sign in sign up
<<
Home , Choriocarcinoma, Embryonal carcinoma, Ovarian germ cell tumors, Seminoma

ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 9, No. 6 Copyright © 1979, Institute for Clinical Science, Inc. Malignant Tumors of the and Testis An Immunohistologic Study of 69 Cases R. J. KURMAN, M.D.,* P. T. SCARDINO, M.D.,t K. R. McINTIRE, M.D.J T. A. WALDMANN, M.D.J N. JAVADPOUR, M.D.J and H. J. NORRIS, M.D.§

* Department of Pathology, U.S.C. School of Medicine and LAC-USC Medical Center, Los Angeles, CA 90033 f Department of , U.C.L.A. Medical Center, Los Angeles, CA 90024 | National Institute, N.I.H., Bethesda, MD 20014 § Division of Gynecologic Pathology, Armed Forces Institute of Pathology, Washington, DC 20007

ABSTRACT

An immunohistologic study of 21 patients with germ cell tumors of the testis with measured serum levels of human chorionic gonadotropin (HCG) and alpha fetoprotein (AFP) was undertaken to correlate the various types of with the presence of these tumor markers in the tissue and serum. The same immunoperoxidase technique was applied to 48 patients with malignant germ cell tumors of the ovary in whom serum was not available. In the testicular tumor group, AFP was demonstrated in mononuclear embryonal cells within and endodermal sinus tumor. HCG was identified within syncytiotrophoblastic giant cells frequently in association with embryonal carcinoma, and rarely with endodermal sinus tumor and , as well as in the syncytiotrophoblastic component of . Eighteen of the 21 patients (86 percent) had elevated tumor markers in their serum; serum HCG alone was elevated in 5 (24 percent), AFP alone in 5 (24 percent) and both were elevated in 8 (38 percent). There was tissue localization of HCG in 12 of 13 patients (92 percent) with elevated serum HCG while AFP was identified in the tumor in eight of the 13 patients (53 percent) with elevated serum AFP levels. In the group, all 15 endodermal sinus tumors examined were positive for AFP and negative for HCG. Seven of 10 embryonal car­ cinomas were positive for AFP and all 10 were positive for HCG. The two cases of choriocarcinoma were both positive for HCG and negative for AFP. In contrast, the 11 and 10 were negative for both AFP and HCG. The results parallel those for malignant germ cell tumors of the testis, affording additional evidence of the analogous nature of germ cell tumors of the gonads. Based on these findings a tentative immunohisto­ logic classification of germ cell tumors utilizing AFP and HCG is proposed. Thus, embryonal carcinoma is frequently associated with both AFP and HCG, endodermal sinus tumor with AFP and choriocarcinoma with HCG, whereas pure seminoma, and are unlikely to be associated with either marker.

* Present Address: Georgetown University, School of Medicine, Department of Pathology, 3800 Reservoir Road, Washington, DC 20007. 462 0091-7370/1100-0462 $00.90 © Institute for Clinical Science, Inc. MALIGNANT GERM CELL TUMORS OF OVARY AND TESTIS 463

Introduction have been confused because terms such as embryonal carcinoma, teratocarcinoma, The value of alpha-fetoprotein (AFP) malignant teratoma, teratoblastoma, en- and human chorionic gonadotropin dodermal sinus tumor, yolk sac tumor, (HCG) as tumor markers in the diagnosis embryonal carcinoma of the infant testis and management of germ cell tumors of have been used interchangeably.9 the testis is now well established.1’2,10,11,14 Clinically, determination of these mar­ 2. Until recently, histopathologic studies of germ cell tumors have not kers has merit for several reasons as fol­ lows: paralleled the advances in tumor immu­ nology. Consequently, a contemporary 1. Successful treatment of patients with hypothesis of AFP and HCG synthesis by germ cell tumors depends on accurate germ cell tumors is compromised by out­ clinical staging. Occult not de­ moded concepts of histopathology. For tectable by lymphangiography can be example, the presence of HCG in the identified by the presence of elevated serum of patients with germ cell tumors levels of AFP and/or HCG in the serum has been regarded as evidence of the using highly sensitive and specific radio­ presence of choriocarcinoma. However, immunoassays (RIA).11,14 the rarity of choriocarcinoma arising in the 2. is most effective male testis (only 18 cases of pure chorio­ when used for the eradication of micro­ carcinoma among 6,000 testis tumors on scopic rather than gross disease.16 The RIA for AFP and HCG has been shown to file in the American Testicular Tumor be effective in detecting early “occult” Registry at the Armed Forces Institute recurrence before tumor can be identified of Pathology and only 5 percent of mixed by clinical examination and can therefore germ cell tumors contain chorio­ be used to direct chemotherapy.11,14 carcinoma)9 is inconsistent with the fact 3. Since certain tumor markers are as­ that approximately 70 percent of non- sociated with particular histologic types, seminomatous testicular germ cell tumors determination of these markers may have are associated with elevated serum levels prognostic import.14 of HCG as measured by RIA.24 4. Similarly, tumor markers, by iden­ 3. F inally, since 40 percent of testicular tifying specific tumor types, may be sig­ germ cell tumors9 and almost 10 percent of nificant in the selection of appropriate ovarian germ cell tumors7 are composed of therapeutic agents.17,25 combinations of two or more histologic For these reasons it is apparent that a patterns, it is manifestly impossible to de­ correlation of the histologic type with the termine which element in the tumor is tumor markers is vital for a rational ap­ responsible for the synthesis of a particu­ proach to accurate diagnosis and man­ lar in the serum using or­ agement of ovarian and testicular germ thodox histologic techniques. In view of cell tumors. Also, a correlative histopatho­ this limitation an investigation was under­ logic and tumor marker study might be taken to identify the specific cellular helpful in shedding light on the histo­ components responsible for the synthesis genesis of this poorly understood group of of AFP and HCG using an immuno- neoplasms. Factors that must be consid­ histologic method. ered in this type of study and problems encountered in the past are the following: Methods 1. In view of their rarity, much of the The analysis of testicular tumors8 was confusion regarding germ cell tumors has based on 21 patients for whom paraffin resulted from a lack of a uniform terminol­ blocks of tissue were available. Serum de­ ogy. Neoplasms with distinctive his­ terminations of AFP and HCG were per­ tologic patterns and biologic behavior formed either prior to or within 30 days of 464 KURMAN, SCARDINO, MCINTIRE, WALDMANN, JAVADPOUR AND NORRIS tissue acquisition, thereby permitting a tile type, characterized by a distinctive correlation of the various histologic pat­ reticular-papillary pattern was found in terns together with the cellular localiza­ four instances. tion of AFP and HCG and the levels of was present in two forms in these markers in the serum. An indirect this group of tumors. One type, corre­ immunoperoxidase technique for the sponding to choriocarcinoma, contained localization of AFP and HCG, previously an intimate admixture of reported by Kurman and Norris,5,6 was and and was found in selected since this method can be used in two cases. The other form was comprised a retrospective fashion to study tumors in of syncytiotrophoblastic giant cells exclu­ paraffin blocks,—frozen tissue not being sively. These were typically multi­ necessary. The tumors were classified in nucleated with abundant eosinophilic or accordance with the classification