ADENO-CARCINOMA OF THE SKIN 
ORIGINATING IN THE COIL GLANDS 
BY 
J. A. FORDYCE, M. D. 
Professor of Dermatology and Syphilology, Bellevue Hospital Medical College ; Visiting Surgeon 
to the City Hospital, etc. 
[Reprinted from the Journal of Cutaneous and Genito-Urinary Diseases 
for February, 1895] 
NEWYORK 
D. APPLETON AND COMPANY 
72 Fifth Avenue 
LONDON ; CAXTON HOUSE, PATERNOSTER SQUARE Journal of Cutaneous and Genito-Urinary Diseases, February, 1895 
Fig. 4.—Spencer in., x 400. 
Fig. s.—Spencer J in., x 400. 
Fig. 6.—Spencer \ in., x 150. 
Fig. A—Spencer £ in., x 400. 
ILLUSTRATING DR. FORDYCE’S ARTICLE ON ADENO-CARCINOMA OF THE COIL GLANDS 
(Photomicrographs by the Author.) ADENO-CARCINOMA OF THE SKIN ORIGINATING IN THE 
COIL GLANDS.* 
Professor of Dermatology and Syphilology, Bellevue Hospital Medical College; Visiting Sur- 
geon to the City Hospital, etc. 
J. A. FORDYCE, M. D., 
CANCERS of internal organs are generally the result of an ab- 
normal proliferation of the glandular epithelium of the part 
affected. In the skin, on the contrary, the epidermis is the 
structure which is most frequently first involved. The squamous- 
celled epithelioma, or the lobulated variety of the French writers, un- 
doubtedly starts in those cells of the epidermis which are destined to 
form the stratum corneum. Certain varieties of skin cancer which 
have their point of departure in the epidermis are made up entirely of 
small epithelial cells which preserve the character of the cells in the 
lowermost rete layer. These cells extend throughout the cutis and 
subcutaneous tissue in the form of tubular processes, showing no dis- 
position to form epidermic “ globes ”or horny tissue. It is difficult to 
explain why the epithelium in one variety of these malignant growths 
should always produce, although in an abnormal manner, the form of 
cell in which it in a physiological manner ends its existence, while 
another variety preserves its embryonic or original character. 
We may suppose that certain epidermic cells have in themselves 
the power to develop into horny tissue, while others do not possess 
such power, or are destined for another purpose, such as the formation 
of the glandular appendages. 
In that rare variety of malignant cancerous new growth in which 
pigment-bearing cells predominate, it is possible that yet another 
variety of epithelial cell is first involved. 
* Read at the Eighteenth Annual Meeting of the American Dermatological Association, 
Washington, D. 0., May, 1894. 2 
Original Communications. 
The tubular variety of skin cancer, to which type the so-called 
rodent ulcer belongs, may originate from the epidermis, from the 
outer root sheath of the liairs, or possibly from the glands of the skin. 
Although many attempts have been made to show that it has a uni- 
form pathology, such efforts have been unsuccessful, as in specimens 
removed from epitheliomata, which are clinically of the same class,, 
various points of origin, either alone or combined, have been noted. 
From whatsoever structure epitheliomata arise, after a certain 
duration the original arrangement of the cells may be lost in the tend- 
ency which they show to arrange themselves in alveola which bear a 
close resemblance to those met with in mammary or other glandular 
carcinomata. I have, during some recent investigations of cutaneous 
new growths, excised a number of small pearly tumors which are seen 
with more or less frequency on the faces of middle-aged individuals. 
Such small tumors not infrequently exist for many years without ulcer- 
ation or any change to indicate a malignant tendency. After excision, 
partial or complete, the wound heals kindly, and no sign has afterward 
been observed which would point to a recurrence of the growths. 
Microscopic examination has shown such tumors to consist of epi- 
thelial downgrowths from the epidermis, which in many cases had 
undergone a partial or complete degeneration of a hyaline nature. 
With this degeneration of the cell the power of proliferation which it 
originally possessed is lost and the new growth comes to a standstill. 
In other instances, however, the center of the tumor will slowly 
ulcerate while extension at the margin takes place, giving rise to an 
open epitheliomatous ulcer whose peripheral cells readily take the mi- 
croscopic stains and show active cell division. These pearly tumors 
have many points of resemblance, both in their clinical features and 
pathological anatomy, to the affection which has been described under 
the various names of hydradenomes eruptifs, adenomes sudoripares, 
syringo-cystadenom, adenoma of the sweat glands, benign cystic epi- 
thelioma, etc. 
This curious affection, at first considered to be a benign epithe- 
lioma having a sweat-gland origin, has been shown to originate in the 
epidermis and to be independent of the coil glands. Its benign na- 
ture, too, has been questioned, as in the case lately reported by Dr. 
