THE ANATOMICAL RECORD 209:491-500 (1984) Responsiveness of Testis Morphology to Chemotherapy in Childhood Leukemia GIULIANA ANELLI, GIUSI COZZI, EMILIA FRANCHI, MARIELLA PARLAVECCHIA, FRANC0 CHIARA, GIUSEPPE MASERA, AND MARINA CAMATINI Department of Biology, Comparative Anatomy (G.A., G.C., E.F., M.P., M.C.), 2nd Obstetrical-Gynaecobgical Clinic (RC.), and Clinic of Pediatry (G.M.), University of Milano, 20133 Milano, Italy ABSTRACT The appearance of seminiferous tubules and interstitial cells of children, aged 2.5 to 13 years, affected by acute lymphoblastic leukemia was analyzed in sections. The testicular biopsies were performed a t the end of therapy (vincristine, prednisone, L-asparaginase, 6-mercaptopurine, intrathecal methotrexate), which was affected for the same period and a t the same doses. Three age groups were considered (I, 2.5 to 5 years; 11, 6 to 9 years; 111, 12 to 13 years). Age groups I and I1 presented damage of some tubules (2535%) and areas of degeneration. Histometric analysis performed for A type spermatogonial population gave a mean value corresponding to controls in age group I and a mean value significantly lower with respect to controls in age group 11. Moreover, age group I1 presented a lack of increase in tubular cross section. These results suggest that there is a vulnerability both of whole tubules and of some areas of Sertoli cells and germ cells to cytotoxic-induced damage. Leydig cells appear to be the cells least sensitive to drugs, and hormonal data indicate that the hypothalamic pituitary function appears to be intact, despite chemotherapy. Long-term prospective studies of reproductive function in children receiving cancer chemotherapy are needed to determine the magnitude and duration of damage resulting from therapeutic treatment. We have previously demonstrated (Camatini et al., 1982) that inter-Sertoli junction formation in prepuberal boys who received antileukemic agents was similar to that of control children of a similar age (Camatini et al., 1981). Our observations have now been extended to the appearance of seminiferous epithelium and interstitial tissues in the same patients. Relatively few studies have been performed on the effect of chemotherapeutic drugs on development of testis in children, whereas there are recent data on the histologic and ultrastructural appearance of human prepuberal testis in normal individuals (Vilar et al., 1970; Hadziselimovic, 1977; Nistal et al, 1982; Muller and Skakkebaek, 1983), so that a comparison of the data reported here with those in the literature is possible. Unfortunately, limited information about the spermatogenic cycle is available in young (Hadziselimovic, 1977) and even adult humans (Rowley et al., 19711, and the stem cell has not been unequivocally identified. 0 1984 ALAN R. LISS, INC. Type A spermatogonia includes three spermatogonial types: A long (Al), A pale (Ap), and A dark (Ad) (Rowley et al., 1971), each of which may be a candidate for stem cell in adults. In children only Ap and Ad spermatogonial types have been described in control testes (Hadziselimovic, 1977). Thus, type A spermatogonial population includes the stem cell, and the histometric count of this spermatogonial population in the developmental stages of the testis may furnish some indication on the cytotoxicity of drugs on these cells.,Obviously, there is no way to predict if the damage caused by the treatment will persist and result in a n increased cell loss during spermatogenesis. The high incidence of testicular relapse in the treatment of childhood acute leukemia, after cessation of therapy, has prompted the requirement of bilateral testicular biopsies prior to discontinuation of therapy. Children have been biopsied following the DEW policy on the protection of human subjects and after informed consent. Received May 20, 1983; accepted February 21, 1984. 