[Reprinted from the Journal of Cutaneous and Genito-Übinary Diseases, October, 1895.] 
XANTHOMA DIABETICORUM (SO CALLED). ITS PLACE AMONG 
THE DERMATOSES. 
JAMES C. JOHNSTON, A. 8., M. D., 
Assistant Physician, Department for Skin and Venereal Diseases, New York Hospital; Clinical 
Instructor in Dermatology, New York Post-Graduate Medical School. 
THE selection of this subject for a thesis* was influenced by two 
considerations : first, the study of one of the very rare examples 
of this interesting affection which has fallen under my personal 
notice ; second, the failure in the latest works on dermatology to recog- 
nize its peculiarities, both in symptomatology and pathology, and its 
claims for a place, distinct and separate from that of ordinary xanthoma, 
among the dermatoses. Among recent treatises, two of the more prom- 
inent may be cited as sufficient to show the general trend. Kaposi,f 
in the fourth German edition of his work, considers the question of the 
identity of the two forms still undecided, but thinks it worthy of 
further study. Heitzmann,:}; in Morrow’s System, without reviewing 
the evidence at hand, claims that xanthoma diabeticorum is not a dis- 
tinct affection, and that the seeming differences are easily accounted 
for. Publication of the article may be excused on the ground of the 
relapse in the case, more extraordinary in many respects than the 
original attack already reported. 
History.—An account of the initial appearance of the disease in 
my case will be found in this Journal (May, 1894, vol. xii, p. 205). 
There is no necessity for a repetition of what has already been said, 
* Presented to the Society of Dermatology and Genito-Urinary Surgery, 
f Pathologic und Therapie der Hawtkrankheiten, p. 748. 
X System of Genito-Urinary Diseases, Syph. and Derm., vol. iii, p. 486. 
COPYRIGHT, 1895, BT D. APPLBTON AND COMPANY. Original Commlln kjatkxn s. 
and I shall pass on to a short account of the second attack, merely re- 
marking that the first presented the classical features of the xanthoma, 
with the additional involvement of the soles of the feet. 
The nodules began their involution in the winter of 1894, simulta- 
neously with the improvement of the patient’s diabetes and general 
health, disappearing in the reverse order of their appearance—viz., legs, 
back, buttocks, arms. The disease has never faded completely since 
its first outbreak. The lesions on the knees and the plaques on the 
soles, the latter the last in order of evolution, have always persisted. 
Ulceration has not at any time been a part of the morbid process, and 
the disappearance of a nodule leaves absolutely not a trace upon the 
skin. In March, 1894, the patient’s urine contained less than one per 
cent of sugar, and the cutaneous disease was then confined to his right 
wrist, knees, and soles. In the fall of the same year he began to drink 
heavily ; the amount of urine passed, and the proportion of contained 
sugar, increased day by day, until the latter reached eight per cent and 
the eruption began to reappear. Its development was far more rapid 
than in the original attack, the recrudescence being complete in De- 
cember. 
The activity of the process was clearly shown by the fiery red color 
of the lesions, in marked contrast to the pale pink seen during involu- 
tion. (This observation confirms Torok’s statement* as to the tint, in 
which he is singularly correct, in view of his never having seen a case. 
At the same time it controverts that of Payne.) The shade may be 
aptly compared to that of raw beef. When stripped, the patient pre- 
sented a remarkable spectacle. Over the knees the nodules had coa- 
lesced into enormous plaques, four by eight inches in size, dotted here 
and there with the yellow central point of the individual lesion. The 
extensor surfaces of the arms were in almost the same condition, the 
chest and abdomen were involved for the first time, and on the buttocks 
and backs of the thighs a ten-cent piece could not be laid on a clear 
surface. The dorsa of the feet and toes showed scattered and confluent 
papules, and the plantar plaques were increased in size. The genitalia 
were still uninvolved, but two surfaces, previously free, were attacked. 
These were the face, on the edge of the alee and tip of the nose (not, 
however, on the lids), and the backs of the hands and fingers, on which 
the nodules coalesced as on the toes. It was impossible, of course, to 
count them, but the lesions could not have numbered less than 
twenty-five hundred. The subjective symptoms were excessively an- 
noying, often preventing rest in the recumbent posture. Any pres- 
* Annales de derm, et de syph., Nos. 11 and 12, 1893. Xanthoma Diabeticorum. 
3 
sure on them produced a sensation of tenderness and prickling, and on 
account of the extreme hardness it seemed to the patient as though he 
were lying or sitting on a bag of shot. 
