- Email: [email protected]
The effect of testicular nongerm cell tumors on local spermatogenesis
Vol. 62, No.1, July 1994
FERTILITY AND STERILITY
Printed on acid-free paper in U. S. A.
Copyright" 1994 The American Fertility Society
The effect of testicular nongerm cell tumors on local spermatogenesis
George T. Ho, M.D.* Humphrey Gardner, M.D.t Kash Mostofi, M.D.:\:
William C. DeWolf, M.D.§ Kevin R. Loughlin, M.D.II Abraham Morgentaler, M.D.§1f
Beth Israel Hospital, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, and Armed Forces Institute of Pathology, Washington, D.C.
Objective: To assess ipsilateral spermatogenesis in men with malignant and benign nongerm cell tumors of the testis. Design, Patients: Histologic review of radical orchiectomy specimens performed for 20 men with malignant nongerm cell tumors and 15 with benign testicular lesions, including five Leydig cell tumors with benign clinical features. Main Outcome Measures: Degree of spermatogenesis was determined on a 1 to 10 scale, with 10 representing mature sperm within a seminiferous tubule. For each patient "near" and "far" scores were determined by obtaining the mean score of 50 tubules adjacent «3 mm) to the tumor and 50 tubules distant (>3 mm) from the tumor, respectively. Results: Total, near, and far scores were all lower for malignant tumors than for benign lesions. Scores for Leydig cell tumors were similar to benign lesions. Malignant tumors demonstrated a gradient effect, with greatest impairment of spermatogenesis occurring adjacent to tumor. In contrast, a distinction between near and far scores was not observed for benign lesions or Leydig cell tumors. Conclusions: Malignant nongerm cell tumors of the testis were associated with significant impairment of ipsilateral spermatogenesis, particularly in areas adjacent to tumor. These findings are similar to those observed for testicular germ cell tumors, suggesting a generalized negative Fertil Steril 1994;62:162-6 influence on ipsilateral spermatogenesis by malignant tumors. Key Words: Spermatogenesis, testis, cancer, infertility, tumor
Clinical subfertility has been reported among men with a variety of malignancies, most notably testicular carcinoma (1-6). The cause of this subfertility, however, is not known. In a histologic
Received October 27, 1993: revised and accepted February 21, 1994. * Harvard Program in Urology (Longwood Area), Harvard Medical School. t Department of Pathology, Beth Israel Hospital. :j: Armed Forces Institute of Pathology. § Division of Urology, Beth Israel Hospital. 1\ Division of Urological Surgery, Brigham and Women's Hospital. 1f Reprint requests: Abraham Morgentaler, M.D., 330 Brookline Avenue, Boston, Massachusetts 02215 (FAX: 617-2787292).
162
Ho et al.
Testis tumors and local spermatogenesis
review of radical orchiectomy specimens for testicular germ cell tumors, Ho et al. (7) observed widespread impairment of spermatogenesis throughout the ipsilateral testis. This impairment was particularly severe in regions adjacent to the tumor, with the degree of impairment dependent on tumor type. Since nongerm cell tumors are also associated with subfertility, the possibility that malignant tumors exert a generalized negative influence on spermatogenesis warrants consideration. To assess the effect of nongerm cell tumors on spermatogenesis, we performed a histologic review of orchiectomy specimens from men with malignant and benign nongerm cell tumors of the testis. Spermatogenesis adjacent to and distant from the primary tumor was examined to determine whether Fertility and Sterility
a gradient effect might be present, as noted with germ cell tumors ofthe testis (7). Clinical parameters such as patient age, tumor type, and tumor size were also evaluated for their influence on spermatogenesis. MATERIALS AND METHODS