of tes­ amphophilic cytoplasm containing one or ticular neoplasms proposed by Mostofi more vacuoles of varying size. These cells and Price.9 were found in association with areas of A double antibody RIA developed by embryonal carcinoma in 11 instances, Vaitukaitis and associates, specific for the with endodermal sinus tumor in one case /3 subunit,22 was utilized for the quantita­ and with seminoma in another. tion of the serum level of HCG and a dou­ With the immunoperoxidase technique, ble antibody RIA developed by HCG was found only in the syncytiotro­ Waldmann and Mclntire23 was used for phoblastic element of choriocarcinoma quantitation of AFP in the serum. (two cases) or in syncytiotrophoblastic Serum samples were not available for giant cells (13 cases). the patients with ovarian tumors. Con­ sequently, this part of the study was lim­ AFP was identified in the mononuclear ited to an immunohistologic analysis of 48 embryonal carcinoma cells of embryonal pure malignant germ cell tumors using an carcinoma in seven of the eleven cases (47 immunoperoxidase technique identical to percent) and within similar type cells in that utilized for the testicular study. The two of the four areas of endodermal sinus ovarian neoplasms were classified in ac­ tumor (50 percent). Cells in which HCG cordance with the WHO classification for was localized did not contain AFP and ovarian tumors adopted in 1973.15 similarly those in which AFP was localized did not contain HCG. Thus, it Results appears that AFP and HCG are found in and probably synthesized by two dis­ T e s t ic u l a r G e r m C e l l T u m o r s tinctly different cell lines. The cellular (21 c a s e s ) constituents of seminoma, teratoma, and Areas of typical seminoma were found the cytotrophoblastic element of chorio­ in seven instances. carcinoma were negative for both AFP Areas of teratoma containing elements and HCG. derived from all three germ cell layers Elevated serum levels of AFP alone were encountered in eight instances. were found in five patients (24 percent), Foci of primitive neuroectoderm and HCG alone in five (24 percent) and both mesoderm were present in five instances. were found elevated in eight patients (38 Embryonal carcinoma was found in 15 percent). There was almost complete cor­ instances and was characterized by sheets relation of elevated levels of HCG in the of large pleomorphic cells arranged in serum with the tissue localization (92 per­ solid, papillary and acinar patterns. cent) while AFP was identified in the tis­ Endodermal sinus tumor (yolk sac car­ sue in 53 percent of patients with elevated cinoma) or embryonal carcinoma, infan­ serum levels. MALIGNANT GERM CELL TUMORS OF OVARY AND TESTIS 465

O v a r ia n G e r m C e l l T u m o r s (48 cases) Tissue Localization of AFP* and HCGt The histologic characteristics of the in Malignant ovarian germ cell tumors closely approx­ imated those in the testicular neoplasms. Tumor Type Total A F P * H C G t The results of the immunocytologic local­ Dysgerminoma 11 0 0 ization are summarized in table I. Immature teratomas 10 0 Ü Embryonal carcinoma 10 7 ( 70%) 10 (100%) Endodermal sinus tumor 15 15 (100%) 0 D isc u ssio n Choriocarcinoma 2 0 2 (100%)

In recent years there has been a *Alpha-fetoprotein tHuman chorionic gonadotropin growing awareness of the homologous nature of germ cell tumors of the ovary marker. It is important that the presence of and testis.6’8’9,15’20,21 The similarity of pattern for AFP and HCG distribution in syncytiotrophoblastic giant cells which were associated with embryonal car­ both testicular and ovarian germ cell cinoma, seminoma and endodermal sinus tumors further corroborates their analo­ tumor be specifically noted in the histo­ gous nature. logic diagnosis, since unlike choriocar­ The immunohistochemical identifica­ cinoma, it is not well