 " 1 11 c ( 2 12 11 2 3 months 26 13 " 1 14 weeks 1 15 " 1 4 months 19 22 " 1 24 " 1 From this table it will be noted that no less than 38 cases relapsed at the end of two months and 26 at the end of three months, so that it would appear that these periods are of special significance as regards the time of relapse. It is also interesting to note that no less than 100 of the 172 cases that relapsed did so by the end of three months and no less than 149 of the 172 cases by the end of six months after cessation of treatment. From this it would appear that the vast majority of cases of syphilis treated as these were, with salvarsan, will relapse within six months so far as the Wassermann reaction is concerned, and that it is during this period that repeated Wassermann tests are most essential. The statement that has already been made regarding the worthlessness of a single negative Wassermann reaction is well illustrated by this table, and even in those cases in which the reaction remains negative for a long period of time, no assurance can be given that it will not again be- come positive, for it will be observed that some of these WASSERMANN REACTION AS CONTROL OF TREATMENT 229 cases had a negative Wassermann reaction for as long as one and two years and became positive. As regards the time of relapse in the various stages of syphilis it may be stated that the Wassermann tests showed that the average time of relapse in primary cases was 4^ months; in the secondary cases, 4 months; in the tertiary cases, 3 months; and in the latent cases, 3^ months. The latent cases included both early and late latent cases, which accounts for the longer period of re- lapse in this class. As will be noted, relapses occur earlier the longer the syphilitic infection has lasted. The Relation of the Method of Administration of Salvar- san to Relapses.-The Wassermann test has shown that the percentage of relapses after salvarsan varies considerably with the method of administration of the drug. The three methods that have been tested in this way have been the intramuscular injection of the drug, the intravenous in- jection, and the combined intramuscular and intravenous injection. Table XLIX gives the number of cases treated which have been thus tested and the number of relapses, together with the percentage of relapse in each class of cases. Table XLIX The Relation of the Method of Administration of Salvarsan to Relapse, as Shown by the Wassermann Test METHOD OF ADMINISTRATION TOTAL NUMBER OF CASES RELAPSES NUMBER % Intramuscular 74 54 72.9 Intravenous 115 106 92.1 Combined intramuscular and intravenous 9 6 66.6 The number of cases tested that were treated by the com- bined intramuscular and intravenous method was so small that the percentage of relapses; i. e., 66.6 per cent, can not be considered as a fair estimate of the percentage that would relapse were a greater number of cases tested. However, the number of cases tested that were given sal- 230 THE WASSERMANN TEST varsan intramuscularly and intravenously is large enough to base fair estimates upon, and it is very evident that a much smaller number of relapses follow the intramuscular method of administering the drug than the intravenous method. In justice to the intravenous method it should be stated that fully 50 per cent of the cases tested had only received one dose of the drug, but even after two injections the results were not as good as after one intramuscular injection. Of 80 cases receiving one intravenous injection of 0.06 gm. of salvarsan 78 relapsed, or 97.5 per cent, while of 67 cases receiving the same dose intramuscularly, 51 relapsed, or 76 per cent. From what has been stated regarding the study of re- lapse in syphilis after the administration of salvarsan some idea may be gained regarding the great value of the Wassermann test when it is used as a control of treatment or as a test of the efficiency of any drug recommended in the treatment of the disease. The results that have fol- lowed the modern method of treatment; i. e., numerous intravenous injections of salvarsan combined with mercu- rials, given in the proper manner, are far superior to those which have been noted in this chapter with either drug alone. There is but one way of determining what these results are and that is by the use of the Wassermann test as a control of the treatment of the disease. Without it, the physician is at as great a disadvantage in judging of the results of his treatment of syphilis as the mariner would be without a compass, and this test may well be called the compass of the physician in the treatment of this disease. The Interpretation of the Reaction when the Test is used in the Control of Treatment.-A Wassermann test should not be performed for the purpose of controlling the treat- ment of syphilis until treatment with salvarsan, mercu- rials, or other agents has been omitted for at least a month and the test should be repeated at monthly intervals re- WASSERMANN REACTION AS CONTROL OF TREATMENT 231 gardless of the reaction if treatment is not again instituted. If the reaction is negative monthly tests should be made in all cases possible. If it remains persistently negative treatment should not be persisted in, but if there is any trace of inhibition beyond a weak plus-minus reaction, the treatment should be continued. It should always be re- membered that a single negative reaction is absolutely val- ueless as indicating the cure of the infection and repeated tests must be made with a similar result before any opin- ion favoring cure is warranted. Too many patients have been assured that they were cured of syphilis on the strength of one or two negative Wassermann reactions, with disastrous results to innocent individuals, as well as to the patient, and the utmost caution should be exercised in expressing any opinion regarding ultimate cure in syph- ilis. It should be remembered that the Wassermann test has shown that the vast majority of treated syphilitic in- fections eventually relapse, even though the treatment has been scientific and thorough, so that the prognosis in every case should be very guarded. In primary infections which have been treated and in which the Wassermann test has become negative, a return of the reaction even to the extent of a plus-minus reaction, when it is persistent, should indicate further treatment. In the secondary cases a plus-minus reaction is not of as much importance, as most of these cases will give a stronger reaction if they relapse, but even in such cases if the plus-minus reaction persists for several weeks, treat- ment should be instituted. In tertiary and latent cases the occurrence of a plus (two- or three-plus of some authors) reaction after treatment has been stopped should be fol- lowed by the resumption of the treatment. Treatment in syphilis should never be discontinued until the Wassermann test has become negative unless it is found that the case is "Wassermann fast"; i. e., that no amount of treatment will cause the reaction to become neg- 232 THE WASSERMANN TEST ative. Wassermann fast cases are frequently met with in practice and occur during every stage of the disease but such cases occur more frequently in tertiary and latent infections than in primary or secondary. In the "Was- sermann fast" cases the clinical symptoms, aside from the positive Wassermann reaction, may, and generally do, dis- appear under treatment, but the reaction never becomes negative. Regarding the significance of these "Wassermann fast" infections, the recent work of Stokes and Busman177 is most valuable and interesting. They found that of 458 syphilitic patients who had received from twelve to twenty-nine ars- phenamine injections combined with inunctions of mercury, the average being fourteen injections of arsphenamine and ninety inunctions distributed over a period of eleven months, no less than 6.6 per cent of primary and secondary cases, and 22 per cent of latent, late and hereditary cases remained persistently positive. A most careful examina- tion of these patients, however, showed that 44 per cent of the "Wassermann fast" cases showed cardiovascular syphilis; 30 per cent osseous lesions; 30 per cent neuro- syphilis; 21 per cent hepatic, splenic or gastric syphilis, and other lesions were found in from 10 to 17 per cent. Most of these patients presented more than one type of syphilitic involvement and in only 10 of the patients was the positive and persistent Wassermann reaction the only evidence of syphilis. In 87 per cent of the "Wassermann fast" patients the treatment resulted in a symptomatic ar- rest of the disease. They could find no evidence that the " Wassermann-fastness" was the result of infection with any special strain of Treponema pallidum, as suggested by some authorities. They regard a persistently positive serum Wassermann reaction as an indication of grave rather than trivial syphilitic infection, for their experience has shown that while obvious symptoms may be absent, the vast majority of patients showing persistent positive reac- PROVOCATIVE WASSERMANN REACTION 233 tions have aortitis, or some form of visceral or nervous syphilis. The Provocative Wassermann Reaction In 1910, Gennerich107 called attention to the fact that cases presenting a negative Wassermann reaction often showed a positive reaction after the administration of a dose of salvarsan, this reaction occurring either within a few hours after the injection or within two weeks follow- ing it. He called attention to the value of this phenomenon in the diagnosis of obscure syphilitic infections, and as a method of determining whether a patient is really cured of the infection. He suggested that this so-called "pro- vocative Wassermann" be used in treated cases that had presented negative reactions for some time, and were pre- sumably cured, as, in such instances he had found the provocative test frequently positive, thus demonstrating that the disease was still present. Milian,108 of Paris, about the same time, called attention to the value of the procedure in the study of the disease, as well as Herx- heimer,109 and Michaelis,110 but to Gennerich belongs the credit of first actually applying the test in the diagnosis of syphilis and as a control of the treatment of the disease. Method of Making a Provocative Wassermann Test.- The writer recommends the following procedure in making a provocative Wassermann test: The suspected patient is given from 0.3 to 0.6 gm. of salvarsan intravenously and the blood is collected and tested 12 hours after the admin- istration of the drug; 24 hours after administration; and every day for at least ten days. If the negative reaction is to become positive, it generally does so within from a few hours to 5 or 6 days after the salvarsan has been given, but delayed reactions have occurred not infre- quently up to 10 days after administration, and it is good practice to make a test upon the 12th and 14th days after administration, in addition to those recommended. Dur- 234 THE WASSERMANN TEST ing the first two days after salvarsan has been given, if it is possible, the blood should be collected at intervals of three or four hours and tested, as sometimes the reaction is very evanescent in character and quickly disappears. However, owing to the labor involved in making so many tests it is seldom practicable to do this, and the procedure first mentioned will generally prove sufficient to detect the reaction, if it occurs. Stokes and O'Leary111 recommend that the blood be drawn at the time of injection and daily thereafter for one week, and that all blood collected be tested with several antigens at one sitting and with the same reagents, in order to avoid any difference in the re- sults that might occur from variations in the reagents used in the test from day to day. Results of the Test.