All radical orchiectomy specimens evaluated at the Beth Israel Hospital of Boston, the Brigham and Women's Hospital Boston, Massachusetts, and the Armed Forces Institute of Pathology Washington, D.C., from 1980 to 1991 were reviewed. Patients with nongerm cell testicular tumors without prior history of chemotherapy or radiation therapy and with adequate medical history and histologic material were included in this study. Patients were divided into three groups as follows: [1] 20 men with malignant testicular nongerm cell tumors (12 large cell lymphomas, 3 Burkitt's lymphoma, 2 rhabdomyosarcoma, 2 gastric carcinoma, 1 pancreatic carcinoma); [2] 10 men with benign testicular tumors (5 cysts, 3 fibrosis, 1 sarcoidosis, 1 splenic rest); and [3] 5 men with Leydig cell tumors. None of the last group manifested clinical evidence of malignancy. Because a majority of the malignant tumors were reviewed as consult slides at the Armed Forces Institute of Pathology, only limited information was available on the general health of these patients. The quality of spermatogenesis was determined by scoring seminiferous tubules according to the method of Johnsen (8). Tubules were scored on a scale of 1 to 10, with 10 indicating more than five mature spermatozoa within the lumen. Two groups of 50 seminiferous tubules were scored for each case: "near" scores were obtained from tubules within 3 mm from the tumor margin, and "far" scores from beyond 3 mm. All tubules in a series of contiguous fields were scored. Scoring for the far group was performed in areas demonstrating the highest degree of local spermatogenesis. Whenever possible, the near and far groups of tubules were scored on the same slide. In a few cases, no slide was available that showed tumor contiguous with uninvolved tubules, and in these cases a near score could not be obtained. The mean far Johnsen scores of patients with malignant nongerm cell testicular tumors were stratified on the basis of patient age and tumor size to determine the possible contribution of these parameters on the degree of spermatogenesis impairment. Testis and tumor sizes were determined from the original pathology reports. Patients were stratiVol. 62, No.1, July 1994
fied into two subgroups on the basis of age with one group consisting of patients 55 years old or younger and a second subgroup consisting of patients older than 55. Patients were also stratified on the basis of absolute tumor size by dividing the patients into two subgroups, one consisting of patients whose tumors were <20 cc in size and a second consisting of patients with tumors ~20 cc in size. Last, the contribution of relative tumor size was evaluated by dividing the patients into two subgroups consisting of one group with tumor involving <50% of the testis and a second subgroup composed of men whose tumors comprised >50% of the testis. Statistical analysis was performed using the mean Johnsen scores for each of 50 tubules, consisting of a mean near and a mean far score for each patient. Comparison of mean Johnsen scores for tumor types was conducted using a Mann-Whitney test. Other parameters were analyzed by t-test (two-tail). Relative tumor size was calculated by dividing the absolute tumor size by the size of the testicle.
RESULTS
Scores for malignant tumors were lower than benign tumors in all categories: near (4.4 versus 8.5, P < 0.0001); far (5.7 versus 8.8, P < 0.0001); and overall score (5.1 versus 8.7, P < 0.0001). Scores for Leydig cell tumors were similar to other nonma1ignant tumors (Table 1) and also differed significantly from the malignant group (near 8.1, P < 0.001; far 8.1, P < 0.01). Score distributions for individual tubules are presented in Figure lA, revealing minimal overlap of scores between benign and malignant tumors. Almost 90% of tubules in the malignant group failed to demonstrate spermatids or
Table 1
Patient Demographics and Tumor Characteristics Group 1 malignant
No. of patients Mean age (y) Mean tumor size (cc) Overall score Near score Far score
64.9 36.3 5.1 4.4 5.7
20 19.8* 11.6 1.5 1.3 1.4
± ± ± ± ±
Group 2 benign 10 25.6 ± 15.8 ± 8.7 ± 8.5 ± 8.8 ±
Group 3 leydig cell 5
4.7 6.6 1.0 0.8 1.2
62.4 25.3 8.1 8.1 8.1
± ± ± ± ±
9.2t 8.8t 1.1:1: 1.1§ 1.4:1:
* Values are means ± SD. t Group 2 differs significantly from groups 1 and 3 (P < 0.05). :I: Group 1 differs significantly from groups 2 and 3 (P < 0.01). § Group 1 differs significantly from groups 2 and 3 (P < 0.001).
Ho et al.
Testis tumors and local spermatogenesis
163
A
% OF TOTAL TUBULES
50 40
30 20
2
3
4 5 6 7 8 JOHNSEN SCORE
9
10
B
% OF TOTAL TUBULES
2
3
4 5 6 7 JOHNSEN SCORE
8
9
10
Figure 1 (A), Distribution of individual tubule scores for malignant and benign tumors in the near category «3 mm from tumor). ~, malignant tumors; 0, benign tumors. (B), Comparison of near and far tubule scores for malignant tumors. ~, tubules within 3 mm from tumor (near); 0, tubules >3 mm from tumor (far).
