THE AMERICAN TOURNAL OF CANCER d A Continuation of The Journal of Cancer Research

THE AND THYMIC HYPERPLASIA IN WITH OBSERVATIONS ON THE GENERAL '

E. H. NORRIS, M.D (From the DcpnrtmPnt oj Pathology, Unizlersity of Minnerota)

Since the comprehensive study by Bell in 1917 no systematic treatise on the thymic tumors of myasthenia gravis has appeared. It might be expected that the past twenty years would have added the records of many more cases to the literature and have extended our knowledge of the pathological anatomy of myasthenia gravis and the associated thymic tumors. On the contrary but few cases with autopsy findings have been reported and our conceptions con- cerning the morbid anatomy of this disease are much the same as they were in 1917. In the period from 1901, the year in which Weigert published the record of the first case of myasthenia gravis with an associated tumor, to 1917, Bell found published reports of 56 autopsied cases of this disease. He wrote: '' The thymus was described as enlarged in seventeen of these cases; and in ten others it contained a tumor. Therefore some form of thymic involvement seems to occur in nearly half the cases of myasthenia gravis." Since 1917 the records of 6 more autopsies have been published and 5 of these showed thymic lesions. Of the 4 cases to be reported here, 2 had thymic lesions which were recognized. I am of the opinion that pathologic changes may be found in the thymus in cases of myasthenia gravis in direct ratio to the care with which they are sought. Thus, prior to 1901, when Weigert first called attention to the association, the presence of changes in the thymus had been completely overlooked. In 1912 Starr reviewed 250 cases of myas- thenia gravis, including cases reported prior to 1901 in which the thymus was probably not carefully investigated. His figures showed thymus involve- ment in only 28 per cent of the cases, while Bell's analysis of 56 autopsies between 1901 and 1917 yielded an incidence of thymic involvement in 50 per cent. W'ith the addition of our 4 cases and the 6 reported elsewhere since Read before the American Association for C;tnccr Research, Boston, Mass, April 8, 1936. For discussion scc pa~e557. 42 1 42 2 E. €1. NORRIS

TanLI< I : Reported Lesions of the Thymus in Myaslheizia Gravis *

I’n t ien t

(‘,lbC No. Eulargetl Age or Per- Tumor sistent _. 1 1901 I.,iquer

* In ii .‘rianer bv, I3utt the ureseuce of a thvmoriia two out of seven autopsied cases is men- t ioned. IIoRever, none of these seven cases is given individual consideration and they cannot, for tIi,il rensori, be inclutletl in this t.ible. 1917 the incidence of thymic lesions remains at 50 per cent. This figure probably expresses the frequency of grossly recognizable lesions of the thymus. It seems quite possible, however, that thymic lesions which have produced little or no macroscopic alteration of the suprapericardial tissue may have been overlooked in some of the reported cases. In the first place, as will be pointed out later on, the thymic lesion may not be recognizable without a microscopic examination. In the second place the thymic region in the post-adolescent ages is frequently passed over in the routine necropsy with the most cursory inspection. These facts have doubtless combined to lower the figure which expresses the frequency of reported lesions of the thymus. THYMOMA AND THYMIC HYPERPLASIA 42 3

The reported autopsy cases of myasthenia gravis in which thymic lesions have been found are listed in Table I, in chronological order. The first 27 cases comprise the 17 which were tabulated by Bell and the group of 10 thymic neoplasms which were analytically studied by the same investigator in 1917. These are followed by the 6 cases recorded since 1917, and finally the author’s two cases are appended. It will be convenient first to summarize the cases which the literature already records.