-In a recent contribution, King,112 as the result of his investigation of the provocative test upon a very small number of patients, came to the conclusion that the provocative Wassermann reaction does not exist, but that the positive result that sometimes follows an in- jection of salvarsan is due either to normal variations in the complement-binding power of syphilitic serum, as shown by the writer, or to variations in the technic or reagents used in the test when made upon different days. His work is very far from convincing and the united ex- perience of hundreds of observers who have used the pro- vocative test in their laboratories and practice is unan- imous as regards the actual occurrence of this phenom- enon and as to the value of the test in the diagnosis of syphilis and the control of its treatment. Nichols,113 after stating that it is a well-established fact that in a syphilitic a negative Wassermann reaction can be converted into a positive one by an injection of salvarsan, gives the following table illustrative of the effect of the provocative test in ten patients in whom a luetin test was also performed at the same time. Table L gives the re- sults that he obtained in these cases: PROVOCATIVE WASSERMANN REACTION 235 Table L Results of the Provocative Wassermann Test and the Luetin Test. (After Nichols.) CASE NO. TIME ORDINARY WASSERMANN REACTION HAD BEEN NEGATIVE LUETIN TEST PROVOCATIVE WASSERMANN TEST 1 16 months 2 21 " - 3 24 " 4 24 " 5 24 " - 6 15 ' ' + + + 7 17 11 + + + 8 18 " + + + 9 19 " + + + 10 ? + + + This table is conclusive proof of the occurrence of the provocative reaction, as it shows complete agreement be- tween the results of this test and the luetin test, for it is hardly probable that, if the positive reaction in these cases was due to variation in the complement-binding power of the blood serum or inaccuracies in technic, the results of the provocative test would be identical with the results of the luetin test in every case. Nichols calls attention to Case 10, in the above table, be- cause of the interesting results obtained as regards a delayed provocative reaction. The patient was infected in 1906 and had two years of treatment by the mouth with mercury and potassium iodide. No symptoms had oc- curred since 1908 and in March, 1913, he gave a negative reaction. On April 18th he was given an intravenous injec- tion of 0.6 gm. of salvarsan and his blood was tested every day with the following results: April 18th, negative; April 19th, negative; April 20th, negative; April 21st, negative; April 22nd, negative; April 23rd, double-plus (four-plus). This case well illustrates the value of the provocative reaction in patients who have had no symptoms of syphilis for several years and who give a negative Wassermann test. Stokes and O'Leary,114 in a very excellent study of the provocative Wassermann reaction, obtained very signifi- 236 THE WASSERMANN TEST cant results. They made 103 provocative tests, of which 19, or 18.4 per cent resulted in a reversal of the ordinary Wassermann from negative to positive. This small per- centage of positive results with the test is not unusual and it has been the writer's experience that it is seldom that a higher percentage than 20 per cent is obtained with the provocative Wassermann test upon limited groups of cases. If thousands of tests were made it is probable that the positive percentage would be somewhat higher but from 18 to 20 per cent of positive results is about all that can be expected from the reaction. As to the character of the cases showing a positive re- action with the test, Stokes and O'Leary state that 26.3 per cent presented no clinical symptoms of syphilis, but that the remainder show suspicious symptoms although the Wassermann was negative. However, they state that "In no one of the positive cases were the signs so indu- bitable that the provocative could be regarded as unneces- sary or merely confirmatory." They say further: "we believe the test has distinct value in determining the status of a patent under treatment." Stokes and O'Leary give Table LI showing the results obtained with the provocative Wassermann test in differ- ent types of syphilis. Table LI Efficiency of the Provocative Test in Various Types of Syphilis TYPE NUMBER OF CASES % POSITIVE Heredosyphilis 2 0 Osseous 5 20 Central nervous system 12 25 Vascular 3 33 Latent 10 40 Late cutaneous 5 60 Late mucous membrane 5 80 While they admit that the number of cases treated is not large enough to base accurate conclusions upon, they state that, from their series, it appears as though the pro- PROVOCATIVE WASSERMANN REACTION 237 vocative test is least useful in the class of cases where it is most needed, but, even if this proves to be true, and we can only expect 20 per cent of positive results in osseous syphilis, 25 per cent in syphilis of the central nervous system, and 33 per cent in vascular syphilis, the writer believes that the test is well worth while, as often it will be the only method available for making a diagnosis in this class of cases. The indications for a provocative Wassermann reaction are given by Stokes and O'Leary in their paper and are so admirable that this section of the paper is here quoted: "1. A definite history of primary or secondary lesions or a suspicious genital sore of any description. (With a negative Wassermann.) "2. Syphilis in husband or wife or a history of a sore in either. "3. Treated cases to determine the fact of cure or need for further treatment. One-third of the cases thus tested by us, gave a positive provocative effect. "4. Obscure bone or joint lesions. "5. Histories of miscarriages unless the anatomical cause is glaringly obvious. "6. Mothers of syphilitic children without clinical signs of the disease. "7. Cases with a history of a positive Wassermann else- where, negative on present examination. "8. Mental deviates and constitutionally inferior indi- viduals with suspicious histories. "9. Certain signs elicited by special examinations, such as decreased bone conduction with normal hearing, chorio- retinitis and retinitis pigmentosa, bilateral dacryocystitis in childhood," etc. In all the above conditions the writer agrees with the authors that a provocative Wassermann test should be made, if possible, for in nearly all cases in which the indi- 238 THE WASSERMANN TEST cations for the test cited have occurred, the writer has found the test of service. As an illustration of the results obtained with the test in cases of syphilis that have not shown symptoms for a long period of time and in which the ordinary Wassermann test has been negative for months and even years, Table LII is given: Table LII Results of Provocative Wassermann Test in Syphilis CASE NO. TIME SINCE TREATMENT TIME SINCE WAS- SERMANN BECAME NEGATIVE RESULT OF WASSERMANN REACTION RESULT OF PROVOCATIVE REACTION 1 12 months 12 months - + + e 13 " 13 " - 3 14 " 14 " - + + 4 15 " 15 " - + + 5 16 11 16 " - 6 17 " 17 " - + 4- 7 18 " 18 " - + + 8 19 " 19 " - + + 9 20 " 20 " - + + 10 21 " 21 " - - 11 24 " 24 " - - 12 24 " 24 " - + + 13 24 " 24 " - - 14 25 " 25 11 - + + The cases given in the above table are selected from a large number because they show the results of the provoc- ative reaction month after month, as regards time, and because the data relating to them is of such a nature as to make the test definite and conclusive. It will be seen that a positive reaction was obtained as long as 25 months after the cessation of treatment, during which time no symptoms of syphilis had been noted and the Wassermann reaction was always negative, when tested. Pease,166as the result of his experience with the provoca- tive Wassermann reaction, regards the test as a most valu- able one in all stages of syphilis, but especially in the ter- tiary and latent stages. He employed the test in 29 cases in which the ordinary Wassermann test gave a negative re- sult, and obtained a positive reaction in 24 of them. In 8 PROVOCATIVE WASSERMANN REACTION 239 primary cases, in which the ordinary Wassermann reaction was either negative or plus-minus, the provocative test gave a positive result in all, and this was also true of 4 congenital infections. In 6 latent cases he obtained a posi- tive result in 3, and in 6 syphilitic infections of the nervous system, the provocative test gave a positive reaction in no less than five, although all gave a negative or weak reaction with the ordinary Wassermann test. Pease states that he regards the provocative Wassermann test as equal in value to the luetin test and urges its more general use in all sus- pected syphilitic infections in which the ordinary comple- ment-fixation test is negative. O'Leary,176 in a further series of experiments with the provocative test, concludes that when properly performed this test adds from 10 to 20 per cent to the acuteness of the ordinary Wassermann test, but that a negative provocative test should never be interpreted as indicating cure, as in numerous instances double or four plus Wassermann reac- tions have been observed after a negative provocative test. The writer is firmly convinced, from personal experience, that a dose of salvarsan or neosalvarsan will frequently convert a negative Wassermann reaction into a positive one and that this phenomenon is not due to any variation occurring normally in the patient's blood serum or to dif- ferences in technic or in the reagents used in making the test. When one has tested cases week after week, always with a negative result, and when these same cases after a dose of salvarsan or neosalvarsan become positive within a few hours or days after the administration of the drug, there can be only one conclusion, and that is that the posi- tive result has been produced by the drug, and the writer has had this experience too many times to allow of the least doubt in his mind of the existence of a provocative reaction following the administration of these drugs. It is true that the test often fails in cases in which one would 240 THE WASSERMANN TEST expect it to be positive, but this is no proof of the non- existence of the phenomenon. Cause of the Provocative Reaction.-The exact nature of the provocative reaction is still undetermined. There are two theories that have been discussed in this connection, the first being that the reaction is due to the stimulation of the treponemas by the drug and the consequent pro- duction of more of the substance or substances which give rise to a positive Wassermann reaction; while the second is that the drug kills sufficient of the treponemas to increase the amount of toxin and thus the syphilitic antibody, as it may be called, is increased. The first theory depends upon the belief that Treponema pallidum, during its life and multiplication in the tissues causes the production of some substance which produces the Wassermann reaction, and that a small dose of salvarsan stimulates the few re- maining organisms in the negative Wassermann cases to overproduction of this substance for a short time, and thus the negative Wassermann reaction becomes positive. The second theory rests upon the belief that the organisms killed by the drug either liberate the substance producing the positive reaction, or a toxin which stimulates the pro- duction of this substance in the tissues of the body. Neither theory has been proved correct as we know nothing cer- tain as to the actual cause of the Wassermann reaction. Some authorities have recommended intramuscular in- jection of mercury for the production of the provocative Wassermann reaction but the reports concerning the effi- cacy of this method are so conflicting that it is not recom- mended. The provocative Wassermann test is a useful adjunct to our diagnostic methods for syphilis and should be re- sorted to whenever it is possible, both in the diagnosis of the disease and in determining the question of the cure of the infection. The large amount of work involved in mak- ing the numerous tests that are necessary in correctly us- PROVOCATIVE WASSERMANN REACTION 241 ing the provocative reaction will always operate to limit its use to comparatively few cases but whenever it is pos- sible to use the test it should be employed. From the evidence that has accumulated it may be stated that in patients who have presented a negative Wasser- mann reaction for long periods of time and in whom symp- toms of syphilis have been absent during this period, the provocative W assermann test often results positively. The same result is sometimes obtained in cases presenting sus- picious symptoms but in which the ordinary W assermann test is negative, or in cases which give a history of infec- tion but with no symptoms and, a negative Wassermann re- action. CHAPTER X THE WASSERMANN TEST UPON THE CEREBRO- SPINAL