White, open epitheliomatous ulcers resulted from a number of the sup- 
posed benign tumors. As the predilection of cancer in general for 
the glandular structures is so strongly marked, the temptation has 
been strong among many pathologists to attribute to certain cutaneous 
epitheliomata such an origin without sufficient proof that such was the 
case. An examination of many of the usual forms of cancer situated Adeno-Carcinoma of the Shin Originating in the Coil Glands. 
at a distance from the mucous membranes will show a structure made 
up of tubules of small epithelial cells, which suggest in a striking man- 
ner overgrown and abnormally proliferated sweat ducts or glands. 
This is the case in the so-called benign epithelioma and in other varie- 
ties of skin cancer. 
Although a wide difference of opinion exists among pathologists 
regarding the rdle played by the sweat glands in tumor formation, 
there are few well-authenticated examples of adenomata which can 
be directly traced to these organs. Yirchow {Die hranhhaften Ge- 
schwulste, Bd. hi) had never seen a single example of such a tumor. 
Rindfleisch (lehrhuch der patholog. Gewebelehre) found a fiat, warty- 
like elevation of the skin to be on section a tumor originating in the 
coil glands. Peterson {Archiv fur Dermat. xtnd Syph., 1892, p. 919) 
reported a case of nsevus verrucosus unius lateris, which proved on mi- 
croscopic examination to be an adenoma of the sweat glands. He 
expressed the view that the tumor originated not from the fully devel- 
oped gland, but from some disturbance in its embryonal development. 
In a later article (Beitrage zur Kenntniss der Schweissdrtisen-Erkrank- 
ungen, Archiv fur Dermat. und Syph., 1893, p. 445) he concludes, 
from a further examination of the same case, that the tumor started 
from the ducts of the fully developed glands. 
Elliot {'Journal of Cutaneous and Genito-Urinary Diseases, 1893, 
p. 168) found a similar pathological condition of the coil glands occur- 
ring in a nsevus unius lateris of the shoulder, which had existed since 
childhood. Elliot preferred to give the name adeno-cystoma to the 
pathological condition which he found, there being no new formation 
of glandular tissue. He further believed the process to be secondary, 
and developed in consequence of the irritation to which the nsevus 
had been subjected. 
While it is not improbable that adenomata of the sweat glands may 
remain in a quiescent state for a long time before giving rise to malig- 
nant epitheliomata, there is very little proof, either in the clinical 
course or pathological anatomy of such malignant growths, to show 
which ones had such a beginning. 
Thiersch and Waldeyer have described primary cancer of the 
sweat glands, but the attempts which have been made by Cornil and 
Ranvier, Thin, and others to prove that all tubular epitheliomata of 
the skin have such an origin have not been successful. Neither the 
shape, size, nor arrangement of the cells are sufficient criteria on which 
to base the diagnosis of sweat-gland epitheliomata. 
Darier (Contributions a I’etude de I’epitheliome des glands 
sudoripares, Archives de med. experimentale et dianat. pat hoi., 1889, 4 
Original Communications. 
pp. 115 and 267) lias reported a peculiar case occurring in the service 
of Besnier, in which multiple, deep-seated, and painful tumors devel- 
oped with great rapidity on the trunk of a man aged seventy-one. 
Darier’s examination showred that each tumor started with a prolifera- 
tion of the sweat-gland epithelium, which filled the lumen of the canal, 
broke through the membrana propria, and infiltrated the connective 
tissue. He gave to his case the name “ diffuse multiple epithelioma 
of the sudoriparous glands,” and believes it to be unique in the num- 
ber, the rapid evolution, and the structure of the tumors. The author 
was able in medical literature to find but two analogous cases, one 
of which was reported by Malherbe, the other by Creighton. In the 
latter case a dog was the subject of the affection. 
Darier {hoc. ait.) has quoted a number of cases collected from litera- 
ture in which a sweat-gland origin of the tumors in question was 
probable. These cases are briefly as follows; 
Henocque and Souchon {Gaz. hebdom., 1866, p. 310) observed a 
tumor on the back of a man aged thirty-four which had existed for 
eight years. At first its growth was slow, later it became more rapid. 
The authors concluded from its microscopic structure that it was an 
adenoma of the sudoriparous glands. 
Cristol {Gaz. hebdom., 1886, p. 364) observed a tumor with a similar 
structure on the back of a woman aged thirty-three. The growth was 
stationary during many years, but finally ulcerated deeply, and was be- 
lieved, on microscopic examination, to have had a sweat-gland origin. 
Chandelux (Arch. de physiol., t, xiv, 1882) published a report of a 
painful subcutaneous tumor the size of a pea which had existed six 
years on the forearm of a woman aged forty-three. The author was 
able to determine a proliferation of the epithelium lining the sweat 
glands, which he believed to be the starting point of this growth. 