(1) BELL, 1917: 2 The patient was a male, fifty-eight years old, with a typical clinical picture of myasthenia gravis and a retrosternal, encapsulated, nodular tumor which meas- ured 6 X 3.5 X 3.6 cm. and weighed 60 gm. The cut surface of the tumor showed soft gray tissue. “ The tumor tissue consists of cells with large vesicular and abundant light-staining cytoplasm, fused together to form a syncytium. Throughout the syncytium are numerous spnces of variable shape and size, containing small lymphocytes. In some areas the lymphoid cells predominate, in other areas the epithelial cells. . . , The cells with large clear nuclei are thymic epithelial cells. . . . No corpuscles of Hassall are to be seen.” (2) BOIJTTIERAKD BERTRAND,1921: The patient was a middle-aged woman with typi- cal myasthenia. The thymic region showed the presence of small grayish red masses, and these on section proved to be thymic tissue. Lymphoid cells, Hassall’s corpuscles and reticulum cells are described. The fact that thymic tissue was present in sufficient amount to be recognized grossly is enough to classify this as a case of so called persistent thymus. (3) MELLA, 1923: A colored male, forty-eight years old, had typical myasthenic symptoms for nine months. In the thymus a discrete mass measuring 7.5 X 5.5 X 2.5 cm. was found. The structure of this thymic mass is described as made up of closely packed lymphocytes with a few Hassall’s corpuscles and with no signs of degeneration or inflam- mation. (4‘) ALTERAND OSNATO,1929: The patient was a woman thirty-one years old. A large thymic tumor measuring 9 X 7 X 6 cm. was present. This tumor had no regular structural arrangement. It was made up of lymphocytes and epithelial cells, and the presence of young capillaries, eosinophiles and polymorphonuclear leukocytes was noted. Degenerative changes, such as hemorrhage, fibrosis, and calcification, are described. In addition to the mediastinal mass a sepnrate cervical nodule was found which had a calcified capsule. The central soft part of this nodule, however, showed the same structure as did the mass in the thymic region. (5) BREMASD W’ECHSLER,1931: A woman of twenty-seven years had typical myas- thenia for three and one-half months. ‘‘ In the anterior mediastinum was a mass, appar- ently composed of two structures, extending down the anterior surface of the pericardium. The upper half had the appearance of a hyperplastic thymus gland, except that it was some- what firmer than normal; the lower half, rising from the inferior pole of the gland, was a firm, oval, encapsulated nodular mass, measuring approximately 5 cm. in its longest and 2.5 cm. in its shortest diameter. On section, it was grayish white and trabecular and con- tained several areas of calcification. “Sections through the upper portion of the mediastirial mass showed it to be the thymus gland. The lower circumscribed mass had the typical structure of a thymoma. . . . Sections from different portions of the tumor showed a somewhat varied picture. For the most part, it was composed of large, anastomotic sheets of polyhedral cells, separated by small amounts of vascular fibrous tissue. These cells had large vesicular nuclei and definite nucleoli. The cytoplasm was acidophilic and granular, and the cell outlines were indistinct. Some of the cells had a flattened appearance. Small round cells were scanty and were situated mainly about the blood vessels. In other areas, large amounts of dense fibrous stroma divided the tumor cells into irregular islands, and round cells were more conspicuous. In still other areas, the structure approached that of the thymus gland, round cells pre- dominating over the thin cords of polyhedral cells. The latter lay adjacent to the ca$- laries. R/litoscs were infrequent. No plasma cells, eosinophils or Hassall’s corpuscles were seen.”

”For analysis of CBSCS prior to 1917, see Bell: J. Nerv. & Ment. Dis. 45: 130, 1917. 424 E. H. NORRIS

(6) RREM AND WECHSLEH,1034: A fifty-four-year-old man had typical myasthenia Tor one month. “The region of the thymus was occupied by a very dense, encapsulated, sausage-shaped mass, measuring 9 X 3 cm. It lay transversely across the trachea, great vessels and upper attachments of the pericardium. It was difficult to cut, and its surface prcscntecl a lobulated structure. The lobules were creamy yellow, and were separated by tlcnse bands of fibrous tissue. Sections through the niediastinal tumor showed it to be completely encapsulated by J thick layer of dense fibrous tissue from which wide trabcculae radiated into its substance,

Fll:S. 1 AND 2. CASE I: VENTRAL. AND DORSAL VIEWS OF THE THYMICMASS AS IT WAS REMOVED AT AUTOPSY Note the cervical horns in the upper part of the photographs. In Fig. 2 the body of the tumor has been sectioned and laid open, showing the dense, gray tumor tissue, the fibrous capsule, and the cyst cavity. The cervical horn on the left in this photograph had a solid lobular structure :ind wits made up of hypcrplastic thymus tissue; thr broader cervical horn to the right was cystic. giving it a lobulated appearance. The trnbcculae had a peculiar arrangement, branching off the main trunk like the limbs of a tree. The parenchyma was composed of large poly- hedral or flat tcned reticular cells. Lymphocytes were scanty, and only an occasional plasma cell wiis present .” Branch, 1936, reported the autopsy findings in a case which showed no hyperplastic changes in the thymus, the patient being a young woman who had suffered severe mynsthenic symptoms for six months. The following 4 cases are taken from the records of the Department of Pathology of the University of Minnesota since 1917. THYMOMA AND THYMIC HYPERPLASIA 425