FLUID. INTERPRETATION OF RE- SULTS. THE COLLOIDAL GOLD TEST. THE CELL COUNT AND TESTS FOR INCREASE IN GLOBULINS IN THE CEREBROSPINAL FLUID Recent observations have shown the great importance of the Wassermann test upon the cerebrospinal fluid and it may be stated that no case of syphilis can be considered as having had the best treatment unless this fluid has been tested, even though the Wassermann test upon the blood is negative and the patient has been free from symptoms for long periods of time. It is now demonstrated beyond doubt that involvement of the central nervous system fre- quently occurs very early in this disease and this demon- stration has been rendered possible by the results of the Wassermann test upon the cerebrospinal fluid during the early latent and secondary stages of the infection. The Wassermann test upon the cerebrospinal fluid often shows involvement of the central nervous system before any clinical symptoms are observed and the use of the test has rendered intensive treatment possible and has resulted in either preventing the development of such symptoms or in greatly delaying their appearance. Formerly it was be- lieved that if the test gave a negative reaction in the cere- brospinal fluid and a positive one in the blood, the diag- nosis of cerebrospinal syphilis should be made, the result serving to distinguish this condition from paresis, in which both the blood and cerebrospinal fluid usually gave a posi- 242 CEREBROSPINAL FLUID 243 tive reaction. However, it has since been shown that the negative result in the cerebrospinal fluid depended en- tirely upon the amount of the fluid used in the test and that if a large enough amount was used a positive result could be obtained in the fluid in almost 100 per cent of cases of cerebrospinal syphilis. Therefore, it follows, that the Wassermann test upon the cerebrospinal fluid, so far as its value in differentiating various syphilitic con- ditions of the central nervous system is concerned, depends upon quantitative factors entirely, so that it is necessary in testing this fluid to test different quantities if one is to secure the best results from the test. The importance of the Wassermann test upon the cerebrospinal fluid is shown by the statement of Fordyce115 that "From 25 to 35 per cent of syphilitic individuals have positive findings in the spinal fluid during the first year of their infection" and during this time symptoms referable to the central nervous system are frequently present. This being so, the impor- tance of ascertaining in every case of syphilis whether the central nervous system is involved at as early a date as possible, can hardly be overestimated, and for this rea- son the cerebrospinal fluid should be tested as well as the blood serum, although the necessity of the test upon the spinal fluid has not, as yet, been fully realized by the medi- cal profession. Carnaz167 examined the spinal fluid in 76 cases of pre- secondary syphilis, of which 35 per cent showed changes and one gave a four-plus complement-fixation reaction in the fluid. This might be expected to occur in rare instances, as Marinesco and Minea168 demonstrated, experimentally, the occurrence of Treponema pallidum in the cerebrospinal fluid in 5 of 8 primary infections and their results have been confirmed by the finding, by dark-field illumination, of the treponema in the spinal fluid in primary and secondary cases by Artz and Kerl169 and Fruhwald and Zaloziecki.170 The recent observations of Fildes, Parnall, and Maitland171 244 THE WASSERMANN TEST are of interest in this connection, as they examined the spinal fluid of 1314 men, most of them in the early stages of syphilis, and found that no less than 236 showed a pleocyto- sis of 10 or more cells per c. mm. and 91 gave a positive complement-fixation reaction, although no signs of nervous disease were present. In one case Treponema pallidum was demonstrated in the spinal fluid by dark-field examina- tion, the infection being only twenty weeks old. The writer has already described the method of obtain- ing the cerebrospinal fluid and the technic required (Chap- ter V) in making the test. As the differentiation of some of the syphilitic diseases of the central nervous system depends upon the quantity of fluid tested, the writer rec- ommends that 0.2, 0.5, 0.7 and 1 c.c. quantities be tested in each case, with his method of performing the reaction, the fluid, of course, not being inactivated before use. The quantity of spinal fluid tested, with any method, should always be at least four times the quantity of blood serum tested with the particular method used, and it is always best to use both smaller and larger amounts than this in order to determine the strength of the positive reaction upon which depends the differential diagnosis of the con- dition present. Hauptmann and Hossli116 were the first to insist upon the use of large quantities of the cerebrospinal fluid in ap- plying the Wassermann reaction and their work changed entirely the prevalent conception of the comparative rarity of a positive reaction in diseases like tabes and demon- strated that almost any syphilitic disease of the central nervous system would give a positive AVassermann re- action with these larger amounts of fluid. Thus, the find- ing of a positive reaction in the cerebrospinal fluid, once looked upon as almost pathognomonic of paresis, has ceased to point to this disease unless it is obtained with small amounts of the fluid, for with large amounts other CEREBROSPINAL FLUID 245 syphilitic diseases of the central nervous system also give a positive reaction. Because of the use of too small amounts of the cerebro- spinal fluid in making the Wassermann test much of our statistical data regarding the occurrence of a positive re- action in the various syphilitic infections of the central nervous system are worthless and a great many divergent results are reported in the literature owing to this fact. Mueller117 obtained 75 per cent of positive reactions in tabes, and 98 per cent in paresis, while Boas found that all of his untreated cases of tabes gave a positive reaction. This is a very good illustration of the confusion in results due to the use of different amounts of the spinal fluid, for Boas used larger amounts of the fluid in his tabetic cases and obtained a much higher percentage of positive results. The writer's experience may be summed up as follows: In paresis the Wassermann test is positive in the blood in nearly 100 per cent of the cases, and if 1 c.c. of the cerebrospinal fluid be used, it is positive in practically 100 per cent of the cases. If smaller amounts of the spinal fluid be tested the percentage of positive reactions will be found to decline, until when but 0.2 c.c. of the fluid is tested only from 75 to 80 per cent of the cases will give a positive reaction. When 0.5 c.c. of the fluid is used practically 98 per cent of cases of paresis will give a positive reaction in the spinal fluid, a result never found in any other syphilitic condition of the central nervous system. In cerebrospinal syphilis, if 1 c.c. of the spinal fluid be tested almost 100 per cent of the cases will prove positive, but if only 0.2 c.c. of the fluid be tested only about 10 per cent will prove positive. The Wassermann on the blood in these cases gives from 60 to 75 per cent of positive re- sults, as reported by different investigators. In tabes, if 1 c.c. of the cerebrospinal fluid be tested, the Wassermann will be positive in about 95 per cent of the cases, but if only 0.2 c.c. of the fluid be tested the positive results will only amount to from 5 to 10 per cent. In this 246 THE WASSERMANN TEST disease the Wassermann test upon the blood gives about 70 per cent of positive reactions. From the above it is apparent that in differentiating be- tween paresis, cerebrospinal syphilis and tabes, the ac- curate titration of the complement-binding power of the cerebrospinal fluid is most important. If 1 c.c. of the fluid is used for the test, the results will be positive in prac- tically all of the cases of either disease, but if smaller amounts of the fluid be used it will be found that results are obtained, which, if taken in conjunction with the clin- ical symptoms present, will generally enable the correct diagnosis to be made. Thus, if a positive reaction is ob- tained with 0.2 c.c. of the fluid, or less, the diagnosis is al- most certainly paresis, for only from 5 to 10 per cent of cases of cerebrospinal syphilis or tabes give a positive result with this amount of fluid. If a positive reaction is obtained with 0.1 c.c. of the fluid the diagnosis of paresis will be found correct in 99 per cent of the cases examined. The Test upon the Cerebrospinal Fluid in the Various Stages of Syphilis.-As stated, recent researches have shown that involvement of the central nervous system in syphilis occurs early in the disease in a considerable pro- portion of the cases, as has been shown by the results of the Wassermann test upon the cerebrospinal fluid. In fact, so good an authority as Fordyce holds that paresis, tabes, and the majority of other manifestations of syphilis on the part of the nervous system originate "at the time of the general infection in the first year" and there is much to support this theory in the biological and clinical char- acteristics of syphilitic infection. Results in Primary Stage.-The writer has no knowledge of a well authenticated positive result of the Wassermann test upon the cerebrospinal fluid in the primary stage of syphilis. Results in the Secondary Stage.-The older writers upon this subject all agreed in stating that cases showing no CEREBROSPINAL FLUID 247 symptoms of involvement of the central nervous system, in the secondary stage of syphilis, were invariably negative to the Wassermann test upon the spinal fluid, or to other tests upon this fluid, while, if symptoms of involvement were present the Wassermann was generally positive. One ex- ception may be made to this statement in the case of Ravaut119 who, in 1903, reported that among 116 cases in which the spinal fluid was tested, no less than 67 per cent showed some abnormality of the fluid, either in an in- creased cell count or an increase in globulins. The more recent investigators have reported numerous instances in which the Wassermann test upon the cerebrospinal fluid was positive early in the secondary stage, even when there were no symptoms of involvement of the central nervous system. Thus Wile120 found 30 per cent of the cases of secondary syphilis that he tested gave a positive result with the Wassermann test upon the spinal fluid, and Craig and Collins121 found that a considerable proportion of their early secondary cases gave a positive Wassermann reac- tion in the spinal fluid. Ellis and Swift122 examined 22 cases of untreated early secondary syphilis and found the spinal fluid negative in all, but in 56 cases of late secondary syphilis, tested by Altman and Dreyfus,123 no less than eleven gave a positive Wassermann reaction in the spinal fluid. Results in the Tertiary Stage.-The results of the Was- sermann test upon the cerebrospinal fluid in paresis, cere- brospinal syphilis, and tabes have already been discussed, but it should be remembered that the examination of the fluid in this stage of the disease from many cases in which there are no discoverable lesions in the central nervous system will result positively, as stated by Kolmer.124 The writer has repeatedly observed cases of tertiary syphilis that gave a positive Wassermann reaction with the cere- brospinal fluid but in which there was absolutely no evi- dence of involvement of the central nervous system, but he would be loath to conclude from this that no lesions due 248 THE WASSERMANN TEST to Treponema pallidum were present. At any rate such a finding is the signal for a very careful scrutiny of the patient and a very guarded prognosis as regards future involvement of the central nervous system. Results in the Latent Stage.