492 G. ANELLI ET AL. TABLE 1. Characteristics of 24 patients with childhood leukemia Group No. of patients I I1 111' 8 12 A B C D Age at biopsy (years) Range Mean 4.50 6.75 2.5-5 6-9 12 12 12 12 No, with testicular infiltrate 1 1 Mean hormone value (mIU/ml) _____ FSH LH 2.03 2.95 2.26 3.46 + + +- 'Patient A, prepuheral. Following hone marrow relapse, he underwent testicular irradiation (Fig. 11). Patient B, pubertal. Patient C, pubertal. Patient D, pubertal. He received a more aggressive chemotherapy (ara-C and cyclophosphamide) (Fig. 10). Since there is a n improved prognosis for children treated for acute lymphoblastic leukemia, attention must be paid to the quality of life of the survivors (Shalet, 1980),and the knowledge of gonadal morphology following chemotherapy represents a n important aspect, which must be taken into consideration. Controversial results (Berry et al., 1972; Rapola et al., 1973; Lendon et al., 1978) are related to the variability of drugs administered and their doses and combinations. Moreover, differences in the sensitivity of the prepuberal, pubertal, and adult testis to alkylating agent chemotherapy have been ren ported, and these results are conflicting (Lendon et al., 1978; Schilsky et al., 1980). The results presented here are significant, since they consider only prepuberal patients who have been treated with the same combination of chemotherapeutic agents, for the same period and at the same doses. Evidence is presented that the combined chemotherapy used resulted in damage to 20% of seminiferous tubules in the first age group considered (2.5 to 5 years) and 35% in the second age group (6 to 9 years). The unaltered tubules of the second group presented a reduction in number of type A spermatogonia with respect to control testis. Leydig cells' morphology and endocrinologic data were not apparently affected by the drugs. MATERIALS AND METHODS Twenty-four bilateral wedge testicular biopsies (Rowley and Heller, 1966) of children aged between 2 and 12 years affected by acute lymphoblastic leukemia were considered (Table 1). Testicular biopsies were performed, under general anaesthesia and following parents' consent, after 30 months of therapy to recognize the presence of lymphoblastic cells in the interstitial tissue. Twenty-three patients received standard chemotherapy (vincristine, prednisone, L-aspar- I age group I I age g roll P Fig. 1. Histogram showing mean spermatogonia number (n) per tubular cross section in group I (range 2.5-5, mean age 4.5 years) and group I1 (range 6-9, mean age 6.75 years), P < 0.08. Fig. 2. Area of preserved tubules in a 5-year-old testis. Spermatogonia (G) and Sertoli cells (S) are easy to distinguish. The appearance of tubules and interstitial tissue correspond to a control testis of the same age. x200. Fig. 3. Area of damaged tubules in a 9-year-old testis. Extensive vacuolized regions are evident (arrows); Sertoli cells (S)present irregularly shaped condensed nuclei; germinal elements are not distinguishable. x 580. CHEMOTHERAPY ON TESTIS LEUKEMIA 493 494 G. ANELLI ET AL. aginase, 6-mercaptopurine) and intrathecal methotrexate and cranial irradiation (1,8002,400 rads) for the prevention of central nervous system leukemia. One patient (D, Table 1) received a more aggressive chemotherapy (arabinofuranosyl-cytosine [ara-C] and cyclophosphamide [CP]) after bone marrow relapse. Five of the 24 patients (Table 1)showed diffuse testicular infiltrate. One of these infiltrated patients (A, Table 11, following bone marrow relapse, underwent testicular radiation and had a second testicular biopsy. For ultrastructural analysis, samples of testicular tissue were fixed in a mixture of glutaraldehyde-paraformaldehyde in 0.2 M cacodylate buffer, pH 7.4 (It0 and Karnovsky, 1968), and postfixed in 1% 0 ~ 0 4 .Tissues were stained en bloc with aqueous 2% uranyl acetate, dehydrated in ethanol, and embedded in Epon-Araldite. Thin sections stained with lead citrate were examined with a Philips 400 electron microscope. One-micronthick sections were stained with methylene blue and saphranine for light microscope and histometric analysis and were photographed with a Leitz Dialux 20 microscope. The values of FSH and LH (Table 1)were measured by radioimmunoassay (Weintraub et al., 1973). type spermatogonia number per tubular cross section was calculated for each patient and for the entire group, counting a mean number of 20 undamaged tubules for each patient in group I and a mean number of 30 undamaged tubules for each patient in group 11. RESULTS Tubules with gonia and Sertoli cells, which appeared a s shown in Figure 2, were considered undamaged, since their histological characteristics corresponded to control tubules. Tubules, such as those in Figure 3, were considered damaged on the basis of disappearance of the spermatogonial population, extensive vacuolated areas (Fig. 3, arrows), and the appearance of the Sertoli cells, whose fine structure is presented in Figure 6. The ultrastructure of the damaged tubules confirmed that both germ cells and Sertoli cells were affected (Fig. 8). The contour of Sertoli cells appeared modified as a consequence of retractions where plasma membranes were juxtaposed (Fig. 8, arrows); the