PREPARATION OF JAPANESE VAC CUTE, CHICK TYPE DRIED, FOR THE U,S*AHMY, 1947 * by The Staffs of the The Department of Virus and Rickettsial Diseases and The Veterinary Division Army, Medical Department Research and Graduate School Washington 12* D*C« * Information contained in this paper has already been reported to the Commission on Immunisation, Army Epidemiological Board, v* Work reported hare resulted from the ocabined efforts of numerous people at the Army Medical Department Research and Graduate School# It is not feasible to mention each individual and his exact contri- bution, since many provided part-time assistance in the preparation of vaccine in addition to performing all of their regular duties* This report was prepared by Joseph E* Smadol, Raymond Randall, and Joel Warren, August 7, 1947© In 1945 and 1946 military personnel in certain areas of Japan, Okinawa and Korea were vaccinated against Japanese encephalitis® A formalinised vaccine commercially prepared from infected mouse brain was used during both years and in addition, a formalinized vaccine prepared from infected chick embryos was employed in 1946* The latter had boon prepared on a develop- mental basis Vt the Army nodical Department Research and Graduate School® House brain (1) and chick embryo vaccines (2) induced in nice a comparable resistance to infection with the virus of Japanese encephalitis* Furthemore, antibodies capable of neutralizing the Japanese virus have been demonstrated in the sera of an equal proportion of human subjects following vaccination with either of the two types of vaccines (3)® In the fall of 1948 it become evident that any vaccine used in the Pacific Theater during the 1947 season would of necessity bo supplied by the Army Hodical Department Research and Graduate School because commercial production had been discontinued early in j A 1946* Facilities for the manufacture of the mouse brain type wore not avail- able at the School* Japanese encephalitis vaccine, chick embryo type, fluid, was not entirely suitable since it was known to lose its immunogenic potency when stored at 5° C® for longer than one month (4)® Various procedures v/oro investigated in an attempt to find a means for retarding this deterioration® Only two of the methods tested provided encouraging results* Urea added to the crude vaccine was moderately effective as a stabilizing agent but vaccines containing Z% urea were no longer of acceptable potency after storage for two to three months* The second method, drying from the frozen state, appeared to offer the best solution to the problem of deterioration® This report deals with the methods employed for tho production of lyophilized Japanese encepha- litis vaccine, chick embryo type, and provides data on certain characteristics of the finaT- product* miSPARATIOH OF THE VAOCIME General«, The vaccine* prepared from the strain* is an uncentrl i iged 20$ suspension of infected chicle enbryo tissue in buffered physiolcgical saline solution containing sufficient formaldehyde to inactivate the virus* then the fresh vaccine Is no longer infectious the residual formaldehyde is neutralised with sodium ■ bisulfite and the fluid preparation is dried from the frozen state under carefully controlled conditions* Preparation of Seed Ixrcula.* The *Nakayama* strain, of virus used for the preparation of vaccine was brought to the School in 1942 as infected mouse brain* For the past two and one-half years a line of this strain has boon ina.inta5.nod by serial transfer in ecibryorated eggs* . The seed virus consists of a 20$ concentration of pooled infected embryos (usually three per pod) which are suspended in buffered, saline, pH 7,*8«3«0 (Sorensen’s buffer)* by mechanical shaking in a thick-vailed bottle containing glass beads* The .sus- pension, after light centrifugation, is ter ted for bacterial sterility and then frozen and stored at -70° C* The inf otious titer of the seed prepara- tion should be in the region of •’ whe fresh and should bo close to this figure after thawed for use* The frozen s>; ©d virus is discorded if not used within ten days* Shortly before the seed v.lrus is employed* it is thawed and diluted with a solution of penicillin in such a manner that the final concentration of tissue is 10$ and of penicillin 100 units per oc*. Approximately 500 egga are inoculated with each pool of seed virus* Preparation and Harvesting of Vaccine* Fertile eggs from white Leghorn hens* free of infection with S« pulloruo* are incubated fox* nine days at 33° C* after which they are injected with the inoculum described above and incu- bated at 35° C® Prior to inoculation., on© hole is made into the air sac and a second through tho shell in tho region