-The results of the test upon the cerebrospinal fluid in the latent stage of syphilis will vary, of course, with the involvement of the central nervous system, but the percentage reported by most ob- servers does not exceed 15 per cent, including both early and late latent infections. The Interpretation of the Results of the Test The interpretation of the results of the Wassermann test upon the cerebrospinal fluid, as regards the intensity of the reaction, are the same as already noted in the dis- cussion of the interpretation of the results upon the blood, with this exception; that a positive Wassermann reaction upon the cerebrospinal fluid is specific of syphilis, in the opinion of practically all observers. The writer has never seen a positive reaction upon the spinal fluid in any other condition than syphilis and his experience is that of all who have studied this phase of the subject. It is also true that more double-plus (four-plus) reac- tions are observed in the spinal fluid than in the blood although occasionally weak reactions are noted with the maximum amount of fluid tested. Of course, in cases of tabes and cerebrospinal syphilis, the smaller amounts of fluid show gradations in the strength of the reaction, but a diagnosis of syphilis of the central nervous system should never be made upon anything less than total inhibition of hemolysis; i. e., a double-plus (four-plus) reaction, when as large an amount of the serum as 1 c.c. is tested. The writer has already given the significance of the posi- tive reaction when obtained with the different amounts of the cerebrospinal fluid tested but will here summarize the interpretation to be given the reaction in this respect: CEREBROSPINAL FLUID 249 If amounts of the spinal fluid as small as 0.05 or 0.1 c.c. give total inhibition of hemolysis the diagnosis is almost certainly paresis; if 0.2 c.c. of the fluid give the same re- sult, the diagnosis is, in all probability, paresis; if larger amounts up to 0.5 c.c. of the fluid are used, the diagnosis is probably paresis; while if still larger amounts are used the diagnosis must rest largely upon the clinical symptoms present, as cerebrospinal syphilis, tabes, and paresis all give positive results with the larger quantities of fluid. In order to illustrate the difference in results obtained by using different amounts of the cerebrospinal fluid, Table LIII is inserted, showing the results of the Wassermann test upon cases of cerebrospinal syphilis, tabes and paresis. A consideration of this table will show the importance Table LIII The Results of the Wassermann Test upon the Cerebrospinal Fluid in Cerebrospinal Syphilis, Tabes, and Paresis NAME OF INFECTION CASE NO. WASSERMANN IN BLOOD AMOUNT OF CEREBROSPINAL FLUID C.C. 0.05 0.1 0.2 0.5 0.7 1 Cerebrospinal syphilis 1 + + + + + + + + + 2 - - - - + + + + + 3 + + - - + - + + + + 4 + - - - - + + + 5 + - - - +- + + + + 6 + + - - + - + + + + + + 7 - - - - + + + + 8 + - - - + + + + + Tabes 1 + + + + + + + + + 2 + + - - - + + + + + 3 - - + + + + + + 4* 4 + + - - + + + + + + + + 5 + - + + + 6 + + - - - + - + + + Paresis 1 + + + + + + + + + + + + + + 2 + + + + + + + + + + + + + + 3 + + + + + + + + + + + + + 4 ++ + + + + + + + + + + + + 5 + + + + + + + + + + + + + + 6 + + + + + + + + + + + + + + 7 + + + + + + + + + + + + 8 + + + + + + + + + + + + + + 250 THE WASSERMANN TEST of testing different amounts of the cerebrospinal fluid and how useful such titrations are in the differentiation of the conditions mentioned, especially the differentiation of pare- sis from cerebrospinal syphilis and tabes. It will be understood that the interpretation of the re- sults of the Wassermann test upon the cerebrospinal fluid should be based not only upon this test but also upon the result of the cell count, the colloidal gold reaction and the tests for the increase in globulins, as all should be taken into consideration in arriving at a just conclusion regard- ing the condition present in any case, but the Wassermann reaction alone is sufficient to base a diagnosis of syphilitic infection of the central nervous system upon, if positive. In using the Wassermann test upon the cerebrospinal fluid in the determination of the cure of syphilis the re- action should be negative with 1 c.c. of the fluid. If this quantity of the fluid gives even a partial positive result the case should not be considered as cured but treatment should be continued. The early occurrence of involvement of the central nerv- ous system, as shown by the results of the Wassermann test, renders this procedure upon the cerebrospinal fluid as essential as upon the blood in syphilitic infections, and the test should be employed much more frequently than it now is as a routine measure. No patient should be assured that he is free from syphilitic infection unless the cerebrospinal fluid has been found to give a negative reaction with quan- tities as large as 1 c.c., but, on the other hand, a diagnosis of involvement of the central nervous system should not be made unless complete inhibition of hemolysis occurs with this quantity of spinal fluid. The Wassermann test upon the blood, in cases showing some nerve condition, if positive, does not prove that the condition is syphilitic in nature, but if the cerebrospinal fluid is also positive, the condition is almost certainly syph- ilitic. While the blood serum is often negative in certain CEREBROSPINAL FLUID 251 syphilitic conditions of the central nervous system, it is but rarely that the cerebrospinal fluid is negative when quantities as large as 0.7 to 1 c.c. are employed in the test, so that the examination of the spinal fluid will detect many syphilitic infections that would be overlooked if one de- pended entirely upon the result of the Wassermann test upon the blood. For this reason the writer urgently rec- ommends the testing of the cerebrospinal fluid in every case of syphilis in which it is possible to do so and this test should never be omitted in the determination of the cure of a syphilitic infection. The operation of lumbar punc- ture, when properly performed, is not dangerous or very painful, and the amount of information that is often se- cured by the examination of the cerebrospinal fluid more than compensates for any discomfort that may be caused by the collection of the fluid. The Colloidal Gold Test The colloidal gold test was recommended by Lange,125 in 1913, as a method of distinguishing between certain syph- ilitic and nonsyphilitic diseases of the central nervous sys- tem, and today this test stands as one of the most useful confirmatory methods that we possess in the diagnosis of syphilitic disease of this system. The exact nature of the reaction in the colloidal gold test is unknown. Lange believes that it is a method which en- ables us to measure the protein content of the cerebro- spinal fluid, for he found that this fluid under conditions in which the protein content was increased caused the pre- cipitation of the gold solution and that, within certain dilutions, this precipitation was characteristic of syphilitic conditions of the central nervous system. Previously, Zsig- mondy126 had demonstrated that the clear red color of solutions of colloidal gold, prepared first by him, changed to a blue with the precipitation of the gold upon the ad- dition of electrolytes, and that solutions of protein in the 252 THE WASSERMANN TEST presence of an electrolyte inhibited the precipitate and that it was possible in this way to determine for various pro- teins the inhibitive point and thus differentiate them. Accordingly, Jaeger and Goldstein127 regard the reaction obtained in the colloidal gold test as a purely physical reaction of an electrolytic nature, but others, especially Zaloziecki,128 believe it to be an immunity reaction. It is true that the reaction is obtained when the protein content of the spinal fluid is increased but it is yet too early to in- sist that it is wholly due to protein increase in the fluid. The literature concerning this test is now very volumi- nous and all investigators are unanimous in their state- ments regarding the value of the test in syphilis of the central nervous system, but there is considerable disagree- ment as to the weight to be placed upon the various findings with the test and their exact significance. However, all agree that the test is practically specific for paresis and some of those having a large experience with it insist that a diagnosis of this condition may be made from the result of this test alone. The writer can not agree with this opin- ion of the value of the test in paresis, although it is un- doubtedly true that in at least 96 to 98 per cent of cases the typical paretic curve is obtained, but the same type of reaction has been observed in other conditions, although very rarely. However, the fact that a paretic curve does occur in other conditions than paresis renders it impos- sible for the test to be considered as absolutely specific of this condition, although it must be regarded as the most valuable of all confirmatory evidences of the presence of the disease. Technic of the Test.-The technic of the colloidal gold reaction is very simple, but the preparation of the solu- tion of colloidal gold is extremely difficult and the utmost care must be used at every stage of the process or results will not be satisfactory. In fact, it may be said that the entire technic of the test really consists in the preparation CEREBROSPINAL FLUID 253 of the solution of colloidal gold, for if a good solution is ob- tained there is little difficulty in performing the test. Preparation of the Colloidal Gold Solution>-It is abso- lutely requisite that all chemicals used in the preparation of the solution be chemically pure and the products of Merck have been found most satisfactory by numerous workers. The following chemicals are used in the preparation of the colloidal gold: 1. Merck's Gold Chloride. Acid. 2. Merck's Blue Label Oxalic Acid. Crystals. 3. Merck's Blue Label Potassium Carbonate. 4. Merck's Formaldehyde. Highest Purity, 40%. All glassware used in making the colloidal gold solution should be thoroughly washed in hot water with ivory soap, rinsed in tap water for several minutes; then with hot bichromate solution, allowing the beakers, pipettes, and test tubes to remain in the bichromate solution for fifteen or twenty minutes; after which they are rinsed in distilled water and finally in triple distilled water. The glassware should be used as promptly as possible after cleaning, the pipettes and test tubes being dried in a hot air oven before use. Thermometers used in preparing the solution should be cleaned in the same manner as the glassware. The distilled water used should be distilled in an appa- ratus in which there are no rubber connections and the water should be distilled three times before it is used. In certain regions, where the water is very heavily impreg- nated with mineral salts, it may have to be distilled four, or even five times, and this is also true if the water used has been heavily impregnated with chlorine in purifying it for domestic use. The following method of preparing the colloid of gold solution is recommended by Kolmer129 who states that it is adapted from the method of Miller, Brush, Hammers, and Felton.130 254 THE WASSERMANN TEST 1. Heat 1000 c.c. of triply distilled water over a proper burner in a properly washed beaker containing a thermometer. 2. When the temperature reaches 60° C. add 10 c.c. of a 1 per cent solution of Merck's gold chloride crystals in triply distilled water and 7 c.c. of a 2 per cent solution of Merck's blue label potassium carbonate in triply distilled water. 3. At 80° C., while stirring briskly, add 10 drops of a 1 per cent solution of Merck's blue label oxalic acid crystals, in triply distilled water. 4. At 90° C. remove tha burner and, while stirring, add 5 c.c. of a solution of 1 c.c. of Merck's highest purity formaldehyde in 40 c.c. of triply distilled water, or enough to produce an initial pink color. 5. The solution must be neutral in reaction when used and to determine the reaction it is titrated with a 1 per cent solution of alizarin red, in 50 per cent alcohol. With this indicator the neutral point is shown by a brownish red tint; an acid solution gives a lemon-yellow, and an alkaline solution a purple-red, color. In determining the reaction the following procedure is carried out: In a properly cleaned beaker is placed 10 c.c. of the colloidal gold solu- tion and two drops of the indicator are added. If a brownish red tint appears the solution is satisfactory; if not, the color is noted, and according to N N whether it is acid or alkaline it is titrated with either - NaOH or - HC1. 