nucleus presented evident chromatin condensation. These aspects are indicative of dysfunction of Sertoli cells. Germ cells, surrounded by such Sertoli cells, were altered too, showing a n unusual vacuolization in the cytoplasm (Fig. 8). Histometric Study About 20% of the total number of tubules The analysis of the general appearance of counted presented the above-mentioned morseminiferous tubules was performed before phology in age group I and about 35%in age the histometric count. Tubules presenting group 11. disappearance of spermatogonial population, The mean diameter of undamaged tubules vacuolated Sertoli cells, and altered contour in group I corresponded to that of the same were considered damaged. We considered age age group reported in the literature for congroup I (mean age 4.5, range 2.5-5 years) trol testis (65 pm, according to Hadziseliand age group I1 (mean age 6.75, range 6-9 movic 119771and Nistal et al. .In group years), since the diameter of seminiferous 11, the mean diameter of normal undamaged tubules, spermatogonia count, and Sertoli tubules corresponded to that of control testis cell appearance are very similar within each group in the analysis of control testis (Hadziselimovic, 1977; Furuya et al., 1980). Age Fig. 4. Tubules in a 9-year-old testis. The diameter of group I11 is not presented in the diagram (Fig. 11, since the appearance of seminiferous the tubules is not increased with respect to the first Gonia (arrows) are less numerous than in group tubules varies greatly during late prepuberal group. I. x300. age; thus, it was not considered as a homoFig. 5. Aspect of tubules in a n 8-year-oldtestis. Many geneous group. In age group I 20% and in age group I1 35% damaged tubules were Sertoli cells ( S )present an irregular profile, a condensed nucleus, and vacuolized areas. In such tubules spermaobserved. togonia cannot be recognized. x220. The number of transversely sectioned unFig. 6. Ultrastructure of a portion of a damaged tubdamaged seminiferous tubules per semithin section and the number of A type spermato- ule in a 9-year-old testis. Degenerated Sertoli cells have their limiting pIasma membrane; the nuclei (N) gonial population per transverse tubular sec- lost present indentations, chromatin condensation, and irtion were calculated using a calibrated ocular regular profiles. Degenerated mitochondria (arrows) and micrometer with a x40 objective. Mean A vacuoles are present in the cytoplasmic area. ~8,800. CHEMOTHERAPY ON TESTIS LEUKEMIA 495 496 G. ANELLI ET AL. only in the case of the 6- to 7-year-old patients. The 9-year-old patients did not present the normal increase in cross-tubular diameter (Fig. 4). In group I, no relevant alteration in the interstitial tissue was observed. Fibroblasts, which represent the most common cell type, were normal, and scattered groups of precursors of Leydig cells, mainly surrounding the tubular wall, were present (Fig. 2). In group 11, peritubular fibrosis and thickening of the basal membrane were frequent: Leydig cells presented some characteristic features of adult cells. Also in group 111, Leydig cells were recognizable even when a heavy infiltrate was present (Figs. 9,121. Postradiation biopsy (Fig. 11)showed a striking interstitial and tubular fibrosis. Irreversibly damaged tubules in which cell types could not be recognized appeared in consequence of testis radiation. Moreover, a complete loss of germ cells (Fig. 10) was observed in another patient of this group treated with ara-C and CP therapy. Leydig cells were still visible. Germ Cells It was previously demonstrated that from birth to puberty Ap, Ad, and B spermatogonia as well a s primary spermatocytes can be discerned (Hadziselimovic, 1977). The A type spermatogonial population includes the stem cells, and the histometric analysis was conducted on this type of gonia (Fig. 11, which could be easily recognized by their size and shape (Fig. 2). In group I (Fig. 2) the mean value (3.5 gonia per tubule) obtained is in agreement with that reported in the literaturefor control testis of the same age group (Hadziselimovic, 1977). A type spermatogonia count carried out on group I1 (Fig. 4) showed that the mean spermatogonia number per tubule (1.6 gonia) was markedly decreased in comparison with that of group I (Fig. 2). Group I11 showed different morphologic patterns (Table 1).Two pubertal patients (B and C, Table 1) presented leukemic infiltrates (Figs. 9, 12) at the end of therapy and displayed tubules with normal diameter, a high number of gonia, and active