CASEI (A-36-58): A white female, thirty-seven years old, had typical myasthenic symptoms during a period of six months, complaining first of difficulty in mastication, then of dysphonia, dysphagia, urinary incontinence, and spasmodic dyspnea. Under clinical ob- servation the temperature and pulse rate were normal; the respirations varied from 40 to 60. The routine blood and spinal fluid examinations were normal. Physical examination yielded negative findings. The patient expired in an attack of spasmodic dyspnea. Upon raising the chest plate a tumor mass presented in the right side of the superior mediastinum. The tumor was not adherent and it was readily dissected from the superior

FIG.J. CASEI. Hicii POWER PIIOTOMICROGRAPHSHOWING THE STRUCTUREOF TIIE FLESHYPORTIOZ~ or TIIE TUMOR The cellular arrangement seen in the upper half of this figure is typical of most of the tumor. Note the large epithelial cells, which are the essential cytological element. Areas with many lymphocytes, such as that shown in the lower half of this figure, were less characteristic.

vena cava and the ascending aorta. Projected upon the chest wall, the tumor extended from the sternoclavicular joint to the level of the third costosternal joint; its location was to the right of the mid-line. Examination of the rest of the mediastinum showed no other masses or enlarged lymph nodes. After removal the thoracic tumor was found to consist of three portions. Above there were two horn-like processes which extended toward the neck (Fig. 1); these occupied the position of the cervical horns of a normal pre-adolescent thymus. The right horn-like process measured 3 X 2 X 1 cm., and the left 4 X 1 S 1 cm. The right horn was cystic and the left was made up of lohules of solid tissue. The lower fleshy portion of the mass measured 5 X 4 X 3 cm. and was well encapsulated by a tough fibrous membrane (Fig. 2). The cut surfaLe of this part showed a grayish white tissue of firm consistence. Between the upper and lower portions was a large simple cyst (Fig. 2) containing clear fluid. This large cyst did not communicate with the cyst in the right horn. The over-all measurements of the thymic mass were 10 X 4 X 3 cm. and the weight 57 gm. Microscopic examination of the fleshy portion of the tumor shows an almost uniform structure throughout. The essential cells of the tumor are large epithelial cells arranged in the form of a syncytium; these have indistinct cell walls, a considerable amount of homo- geneous, aridophilic cytoplasm, and large ovoid or irregularly block-shaped vesicular nuclei in which thca chromdin is distributed in a fine network. A large nucleolus is regularly present. In the interstices of this epithelial syncytium a relatively small number of small lynphocytes are rather regularly distributed (Fig. 3) ; but, less characteristically, in a few places the lymphocytes have accumulated in greater numbers. No Hassall's corpuscles 426 E. H. NORRIS