50 50 If acid the NaOH is used and if alkaline the HC1. When the neutral point is reached; i. e., a brownish red tint appears, the amount of reagent neces- sary to neutralize the 10 c.c. of the gold solution is noted and the amount necessary to neutralize 1000 c.c. is readily calculated. Neutralization should be made w'ith normal or decinormal solutions of the acid or alkali, as the case may be. Kolmer gives the following standards for an efficient col- loidal gold solution: 1. It must be absolutely transparent and of a brilliant red orange or sal- mon red color. (The writer has found that if there is the least trace of cloudiness in the solution, it will be found unserviceable.) 2. Five cubic centimeters of the solution should be completely precipitated in one hour by 1.7 c.c. of a 1 per cent solution of sodium chloride in distilled water. 3. The solution must be neutral in reaction when used. 4. The solution must not produce a reaction greater than a red blue dis- coloration (No. 1 reaction) with normal cerebrospinal fluid, and should give a typical curve with a known paretic fluid. To the above the writer would add that the colloidal gold solution should show no evidence of having been deposited CEREBROSPINAL FLUID 255 on the glass flask in which it is stored or any evidence of a precipitate within the flask. It has been noted by the writer that some glassware is entirely unsuitable for use with colloidal gold as it apparently is attacked by it with a resulting precipitate in the solution or deposited upon the sides and bottom of the storage flask. Jena glassware is the best but Pyrex has been found satisfactory. The colloidal gold should be kept in the flask in the dark and will keep for months unchanged if properly handled. The fact that the colloidal gold solution is perfectly clear does not indicate that it will prove effective in actual prac- tice for it has been found that perfectly clear solutions are sometimes worthless. Such solutions are called "pro- tected solutions" because no precipitation will take place after the addition of any electrolyte, while "unprotected solutions," as they are called, act efficiently in the test, and are those that can be completely precipitated by 1.7 c.c. of a 1 per cent sodium chloride solution in one hour. This test of the efficiency of a colloidal gold solution is very important and should always be carried out before any such solution is depended upon for use in the colloidal gold test. It has also been determined that alkaline col- loidal gold solutions are practically inert when used with a positive cerebrospinal fluid, while acid solutions give little reaction with a positive fluid, and an abnormal reac- tion with a normal spinal fluid. If the colloidal gold so- lution is only slightly acid it will react well with paretic fluids, but will give a reaction with normal fluids similar to that obtained in the luetic zone with fluid from some syph- ilitic conditions. In order to secure reliable results the colloidal gold solution must be neutral in reaction. Method of Performing the Test.-The technic of the col- loidal gold test is very simple but to secure proper results each step must be carefully performed. All glassware must be prepared as already described and pipettes must be perfectly dry before being used. 256 THE WASSERMANN TEST Eleven test tubes, cleaned as described and perfectly dry, are arranged in a rack and in the first tube is placed 1.8 c.c. of freshly prepared, sterile 0.4 per cent sodium chlo- ride solution prepared from chemically pure sodium chlo- ride. Into each of the ten remaining test tubes place 1 c.c. of the salt solution. With a dry pipette add 0.2 c.c. of the cerebrospinal fluid to be tested, which must be free from blood, to the first tube and mix thoroughly. After mixing, 1 c.c. of the mixture is transferred to the second tube, mixed, and 1 c.c. of this mixture transferred to the third tube, and this process repeated until the tenth tube is filled, when from that tube 1 c.c. is discarded. The elev- enth tube acts as a control and will contain no cerebro- spinal fluid. The dilutions which have thus been made are as follows: 1-10, 1-20, 1-40, 1-80, 1-160, 1-320, 1-640, 1-1280, 1-2560, and 1-5120. After the dilutions of the cerebrospinal fluid have been made, as indicated, there is added to each tube 5 c.c. of the colloidal gold solution. This is thoroughly mixed with the fluid in the tubes and the rack kept at room temperature overnight, the test being read upon the next day and re- corded by the numbers corresponding to the changes in the color of the mixture observed, as follows: 5. Complete precipitation. The supernatant liquid is colorless. 4. Pale blue. 3. Blue. 2. Lilac or purple. 1. Red-blue. 0. No change. Reading the Results.-The results of the test are read by the numbers corresponding to the color changes, as noted in the preceding table. Thus, if the first four tubes show complete precipitation, with a colorless supernatant fluid, the fifth and sixth tubes a pale blue color, the seventh tube 1:10 P20 |:M0 U8O IrJA® «====* |:|0 1^20 l-'4O f:80 J-'|6o 1-320 1:640 I • I 280 1: 2560 l: 5120 Paretic Reaction. Tabetic Reaction. Plate III. Paretic Reaction.-The upoer row of tubes illustrates the reaction, or curve, obtained with the colloidal gold test in the cerebrospinal fluid of paresis and is known as the paretic curve or reaction. Tabetic Reaction.-The lower row of tubes illustrates the reaction, or curve, obtained with the colloidal gold test in the cerebrospinal fluid of tabes and is known as the tabetic reaction. (Luetic-zone curve.) CEREBROSPINAL FLUID 257 lilac or purple, and the remaining tubes no change, the reaction would be read as follows: 5555442000. The Results of the Test in Various Syphilitic Condi- tions.-The change in color in the various tubes will differ with different syphilitic conditions of the central nervous system, the color reaction with paresis being the most char- acteristic and diagnostic of all, and the changes observed are called 41 curves. ' ' Thus, we have a ' ' paretic curve ' ' and a "luetic zone curve," as well as a "meningitic zone curve." The "Paretic Curve."-The reaction observed in the cerebrospinal fluid in paresis is so constant and typical that it may be considered as the most valuable of all the confirmatory signs that we possess of this type of involve- ment of the central nervous system. It is to this type of the reaction that Miller and Levy131 gave the name "pa- retic curve" and it is this type of reaction that is met with in practically 100 per cent of cases of paresis. In report- ing the colloidal gold test it is usually customary to record the reaction upon a chart and when this is done, the pa- retic curve will appear substantially as in Table LIV. Table LIV Illustrating Type of Reaction Obtained with the Colloidal Gold Test in Paresis (Paretic Curve) COLOR |COLOR|DILUTIONS OF CEREBROSPINAL FLUID WITH 0.4 PER CENT NACL REACTIONS | no. |l-10|l-20|l-40|l-80|l-160|l-320 1-64011 -128011-256011-5120 Colorless | 5 | 1 1 1 Pale blue 1 4 | | | | | | 1 Blue 1 3 | | | | | | Lilac or purple 1 2 | | | | | | 1 1 Red-blue 1 1 1 1 1 1 1 1 1 1 No change 1 0 | | | | | | i i r* Recorded in numbers this reaction would be thus re- ported: 5555554210. Often the first three, four, or five tubes are decolorized, while the remaining ones show some changes, as would be found in the following combinations: 5554432200; 5555422100, 5555542210, or as many as eight 258 THE WASSERMANN TEST of the tubes may be completely decolorized. Curves of this type are characteristic of paresis and are not found in other syphilitic diseases of the central nervous system, although a few reports are on record of cases of menin- gitis and multiple sclerosis giving similar curves with this test. A paretic curve is also observed in taboparesis. The Luetic-Zone Curve.-The luetic-zone curve is ob- tained in cases of cerebrospinal syphilis and tabes, as well as in the fluid in the tertiary stage of syphilis. In tabes the colloidal gold test usually gives a luetic-zone curve of No. 4 intensity, the precipitation being only partial in the first two or three tubes, complete in the next two or more tubes, and gradually shading off into normal. However, it should be distinctly understood that there is nothing char- acteristic in the curves noted in tabetic patients and there is no constancy in the type of curve observed, except that the "paretic curve" is never noted. Table LV illustrates the ' ' luetic-zone curve ' ' obtained in a case of tabes. Table LV Illustrating Type of Reaction Obtained with the Colloidal Gold Test in Tabes (Luetic Zone Curve) COLOR |COLOR|DILUTIONS OF CEREBROSPINAL FLUID WITH 0.4 PER CENT NACL REACTIONS | no. |l-10| l-20|l-40| l-80|l-160jl-320|l-640|l-1280|l-2560|l-5120 Colorless 5 1 1 1 L-I--l-'vl 1 1 1 Pale blue 4 I L.'| 1 1 I"-' 1 1 1 Blue 3 i i i i i i i \ r i Lilac or purple | 2 1 1 1 1 1 1 1 1 x 1 I Red-blue i i i i i i i i i h No change 0 1 1 1 1 1 1 1 1 1'--1- Recorded by numbers the above reaction would be thus reported: 4445554200. In cerebrospinal syphilis the fluid usually shows weaker reactions than that for tabes but al- ways in the luetic-zone, while the reactions obtained in the fluid during the tertiary stage of syphilis when symp- toms of involvement of the central nervous system are absent are still weaker, only slight changes being observed, as a rule, in the luetic zone of the test. CEREBROSPINAL ELUID 259 In nonsyphilitic disease of the central nervous system, as in meningitis, acute poliomyelitis, multiple sclerosis, etc., the findings with the colloidal gold test are often con- fusing and of little practical value. In meningitis the so- called "meningitic zone" curve is encountered, in which the color changes occur in the highest dilutions, complete precipitation occurring at the opposite end of the chart from paresis. Table LVI illustrates the type of curve ob- tained in tuberculous meningitis or purulent meningitis. Table LVI Illustrating Type of Reaction Obtained with the Colloidal Gold Test in Meningitis (Meningitic Curve) COLOR. | COLOR) REACTIONS | NO. j Colorless | 5 | Pale blue ) 4 | Blue j 3~| Lilac or purple | 2 Red-blue | 1 No change | 0 [DILUTIONS OF CEREBROSPINAL FLUID WITH 0.4 PER CENT NACL |l-10|l-20|l-40|l-80|l-160|l-320|l-640|l-1280|l-2560|l-5120 Recorded by numbers the above reaction would be thus reported: 0001223550. Interpretation of the Results of the Colloidal Gold Test.