recognized that tion of HCG within syncytiotrophoblastic these cells are responsible for elevated giant cells indicates a relationship be­ serum levels of HCG. Terms such as tween choriocarcinoma and embryonal “teratocarcinoma” or “mixed germ cell carcinoma, while a relationship between tumor” should be amplified to include an endodermal sinus tumor (yolk sac tumors) enumeration of the various germ cell and embryonal carcinoma can be illus­ tumor components so that a more accurate trated by the presence of AFP within correlation of tumor histology and tumor embryonal carcinoma cells in both these markers might be achieved. neoplasms. Taken in conjunction with the ultrastructural studies of Pierce and his The importance of this approach lies in the fact that it provides a basis for a com­ associates12,13 showing the common origin bined immunologic and histologic clas­ of all germ cell tumors and the experimen­ tal production of teratomas using embry­ sification of these tumors, thereby clarify­ ing some of the clinical aspects that have onal carcinoma explants by Stevens and his co-workers,18,19 the immunohisto­ chemical findings lend support to the PRIMORDIAL GERM CELL concept that embryonal carcinoma may be \ EMBRYONAL CARCINOMA the neoplastic progenitor of the other Seminoma or AFP (+) germ cell tumors. This is further substan­ Dysgerminoma HCG (+) AFP (-) tiated by the in vivo and in vitro synthesis HCG (-) \ Extraembryonic Embryonic of AFP by these transplantable murine Differentiation Differentiation embryonal carcinomas.4 V In figure 1 is illustrated a tentative Trophoblast Immature histogenic and immunohistologic classifi­ Teratoma Choriocarcinoma Endodermal AFP (-) cation for germ cell tumors that incorpo­ AFP (-) Sinus Tumor HCG (-) rates AFP and HCG tumor markers. Thus, HCG (+) AFP (+) HCG (-) 'L embryonal carcinoma is frequently as­ Mature Teratoma sociated with both AFP and HCG, endo­ AFP (-) dermal sinus tumor with AFP and chorio­ HCG (-) carcinoma with HCG, whereas pure F ig u r e 1. Histogenetic and immunohistologic seminoma, dysgerminoma and teratoma classification of germ cell tumors. AFP = alpha-feto- are unlikely to be associated with either protein; HCG = human chorionic gonadotropin. 466 KURMAN, SCARDINO, MCINTIRE, WALDMANN, JAVADPOUR AND NORRIS been observed. Discordance of AFP and and T e i lu m , G.: Serum alpha-fetoprotein as a marker for endodermal sinus tumor (yolk sac HCG can occur after treatment, and either tumor) or a vitelline component of teratocar- AFP alone, HCG alone or both markers cinoma. Acta Path. Microbiol. Scand. Sect. A, may rise with the development of recur­ 53:573-589, 1975. rent tumor depending on which compo­ 11. Perlin, E., E ngeler, J. E„ Jr., Edson, M., K a r p , D., M cIntire, K. R., and W a ld m a n n , T. nent of the tumor recurs.3,11 As further in­ A.: The value of serial measurement of both vestigation identifies additional markers human chorionic gonadotropin and alpha-feto­ protein for monitoring germinal cell tumors. associated with germ cell tumors, it may Cancer 37:215-219, 1976. become necessary to modify this classifi­ 12. PIERCE, G. B ., J r .: Ultrastructure of human tes­ cation. ticular tumors. Cancer 19:1963, 1966. Thus, tumor histopathology stands at 13. P i e r c e , G. B., Jr., Stevens, L. C., and NaKane, P. K.: Ultrastructural analysis of the the threshold of a new era. The limits of early development of teratocarcinomas. J. N at. morphologic characterization have now Cancer Inst. 39:755, 1967. been realized and the functional aspects 14. S c a r d i n o , P. T., C o x , H. D., W a ld m a n n , T. A., M cIntire, K. R., M ittem eyer, B., and of tumors must be considered. In the fu­ Javadpour, N.: The value of serum tumor mar­ ture, both a morphologic and functional kers in the staging and prognosis of germ cell classification of neoplastic disease should tumors of the testis. J. Urol. 118:994-999,1977. 