The disease remained in this condition for two or three months, 
but in April, 1895, began slowly to grow paler, after a trip to the sea- 
shore, following as before an improvement of the diabetes and the 
health generally. It is a curious and notable fact that, while whisky 
was kept from the man, no involution took place, but, on allowing him 
a regular daily portion (a pint and a half, ordinarily), improvement 
began. It seems likely that every exacerbation of diabetes will be fol- 
lowed by an outbreak of the skin affection. No local treatment has 
ever been attempted, for the reason that I have little faith in local ap- 
plications except electrolysis, which the patient declines and because 
general improvement is surely followed by fading of the cutaneous 
disease. 
DEtiology.—The origin of the disease is, like many others in the 
held of dermatology, more or less shrouded in obscurity. The connec- 
tion, causal or otherwise, between diabetes and this form of xanthoma 
is not understood ; its very existence is denied, and with apparent rea- 
son, in view of the cases with an identical clinical history and histologi- 
cal characters bearing close resemblance to those found in so typical 
a case as mine. 
Twenty-one cases have been reported up to this time (see appended 
bibliography). The light which they shed upon the aetiology of the 
affection is not brilliant. Twenty of them have occurred in men ; one 
only in a woman, reported by Hillairet. These patients, with the sin- 
gle exception of the first—that of Addison, a young man of twenty-six 
—had arrived at middle life. In seventeen, diabetes mellitus was pres- 
ent; in Bristowe’s case it followed the eruption; in Cavafy’s and my 
own it preceded the xanthoma. Hutchinson, Besnier, Tidal, and Payne 
failed to discover a glycosuria in their cases. The eruption itself led 
Crocker to the discovery of diabetes in the man under his care. It 
seems to me that too much has been made of the absence of sugar in 
the urine of these people attacked by a transient xanthoma. The gly- 
cosuria (I use the word here in its most limited sense) may have been 
temporarily absent, but it is not necessary now to have recourse to this 
explanation. Ton Noorden and Oertel, in their respective articles on 
Diabetes and Obesity recently published in vol. ii, Twentieth Century 
Practice, give the weight of their opinion to the theory of the identity, 
setiological at least, of these two conditions. Sir Dyce Duckworth * in- 
* Chronic Glycosuria in Middle and Advanced Life. Brit. Med. JournOctober V, 1893. 4 
Original Communications. 
dines evidently to the same belief from his recent utterances. A dis- 
cussion of this belief would be out of place here, but it is interesting to 
note how often obesity in varying degrees occurred in the twenty-one 
cases. Including those frankly described as diabetic (Barlow’s case is 
omitted for the reason that much doubt attaches to it), no less than 
eleven are described as remarkably stout, of florid appearance, or obese; 
in two the condition was well nourished, stoutness nut being marked; 
two more were cooks and presumably suffered from their occupation 
and its temptations; two were drinkers and addicted to the pleasures 
of the table ; in the remaining three nothing is reported concerning the 
nutrition, but the number includes Addison’s case, a diabetic. Payne, 
who failed to discover diabetes, says his man “ was inclined to be stout 
and florid.” In view of their all but uniform occurrence there seem- 
ingly must be a connection between the vice of constitution expressed 
in the obesity and diabetes and the transient xanthoma occurring in 
these cases. 
Charabard * believes that the disease is due to a special diathesis— 
i. c., it is a xanthomatosis. This is an explanation which explains noth- 
ing. There is certainly a tendency in the constitution of the individual 
to the formation of these lesions, but that offers no solution of the dif- 
ficulty. The theory that they owe their origin to the same factors 
which give rise to the diabetes (nervous, pancreatic, gouty, or rheu- 
matic [Duckworth], poisoning by phbridzin [von Noorden]) is difficult 
if not impossible of proof. The most, and to my mind the only, rea- 
sonable explanation is that xanthoma diabeticorum is an irritative 
process, the irritation being supplied by tbe excess of glucose or some 
faulty product of metabolism circulating in the blood—a theory sug- 
gested by Torok and Kaposi. What the particular substance may be 
will doubtless be discovered when the pathogeny of diabetes is fully 
elucidated. Color is lent to this proposition by the fact that the nod- 
ules of the disease begin in the corium in the neighborhood of the sweat 
glands and hair follicles with their attached sebaceous structures, all of 
which are supplied by the same set ot vessels, part of the excretory ap- 
paratus of the skin. It has occurred to me that the irritation might be 
occasioned by the effort of the organism to rid itself of the sugar excess 
through this channel, as well as through the renal filter, an analogue 
of the process in scarlatina. 