-There is only one condition in which the colloidal gold test may be practically relied upon for diagnosis and that is paresis. In the overwhelming number of instances, if a typical paretic curve is obtained with this test it is safe to make a diagnosis of this condition and the Wasser- mann test upon the blood and cerebrospinal fluid will gen- erally confirm the diagnosis. However, the writer is still doubtful if we should rely upon the colloidal gold reaction alone in making a diagnosis of paresis, especially if the other diagnostic aids that we possess do not confirm the reaction. It must be admitted, however, that the first evi- dence we have in many cases of incipient paresis is a pos- itive result with this test and whenever such a result is 260 THE WASSERMANN TEST obtained it should be given great weight and the most in- tensive specific treatment at once instituted. In cerebrospinal syphilis and tabes the colloidal gold re- action is of no diagnostic importance and the results obtained should never be considered as more than confir- matory of the diagnosis in cases which are very suspicious in other respects. With the exception of paresis and tabo- paresis the colloidal gold test is of value only as an aid in the diagnosis of syphilis of the central nervous system. In congenital syphilis, cerebrospinal syphilis, tabes, and sec- ondary and tertiary syphilis the results are inconstant and only of confirmatory value, but its appearance in pare- sis, in many cases, before symptoms of the condition de- velop, renders the test of great value in early diagnosis and as this complication may occur in any case of syph- ilitic infection the routine testing of the cerebrospinal fluid with the colloidal gold test is to be encouraged. The Cell Count It is well known that in most inflammatory conditions of the central nervous system there is an increase in the number of cells found in the cerebrospinal fluid and this increase occurs to such an extent in some conditions as to be of considerable service in their differential diagnosis. The cell count, however, so far as syphilitic disease of the central nervous system is concerned, is seldom of more than confirmatory value of results obtained by other meth- ods of laboratory diagnosis. In no syphilitic condition of the central nervous system is the cell count specific nor can one differentiate syphilitic disease from other diseases of this system by its means. However, the value of the cell count, in detecting early involvement of the central nervous system in syphilitic in- fection, can hardly be overestimated, for numerous inves- tigators have shown that an increase in the number of cells in the cerebrospinal fluid is the earliest evidence of infec- CEREBROSPINAL FLUID 261 tion of the central nervous system and that a considerable proportion of patients in the primary, secondary, and early latent stages of the disease, show such an increase. Thus, Wile and Stokes172 found the cell count increased in 2 of 6 cases of primary syphilis, one of four weeks ' duration, and one of only one week's duration, while Andry and Lavan173 observed a primary case with a cell count of 40 cells and With174 observed 4 primary cases with an increased cell count. In this connection the recent observations of Fildes, Par- nail, and Maitland175 are of great interest and value. They examined the cerebrospinal fluid of 1314 cases of syphilis of which 236 showed cell counts above 10, or 18 per cent of those examined. Of these, 46 were in the primary stage of syphilis, 110 in the secondary stage, and 80 in the tertiary or latent stages. In practically all the cases there were no symptoms of involvement of the nervous system and very few showed any signs upon the most careful examination, but a considerable proportion presented lesions of the eye- grounds indicating a commencing retinitis. An increase in the number of cells in the cerebrospinal fluid, called pleocytosis, occurs in practically 95 per cent of syphilitic affections of the central nervous system but also occurs in other diseases of this system, and in acute infections, as cerebrospinal meningitis, although in non- syphilitic conditions the cell count is seldom as high as in conditions like tabes, cerebrospinal syphilis, paresis, and syphilitic meningitis. According to Nonne132 the cell count in the cerebrospinal fluid may be appreciably increased in from 30 to 40 per cent of cases of syphilis even in the ab- sence of symptoms of organic nervous disease. In old cases of tabes, or cases in which the symptoms are not pro- gressive, there may be an absence of pleocytosis and even in progressive cases this may sometimes occur. Efforts have been made to differentiate the kinds of cells increased in the pleocytosis of syphilitic disease of the cen- 262 THE WASSERMANN TEST tral nervous system, and Alzheimer has rendered it pos- sible to recognize the different types of cells that occur. However, from a practical standpoint, the differentiation of the various cells found is of very little, if any, assistance in differentiating the various syphilitic diseases of the nervous system, a simple count of the cells present being as useful as any differential count. As Nonne well says:133 "The attempt to recognize in the spinal fluid a character- istic cell-picture for the development of a paresis, or to foretell the time of progression of a syphilogenetic affec- tion, has not as yet been crowned with success." The cells which are increased in syphilitic disease of the central nervous system comprise both lymphocytes and leucocytes, but their origin is not absolutely determined. Whether they originate from the blood or from the connec- tive tissue of the leptomeninges is unsettled but the pleo- cytosis is the direct result, undoubtedly, of an inflamma- tory process in the leptomeninges. This is the conclusion reached by Nonne and it is believed that it is the one gen- erally held by the best students of the subject. Technic of Counting the Cells.-In counting the cells in the cerebrospinal fluid the most generally used method is with the Fuchs-Rosenthal counting chamber. The method of making the dilution of spinal fluid and counting is as follows: The staining fluid consists of 0.1 gram gentian violet; 2.0 grams glacial acetic acid; and 50 c.c. distilled water. This fluid is drawn up into the ordinary pipette used for counting the white cells of the blood, to the point marked I and then the cerebrospinal fluid drawn up to the point marked II. The pipette is now shaken very thor- oughly for at least two minutes, when a drop of the mix- ture is placed on the counting chamber, a few drops being expelled from the pipette before doing so. All of the lymphocytes and leucocytes are counted in all the squares and the total number divided by three. Nonne regards CEREBROSPINAL FLUID 263 from 0 to 5 cells as normal; 6 to 10 as borderland cases; and more than 10 cells as pathological. The Fuchs-Rosenthal counting chamber is not necessary in making a count of the cells in the spinal fluid, as the Zappert, Turk, or Levy counting chamber may be used just as well, the spinal fluid being undiluted or, if the cells are very numerous, diluted as for making a white blood cell count. The staining of the cells is not necessary if the fluid be diluted with a 0.5 per cent solution in distilled water of glacial acetic acid, when the Fuchs-Rosenthal chamber for counting is used, while if either the Turk or Zappert count- ing chambers are used the fluid should be used undiluted, unless the fluid is purulent or contains an excessive num- ber of cells. The number of cells found in the cerebrospinal fluid in the various syphilitic diseases of the nervous system varies very greatly and no diagnostic data can be based upon the number of the cells alone. The increase in the number of cells is the important diagnostic feature, but this increase is not of such a nature that one can say that paresis exists in this case, and tabes in that. In illustration of this, Table LVII shows the counts observed in paresis, cerebro- spinal syphilis, and tabes. Table LVII Cell Counts in the Cerebrospinal Fluid Observed in Paresis, Cerebro- spinal Syphilis, and Tabes PARESIS CEREBROSPINAL SYPHILIS TABES CASE NO. COUNT CASE NO. COUNT CASE NO. COUNT 1 18 1 13 1 9 2 45 2 64 2 8 3 67 3 243 3 43 4 16 4 11 4 44 5 29 5 86 5 55 6 88 6 65 6 69 7 12 7 50 7 186 8 36 8 590 8 237 9 168 9 344 9 36 10 243 10 156 10 110 11 507 11 6 11 124 12 323 12 402 12 116 264 THE WASSERMANN TEST It will be seen from the above table that there is nothing characteristic in the number of cells found in the cerebro- spinal fluid in the three conditions, the characteristic fea- ture being that there is an increase above normal in all the cases cited. Interpretation of Results of Count.-The increase in the number of cells in the cerebrospinal fluid, as has been stated, is due to inflammatory conditions of the leptomen- inges, but an increase does not always indicate that the condition causing it is syphilitic. Of the syphilitic con- ditions of the central nervous system the highest counts are generally observed in syphilitic meningitis, the count frequently exceeding one thousand cells; in tabes the counts are apt to be high, running into the hundreds; in paresis, the average count is not above 100, generally run- ning between 20 and 60, but counts above 100 are often ob- tained, and the same statement is true of cases of cerebro- spinal syphilis. However, the mere number of cells found is no criterion of the condition present, and the cell count has to be interpreted in conjunction with the clinical symp- toms present and the results of the other methods of lab- oratory diagnosis in syphilis; i. e., the Wassermann test, the colloidal gold test, and the test for increase in proteins in the cerebrospinal fluid. Tests for Increase in Globulins in the Cerebrospinal Fluid.-In syphilitic disease of the central nervous system, as well as in other organic diseases of this system, the pro- tein content of the cerebrospinal fluid is generally increased, and this increase can be detected by tests which have been perfected for demonstrating an increased globulin content in this fluid. Three methods are in general use, known as the Nonne and Apelt method, the Noguchi method, and the Pandy method. The Nonne-Apelt Globulin Test.-This test is thus de- scribed by Nonne:134 "The simplest method of testing the cerebrospinal fluid for albuminous bodies is to search for CEREBROSPINAL FLUID 265 an increase of globulin-the so-called 'Phase I' reaction, introduced by Apelt and myself. For this purpose there is added to a hot saturated solution of sulphate of ammo- nium, which has been permitted to cool, an equal amount of cerebrospinal fluid; 1 c.c. of each is quite sufficient. It is advisable to pour one liquid gently on top of the other; if the globulins are increased, there occurs a more or less distinct gray ring at the plane of contact. After this pre- liminary observation the mixture is well shaken and the result may be read off within three minutes. If it is dis- tinctly opalescent or cloudy, we call it a positive ' Phase I ' reaction. ' ' In making the above test it is best to add the cerebro- spinal fluid from a 1 c.c. pipette very slowly to the ammo- nium sulphate solution, as in this way the gray ring indi- cating increased globulin is rendered much more distinct than when one fluid is poured into the other. This method is still in general use but has been very largely displaced by the Noguchi and Pandy methods. The Noguchi Butyric Acid Test for Globulin.-This method was devised by Noguchi,135 is simple, and accu- rate, and the one that the writer has generally used in his laboratories. The method of making the test is thus de- scribed by Noguchi:13*5 "Two parts of the cerebrospinal fluid to be examined are mixed with 5 parts of a 10 per cent butyric acid solu- tion in physiological salt solution, and are heated over a flame and boiled for a brief period. One part of a normal solution of NaOH is then added quickly to the heated mix- ture, and the whole boiled once more for a few seconds. The actual quantities of these three reagents that I pre- fer are 0.1 or 0.2 c.c. of the spinal fluid, 0.5 c.c. of the butyric acid solution, and 0.1 c.c. of the normal sodium hydrate. It is necessary to take the precaution to employ for this test only cerebrospinal fluid entirely free from blood." 266 THE WASSERMANN TEST A positive reaction with this test is indicated by the ap- pearance of a granular or flocculent white or gray precip- itate which settles slowly to the bottom of the test tube. The amount of this precipitate and the rapidity with which it is formed will vary, as Noguchi observed, with the quan- tity of protein in the fluid examined, the greater the amount of protein the quicker the reaction will appear and the greater the amount of precipitate, in the one case the reaction appearing within a few minutes, while, if the amount of protein increase is small the reaction may take an hour or more to appear. Two hours should be the time limit for the reaction. The Pandy Method for Globulin Increase.-A method which is spoken very highly of by those who have used it is the Pandy method. It is the most simple of all methods that have so far been devised for the purpose and is un- doubtedly reliable. Miller,137 who has had a large experi- ence with this method thus describes its use: " Pandy's test has not received the attention it deserves. None of the other reactions used to reveal an excess of globulin is so simple in execution or so quickly decisive in its results. The reagent consists of a saturated aqueous solution of carbolic acid; ten parts of pure crystals are added to 100 parts of hot distilled water; the mixture is kept at room temperature for three or four days, during which time it should be frequently shaken. At the end of this time the clear supernatant fluid is drawn off into an- other bottle. To approximately 1 c.c. of the reagent is added one drop of the spinal fluid. Normally no change occurs or at most an extremely faint opalescence; with a fluid abnormal in its protein content there develops in- stantly at the point of contact a bluish white cloud, often resembling a ring of smoke, which gradually settles to the bottom of the tube." The writer has had very little experience with this test but it has proved accurate in those cases in which it has been used. ✓ CEREBROSPINAL FLUID 267 Interpretation of the Results of the Globulin Tests.