spermatogenesis. One patient (A, Table 1) was still prepuberal; he underwent testicular radiation and presented completely damaged seminiferous tubules (Fig. 11).The last one was pubertal (D, Table l), but only Sertoli cells were present in tubules (Fig. 10). This patient had been treated with ara-C and CP. Sertoli Cells Seminiferous tubules of testes from groups I and I1 showed immature Sertoli cells. These presented a pseudostratified disposition and, in unaffected tubules, could be easily identified among germ cells (Figs. 2, 4). Their normal ultrastructural features are presented in Figure 7. The nucleus had a small nucleolus and dispersed chromatin except for a peripherally placed chromatin band. The cytoplasm was characterized by small mitochondria, abundant endoplasmic reticulum cisternae, and vesicles. The appearance of Sertoli cells was completely modified in damaged tubules (Fig. 3) and also in scattered areas of some preserved tubules (Fig. 5). Thick sections evidenced irregularly shaped, intensely stained nuclei, surrounded by vacuolized regions (Fig. 3). The contour of these cells could not be recognized. At ultrastructural level the nucleus of Sertoli cells presented numerous indentations and irregular masses of condensed chromatin; the cytoplasm was filled with a large number of vacuoles and altered mitochondria, which lost their internal structure (Fig. 6). DISCUSSION There are many difficulties in performing human studies concerning the effect of combined chemotherapeutic drugs on fertility, and scanty and controversial are reports concerning damage on gonadal function of treated children (Rapola et al., 1973; Lendon et al., 1978). The action of single tested drugs on the seminiferous tubules of adult rodents (Lu and Meistrich, 1979) is reported in literature. The most apparent effect of Vinca alkaloids (such as vincristine [VCR] and vinblastine [VLB]) on testicular cells was a partial killing of type B spermatogonia, and these drugs apparently produced less damage than did ara-C and CP. Despite these results on adult rodents, it is well known that the reproductive function of Fig. 7. Aspect of normal, immature Sertoli cells (S) in a 5-year-old testis, showing round-shaped nuclei (N) with dispersed chromatin. The cytoplasm is characterized by the usual components. Sertoli-Sertoli contacts (arrows) are well preserved. X3,500. Fig. 8. Ultrastructure of a portion of a tubule in which Sertoli cells (S) present a beginning of a degenerative process: altered contacts (arrows) and nuclei (N) with partially condensed chromatin. The germ cell (G) presents an extensive vacuolization in the cytoplasm. ~4,700. CHEMOTHERAPY ON TESTIS LEUKEMIA 497 498 G. ANELLI ET AL. CHEMOTHERAPY ON TESTIS LEUKEMIA man is seriously affected by VLB Wilar, 1975) and CP (Pairley et al., 1972). The results herein reported on the effect of a combined therapy on testicular morphology of children suggest that there is evident damage which involves less than 35%of the observed tubules. The reason why tubules display different sensitivity to drugs is unknown: drug penetration would be the same for all tubules, since during this period no compartmentalization of seminiferous tubules exists. Moreover, it is impossible to establish if the combined therapy has a preferential effect upon stem cells or whether it is secondary to a n effect upon Sertoli cells. A progressive increase in the number of A type spermatogonia per transverse tubular section has been reported in 4- to 10-year-old control testis (Hadziselimovic, 1977; Miiller and Skakkebaek, 1983).In group 11, the number of type A spermatogonia appears reduced in comparison with control testis of the same age. This result may be related with some spermatogonial degeneration or to a mitotic arrest in the existing spermatogonial population. Thus, age group I1 appears the most affected. In age group 111, two patients presented some preserved tubules with spermatogenesis, suggesting that the cytotoxic effect of drugs did not prevent the development of progeny of some stem cells. Leydig cells appear to be the cells least sensitive to drugs in prepuberal testes studied. Their morphology correlates well with the gonadotropins mean values (Table 1) which are in agreement with the data of the Fig. 9. Seminiferous tubules in a 13-year-oldpubertal boy. Active spermatogenesis is evident; leukemic cells (arrows) are present in the interstitial tissue among Leydig cells. X420. Fig. 10. Seminiferous tubules of a pubertal 13-yearold boy who received ara-C therapy. The loss of germ cells is evident, and only Sertoli cells occupy the tubules. The irregular profile of tubules and thickening of the basement membrane (arrows) evidence the damage of this therapy. X420. Fig. 11. Appearance of tubules, following testis radiation, in a 12-year-old testis. It is possible to recognize only the profile of the tubules, which are completely damaged. A Leydig cell (L)is morphologically identifiable in the interstitial tissue. X420. Fig. 12. Leydig cells (L) in a pubertal testis with lymphoblasts. Among blast cells (arrows) a Leydig cell presents the typical morphology, with numerous lipid inclusions. x 1,280. 