Hallopeau,]' after studying his case of transient xanthoma, still be- 
lieves that xanthomata are nsevi localized in the same way. The gly- 
cosuria in the particular affection under discussion he regards as due to 
* Annales de derm, et de syph., vol. v, p. 348. 
\ Ibid., vol. iv, No. 8. Xanthoma Diabeticorum. 
5 
a visceral localization of the lesions in the pancreas, and consequently 
holds that the xanthoma is even here the primitive process. He thinks 
that the possibility of involution and change of size according to the 
degree of repletion of the vascular system explains the intermittences 
of diabetic symptoms. There is no report of an autopsy on these cases, 
so that his theory, ingenious as it is, lacks confirmation. His reason- 
ing, it should be said, is flora analogy with icteric xanthoma, in which 
lesions have been found in the bile ducts (Kaposi). 
It will be noticed in the bibliography that eleven of the cases 
had been reported in the last five years, wdiile forty were required to 
collect the first ten. It seems likely that with a wider spread of knowl- 
edge of dermatology the instances will be multiplied until xanthoma 
diabeticorum ceases to be classed among the rare dermatoses. 
Histology.—The pathological anatomy has been investigated by 
Robinson, Crocker, Clarke in the second of Morris’s cases, Payne, and 
recently by Scharnberg, There is little to add to their reports. The 
general consensus of opinion seems to be that the diseased process 
bears some resemblance under the microscope to that of ordinary xan- 
thoma, but that inflammatory changes are more marked and connective- 
tissue increase less noticeable in the diabetic form. The papule begins 
and remains throughout its whole course localized in the corium in the 
neighborhood of the sweat glands and follicles chiefly, as has been said. 
The corium is at first attacked superficially or toward the center, the 
process later involving the whole depth. The epidermis remains un- 
affected, except that the rete pegs disappear to a great extent and some 
of the cells become vacuolated. At the outset vascular dilatation takes 
place and persists, a slight oedema separates the tissue elements, and is 
followed by an infiltration of round cells, lymphoid corpuscles, and, 
according to Robinson, a proliferation of the connective-tissue cells. 
“ Xanthoma cells,” large, multinucleated cells with a fine membrane, 
including tiny droplets of fat, are found in xanthoma diabeticorum as 
in ordinary xanthoma, but are very few in a beginning lesion. They 
increase later and are proportionate in number to the size of the pap- 
ule. Crocker describes them as epithelioid or giant cells.* The bun- 
dles of fibrous tissue swell, thicken, become translucent, and finally 
disappear or “ appear rarefied,” with large intervals. In the meshes a 
granular material is seen, composed chiefly of minute oil drops, not 
solely contained in the xanthoma cells, but resulting from their degen- 
eration, finally taking their place, together with that of the connective 
tissue, and forming a confluent mass in the center. This mass of fatty 
* Such they are undoubtedly, and in consequence little importance can be attached to them 
in view of their presence in numerous other pathological conditions. 6 
Original Communications. 
granules gives the yellow center to the xanthoma nodule. There is 
then an entire central degeneration, with little or no connective-tissue 
increase, according to all investigators except Payne, who says that the 
elastic tissue u appeared to be much increased in amount, with little 
intervening tissue.” He does not, however, himself regard it as a 
hyperplasia, but as a preponderance of elastic tissue in consequence of 
degeneration and absorption of the other constituents of the corium. 
Resolution presumably takes place by the absorption of the central 
granular mass and a regeneration of the fibrous tissue of the corium by 
the proliferated fixed cells already mentioned. In the words of Torok 
{loc. cit.), “ the lesions of xanthoma diabeticorum are caused by a local 
irritative process resulting in a granulo-fatty degeneration.” 
Classification.—I took occasion in the opening remarks in this 
paper to call attention to the place accorded xanthoma diabeticorum 
among skin affections. Opinion on the subject is fairly divided. Bes- 
nier and Doyon* unequivocally state that the sharp clinical points of 
differentiation and the histological findings in xanthoma vulgare and 
xanthoma diabeticorum establish only the variety, not the nonidentity 
of the two forms of the disease. Morris, Crocker, and Payne give this 
dictum only half-hearted support. Payne thinks that “ the association 
of morbid changes seems to suggest that there is at least a common 
element in the three forms (including elastic xanthoma), and that the 
differences may not be essential but may depend on necessary circum- 
stances.*’ Torok {loc. cit.) advances a diametrically opposite view, 
classing the glycosuric as a distinct affection with no relation to the 
common form. 