between the equator and the air aao® A short, curved needle is introduced through tho shell membrane underlying tho second hole, 0*1 oo of the inoculum is injected, and tho holes scaled. Tills teclmiquo, which had previously proved satisfactory in the preparation of vaccine for equine encephalomyelitis (5) almost invariably deposits the inoculum on tho ohorio-allantoio membrane without backflow of fluid. Inocu- lated eggs are candled daily and those with dead or sluggish embryos are dis- carded up to 48 hours. After incubation for 65-66 hours, at a tine when approximately 40*;. of the embryos are dead and 30% ore moribund, all of the embryos ore harvested. At the time of harvesting, eggs are dipped in a 3 per cent iodine tincture for five minutes, rinsed in a 1/2000 solution of Roccala and drained® Under aseptio conditions the eggs are cracked on a breaking-bar in a manner so as not to rupture the amniotic sac® T/hile still in the shell, the sac is ruptured and the embryo is removed with Allis forceps® This method was adopted from the procedure employed for equine encephalomyelitis vaooino (5)e Embryos are pooled in lots of about 50 end then tho ©yes and feet are removed by dissection® Pools of embryos weighing 1600-1700 grams are ground in colloid mills (Eppenbaoh)o The material is coarsely ground in the first of two mills and finely ground in tho second® A sufficient amount of buffered physiological saline solution, pH 7® 8, is added slowly during tho process to make a final concentration of 20,embryo tissue® The entire grinding procedure requires 10-12 minutes for each batch of 1600 grsms of tissue* A representative somple is taken from the first pool of ground tissue obtained on each day of harvest and its inf activity titered in nice® After tho sampling, sufficient formalde- hyde solution U®S®P® is added to the suspension during tho milling process to give a final concentration of 0*2$« The formalinized material is stored in the dark at 3-4° G» in nine liter bottles for 14-18 days % each bottle is shaken daily by hand to resuspend the sediment which forms* Between the X4th and 18th day the supernatant fluids from bottles of one or more batches are aspirated from the sedimented sludge and pooled* Residual coarse particles in the supernatant fluids are removed by passage through a fine nosh screen* Penicillin is now added to th© pool in sufficient amounts to make a final concentration of 5 units per cc* At this stage a sample of each pool is taken for test of bacterial sterility, for safety test in animals and for an immunogenic assay* LyophiXisation of Vaccine* The lyophilization apparatus available had a capacity for drying 15 liters of vaccine in a 24 hour period* Therefore* complete lyophilization of the vaccine produced during one week would require 4-7 days* Approximately 16 liters of fomalinized* pooled vaccine* an amount which was sufficient for each day’s lyopldlizing* was removed from storage and processed as outlined below* a© neutralization of formaldehyde* The formaldehyde it the pooled fluid vaccino is neutralised by moons of sodium bisulfite0 A 100 oc» sample is re- moved from the vaccine destined for drying on that day and the amount of sodium bisulfite solution required to neutralise the formaldehyde in this aliquot is determined* The technique previously employed with mouse brain vaccine (1) is used in the present procedure* This consists of adding suf- ficient bisulfite until the Scbryvor tost formaldehydes boo or.© negative* after which the pH of the sample is adjusted to about 7*5* On the basis of the data obtained for the 100 co* a 16 liter volume is neutralized and adjusted* Samples of the larger pool alter neutralization are retested to insure that the added materials give the expected effect. b° Bottling and Freezing. Sufficient ncrthiolato is added to bring the final concentration to 1:10,000. The fluid material is distributed in 25 cc. amounts in 60 oo. pyres glass ampoules and immediately shell-frozen by mechani- cal rotation in a bath of alcohol and solid CO,,. The ampoules are then stored •70° C. for one or two hours when they are attached to the lyophilising machine. lyophilizing Apparatus* A diagram of the apparatus used for drying the vacoine is presented in Figure 1* Thrwi units of the type obtained" frcu the Xutgrnatioaiiii-Health-■Oivision of the Roolsefeller Foundation wlilgh had"used Uma previously fear drying influenza and yullov/ fuuyr uwwtogso Thoso were built according to designs modified from those previously pub- lished by Pickels and Bauer (6) who, in turn, modified the apparatus of Flosdorf and Hudd (7). Process of lyophilization. For the successful drying of Japanese encephalitis vaccine it is essential