Lienaux {Ann. de med. veterinaire, avril, 1888) examined a tumor 
the size of an egg which affected the skin of a dog. His examination 
showed pretty conclusively that the glands under consideration gave 
rise to the new growth. 
Cornil and Banvier {Jour, de Vanat. et de la phys., 1865, pp. 256 
and 466) were able to demonstrate the development of a cancer of the 
leg from the coil glands. The careful study of this case by the authors 
leaves little doubt regarding the structure first involved, which is ad- 
mitted by Darier to be incontestably an epithelioma of the sweat 
glands. 
ZD 
Equally good pathologists differ regarding the part played by the 
sweat glands in the production of tubular epithelioma. While it is 
admitted by many that their epithelium may proliferate in many Adeno- Carcinoma of the Skin Originating in the Coil Glands. 
forms of skin cancer, it seems difficult to determine in what class of 
cases this proliferation is primary. The slow clinical course and non- 
implication of the lymph glands were supposed, among other symp- 
toms, to indicate a growth which, took place at the expense of the 
sudoriparous glands. When in a growth of this character epithelial 
processes recalling the shape of the sweat ducts were encountered, the 
conclusion was reached that these glands were their original point of 
departure. It is an error to suppose that any universal law can be 
formulated from either the clinical course or histological structure of 
such growths ; their location on the skin away from the mucous ori- 
fices, where there is less irritation and less vascularity and lymphatic 
supply, as has been observed by Besnier, Hutchinson, and others, 
seems to have more to do with their benign course than a fixed origin 
or structure. 1 have examined many cases of such epitheliomata from 
the face, presenting typical tubular prolongations of epithelium, with 
no tendency to the formation of horny epithelium, and very frequently 
could trace their origin to the interpapillary processes of the rete or to 
the hair follicles, both the sweat and sebaceous glands being unaltered. 
While it is not denied that either or both of these glands may give rise 
to the tubular epithelioma, the proof that such is the case is difficult 
to obtain, as the ulcerative process so soon destroys the tissues which 
may be first involved. 
It is probable that rodent ulcer and tubular epithelioma are patho- 
logically identical. Thin {On Cancerous Affections of the Skin, 
London, 1886) has expressed this view, and has endeavored to associate 
it with a sweat-gland origin. lie does not believe it to be identical 
with an epithelioma starting from the rete Malpighii, though admit- 
ting that the clinical diagnosis is not easy to make. Walker {British 
Journal of Dermatology, September, 1893) is the most recent advo- 
cate of the glandular origin of rodent ulcer. His argument is that in 
its structure it corresponds to glandular cancer, and that when car- 
cinoma of the breast secondarily attacks the skin it is almost identical 
with rodent ulcer. 
In most of his cases he was able to trace an origin to the sweat 
glands. The slow course of the ulceration is to be attributed, accord- 
ing to this author, to the normal slow growth of the glandular epithe- 
lium, the rete cells having a rapid growth. 
The case which forms the subject of this report differed in its clini- 
cal features from the usual type of cutaneous cancer in which a sweat- 
gland origin has been claimed. Unfortunately, the duration of the 
tumor before the stage of ulceration was not noted. It may be said 
to have some clinical resemblance to the cases reported by Henocque and Original Communications. 
Souchon and by Cristol. It would have been a matter of some 
interest to have determined whether it began as a deep or surface 
affection. 
Dr. Bronson, in whose service at the City Hospital the case oc- 
curred, excised the tumor from the anterior surface of the leg of a 
Fig. ].—(Spencer 1 in. x 60.) Showing the distribution of the epithelial tissue and its 
adenomatous structure. 
man thirty-five years old, and kindly gave it to me for examina- 
tion. 
It is interesting to note that the cancerous nature of the new 
growth was not suspected prior to a microscopic examination made 
before the operation of excision. 
The growth was composed of distinct lobules, separated by the 
ingrowth of connective tissue; it ivas soft to the feel, about half the Adeno- Carcinoma of the Skin Originating in the Coil Glands. 
7 
size of a small egg, and defined from the surrounding tissue without 
being distinctly encapsulated. It extended about one inch below the 
skin, and measured an inch and a half in its long diameter. An open 
ulcer, somewhat larger than a silver dime, was present over the tumor. 
Portions of the growth were fixed in Flemming’s solution, Foa’s solu- 
tion, and in absolute alcohol. Sections were stained in hsematoxylon, 
safrarine, Biondi’s mixture, and in other ways. 
Under a low power (Spencer one inch, Fig. 1) the derma is seen to 
be occupied by bands, masses, and glandlike arrangements of small 
epithelial cells which stain imperfectly in places. The cell masses and 
bands extend from just below the epidermis, which is unchanged ex- 
Fig. 2.—(Spencer in. x 150.) More highly magnified field of a portion of Fig. 1 
cept at the location of the ulceration previously mentioned, to the sub- 
cutaneous connective tissue which is penetrated by them. 