spermatozoa, suggesting a maturation arrest at the spermatocyte level of differentiation. In contrast, spermatids or spermatozoa were present in 90% of tubules in the benign group. Among men with malignant tumors, the near score was significantly lower than the far score (4.4 versus 5.7, P < 0.0001). A histogram of tubule scores reveals similar curves with a shift toward lower scores for the near category (Fig. IB). Spermatogenesis was consistently poor in regions adjacent to these tumors regardless of tumor type. In contrast, spermatogenesis distant from the tumor appeared to vary with tumor type, with far scores ranging from 5.5 in men with large cell lymphomas to 6.9 in men with non lymphomatous metastases to the testis, such as gastric carcinoma. Marked intrapatient variation of score counts was also observed among tubules in these distant regions. For example, small islands of tubules with high scores could be found amid tubules with lower scores. In addition, moderate lymphocytic infiltrates were noted routinely in areas adjacent to malignant tumors. 164
Ho et al.
Testis tumors and local spermatogenesis
Among the benign and Leydig cell groups, there was little variation in scores, and lymphocytic infiltrates were not observed. Two subgroups merit special mention. The first consists of two patients with large cell lymphomas who exhibited completely normal spermatogenesis throughout the specimen. The second subgroup is comprised ofthe three patients with Burkitt's lymphoma. Although they were among the youngest men in this study, they demonstrated a diffuse, homogeneous impairment of spermatogenesis throughout the testis, with similar near (mean, 4.9) and far (mean, 5.1) scores. Statistical analysis of tumor size and patient age failed to identify features contributing significantly to score variation. Although larger malignant tumors demonstrated lower scores than smaller tumors (5.7 versus 6.5, P = 0.07), this difference did not reach statistical significance. No statistical difference was noted between tumors occupying more than half of the testicular volume compared with those involving less than half the testis. Patient age did not appear to be a factor in Johnsen score. The age of the Leydig cell group was similar to that of the malignant group (62.4 years versus 64.9 years), yet their overall scores differed considerably (8.1 versus 5.1, P < 0.0001). Further, despite a large difference in age between the Leydig cell group and the benign group (62.4 years versus 25.6 years), comparison of Johnsen scores revealed no significant difference (8.1 versus 8.7, P> 0.05). Overall, men :2:55 years of age scored 5.7, and men <55 scored 5.8. DISCUSSION
Subfertility has been noted among young men with malignancies ranging from testicular germ cell tumors to nontesticular tumors such as lymphoma, malignant melanoma, and thyroid cancer (1-6). In studies evaluating the feasibility of semen cryopreservation in patients with malignant disease only, 23% of men with either lymphomas or testicular germ cell tumors demonstrated sperm density and motility comparable to an age-matched control group (1). The etiology or mechanism for this cancer-associated subfertility has not been established. To directly evaluate the influence oftesticular cancer on ipsilateral spermatogenesis, Ho et al. (7) performed a histologic review of radical orchiectomy specimens. This revealed marked impairment of spermatogenesis, particularly in the immediate vicinity of malignant lesions. Tumor type was also associated with different degrees of impairment. Fertility and Sterility
Those results suggested that impairment of spermatogenesis may be mediated by humoral or local factors, either produced directly by malignant germ cell tumors or in response to the presence of malignancy. To determine whether depressed spermatogenesis represents a specific response to germ cell tumors or a general response to malignancy within the testis, a similar histologic review of radical orchiectomy specimens was performed for three groups of men with nongerm cell tumors of the testis: malignant nongerm cell tumors, benign tumors, and Leydig cell tumors. Leydig cell tumors are generally considered benign but may metastasize in up to 10% of affected men (9). The clinical presentation of all men with Leydig cell tumors in this study was consistent with benign pathology. Spermatogenesis was relatively uniform and complete in men with either benign lesions or Leydig cell tumors, and scores in these two groups were similar. In contrast, malignant tumors demonstrated markedly impaired spermatogenesis throughout the testis. This impairment was particularly severe in regions adjacent to tumor, regardless of tumor type. Spermatogenesis was less severely depressed in regions distant from tumor and was influenced by tumor type. For example, men with Burkitt's lymphoma demonstrated particularly low Johnsen scores. These observations regarding malignant nongerm cell tumors are identical to those previously reported for testicular germ cell cancers (7). Although the group with malignant tumors was older than the group with benign lesions, age was not a significant factor