Lz’erc lounti within the encapsulated tumor mass. The structure of this tumor corresponds :ip~)roximatclyto that of the thymus of a human fetus of 80 or 90 mm. crown-rump length. ‘I’hc vascularization of the parenchyma is accomplished by large numbers of capillaries antl fewer t hin-walled vessels. The capillaries form ;t close network which appears to provide ndequnte nourishment. for there is no necrosis in the tumor. There is almost no fibrous strom:t and no septa divide the tumor. The fibrillar reticulum was studied with I’erdrau’s modification of Uielschowsky’s method. Only a small amount of reticulum was present, ;ind the intimate association of the reticular fibers with the capillaries and blood vessels was t hc striking feature. Sections through the walls of the large cyst and the cystic cervical process show the 1)rcseiice of a flat endothelium-like epithelium lining the cavities. In the dense fibrous tissue which constitutes the outer wall of the cyst, scattered islands of lymphoid tissue are present; none of these show germinal centers, and certain of thcm may be recognized as thymic tissuc by the presence in them of masses of hyperplustic thymic epithelium (Fig. 4). The epithelid cells of these hyperplastic regions (Fig. 5 ) are quite similar in form and structure to the essential cells of the main part of the tumor. Microscopic sections taken through the left cervical horn and through the tissue which I;iy on the outside of the fibrous capsule of the fleshy tumor mass show the prescnce of tissue which has the appearance of a persistent thymus (Figs. 6 and 7). Lohules of thymic tissur in which medullary and cortical zones are differentiated are present. In these lobules 1;irgr Himall’s corpuscles are found, and these have the structure of those of ;in adult thymus (Fig. 6). The medullary portions of the lobules were made up of closely picked (~pithelialcells ; in some instances this medullary structure did not corres;iontl to the centers of the loliules but was found spreading out peripherally in sheets into the surrounding areohr tissue (Fig. 6). The histologic appearance of the tissue from these two regions is that of the adult thymus in which an hyperplastic proliferation of the thymic epithelium is going on (Fig. 8 ), Exmination of muscle from the diaphragm and gastrocnemius show the presence of tmny Iymphorrhages (Figs. 11, 12, 13). No evidence of atrophic or degenerative changes could he made out in the fibers of ;my of the muscle examined, antl the striations were rcgularly apparent even in fibers contiguous to the Iymphorrhages. The brain, hcnrt, lungs, liver, . and kidneys showed no gross or microscopic evi- dence of disease. Several small adenomyomas were present in the uterus. There was a large corpus hemorrhagicuin present in the right ovary. The thyroid apl)e;ired normal in size; the pnrenchyma was made up of small atrophic follicles md there were considerable interfollicu!nr fibrosis antl lymphocytic infiltration-the picture usuully regarded as chronic thyroiditis. In addition, localized aggregations of I!mphoitl tissue were organized in the form of germinal centers (Fig. 14). The atlrenals :q)pearetl normal except for the presence of some perivascular, lympho- cytic infiltration in the metlulln. ‘The hypophysk showed hyperp1asi;i of the chromophohes in the peripheral parts of thct anterior lobe, with :I tentlency for these to become acidophilic; the cell ratio was normal. For these observations on the hypophysis I am indebted to Ur. A. T. Rasmussen. CASE I1 (A-2s-669): 11 white male, thirty-three years old. had typical symptoms of m\asthenia gravis for two years and ;I half. During the last six months of life the patient was troubled with epigastric distress and vomiting. From the x-ray findings these abdominal symptoms were thought to he due to duodenal atony. In the anterior mcdinstinum a small, firm, non-encapsulated mass was found. The mass Imsentetl several nodular elevations posteriorly antl R large nodule anteriorly, which meas- ured 1.5 cm. in diameter. The tumor mass measured 6 X 5 X 1.5 cm. and weighed 17 gin. On scction it wis composed of firm, grayish-pink tissue of rather uniform structure. 3licroscopic scictions of this tumor show that a part of it has a lobular arranpient; tlitw. lol~ulcs;ire tongue-like in outlinc and suggest the form and arrangement ol those of t hi’ norii1;iI I h!mius. Thcw t ongue-like lohules are sepur;ttetl hy a small amount of loo st^ ;irtwl:ir lissut. ( Fig. 9 ), A further mtl more irregular division of other parts of the tumor ih :ic.c.oml)lisIictl I)y the pr~wnc~~of kmad septa of dense fibrous tissue. The impression is strong that these dense fibrous septa do not represent essential parts of the growth but that thry ni;iy 1)c only accidental inclusions about which the tumor may have grown. FIGS. 4-8. CASE 1: PIIOTOMICHOGRAPITS OF TUMOR Fig. 4. Transverse section through the right cervical horn of the thymic mass (low poaer). Vote the cavity oi the cyst (C) lined by flat epithelium. In the outer wall of the cyst two masses of thymic tissue are seen. H.T. = Hyperplastic thymus epithelium. Fig. 5. Region at the end of the leader marked H.T. in Fig. 4 (high power). Note the large syncytial epithelial cells. Fig. 6. X transverse section through the left cervical horn of the thymic mass (low power). The thymus tissue is definitely hyperplastic, and cortical and medullary zones may be recognized. H.C. - Hassall's corpuscle. High power views of the hyperplastic portions oi this section show ;i structure similar to that shown in Figs. 5 and 8. Fig. 7. Section taken through the fleshy portion of the tumor, in which the fibrous capsule of the tumor and some of the overlying tissuc are shown (low power). Note the lobule. oi hyperplastic thymic tissue. F.C. = Fibrous capsule. H.T. = Hyperplastic thymus lobule. T. = Fleshv portion of thymic tumor. Fig. 8. Region marked H.T. in Fig. 7 (high power). Note large syncytial epithelid cells. 42 7 428 E. 11. NORRIS