-In organic disease of the central nervous system an increase in the protein content of the cerebrospinal fluid commonly occurs and this increase is present in the vast majority of cases of syphilitic disease of this system. However, as this protein increase, indicated by the increase in globu- lins, is not confined to syphilitic disease, a positive reac- tion with any of the tests mentioned does not prove that the condition present is syphilitic nor is it possible to dif- ferentiate the various syphilitic diseases of the central nervous system by their use. The tests are often positive in infections of the meninges due to the pneumococcus, tu- bercle bacillus, influenza bacillus, meningococcus, and in acute poliomyelitis, as well as in syphilitic disease, and are also present in chronic organic disease of the central nervous system, so that a positive reaction can only be re- garded as confirmatory of the results of other laboratory methods of diagnosis of syphilitic involvement of this sys- tem and of the nature of the clinical symptoms which may be present. Summary of the "Four Tests."-In diagnosing and in differentiating syphilitic disease of the central nervous system, it is necessary to carefully consider the results of the "Four tests" as they may be called, upon the cerebro- spinal fluid. These are 1, the Wassermann test upon the fluid; 2, the colloidal gold test; 3, the cell count; 4, the test for increase in globulins. As an illustration of the results Table LVIII Showing Average Frequency of the Various Reactions in Syphilis of the Central Nervous System TEST PARESIS % TABES % CEREBROSPINAL SYPHILIS % Spinal fluid Wassermann 97 60-89 85-90 Globulin test 100 90-95 90-95 Cell count 98 85-90 85-90 Colloidal gold test 98-100 85-90 85-90 Paretic Luetic Luetic curve curve curve 268 THE WASSERMANN TEST that are commonly obtained with these four tests, Table LVIII, adapted from Miller138 is inserted. The blood Wassermann in these diseases gives approx- imately the following results: Paresis, 98-100 per cent; tabes, 70 per cent; and cerebrospinal syphilis, 70-80 per cent. A careful study of the cerebrospinal fluid by means of these four reactions will often result in enabling one to differentiate the various syphilitic diseases of the central nervous system and, in the vast majority of our cases, if the tests are negative, to exclude syphilitic disease in a suspected individual. If the globulin test is negative in a patient suffering from disease of the central nervous sys- tem it is almost certain that the condition is not syphilitic, although, if positive, it can not be said that it is syphilitic. A fluid showing a positive Wassermann reaction in amounts as small as 0.2 c.c., together with a paretic curve with the colloidal gold test, is sufficient upon which to base a diagnosis of paresis, while a high cell count, a luetic curve with the colloidal gold test, and a negative Wassermann, or one only positive in large amounts of the spinal fluid, would point to tabes. By a careful compar- ison of the results of these various tests with the clinical symptoms present it should be possible, in the vast ma- jority of instances, to make a diagnosis of the form of syphilitic disease of the central nervous system which may be present, and the value of these tests in diagnosing very early involvement of this system has already been empha- sized. In no case can it be stated that the syphilitic patient has been given the best that it is possible for medical science to give him, either in the way of diagnosis or treatment, if the thorough examination of the cerebrospinal fluid has been omitted and this examination should become as much of a routine in the diagnosis and treatment of syphilis as is the Wassermann test upon the blood serum. BIBLIOGRAPHY The references given below do not pretend to cover the literature of the subject, which is now immense, but only cover the papers actually referred to in the text. iNuttall: Ztschr. f. Hyg., 1888, iv, 353. aFodor, von: Deutsch, med. Wchnschr., 1888, vi, 460. 3Buchner: Centralbl. f. Bakt., 1889, v, 252. 4Buchner: Arch. f. Hyg., 1891, x, 84. 5Pfeiffer: Ztschr. f. Hyg., 1894, xvii, 355. 6Bordet: Ann. d. 1'Inst. Pasteur, 1898, xii, 688; Ibid., 1899, xiii, 682. ^Bordet: Studies in Immunity, Trans., Gay, New York, 1909. sBelfanti and Carbone: Gior. d. r. Accad. di. med. di Torino, July, 1898. 9Bordet and Gengou: Ann. d. 1'Inst. Pasteur, 1901, xv, 289. i°Wassermann, Neisser, and Bruck: Deutsch, med. Wchnschr., 1906, xxxii, 745. uWassermann, Neisser, Bruck, and Schucht: Ztschr. f. Hyg. u. Infections- krankh., 1906, Iv, 451. i2Detre: Wien. klin. Wchnschr., 1906, xix, 619. isWeygandt: Munchen, med. Wchnschr., 1907, liv, 1557. i*Marie and Levaditi: Ann. d. 1'Inst. Pasteur, 1907, xxi, 138. i5Plaut: Centralbl. f. Nervenh. u. Psychiat., 1908, xxxi, 289. isLandsteiner: Centralbl. f. Bakt., orig. abstr. 1, 1906, p. 42. iTLandsteiner, Muller and Potzl: Wien. klin. Wchnschr., xx, 1421, 1565. isNoguchi: Jour. Exper. Med., 1909, xi, 84. isBrowning, Cruickshank, and McKenzie: Biochem. Ztschr., 1910, xxv, 85. 2oWalker and Swift: Jour. Exper. Med., 1913, xviii, 566. 2iNoguchi: Serum Diagnosis of Syphilis, ed. 3, Philadelphia, 1912, p. 159. 22Craig and Nichols: Jour. Exper. Med., 1912, xvi, 336. asNoguchi: Jour. Exper. Med., 1909, xi, 84. 24Citron: Immunity, Trans., Garbat, ed. 2, Philadelphia, 1914, p. 166. 2i5Schmidt: Ztschr. f. Hyg. u. Infectionskrankh., 1911, Ixix, 186. 26Levaditi and Yamanouchi: C. R. Soft., Biol., 1907, Ixiii, 740. 2'TWeil and Braun: Wien. klin. Wchnschr., 1908, xxi, 151. 28Bergel: Ztschr. f. Immun., 1913, xvii, 234. as Manwaring: Ztschr. f. Immun., 1909, iii, 309. soBruck and Stern: Ztschr. f. Immun., 1910, xxiii, 534. 3iPeritz: Ztschr. f. Path. u. Ther., 1908-1909, v, 607. 32Schereschewsky: Deutsch, med. Wchnschr., 1909, xxxv, 1652. 33Noguchi: Jour. Am. Med. Assn., 1912, Iviii, 1163. 3<Noguchi: Ibid. 35Noguchi: Serum Diagnosis of Syphilis, ed. 2, Philadelphia, 1912, p. 169. 36Craig and Nichols: Jour. Exper. Med., 1912, xvi, 336. 37Kolmer, Williams, and Laubaugh: Jour. Med. Research, 1913, xxviii, 345. 38Craig and Nichols: Jour. Am. Med. Assn., 1911, Ivii, 474. 39Craig: Jour. Exper. Med., 1911, xiii, 521. 4oCraig: Jour. Am. Med. Assn., 1914, Ixii, 1232. 4iNoguchi: Serum Diagnosis of Syphilis, ed. 3, Philadelphia, 1912, p. 72. 42Kolmer: Infection, Immunity, and Serum Therapy, ed. 2, Philadelphia, 1917, p. 437. 43Noguchi: Jour. Exper. Med., 1909, xi, 392. 269 270 THE WASSERMANN TEST ^Noguchi: Serum Diagnosis of Syphilis, ed. 3, Philadelphia, 1912, p. 75. 45Marie and Levaditi: Am. de 1'Inst. Pasteur, 1907, xxi, 138. ^Landsteiner, Muller and Potzl: Wien. klin. Wchnschr., 1907, xx, 1421-1565. 4?Porges and Meier: Berl. klin. Wchnschr., 1908, xlv, 731. ^sBrowning, Cruickshank, and McKenzie: Biochem. Ztschr., 1910, xxv, 85. 49Vedder: Bull. No. 8, War Dept., Office Surgeon-General, U. S. A., June, 1915. "Noguchi: Serum Diagnosis of Syphilis, ed. 3, Philadelphia, 1912, p. 95. siNonne: Syphilis and the Nervous System, Trans., Ball, Philadelphia, 1913. 52Noguchi: Serum Diagnosis of Syphilis, ed. 3, Philadelphia, 1912, p. 59. 53Hecht and Weinberg: Ztschr. f. Immun., 1910, xxxvi, 317. 54Gradwohl: Am. Jour. Syph., 1917, i, 450. 55Gradwohl: Ibid., 461. ssKolmer: Infection, Immunity, and Serum Therapy, ed. 2, Philadelphia, 1917, p. 459. 57Tschernogubow: Berl. klin. Wchnschr., 1908, xlv, 2107. ssDetre and Brezovsky: Wien. klin. Wchnschr., 1908, xxi, 1700. "Stern, Margarette: Berl. klin. Wchnschr., 1908, xlv, 1489. eoBauer: Deutsch, med. Wchnschr., 1908, xxxv, 698. clBrowning and McKenzie: Ztschr. f. Immun. Orig., 1909, ii, 459. eaThompson: Am. Jour. Syph., 1917, i, 555. 63Zinsser: Infection and Resistance, New York, 1917, 208. e^Smith and MacNeal: Jour. Infect. Dis., 1917, xxi, 233. 63Vedder: Bull. No. 8, War Dept., Office Surg. General, U. S. A., 1915, p. 109. "Schereschewsky: Deutsch, med. Wchnschr., 1909, xxxv, 1652. erNoguchi: Jour. Am. Med. Assn., 1912, Iviii, 1163. 68Craig and Nichols: Jour. Exp. Med., 1912, xvi, 336. 69Craig and Nichols: Bull. No. 3, War Dept., Office Surg. Genl., 1913, 66. "°Kolmer, Williams, and Laubaugh: Jour. Med. Research, 1913, xxviii, 345. ■^Zinsser, Hopkins, and McBurney: Jour. Exper. Med., 1916, xxiv, 561. 72Noguchi and Akatsu: Jour. Exper. Med., 1917, xxv, 765. 73Vedder: Bulletin No. 8, War Dept., Office Surg. Genl., U. S. A., 1915. 74Vedder: Ibid., 96. 75Craig: Jour. Exp. Med., 1910, xii, 726. "6Kolmer: Infection, Immunity, and Specific Therapy, ed. 2, Philadelphia, 1917, p. 486. 77Vedder: Bull. No. 8, Wrar Dept., Office Surgeon General, U. S. A., 1915, p. 9. 78Woods and Morris: Jour. Infect. Dis., 1918, xxii, 43. 79Kolmer: Infection, Immunity, and Serum Therapy, ed. 2, Philadelphia, 1917, p. 486. 8°Boas: The Wassermann Reaction, Berlin, 1911. 81Vedder: Bull. No. 8, War Dept., Office Surgeon General, U. S. A., 1915, p. 9. 82Vedder: Ibid. 9. 83Vedder: Ibid. 9. 84Boas: The Wassermann Reaction, Berlin, 1911. 83Veeder: Am. Jour. Med. Sc., 1916, clii, 522. 88Hecht, Lateiner, and Wilenko: Wien. klin. Wchnschr., 1909, xxii, 523. 87Kolmer: Infection, Immunity, and Serum Therapy, ed. 2, Philadelphia, 1917, 491. 88Browning and McKenzie: Diagnosis and Treatment of Syphilis, Phila- delphia, 1912, 100. 89Thompson: Syphilis, Diagnosis and Treatment, Philadelphia, 1916, 157. "Vedder: Bull. No. 8, War Dept., Office Surgeon General, U. S. A., 1915, p. 10. 91Vedder: Ibid., 109. BIBLIOGRAPHY 271 92Southard: Boston Med. and Surg. Jour., 1914, clxx, 25. "Hammond: Am. Jour. Insan., 1913, Ixx, 107. 9<Thompson: Syphilis, Diagnosis and Treatment, Philadelphia, 1916, p. 27. 95Rosenberger: New York Med. Jour., 1917, cv, 1233. "Johnson: Am. Jour. Syph., 1917, i, 606. "Vedder: Bull. No. 8, War Dept., Office Surgeon General, U. S. A., 1915. "Irvine: Journal-Lancet, April 1, 1917. "Peterson: Am. Jour. Syph., 1917, i, 211. 10(>Ladd: New York Med. Jour., 1917, cvi, 1211. loiCraig: Jour. Am. Med. Assn., 1913, lx, 565; Bull. No. 3, War Dept., Office Surgeon General, U. S. A., 1913, 80. io2Nichols and Fordyce: Jour. Am. Med. Assn., 1910, Iv, 1171. 193Michaelis: Berl. klin. Wchnschr., 1910, xliii, 183. io4Fordyce: New York Med. Jour., 1911, xcii, 897. losNoguchi: Serum Diagnosis of Syphilis, ed. 3, 1912. loeCraig: Bull. No. 3, War Dept., Surgeon General Office, U. S. A., 1913, p. 99. i"Gennerich: Berl. klin. Wchnschr., 1910, Sept. 19, No. 38. i"Milian: Paris, Dermatological Gesellschaft, Dec. 1, 1910. i"Herxheimer: Deutsch, med. Wchnschr., 1910, xxxvii, 1517. iioMichaelis: Berl. klin. Wchnschr., 1910, xlv, 1195. mStokes and O'Leary: Am. Jour. Syph., 1917, i, 629. n2King: Jour. Am. Med. Assn., 1916, Ixvii, 1669. nsNichols: Bull. No. 3, War Dept., Office Surgeon General, U. S. A., 1913, p. 28. 114Stokes and O'Leary: Am. Jour. Syph., 1917, i, 629. iisFordyce: Med. Record, New York, 1917, xci, 927. 116Hauptmann and Hossli: Munchen med. Wchnschr., 1910, Ivii, 1581. n'TMueller: Die Serodiagnose der Syphilis, Berlin, 1913. nsFordyce: Med. Record, New York, 1917, xci, 927. nsRavaut: Ann. de dermat. et de syph., 1903, iv, 1. i20Wile: Quoted by Kolmer. i^iOaig and Collins: Jour. Am. Med. Assn., 1914, Ixii, 1955. i22Ellis and Swift: Jour. Exper. Med., 1913, xviii, 162. !23Altman and Dreyfus: Munchen med. Wchnschr., 1913, lx, 464. i24Kolmer: Infection, Immunity, and Serum Therapy, ed. 2, Philadelphia, 1917, p. 488. i25Lange: Ztschr. f. Chemotherap., 1912-1913, i, 44; Berl. klin. Wchnschr., 1912, xlix, 897. i26Zsigmondy: Ztschr. f. anal. Chem., 1901, xl, 697. i27jaeger and Goldstein: Ztschr. f. d. ges. Neur. u. Psych., 1913, xvi, 391. i28Zaloziecki: Deutsch, ztschr. f. Nervenh., 1913, xlvii, 783. i29Kolmer: Infection, Immunity, and Serum Therapy, ed. 2, Philadelphia, 1917, p. 556. i3oMiller, Brush, Hammers, and Felton: Bull. Johns Hopkins Hosp., 1915, xxvi, 391. i3iMiller and Levy: Bull. Johns Hopkins Hosp., 1914, xxv, 123. !32Nonne: Syphilis and the Nervous System, Trans. Ball, Philadelphia, 1913, p. 345. k i33Nonne: Ibid., 346. i34Nonne: 341. i35Noguchi: Proc. Soc. Exper. Biol, and Med., 1909, vi, 51. i36Noguchi: Serum Diagnosis of Syphilis, ed. 3, Philadelphia, 1912, p. 223. i37Miller: Southern Med. Jour., 1915, viii, 940. i38Miller: Ibid. issKyntoku: Jour., Immunology, 1919, iv, 239. J4oKolmer, Matsunami, and Trist: Am. Jour. Syph., 1919, iii, 407. i4iWilliams: Jour. Exper. Med., 1920, xxxii, 159. 