that the frozen biological not be permitted to thaw at any stage during the procedure. The initial period in the process is a highly critical one, and the interval between the removal of the ampoules from -70° C* storage and the tine the frozen vaccine is subjected to a high vacuum must be shortj in practice 15 minutes is required for loading and obtaining a vacuum of 30-50 microns with tho International Health Division equipment. The drying cycle, which is completed in 24 hours, begins with a period lasting several hours during which the temperature of t he air sur- rounding the outside of the ampoules of vaccine is in the region of -20° C. The air temperature slowly rises and reaches 0° C. 18 to 20 hours after starting the run. Then the temperature is elevated rapidly to -f* 30° C. and maintained here during the 22nd to 24th hours of the cycle. The temperatures designated above are obtained in the early stage by mechanical refrigeration of the compartment mill which surrounds the ampoules (see Figure 1) and in the last stage by the use of electric hot air blowers (hair dryers)o After the second hour3 circulation of the air is maintained in each compartment by moans of electric blowers (hair dryers with heating olement switched off)* thus insuring uniform air temperature throughout a compartment« At the ond of the cycle the vacuum in the system is between microns of moroury. A graphic illustration of the temperature and pressure readings from a typical run Is presented in Figure 2. At the end of the cycle* the line to the vacuum pump is closed and "Super-dry"* oil-pumped, nitrogen (moisture content less than 0.002/0 is passed through a column of a chemical drying agent and thonco into the evacu- ated system until atmospheric pressure is reached. The manifolds with nitrogen charged ampoules attached are removed from the machine and the necks of the ampoules ore sealed with a gas-oxygon torch* Ampoules of lyophilized vaccine are inspected and then stored at 5° C .while control tests are completed. Shipment of Vaccine. Ampoules containing lyophilized Japanese encepha- litis vaccine* chick embryo type* are supplied in individual boxes containing instructions on the method of rehydration together with a rubber stoppered bottle containing 25 oo. of sterile distilled water to be used for its rohy- dration© The dried vaccine is maintained at 5° C. Rohydratod vaccine is also stored at 5° C. when not in actual use and unused portions are discarded after eight hours. Control Tests on Vaccine. a* Safety and sterility tests♦ Tests for inactivation of virus and other safety teste are made in accordance with the Federal Security Agency, National Institute of Health, Ilinimum Requirements, Japanese Encephalitis Vaccine, Chick liabryo Type, dried, August 7, 1947© Sterility tests are per- formed in general accordance with the accepted methods except that the media used for the bacterial cultures contains clarase as a source of penicillinase® This is added to destroy the penicillin which is incorporated in the vaccine at the time of pooling® bo Residual molstureo Estimations of the residual moisture content are made on samples of dried vaccine from each day’s run by weighing the materials before and after exposure to phosphorus pentoxide under partial vacuum at 37° Co for 48 hours* A moisture content of less than 1;% is acceptable for a dried vaccine® The residual moisture content of 49 sub-lots averaged 0o62$ with a maximum of 1®46$ and a minimum of 0®15$® Co Immunogenic potency tost® The mouse protection test employed is that described in Federal Security Agency, National Institute of Health, Llinimum Requirements, Japanese Encephalitis Vaccine, Chick iiabryo Type, dried, August 7, 19470 Potency tests are performed on one or more sub-lots of each lot of i dried vaccine® Freshly rehydrated material is used for each series of immuniz- ing injections of mice® If a number of sub-lots are dried over a period longer than three days, then assays are performed on samples from the first and last sub-lot® Quantity of Vaooine Produced® During the period from early February to mid-&ay, 1947, two batches of 2600 eggs each were inoculated and harvested weekly and these together yielded about 90,000 cc« of a 20$ suspension of chick embryo tissue® In general, the material produced eaoh week was pooled and designated by a lot number® During this period 1,346,369 cc« of 20$ tissue suspension were prepared® Of this amount 15$ was discarded as sedi- mented material during preliminary processing* The major portion of the, formalinized v&ooine olosu'od of the sediment wb& lyophilized; this amounted to 707*550 oc«* (50*702 ampoules)0 Because of the limited facilities for lyophilization a