The fundamental tumor tissue is made up of sheets and cords of syncytial epithelial tells In the interstices between the anastoniotic sheets and cords a few lymphocytes are sprstxly scattercd (Fig. 10 ). The number of lyniphocytes per field is distinctly less th;u~ 111 the tumor tissue of C"ise I. '4 few Ilussall's corpuscles are present. The epithelial cells oi thc tumor arc in ,111 respects similar to those described in Case I. Apparently the \lructure of this tumor corresponds to a slightly earlier fetal stage than the one described in Case I, for there is less lymphocytic infiltration and the epithelial syncytium resembles that oi an earlier period of orpnogenesis. There was a marked atonic dilatation of the entire duodenum. The lungs showed a bronchopneumonia and old healed apical tuberculosis. The brain, thyroid, heart, liver, spleen, kidneys, and adrenals showed no gross or microscopic evidence of disease. Sections from the rectus abdominis muscles show a few lymphorrhages. No evidence of atrophy or degeneration of any muscle fibers is to be seen. CASE 111 (A-33-1373): A white male, sixty-nine years old, had typical symptoms of myasthenia gravis for six months. At the autopsy nothing but fat was found in the region of the thymus. No microscopic examination of tissue from the thymic region was made. The brain, heart, lungs, liver, spleen, kidneys, and adrenals showed no gross or micro- scopic evidence of disease. The skeletal musculature was not examined microscopically. CASEIV (A-33-1973) : A white female, thirty-one years old, had typical symptoms of myasthenia gravis for two years. At autopsy no abnormal mass was found in the thymic region by palpation. No tissue was taken from the thymic region for microscopic examina- tion. The hrain, heart, liver, spleen, kidneys, and adrenals showed no gross or microscopic evidence of disease. The lungs showed a bronchopneumonia. A corpus hemorrhagicum was preaent in one ovary. The skeletal musculature was not examined microscopically.

DIscussIoN A comparison of the descriptive findings set out in the last two columns of Table I leads to an evident conclusion. Of the 35 cases tabulated, 18 were classified by the various authors as representing instances of an enlarged or persistent thymus and 17 cases were classified as tumors of the thymus. These reported cases, therefore, are divided into two almost equal groups. .4 casual survey of the descriptive data is sufficient to convince one that the line of separation is not sharp and that the bases for distinctive classification are uncertain. It will be seen that in certain of the cases listed in the first column the thymic inass was as large or larger than certain of those designated as tumors and placed in the last column. The difficulties with which one may be confronted in differentiating be- tween a benign thymoma and an enlarged thymus in which there is an exten- sive epithelial hyperplasia is well illustrated by Cases I and I1 of our group. Even in the same case, as typified by the findings in Case I, uncertainty niay be encountered if a forced distinction must be drawn. In this case a discrete, encapsulated mass of fetal thymic tissue was present whose structure answers all the requirements for the diagnosis of a benign thynioma. On the other hand, the presence of hyperplastic thymic epithelium, resembling in all mor- phological respects the epithelial elements of the encapsulated mass, was found in the one cervical pracess of the thymic mass and also in the lobules of thymus tissue overlying the capsule of the tumor. Again in Case 11, both the size and the histological findings in the thymic mass justify the diagnosis of thymoma, notwithstanding the facts that the mass lacked a capsule and that its structural arrangement suggested the lobulations of a normal thymus. THYMOMA AND THYMIC HYPERPLASIA 42 9

In order to understand properly the pathogenesis of the lesions with which we are here concerned, it is necessary to have in mind the histogenesis of the thymus. It will be remembered that until the fetus is approximately 75 nini. in length the thymus is made up entirely of epithelium and that only in later development does the epithelial body become invaded by lymphocytes, and a differentiation of cortical and medullary zones become apparent. I am

FIGS.9 AND 10. CASE 11: PHOTOMICROGRAPHS(Low AND HIGHPOWER) OF THE TUMOR Note the lobular structure of the tumor as shown in Fig. 9. In Fig. 10 note the large, syncy- tial, epithelial cells and the sparsely scattered lymphocytes. of the opinion that the pathologic findings which are present in the thymus in niyasthenia gravis are best interpreted as conditions of greater or less degrees of epithelial hyperplasia. When the hyperplasia is extreme a localized, and at times encapsulated, tumor-like mass is formed. In these instances of more extreme hyperplasia the usual lobular structure of the thymus may be com- pletely obliterated and only very few lymphocytes and almost no Hassall’s 430 E. H. NORRIS