272 THE WASSERMANN TEST i-*Kolmer, Matsunami, and Trist: Ibid. i43Sands and West: Jour. Immunology, 1919, iv, 274. i44Williams: Ibid. i45Ruediger: Jour. Infec. Dis., 1919, iii, 213. i46Noguchi: Jour. Exper. Med., 1918, xxviii, 43. i47Simon, C. E.: Jour. Am. Med. Assn., 1919, Ixxii, 1535. i4»Kolmer and Rule: Am. Jour. Syph., 1920, iv, 154. i49Brown and Kolmer: Am. Jour. Syph., 1919, iii, 8. isoManwaring: Jour. Infec. Dis., 1904, i, 112. isiHektoen and Ruediger: Jour. Infec. Dis., 1904, i, 379. i52Cumming: Jour. Infec. Dis., 1916, xviii, 151-179. issRuediger: Jour. Infec. Dis., 1919, xxiv, 43. i54Noguchi: Jour. Exper. Med., 1918, xxvii, 43. i55Thompson: Am. Jour. Syph., 1917, i, 555. isoBartlett and O'Shansky: Am. Jour. Syph., 1917, i, 776. isiKolmer and Flick: Am. Jour. Syph., 1919, iii, 541. issRuediger: Jour. Infec. Dis., 1919, xxiv, 405. isoKlauder: Jour. Am. Med. Assn., 1919, Ixxii, 694. looTurnbull: Med. Research. Com., Sp. Rep. Ser. No. 47, 1920, p. 89. loiMcConnell: Jour. Lab. and Clin. Med., 1919, v, 43. i62Levin: Jour. Lab. and Clin. Med., 1919, v, 93. i63Warthin: Am. Jour. Syph., 1919, iii, 425. i64Nelson and Haines: Jour. Am. Med. Assn., 1914, Ixiii, 2277. lesNelson and Anderson: Jour. Am. Med. Assn., 1915, Ixv, 1905. leepease: Med. Record, 1914, Ixxxv, 982. io7Carnaz: Jour. Nerv. and Mental Dis., 1919, xlix, 282. losMarinesco and Minea: Acad, de Soc. Sem. Med., 1914, 357. isoArtz and Kerl: Wien. klin. Wchnschr., 1914, xxvii, 787. i70Fruhwald and Zaloziecki: Berl. klin. Wchnschr., 1916, liii, 1-28. i7iFildes, Parnall and Maitland: Med. Research Com., Sp. Rep. Ser. 45^ 1920. i72Wile and Stokes: Jour. Am. Med. Assn., 1915, Ixiv, 979. J73Audry and Lavan: Ann. d. Dermat. et de Syph., 1914, xv, 5, 29. i74With: Brain, 1918, lx, 403. i75Fi]des, Parnall, and Maitland: Ibid. 176Q'Leary: Arch. f. Dermat. u. Syph., 1920, ii, 348. urStokos and Busman: Am. Jour. Med. Sc., 1920, cxl, 658. INDEX A Alcohol, influence of, 40 Agglutinins, 73 Amboceptor: character of, 21, 69 definition of, 21 filtration of, 74 hemolytic, 23 preparation of, 68, 72 preservation of, 73 titration of, 69, 91 Anesthesia, Wassermann reaction in, 183 Antieomplementary bodies, filtration of, 49 Antigen, acetone insoluble lipoids as, 28, 83 alcoholic extracts as, 28, 82 aqueous extracts as, 28, 84 character of, 21 chemicals as, 28 cholesterinized, 82, 85 cultures of T. pallidum as, 29, 37, 146 definition of, 21 dilution of, 94, 109 preparation of, 81 titrations of, 93 Apparatus, for Wassermann test, 61 Arsphenamine and Wassermann test, 206 Asepsis, importance of, 50 B Bacillus typhosus, 25, 154 extracts of, 38, 154 as antigen, 154 Bacteria, action of, 45 Bacteriolysin, 22 Bacteriolysis, 17 definition of, 21 discovery of, 17 nature of, 20 production of, 21 specific nature of, 20 Bacteriolytic system, 22 Blood cell suspension, preparation of, 76 Blood corpuscles, washing of, 76 Blood groups, importance of, 76 Blood grouping, technic of, 78 Blood serum: bacteriolytic substances in, 17 collection of, 103 colloids in, 35 hemolytic substances in, 22, 72 inactivation of, 20, 106 lipotropic substances in, 34, 36 thermolabile substances in, 20 thermostable substances in, 20 titration of, 50, 144 variations in strength of, 50, 144 Bordet-Gengou phenomenon, 24 Butyric acid test, 265 C Carcinoma, Wassermann test in, 178, 181 Cell count, 244, 260 results of, 263 technic of, 262 value of, 260 Cerebrospinal fluid, 111, 242 collection of, 111 in latent stages of syphilis, 248 in primary stage of syphilis, 246 in secondary stage of syphilis, 246 in tertiary stage of syphilis, 247 interpretation of results of test, 248 Wassermann test on, 111, 246 Cerebrospinal syphilis, 245, 249 cell count in, 264 colloidal gold test in, 258 globulin tests in, 267 Wassermann test in, 245, 249 Cholera spirillum, bacteriolysis of, 18 extracts of, as antigen, 154 Cholesterinized antigens, 85 Colles' law, 176 Colloidal gold test, 251 in cerebrospinal syphilis, 258 in paresis, 258 273 274 INDEX Colloidal gold test-Cont'd. in tabes, 258 interpretation of results of, 259 results of, 257 technic of, 252, 255 Colloids, 35 Complement, 22 character of, 22 definition of, 22 preparation of, 63 preservation of, 67 titration of, 86 Complement fixation, 25 alcohol and, 40 discovery of, 24 illustration of, 25 in man, 27 in monkeys, 27 nature of, 25, 34, 36 Wassermann test and, 26 Cultures, as antigens, 146 Cytolysins, 23 D Dark field, diagnosis with, 160, 167 Dementia paralytica, Wassermann test in, 245, 249 Diabetes mellitus, Wassermann test in, 131 E Extracts, as antigens, 28 alcoholic, 28, 82 aqueous, 28, 84 cholesterinized, 82, 85 culture, 146 human heart, 82 F Filter paper, amboceptor, 72 Advantages of, 73 impregnation of, 73 preservation of, 73 titration of, 98 Frambesia, Wassermann test in, 179 G Glass-ware, preparation of, 101 Globulin tests, 264 Nonne-Apelt, 264 Noguchi, 265 Pandy, 228, 266 H Hemolysin, definition of, 22 Hemolysis, 22 Hemolysis-Cont'd. discovery of, 22 nature of, 22 Hemolytic amboceptor, 23 preparation of, 68, 72 preservation of, 73 titration of, 69, 93 Hemolytic antigen, 76 Hemolytic system, 24 I Ice-box incubation, 141 Inactivation of blood serum, 20, 106 Interpretation of reaction, 44, 184, 230, 248 Isohemolysins, 76 L Leprosy, Wassermann test in, 179 Luetic zone curve, 258 Lumbar puncture, 112 precautions in, 114 M Malaria, Wassermann test in, 179 Materials used in Wassermann test, 61 Mercury, effect of, on Wassermann test, 217, 222 Modifications of Wassermann test, 104, 118 N Noguchi's butyric acid test, 265 Noguchi's modification, 132, 139 Nomenclature of results of Wasser- mann test, 115 Normal salt solution, preparation of, 63 P Pandy's test, 266 Paresis, cell count in, 263 colloidal gold test in, 257 globulin tests in, 267 Wassermann test in, 245, 249 Paretic curve, 257 Preparation of antigen, 81 of amboceptor, 68 of blood corpuscles, 76 of blood serum, 103 of complement, 63 Profeta's law, 176 Provocative Wassermann test, 233 cause of, 240 indications for, 237 INDEX 275 Provocative Wasserman test-Cont'd. results of, 234 technic of, 233 Pulmonary tuberculosis, Wassermann test in, 179 R Reaction, time of appearance of, 163 delayed reaction, 196 time of disappearance of, 210 Reagents for Wassermann test, 62 preparation of, 63 titration of, 86 Relapse in syphilis, 226 time of, 227 Relapse in Wassermann reaction, 226 Relapsing fever, Wassermann test in, 179 Results of Wassermann test, 157, 184 S Salvarsan, effect of, on Wassermann test, 205 method of administration of, 205 Scarlet fever, Wassermann test in, 178 Septicemia, Wassermann test in, 178 Spirillum cholerae, bacteriolysis of, 18 cultures of, as antigens, 38, 154 Spirochaeta refringens, 38, 154 Staphylococcus albus, 46 aureus, 46 citreus, 46 Streptococcus pyogenes, 46 Syphilis, cell count in, 244 colloidal gold test in, 251 globulin test in, 264 prevalence of, 197 relapse in, 226 variations in blood serum in latent stage of, 56 in blood serum in primary stage of, 53 in blood serum in secondary stage of, 54 Wassermann test in congenital, 175 in latent stage of, 172, 248 in primary stage of, 160, 246 in secondary stage of, 169, 246 in tertiary stage of, 171, 247 T Tabes, cell count in, 263 globulins in, 267 colloidal gold test in, 258 Wassermann test in, 245, 249 Temperature, relation of, 48, 141 Titration of reagents, 86 of blood serum, 51, 58 Toxolipoids, 34 Treponema calligyrum, 155 mierodentium, 38, 149 mucosum, 155 pallidum, 27, 149 antigens prepared from, 37, 146 cultures of, 37, 146 discovery of, 27 pertenue, 38, 149 Tuberculosis, Wassermann test in, 178 Typhoid complement fixation in, 25 W Wassermann fast cases, 231 Wassermann test, 116 alcohol and, 40 amboceptor for, 73, 91 antigens for, 27, 28, 81, 146 apparatus used in, 61 cerebrospinal fluid and, 111, 246 discovery of, 26 factors influencing results of, 40 general description of, 31 influence of bacteria on, 45 influence of variation in blood se- rum on, 50, 144 interpretation of results of, 44, 184 after specific treatment, 194, 230 in cerebrospinal fluid, 248 in control of treatment, 230 in latent stage, 193 in nervous diseases, 194 in primary stage, 187 in secondary stage, 190 in tertiary stage, 191 modifications of, 118 Gradwohl's, 135 Hecht-Weinberg's, 134 Hygienic Laboratory, 118 Kolmer's, 137 Noguchi's 132, 139 nature of, 27, 30, 33, 36, 38 original technic of, 116 prevalence of syphilis and, 197 provocative, 233 results of, 110, 157 in congenital syphilis, 175 in latent syphilis, 172, 248 in primary syphilis, 160, 246 in secondary syphilis, 169, 246 in tertiary syphilis, 171, 247 276 INDEX Wassermann test-Cont'd. specific antigens in, 28, 33, 36, 146 specificity of, 177 strength of, variations in, 50, 144 technic of, 107 titration of reagents used in, 86 Wassermann test-Cont'd. treatment and, 205 control of, with, 223 Y Yaws, Wassermann test in, 179 INDEX OF AUTHORS A Akatsu, 155 Altman, 143, 247 Alzheimer, 262 Anderson, 222 Apelt, 264, 265 Arning, 162, 170, 171, 175 Artz, 243, Audry, 261 B Bartlett, 140 Bauer, 141 Belfanti, 22 Bergenon, 202 Bergel, 35 Bering, 162, 170, 171 Blumenthal, 170 Boas, 162, 169, 245 Bordet, 20, 22, 24, 25, 26, 30 Braun, 35 Brezovsky, 141 Bronfenbrenner, 64 Brown, 102 Bruck, 26, 27, 35, 162, 170 Brush, 255 Browning, 28, 35, 141, 178 Bushner, 18 Busman, 232 C Carbone, 22 Carnoz, 243 Citron, 34, 35 Collin, 247 Craig, 29, 38, 40, 41, 45, 50, 91, 99, 146, 148, 154, 162, 170, 171, 173, 175, 225 Cruickshank, 28, 85 Cumming, 102 D Detre, 27, 141 Dreyfus, 247 E Ehrlich, 18, 21 Ellis, 247 F Felton, 255 Fildes, 243, 261 Fischer, 162 Fleishman, 171 Flick, 140 Fodor, von, 18 Fordyce, 207, 243 Fox, 173 Fruhwald, 243 G Gengou, 24, 25, 26, 30 Gennerich, 233 Goldstein, 252 Gradwohl, 135 Grosser, 162, 173 H Haines, 222 Hammers, 255 Hammond, 198 Hanchen, 162 Hauptmann, 244 Hecht, 134, 178 Hektoen, 102 Herxheimer, 233 Hohne, 170 Hopkins, 155 Hossli, 244 I Irvine, 202 J Jacobstahl, 143 Jaeger, 252 Jesionek, 175 Johnson, 198 277 278 INDEX OF AUTHORS K Kaplan, 162 Kerl, 243 King, 234 Klauder, 167 Kolmer, 38, 42, 64, 67, 101, 102, 137, 140, 155, 169, 178, 247, 255 Kyntoku, 49 L Ladd, 203 Landois, 22 Landsteiner, 28, 85 Lange, 251 Lateiner, 178 Laubaugh, 38, 155 Lavan, 261 Lederman, 170, 173 Leredde, 143 Lepine, 202 Lesser, 170, 171 Letulle, 202 Levaditi, 28, 35, 85 Levin, 201 Levy, 257 M McBurney, 155 McConnell, 196 McKenzie, 35, 141, 178 MacNeal, 142 Maitland, 243, 261 Manwaring, 35, 102 Marie, 28, 85 Marinesco, 243 Matsunami, 64, 67 Meier, 85, 162 Meierosky, 175 Merz, 170, 171 Michaelis, 207, 233 Milian, 233 Miller, 255, 266, 268 Minea, 243 Morris, 168 Muller, 28, 85, 245 Munck, 34 N Neill, 67 Neisser, 26, 37 Nelson, 222 Nichols, 20, 38, 40, 41, 146, 149, 154, 207, 234 Noguchi, 28, 29, 34, 36, 37, 62, 67, 71, 88, 100, 132, 139, 147, 151, 155, 162, 170, 173, 175, 207, 265 Nonne, 261, 264 Nuttall, 17 O O 'Leary, 59, 234, 235, 239 O'Shansky, 140 Ottenberg, 143 P Tandy, 266 Parnall, 243, 261 Pease, 238 Peritz, 35 Peterson, 202 Pfeiffer, 18, 19 Plaut, 28 PORGES, 85 Potzl, 28, 85 R Ravaut, 247 Roscher, 170 Rosenberger, 198 Rubenstein, 143 Ruediger, 86, 102, 109 Rule, 101 8 Sands, 73 Schaudinn, 27 Schmidt, 34 SCHERESCHEWSKY, 36, 146 SCHONNEFELD, 170 Schucht, 27 Simon, 100 Smith, 142 Southard, 198 Stern, 35, 141, 162, 170, 171 Stokes, 232, 234, 245, 261 Swift, 29, 162, 170, 173, 247 T Thompson, 67, 140, 141, 180, 198 Trist, 64, 67 Tschernogubow, 141 Turnbull, 183 V Vedder, 99, 157, 158, 162, 164, 169, 170, 173, 175, 182, 199, 202 Veeder, 175 INDEX OF AUTHORS 279 W Walker, 29 Warthin, 172, 181, 183 Wassermann, 26, 27, 29, 30, 107 Weil, 35 Weinberg, 134 West, 73 Weygandt, 28 Wile, 247, 261 WlLENKO, 178 Williams, 38, 64, 72, 76, 155 Witte, 261 Woods, 168 Y Yamanouchi, 35 Z Zaloziecki, 243, 252 Zinsser, 141, 155 Zsigmondy, 251