variable amount of the excess fluid vaccine was treated with urea each week in order to retard antigenic deterioration* During the above period approximately 250*000 oo* were so treated* however* relatively little of this web subsequently used for the immunization of military personnel* During the three and a half months of production one entire lot and two sub lots wore discarded because of bacterial contamination; these totaled 105*000 cc Five lots* totaling 302*000 co« produced during this period wore discarded be- cause of poor immunogenic activity* It may be noted that the infectivity of these five unacceptable lots ranged between 10“7 and 10~7°3* ch is below the titer regarded hero as a prerequisite for a potent vaccine* A total of 587*950 co (25*518 ampoules) of lyophilized vaccine was shipped between 19 April and 26 Juno* 1947 for the immunization of military personnel in the Far East* STABILITY OF THE VACCINE Japanese encephalitis vaccine* chick embryo type* when in the fluid state deteriorates rapidly* The majority of such preparations are no longer of acceptable potency after 40 days at C* When held at 37° Ca* three hours is sufficient to destroy most of the immunogenic material in the vaccine (See Table l)e In contrast* dried chick embryo vaccine is quite stable* It may be stored at 6° C* for at least 77 days without appreciable change in its assay value (See Table 2)® Furthermore* dried vaccine stored at room temperature for two weeks displays no loss of activity* whereas when held at 37° G„ for two weeks it shows some loss but still contains on appreciable amount of immunogenic activity* see Table 10 liECOr.ILiK-JDAT I01TS Th© quantitative reproducibility of results with the present assay method employed for Japanese encephalitis vaccine leaves something to be de.- sired# Potent vaccines with a lew minimal immunogenic dosage usually have low values on repeated assays, and poor vaccines consistently give high values® Nevertheless, in the critical range near the present level of ac- ceptability, marked variation occurs in repeated tests® Two examples among many may be cited# Sub-lot A of one lyophilized vaccine when tested on 23 April 1947 had an assay value of 0®031 oo# and when retested on 26 May the value was 0#015 poj sub-lot C of this same vaccine had a value of 0o0095 cc# when tested on 29 April# Sub-lot A of another lyophilised vac- cine when tested on 3 June and retested on 6 June gave values of 0®011 and 0o025 oc», respectively® Three assays performed on other sub-lbts of this same vaccine had values of 0«012, 0®016 and 0®025 co# In all 6 of the tests in which the above mentioned assays were included, the intraperitoneal titer of tlie challenge virus was between io*"l«96 and and the mortality in the non-vaccinat©d control mice 100$® Further evidence indicating that minor variations in the range between 0.005 and 0#020 oc# in assay values on dried vaccines are of little consequence is found in studios on the capacity of two lyophilised chick embryo vaccine with assay values of 0#006 or 0.015 cc« to elicit neutralising antibodies against Japanese encephalitis virus in human beingsi little if any difference was observed (3)® Efforts should be continued to improve the current method of assay of potency® la the mean- time it is recommended that Japanese encephalitis vaccine, chick embryo type, dried, be considered acceptable if the minimal immunogenic dosage, as measured by the current method, is less than 0®020 oo® Non-pathogenic, aerobic, spore-forming bacteria frequently are found in biological materials prepared from embryonated eggs* Those generally resist the bactericidal action of 0*2% formaldehyde even if the biological is hold at 5° Co for long periods of time# Fortunately* these organisms are highly susceptible to the action of small amounts of penicillin while the virus of Japanese encephalitis is not affected by such concentrations of the drug,. Losses associated with contamination by organisms of this group are practi« cally eliminated when penicillin is added to the seed virus and to the fluid vaccine* The addition of small amounts of penicillin to the seed virus is probably the more Important step since it apparently inhibits the growth of these organisms in those occasional eggs in which bacteria are present ot are introduced at the time of inoculation* The employment of an antibiotic for the control of bacterial contamination in the production of Japanese encephal- itis vaccine lias proved satisfactory and should b© continued*, During the past year all lots of vaccine having an infectious titer of or greater were satisfactory and none with titers below this level were acceptable® The average infective titer for 12 acceptable