FIGS.11-13. CASE 1: PiIOTOhiiCHoOH,\iliis OF MIWLE SECTIONS Fig. 11. Section of muscle from the diaphragm to show a lymphorrhage (high power). Musclc fibers cut in transverse plane. Note how the lymphocytes surround and spread out nar- rowly from ii fcw capillaries. Observe the normal :ippearance of the muscle fibers. Hematoxylin and eosin. Fig. 12. Section taken from the gastrocnemius muscle, to show a Ipmphorrhage (high power). Muscle fibers cut obliquely. Note how the lymphocytes spread out from the capillary. Observe the striations and normal appcarancr of the muscle fibers. Iron-hematoxylin and eosin. Fig. 1.3. Section of musclc from the diaphragm to show a lymphorrhage (high power). Muscle fibers cut longitudinally. Note the striations and normal appearance of the muscle fibrrs. Iron-hematoxylin and eosin.

corpuscles may be found in the epithelial mass. On the other hand, when the hyperplasia is considerable but less extreme, the lobulation of the the thymus may persist and the corpuscles of Hassall and lymphocytes may be relatively more numerous. In cases where only little hyperplasia is evident the gland THYMOMA AND THYMIC: HYPERPLASIA 43 1

FIG. 14. CASE 1: PIIOTOMICROGRAPH (Loll: POWER) PROM A SECTION OF THE THYROID Notc thr two large germinal centers, the relativr increase of stroma, and the small follicles. Iicniato\) lin and eosin. structure may simulate that of the pre-adolescent period, in which broad sheets of hyperplastic epithelium may be found and in which cortical and medullary zones may persist. For these reasons, therefore, it seems that the benign thymoma of myas- thenia gravis should not properly be regarded as a true neoplasm, but rather as the result of a marked degree of hyperplasia. That tissue is formed in the benign thyinoina which resembles that of stages in the organogenesis of the thymus might seem to justify the use of the term fetal adenoma. The term persistent thymus does not seem satisfactory, for it commits us to the be- lief of incomplete adolescent involution. Because all of the conditions which have been described blend so well, when thought of as stages or degrees of hyperplasia, it seems best to refer to them as hyperplasias of the thymus, with the addition of a qualifying term to express degree. A consideration of the findings in our cases together with those presented by the literature leads us to suggest that the anatomical changes observed in the thymus in cases of myasthenia gravis may well be thought of as altera- tions in the gland which may have been produced in response to some stimulus from outside the gland; or possibly they may be regarded as alterations re- sulting froni some physiologic demand. The findings described by Alter and Osnato seem to support such deductive conclusions. These authors found a mediastinal tumor which presented the histological structure of a thymoma. In this same case a separate cervical nodule, which had a calcified capsule, was found, and the pulpy center of this cervical nodule showed the same histologic structure as did the mediastinal mass. From the anatomical point of view the reported findings in the general organs in cases of myasthenia gravis show great variability both as regards type and location. One is impressed, however, by the relative insignificance 432 E. 11. NORRIS of most of the lesions described and by the common reluctance of the author to attach much importance to them. Most have found no changes in the central or peripheral nervous systems, but on occasion a bit of perivascular lyniphocytic infiltration, or scnie degenerative manifestation such as chroma- tolysis or pigment formation in the nerve cells, has been noted. Such changes i~sthese, however, are not of any significance as far as the pathogenesis of the disease is concerned. In the first place, they are not sufficiently constant; in the second place, mere cellular changes without reaction of the neuroglia tissue are insufficient explanation for the clinical picture of myasthenia gravis. By far the majority of cases show no lesions in the central or peripheral nervous systems, and such as have been described are not more than incidental findings. Most investigators have observed no parenchymal changes in the skeletal musculature, but there are occasional reports of slight degrees of vacuolar or other forms of degeneration, and a few have described what they have considered as atrophic muscle fibers. It should be emphasized in this con- nection that great care is necessary before the certain recognition of atrophy of individual fibers in a niuscle fasiculus can be made. Cases of long duration may show atrophy, but it is generalized, non-degenerative in nature, and most probably caused by disuse. In any case I am convinced that such of these minor parenchymal changes in the muscles as have so far been mentioned occur too infrequently and are of too limited extent to constitute a part of the underlying anatomical picture of myasthenia gravis. The lymphorrhages, on the other hand, are quite characteristic lesions and have been described only in the muscles of myasthenia gravis. Morpho- logically they are quite different from the picture presented by the lympho- cytic infiltrations of other diseases, and the distinctive and descriptive term originally applied to them by Buzzard has significance. They have been found in nearly two-thirds of the cases, and the frequency of their occurrence makes them of supreme importance among the findings available for labora- tory diagnosis. That they have not been found in all cases may be ex- plained by the fact that it is obviously impossible to examine all of the musculature; for this reason negative findings have little importance, while the finding of typical lyniphorrhages in a niuscle biopsy may, on the contrary, be considered pathognomonic. From the point of view of the morbid anatomy of the disease the lymphorrhages are to be thought of as rather constant accompaniments of the malady and not as the cause of the disease or as an explanation for the clinical picture. Historically myasthenia gravis has been regarded from widely differing points of view. At first it was believed to be a disease of the central nervous system, then regarded as an entity whose picture was produced by metastases from a malignant thymus, and then as a primary disease of the striated muscu- lature of the body. Our point of view then seemed to be changing toward classifying it among the diseases of general metabolism, and finally the malady serms to be best regarded as a disturbance of the neuro-muscular mechanism, possibly upon an endocrine basis. The recent report of Butt, in which he described the finding of bacteria in microscopic preparations of the muscles from patients who died with myas- thenia, held our attention for a time. I also found bacteria in the muscles THYMOMA AND THYMIC HYPERPLASIA 43 3 from a case of myasthenia; but I found similar bacteria in muscles from a typical case of Addison’s disease, as well as in several other control cases. The work of Burn and others who have shown the frequent presence of micro- organisms in post-mortem tissues needs to be kept in mind. The further fact that no cellular reaction is found about these organisms in the muscles argues strongly against their having any essential relationship to the pathogenesis of the disease. A number of investigators have described lesions in one or more of the ductless glands in cases of myasthenia. Here again, even when the consid- erable frequency of thymus hyperplasia is taken into account, the morbid anatomy of the endocrine system in myasthenia gravis lacks constancy. A review of the literature has revealed the records of only 80 cases which have come to autopsy, and not a few of these have been incompletely reported. The records of thorough and complete autopsies in this disease are few, and the pathogenesis and morbid anatomy are still subjects of a great deal of controversy. It is important that all cases of myasthenia gravis with definite pathologic observations should be recorded. Recent advances of our knowl- edge from both the chemical and therapeutic points of view have shed a new light in which our ideas of myasthenia gravis are taking on better definition.