15. S e r o v , S. F., S c u l l y , R. E., and S o b in , L. H.: be established. Histological Typing of Ovarian Tumors. Inter­ national Histological Classification of Tumors, References No. 9, Geneva, World Health Organization, 1973. 1. Abelev, G. I., A ssecritova, I. V., Kraevsky, 16. S k ip p e r , H. E., S c h a b e l , F. M., B e l l , M., e t N. A., Perova, S. D., and Perevodchikova, N. A L .: On the curability of experimental I.: Embryonal serum alpha globulin in cancer neoplasms. I. A-methopterin and mouse patients: Diagnostic value. Internat. J. Cancer Ieukemias. Cancer Res. 17:717-726, 1957. 2:551-558, 1967. 17. S m ith , J. P. and R u t l e d g e , F.: Advances in 2. ABELEV, G. I.: Alpha-fetoprotein as a marker of chemotherapy for gynecologic cancer. Cancer embryo-specific differentiations in normal and 36:669-674, 1975. tumor tissues. Transplant. Rev. 20:3-37, 1974. 18. S t e v e n s , L. C. and H u m m e l, K. P.: A descrip­ 3. Braunstein, G. D., M c I n t i r e , K. R., and tion of spontaneous congenital testicular W ALDM ANN, T. A.: Discordance of human teratomas in strain 129 mice. J, Nat. Cancer Inst. chorionic gonadotropin and alpha-fetoprotein 18:719, 1957. in testicular teratocarcinomas. Cancer 19. S t e v e n s , L. C.: Experimental production of 31:1065-1068, 1973. testicular teratomas in mice. Proc. Nat. Acad., 4. Kahan, B. and Levine, L.: The occurrence of 52:661, 1964. serum fetal alpha-protein in developing mice 20. T e i l u m , G.: “Mesonephromaovarii” (Schiller). and murine hepatomas and teratomas. Cancer An extra-embryonic mesoblastoma of germ cell Res. 31:930-936, 1971. origin in the ovary and the testis. Acta Pathol. 5. K u r m a n , R. J. and N o r r i s , H. J.: Endodermal Microbiol. Scand. 27:249-261, 1950. sinus tumor of the ovary. A clinical and 21. TEILUM, G.: Endodermal sinus tumors of the pathologic analysis of 71 cases. Cancer ovary and testis. Comparative morphogenesis 38:2404-2419, 1976. of the so-called mesonephroma ovarii (Schiller) 6. K u r m a n , R. J. and NO RRIS, H. J.: Embryonal and extraembryonic (yolk-sac-allantoic) struc­ carcinoma of the ovary. A clinicopathologic en­ tures of the rat’s . Cancer 12:1092- tity distinct from endodermal sinus tumor re­ 1105, 1959. sembling embryonal carcinoma of the adult tes­ 22. Vaitukaitis, J. L., Braunstein, G. D., and tis. Cancer 38:2420-2433, 1976. R o s s , G. T.: A radioimmunoassay which 7. Kurm an, R. J. and N orris, H. J.: Malignant specifically measures human chorionic germ cell tumors of the ovary. Human Path. gonadotropin in the presence of human 5:551-564, 1977. luteinizing hormone. Amer. J. Obstet. 8. Kurman, R. J., Scardino, P. T., M cIntire, K. Gynecol. 113:751-758, 1972. R., W a ld m a n n , T. A., and Javadpour, N.: Cel­ 23. W a ld m a n n , T. A. and M c I n t i r e , K. R.: Serum lular localization of alpha-fetoprotein and alpha-fetoprotein levels in patients with ataxia- human chorionic gonadotropin in germ cell telangiectasia. Lancet 2:1112-1115, 1972. tumors of the testis using an indirect immuno- 24. W a ld m a n n , T. A. and M cIntire, K. R.: The peroxidase technique. A new approach to clas­ use of a radioimmunoassay for alpha-feto­ sification utilizing tumor markers. Cancer protein in the diagnosis of malignancy. Cancer 40:2136-2151, 1977. 34:1510-1515, 1974. 9. M O STO FI, F. K. and PR IC E, E. B.: Tumors of the 25. W id e r , J. A., M a r s h a l l , J. R., Bardin, C. U., male genital system. Atlas of Tumor Pathology, LlPSETT, M. B. and ROSS, G. T.: Sustained re­ Second series, Fascicle 8. Washington, D.C., missions after chemotherapy for primary ova­ Armed Forces Institute of Pathology, 1973. rian containing choriocarcinoma. New 10. Norgaard-Pedersen, B., Albrechtsen, R., Eng. J. Med. 280:1439-1442, 1969.