lots was 10«^.»82a jt is recommended that only lots with infective titers of or greater be considered for further processing and testings Comparative studies with throe other strains of virus* namely "Kalirdna" "Matsunaga" and nRoumw* (the latter isolated from a fatal human case in Korea in 1943 by members of the Virus and Rickettsial Disease Commission indicated that none of these strains multiplied in the chiok embryo as well as did the "Wakayama Furthermore* chick embryo vaccines prepared from these strains wore of inferior potency when assayed in mice and by the standard method® Therefore* if is recommended that the "Nakayama* strain continue to bo used for the preparation of Japanese encephalitis vaccine® Th© immunogenic material in Japanese onceplialitis vaccine is relatively labile and deteriorates under conditions which oro not deleterious to many biologicalSo Under proper conditions of lyophilisation, none of the im- munogenic activity of the fluid vaccine is lost in drying* Furthermore, th© lyophilissed product is of relatively stable potency* The method for lyophilisation described in the teoct has given satisfactory results with Japanese encephalitis vaccine both in regards stability and residual moisture content* It is recommended that until more improved methods are developed the principles of drying embodied in the present work be used in subsequent manufacture of this vaccine* It is further recommended that a residual moisture content of less than 1% be required for dried Japanese encephalitis vaccine* TABLE 1 Deterioration of fluid and Dried Vaccines at 57° Cc Vaccine Type Number Treatment Assay I-UIaD* Fluid Lot 12 60 days at 4° C 0*0125 co« « 60 days at 4° c / S lirs«. at 37 °C 0^095 CO* n 60 days at 40 c / S hrso i at 37°C 0*10 oc» M 60 days at 4° c / 24 hrs® at 37°C 0*10 00 o Dried Lot 1A 60 days at 4° c 0*011 CO* « 15 days at 4° c / 15 days at 25°C 0«015 oc* n 15 days at i° a f 16 days at 37 °C 0*025 CO® Dried Lot 6 A 35 days at 4° 0 0*012 CO O « 21 days at 4° 0 / 14 days o o to c\» % 0*013 OCo w 21 days at 4° 0 £ 14 days at 37°C 0o022 COo TABLE 2 Stability of Vaccines Stored at 4° C« Vaccine Lot Interval Between Time of Drying and Assay Assay M«I*D« Vaccine Lot interval between 'itme of Drying and Assay Assay Me X* D« B«66 8 days 0*015 co* 5D 3 days 0*0X5 co* n 60 days 0*0098 oc* « 28 days 0*019 00a 1A 1 day 0*006 co* n 120 days 0*015 cc* 30 days 0*0U cco 6A 4 days 0*0083 co n 77 days 0*015 oo* n 35 days 0*012 oc* » 135 days 0*019 CO* 9A 1 day 0*008 co* 3A 7 days 0*009 oc* it 25 days 0*018 COa « 124 days 0*019 co* Vaccines wore lyophilized 18-22 days after harvest o;; the infected eobryoso FIGURE I Description of lyophilizing Apparatus (B&uer & Piokele) The brass manifolds (A) are attached to a condenser (B) which is im- mersed in a large Dewar flask (C) containing ethyl alcohol and solid 00^, A high vacuum rotary pump (D) is attached to the condenser and serves to eva- cuate the system. The manifolds are fitted with nipples to which the ampoules are attached by rubber tubingo The manifolds and ampoules are placed within an insulated compartment (e) which can bo cooled by mechanical refrigerab ion (p) to «20° C0 Drawing reproduced with pomission of the International Health Division of the Rockefeller Foundation® REFERENCES 1« Sabin* A* B.* Daffy* G. E.* Warren* J** Ward* R.* Peck, J, L* and Ruohman* I* 1943* The St. Louis and Japanese B Types of Epidemic Encephalitis* J*A*l!*A* 126 : 477-486 2* Warren* J# and Hough* R. 6* 1946* A Vaccine Against Japanese B Encephalitis Prepared From Infected Chick ISabryos* Proo* See* liepor* Biol* Lied* 61* 109-113* 3© Warren* Ja* Smadel* J* B* and Rasmusrea* A» P* Jr* The Antibody Response in Human Beings Inoculated with Japanese Encephalitis Vaccine* Embryo Type* In press 4® Monthly Report of Department of Virus and Rickettsial Diseases, September 1946, circulated In Med* Dept* Research & Development Program, 1 July - SO September 1946, Army Med* Research & Development Board, Office of the Surgeon General, U«S*Amy, Page 95* 5* Randall, R.> 1940* The Preparation of Equine Encephalomyelitis Vaco in© (chick) by the Army Veterinary School. Vet* Bull* 34s 7-12* 6* Bauer, J« 31* and Pickels, E* G« 1940* Apparatus for Freezing and Drying Vims in Type Quantities under Uniform Conditions* Jour* Exper Med* 71* 83-88* 7* Slosdorf, E* W* and Mudd, S* 1935* Procedure and Apparatus for Preservaticaa in "lyophile" Form of Serum and other Biological Sub stances* Jour* Immunol* 29* 389-425* D E G R E E S C T E M R FIGURE 2 TEMPERATURE AND VACUUM DATA OBTAINED DURING LYOPH1LIZATION OF JAPANESE ENCEPHALITIS VACCINE, CHICK EMBRYO TYPE (Lot 114 C, Unit 2, 6-4-47) TIME IN HOURS V A C U U M M I C R 0 N S 0 F H G.