Su M MARY 1. Four cases of myasthenia gravis with autopsies are reported. In two of these a high degree of hyperplasia of the thymus was observed; the other two showed no gross evidence of thymic involvement. 2. Fifty per cent of the reported autopsied cases of myasthenia gravis have shown some form of thymic involvement. 3. Pathologic conditions which have been regarded as benign tumors of the thymus have been found in one-half of the cases of myasthenia gravis in which demonstrable changes in the thymus were recognized. Such cases I have interpreted as conditions of cxtrcmc epithelial hyperplasia. 4. Pathologic conditions which have been regarded as instances of enlarge- ment or persistence of the thymus have been found in one-half of the cases of myasthenia gravis in which demonstrable changes in the thymus were recognized. These I have interpreted as conditions of moderate epithelial hyperplasia. BIBLIOGRAPHY

ALTER,N. M., AND OSNATO,M.: Tr. Am. Neurol. Assoc. 55: 303, 1929. Also in Arch. Neurol. k Psychiat. 23: 345, 1930. BELL,E. T.: J. Nerv. & Ment. Dis. 45: 130, 1917. BRANCH,C. F.: Presented at meeting of Am. Assoc. of Path. and Bact., Boston, April, 1936. I~REM,J., AXD WECHSLER,H. F.: Arch. Int. Med. 54: 901, 1934. ROUTTIER, M. H., BERTRAND,I., AND MARIE,P.: Ann. de mird. 10. 173, 1921. BURN,C. G.: J. Infect. Dis. 54: 395, 1934. BUTT,H. R.: Arch. Path. 21: 27, 1936. BUZZARD,E. F.: Brain 28: 438, 1905. MELLA,H.: M. Clinics North America 7: 939, 1023. STARR,M A,: J. Nerv. k Ment. Dis. 39: 721, 1912. WEIGERT,C.: Neurol. Centralbl. 20: 597, 1901.