It is necessary, therefore to review the clinical and microscopic 
points of diagnosis between the two varieties. The former have been 
embodied in a report by Sangster and Crocker f to the Pathological 
Society of London on the first of Malcolm Morris’s two cases, a report 
which has since been extensively quoted. The clinical differential 
points are these : (1) The most important, and a divergence sufficient- 
ly great of itself to demonstrate the dissimilarity of the two forms, viz., 
the sudden evolution and involution of xanthoma diabeticorum, the 
latter invariably taking place spontaneously at some time in the history. 
In ordinary xanthoma involution is extremely rare, and when it does 
occur takes place almost imperceptibly. In fact, such an event is very 
doubtful. (2) The lesions in xanthoma diabeticorum are firm and 
solid, inflammatory, discrete or confluent papules; in xanthoma they 
are soft, inflammatory phenomena are absent, and the disease occurs 
* Kaposi, Maladies de la peau. Traduction, second edition, vol. ii, p. 335. 
\ Path. Tram., 1883 p. 284 Xanthoma Diabeticorum. 
chiefly in flat patches or stride, which are never found in the first form, 
(3) The yellow center is not visible at first in the diabetic affection nor 
in all the papules at any time; in xanthoma vulgare the chamois- 
leather appearance is always to be seen. (4) An unimportant distinc- 
tion—xanthoma diabeticorum is never found upon the eyelids; its pref- 
erence is for the joints and extensor surfaces generally. (5) Subjective 
symptoms, though they may be slight, are the rule in the first variety. 
They are totally absent, except in case of ulceration, in ordinary xan- 
thoma. (6) Jaundice has never been found in xanthoma diabeticorum. 
(7) Glycosuria, often transitory, has been found in a large majority of 
the cases, and those in which it was absent showed at least a related 
condition of plethora and obesity. Moreover, it is possible that the 
nutritive disorder was in abeyance at the time of examination. Dia- 
betes has never been found with the ordinary form except in a case of 
Besnier’s, to which little importance need be attached, since it is an 
isolated instance of a concurrence of two totally unrelated diseases.. 
(8) The lesions of xanthoma diabeticorum occur in the neighborhood 
of the hair follicles and skin glands, a localization not observed in the 
common disease. To these may be added: (9) Ulceration is never a 
part of the process in diabetes; it has been seen in xanthoma multi- 
plex (Morrow’s case). 
Histologically, the only point of resemblance in the two processes 
is the presence of so-called xanthoma cells, giant cells in reality, in 
xanthoma diabeticorum containing minute droplets of oil. Their im- 
portance in diagnosis has already been discussed. I shall refer here 
again to Torbk’s able paper (1. c.) for a comparison of the pathological 
conditions. The tissue of xanthoma multiplex presents a striking anal- 
ogy to adipose tissue, the analogy being borne out by the cells, which 
are identical with those of physiological fat in process of formation ;• 
further, by the formation of these cells around the vessels, in their ad- 
ventitia even ; by the discovery of cells analogous to those of xanthoma 
in imperfectly developed adipose tissue. The fat formation does not 
proceed to the point of occupying the whole cell as in true fat, nor 
does the cell break down into a granulo-fatty detritus as in xanthoma 
diabeticorum. To explain this arrested development, Torbk concludes 
that “ ordinary xanthoma * is formed of adipose tissue in a heterotopic 
situation, and that it is constituted, by reason of that heterotopia, of 
fat cells of incomplete, interrupted evolution.” To the formation of 
fat must be added, in some cases, that of fibrous tissue. The cells are 
of connective-tissue origin, as are those of fat in other situations.. 
* Abstract, Jotjrn. or Cut. and Gen.-Urin. Diseases, vol. xi, p. 225. 8 
Original Communications. 
Xanthoma vulgare is then “ an anomaly produced under the influence 
of congenital and hereditary conditions by proliferation of connective- 
tissue cells and by their transformation into fat cells in places which 
are normally free from adipose tissue.* It is an anomaly of forma- 
tion.” In other words, it is a benign tumor, and its place is with its 
analogue—lipoma. Contrast this with a formation showing every evi- 
dence, clinically and microscopically, of inflammation resulting in a 
degeneration en masse of the central portion of the papule. Does it 
seem proper to class them together as varieties of the same disease, 
ignoring the fact that xanthoma multiplex is an anomaly of formation, 
and xanthoma diabeticorum an inflammation due to an irritative 
process ending in a granulo-fatty degeneration ? It seems to me that 
it has ceased to be a matter of opinion; one must recognize facts. 