499 literature (Fischer and Aur, 1982). Thus, despite chemotherapy and radiation, the hypothalamic pituitary function appears intact in acute lymphoblastic leukemia. The results presented demonstrate that the combined chemotherapy used appears to be considerably less destructive than other therapies, such as CP, ara-C (Pairley et al., 1972; Buchanan et al., 1975) (Fig. lo), and radiation (Fig. 11).It can be foreseen that suspension of the therapy, in cases with favorable prognosis, will result in reestablishment of development conditions, at least in apparently undamaged tubules. ACKNOWLEDGMENTS This research was supported by MPI 1982. LITERATURE CITED Berry, C.L., J.S. Cameron, C.S. Ogg, D.Q. Trounce, and D.R. Turner (1972) Cyclophosphamide and prepubertal testis. Lancet, 2: 1033-1034. Buchanan, J.D., K.F. Fairley, and J.V. Barrie (1975) Return of spermatogenesis after stopping cyclophosphamide therapy. Lancet, 2156-157. Camatini, M., E. Franchi, and I. De Curtis (1981) Differentiation of inter-Sertoli junctions in human testis. Cell Biol. Int. Rep., 5109-110. Camatini, M., E. Franchi, I. De Curtis, G. Anelli, and G. Masera (1982)Chemotherapy does not affect the development of inter-Sertoli junctions in childhood leukaemia. Anat. Rec., 203:396-406. Fisher, J.N., and R.J.A. Aur (1982) Endocrine assessment in childhood acute lymphocytie leukaemia. Cancer, 49r145-151. Furuya, S., Y. Kumamoto, M. Mori, and S. Sugiyama (1980) The blood-testis barrier. In: Descended and Cryptorchid Testis, Vol 3. E.S.E. Hafez, ed. Martinus Nijhoff, The Hague, pp. 73-93. Hadziselimovic, F. (1977) Cryptorchidism: Ultrastructure of normal and cryptorchid testis development. Adv. Anat. Embryol. Cell Biol., 53:l-69. Ito, S., and H.J. Karnovsky (1968) Formaldehyde-glutaraldehyde fixatives containing trinitro compounds. J. Cell Biol., 39t168a. Lendon, M., I.M. Hann, M.K. Palmer, S.M. Shalet, and P.H. Morris Jones (1978) Testicular histology after combination chemotherapy in childhood for acute lymphoblastic leukaemia. Lancet, 2 4 3 9 4 4 1 . Lu, M.C., and M.L. Meistrich (1979) Cytotoxic effects of chemotherapeutic drugs on mouse testis cells. Cancer Res., 39r3575-3581. Miiller, J., and N.E. Skakkebaek (1983) Quantification of germ cells and seminiferous tubules by stereological examination of testicles from 50 boys who suffered from sudden death. Int. J. Androl., 6r143-156. Nistal, M., M.A. Abaurrea, and R. Paniagua (1982) Morphological and histometric study on the human Sertoli cell from birth to the onset of puberty. J. Anat., 14:351363. Pairley, K.F., J.V. Barrie, and J. Warren (1972) Sterility and testicular atrophy related to cyclophosphamide therapy. Lancet, 1568-569. Rapola, J., 0. Koskinies, N.P. Huttunen, P. Floman, J. Vilska, and N. Hallman (1973) Cyclophosphamide and the pubertal testis. Lancet, 1:98-99. 500 G. ANELLI ET AL. Rowley, M.J., J.D. Berlin, and C.G. Heller (1971) The ultrastructure of fan types of human spermatogonia. Cell Tissue Res., 112r139-157. Rowley, M.J., and C.G. Heller (1966) The testicular biopsy: Surgical procedure, fixation and staining techniques. Fertil. Steril., 17:177-186. Schilsky, R.L., B.J. Lewis, R.J. Sherins, and R.C. Young (1980) Gonadal dysfunction in patients receiving chemotherapy for cancer. Ann. Inter. Med., 93t109-114. Shalet, S.M. (1980) Effects of cancer chemotherapy on gonadal function of patients. Cancer Treat. Rev., 7t141152. Vilar, 0. (1975) Effect of cytostatic drugs on human testicular function. In: Male Fertility and Sterility. R.E. Mancini and L. Martini, eds. Academic Press, London, pp. 423-440. Vilar, O., C.A. Paulsen, and D.J. Moore (1970) Electron microscopy of the human seminiferous tubules. In: The Human Testis. E. Rosemberg and C.A. Paulsen, eds. Plenum Press, New York, pp. 127-138. Weintraub, B.D., Y.A. McCamon, and R.L. Berlman (1973) Apparent cooperativity in radioimmunoassay of human chorionic gonadotrophin. Endocrinology, 92: 1250-1258.