They differ in their arrangement from that met with in other 
varieties of skin cancer, suggesting to the observer at the first glance 
an attempted reproduction of glandular tissue, which becomes still 
more noticeable when higher amplification is employed (Fig. 2). 8 
Original Communications. 
Here a highly complicated network has been formed by the inter- 
lacement of bands of epithelial tissue, inclosing cavities which are 
partially or completely filled with degenerated cells or with a homo- 
geneous substance allied to colloid matter. 
Cystlike cavities are seen in other parts of the microscopic field 
which are lined by a single layer of columnar epithelial cells which 
present a striking resemblance to overgrown sweat ducts. 
In other instances the cell layer is double and quite regular in 
arrangement; while, again, the overgrown ducts or glands are com- 
pletely filled with cells or show a localized overgrowth at one or more 
points. In Fig. 3 an overgrowth of columnar epithelium is shown 
which probably represents an early stage of the malignant process. 
Fig. 3.—( x 300.) Showing the beginning proliferation of the glandular epithelium. 
Fig. 4 (plate) represents a group of enlarged and proliferating 
sweat ducts cut transversely. The columnar shape and regular ar- 
rangement of the peripherally located cells are well shown. These 
cells are situated upon a basement membrane, and are an almost exact 
reproduction of those normally met with in the sweat glands, although 
here very much hypertrophied. The lumen of the tubules is com- 
pletely filled by the growing and multiplying cells, and in places the 
basement membrane is penetrated and the surrounding connective tissue 
infected by them. This photomicrograph is made from a section taken Adeno-Carcinoma of the Skin Originating in the Coil Glands. 
9 
from a portion of the tumor fixed in Flemming’s solution. Conse- 
quently the cells are much less deeply stained. 
In Fig. 5 (plate), where the same amplification is employed, the 
tubules are larger, of irregular size, and their caliber completely occu- 
pied by the newly formed cells, which are also seen in places breaking 
through the membrana propria and infiltrating the connective-tissue 
stroma. This infection of the surrounding connective tissue may be 
accepted as positive evidence of the malignant character of the new 
growth. 
Figs. 6 and 1 (plate) represent in a striking manner the adenoma- 
like structure of the new growth, together with the intracanalicular 
proliferation of the basement cells. The cavities are lined by a double 
row of cubical cells, which are united in places by connecting bands, 
dividing the lumen of the tubes into various sized and shaped cavities. 
The outlying portions of the tumor have lost their glandlike struc- 
ture, and assume the appearance of an alveolar carcinoma. 
The other anatomical structures of the skin were not implicated in 
the cancerous process, neither was there any evidence of a tendency to 
cell-nesting or the formation of epidermic “ globes,” The epithelial 
processes met with in the small-celled cancer of the skin, starting in 
the rete Malpighii, resemble at times some of the cell arrangements in 
the case under consideration. In the epidermic cancer, however, the 
cells are larger, and do not rest upon a distinct basement membrane, 
as in this case. Furthermore, the epithelial bands or tubules starting 
from the rete have no distinct caliber, and do not reproduce so per- 
fectly the glandular structure. 
Frequently, when a new growth has reached the size of this one, 
its point of departure has been destroyed by ulceration or in some way 
obliterated, so that it is not possible to determine from what structure 
the epithelial proliferation was primary. The adenomatous character 
of the new tissue is so apparent here, and differs so radically from 
those cases starting from the other anatomical divisions of the skin, 
that an error of this kind could scarcely be made. Active cell division 
(karyokinesis) was well shown in the Flemming preparations, stained 
with safranine, denoting a rapid growth of the new tissue. 
The intracellular bodies (so-called parasites of cancer) will not be 
considered at this time, as I desire on some future occasion to discuss 
their relationship to the cancerous process. 
While believing that the manifestations of the malignant disease 
can best be explained by the parasitic theory, I am not prepared to 
admit that its cause is found in these bodies. 
Prognosis.—Unfortunately, the case here presented has not been 10 
Original Communications. 
seen since the tumor was excised, so that the question of recurrence of 
the growth can not be considered. 
Such cases are too rare to permit any expression of opinion re- 
garding their malignancy as compared to other cutaneous.cancers.* 
66 Park Avenue. 
* The photomicrographs used in illustrating the article were taken with the lenses made 
by the Spencer & Smith Optical Co., of Buffalo, N. Y., especially corrected for photography 
with yellow or yellowish-green light as suggested by Dr. H. G. Piffard, of this city. They are 
reproduced by the half-tone process, and necessarily lose some of the details seen in the 
originals.