In accordance, then, with the differences pointed out, xanthoma 
multiplex is properly placed and named—there is no necessity for the 
name “ordinary or common xanthoma”—xanthoma vulgare. It is 
xanthoma par excellence, a yellow tumor. Xanthoma diabeticorum 
has never been satisfactory. Lichen, proposed by Sangster and 
Crocker, is objectionable on many grounds. Torok’s term—u a papu- 
lar eruption of diabetics due to a granulo-fatty degeneration”-—is too 
long for convenient use. I would propose as a substitute dermatitis 
xanthomatoides diabeticorum—in other words, an inflammatory skin 
disease which occurs in diabetics who have a personal idiosyncrasy 
leading to the formation, and which resembles xanthoma. That seems 
to cover the ground. Further, the affection should be classed among 
the inflammations with the other “ dermatoses diabeticse,” not with 
tumors, tumors even of inflammatory origin, such as the syphilides 
and lupus, because dermatitis xanthomatoides has a tendency to spon- 
taneous resolution. I should retain the term diabeticorum, for the 
reason that I believe diabetes is the true underlying condition, and see 
no object in a change to glycosuricum, even if the disease is manifested 
in a transitory fashion. 
BIBLIOGRAPHY. 
CASES. 
1. Addison and Gull. On a Certain Alteration in the Skin, Vitiligoidea, 
etc. Guy's Hospital Reports, vol. ii, p. 265. 
2. Bristowe. Keloid of a Rare Form. Pathological Transactions, vol. 
xvii, 1866. 
* Whatever may be the individual opinion of the adequacy of Torok’s explanation of the 
arrested development in xanthoma cells, the fact remains that they are of connective-tissue 
origin. Hence the objection in no wise weakens my argument. Xanthoma Diabeticorum. 
9 
3. Hillairet—Gendre. Etude clin. et hist, sur le xanthome. These de 
Paris, 1879. 
4. Morris (first case). Transactions of the Path. Soc. of London, 1883, 
plate xx. 
5. Hardaway. A Case of Multiple Xanthoma. St. Louis Courier of 
Medicine, October, 1894. 
6. Barlow. Brit. Journ. of Derm., vol. i, p. 3,1888. (Some doubt attaches, 
to this case.) 
7. Colcott Fox. Case shown to Dermat. Soc. of London. 
8. Cavafy. Case shown to Dermat. Soc. of London. Also Brit. Journ. 
of Derm., vol. i, p. 76, 1889. 
9. Besnier. Case shown (March, 1889) to physicians of THopital Saint- 
Louis. 
10. Robinson. Journ. of Cut. and Gen.-Urin. Diseases, vol. vii, p. 219. 
Atlas of Rare Skin Diseases, plate xiii. 
11. Vidal. Mentioned by Crocker, p. 497. 
12. Hutchinson. Mentioned by Crocker, p. 497 
13. Morris (second case). Brit. Journ. of Derm., August, 1892. 
14. Crocker. Brit. Journ. of Derm., August, 1892. 
15. Payne. Brit. Journ. of Derm., November, 1893. 
16. Tims. Brit. Journ. of Derm., November, 1893. 
17. Pollitzer. Brit. Journ. of Derm., December, 1893. 
18. Johnston. Journ. of Cut. and Gen.-Urin. Diseases, May, 1894, vol. 
xii, p. 205. 
19. Jameison. Brit. Journ. of Derm., October, 1894, 
20. Hallopeau. Annates de dermatologie, 1893, vol. iv, p. 935. 
21. Schamberg. Journ. of Cut. and Gen.-Urin. Dis., May, 1895. 
The report of Vidal’s and Hutchinson’s cases I have been unable to find ; 
the case reported, according to Besnier, to Chambard by Hillairet, a glycosuric 
woman, is undoubtedly identical with that published by Gendre in his These 
de Paris. Barlow’s case should really be omitted, since so much doubt has 
been cast upon it, but I include it for the sake of completeness. 
Other Literature. 
Tdrok. De la nature des xanthomes. Annates de derm, et de syph., vol. 
iv, Nos. 11 and 12,1893. 
Hallopeau. De la nature des xanthomes. Annates de derm, et de syph., 
vol. iv, No. 8, 1893. 
Besnier. Annates de derm, et de syph., No. 5, 1889. 
Chambard. Annates de derm, et de syph., vol. v, p. 348. 
Treatises of Kaposi, Morrow, Crocker, Brocq, Besnier and Doyon-Kaposi. 
Clarke. Pathological Transactions, 1892, plate xliii. 
Sangster and Crocker. Path. Trans., 1883, p. 284. 
Chambard. Des formes anatomiques de xanthelasma cutane. Archives 
de physiologic, 1879, p. 641. 
115 West Eighty-fourth Street.