in this study. The three individuals with Burkitt's lymphoma had severely depressed spermatogenesis, despite being among the youngest patients in the study. Further, the group with Leydig cell tumors demonstrated Johnsen scores very similar to those with benign lesions, despite a large age difference of 37 years. Finally, the malignant group in this study had a mean age of 64.9 years compared with a mean age of 30.1 years for men with testicular cancer (7), yet Johnsen scores were quite similar. Testicular cancer scores were 4.1 near and 5.9 far, and malignant nongerm cell tumors in this study were 4.4 near and 5.7 far. This indicates that malignancy and not age is the critical determinant of Johnsen score. This histologic review demonstrates that malignancy within the testis is associated with severe impairment of spermatogenesis. The presence of a gradient effect, with the greatest impairment occurring adjacent to malignant tumors and a milder deVol. 62, No.1, July 1994
rangement at a distance, suggests a direct local influence of tumor on spermatogenesis. A mass effect with compression of seminiferous tubules against the relatively inelastic tunica albuginea of the testis seems unlikely because larger tumors did not demonstrate significantly greater impairment of spermatogenesis, and some of the smallest tumors were associated with the lowest Johnsen scores. Further, benign tumors failed to exhibit any difference between near and far scores, suggesting that the mere presence of a mass cannot account for the altered germ cell development seen with malignant lesions. Cachexia associated with malignancy could certainly contribute to impaired spermatogenesis but would not account for the differences between near and far tubules noted in this study. It has also been proposed that the subfertility noted among men with testicular cancer may be due to an inherent testicular defect, resulting in both a predisposition to cancer and impaired spermatogenesis. The spermatogenic impairment noted in this study with nongerm cell metastases to the testis renders this argument untenable. The gradient effect observed for malignant tumors, whether of germ cell or nongerm cell origin, suggests that one or more factors may be secreted by these tumors, which then act locally to interfere with normal spermatogenesis. Alternatively, humoral factors may be elicited in response to malignancy. In this regard, it is noteworthy that lymphocytic infiltrates were routinely observed in the vicinity of malignant but not benign tumors, raising the possibility of an immune phenomenon, perhaps mediated by cytokines. Both tumor necrosis factor and interleukin-1 have been demonstrated in vitro to influence testosterone secretion by adult rat Leydig cells (10), providing a rationale for impaired spermatogenesis in the vicinity of malignant tumors. Tumor-secreted factors, such as a-fetoprotein (AFP) and hCG in testicular cancer, have been proposed as candidates for mediating subfertility in affected men (6). However, no correlation was noted between Johnsen scores and serum AFP or hCG levels among men with testicular carcinoma (7). Furthermore, these particular agents are secreted by only a limited number of tumors and could not be expected to account for the impaired spermatogenesis demonstrated by the variety of tumors in this study. However, the possibility remains that some other unidentified tumor product plays a role in this process. Because the histologic appearance of seminiferous tubules in the presence of maligHo et al.
Testis tumors and local spermatogenesis
165
nancy resembles that of maturation arrest, it is possible that the effect of cancer on spermatogenesis is related directly to the differentiation process itself. Elucidation of the mechanism by which malignancy results in impaired spermatogenesis may therefore provide valuable clinical insights into the biological controls of spermatogenesis. In summary, malignant nongerm cell tumors of the testis were associated with marked ipsilateral impairment of spermatogenesis, particularly in areas adjacent to the tumor. In contrast, normal spermatogenesis was observed among specimens with benign tumors. These results are similar to those observed with germ cell tumors and suggest that malignant tumors exert a local negative effect on spermatogenesis as a general phenomenon. The mechanism for this effect is yet to be determined. REFERENCES 1. Sanger WG, Armitage JO, Schmidt MA. Feasibility of se-
men cryopreservation in patients with malignant disease. JAMA 1980;244:789-90.
166
Ho et al.
Testis tumors and local spermatogenesis
2. Thacil JV, Jewett MA, Rider WD. The effects of cancer and cancer therapy on male fertility. J UroI1981;126:141-5. 3. Fossa SD, Aass N, MoIne K. Is routine pre-treatment cryopreservation of semen worthwhile in the management of patients with testicular cancer? Br J Urol 1989;64:524-9. 4. Berthelsen JG. Sperm counts and serum follicle-stimulating hormone levels before and after radiotherapy and chemotherapy in men with testicular germ cell cancer. Fertil Steril 1992;41:281-5. 5. Banth V, Schantelder M. Fertility studies in malignancy (tumors of the testicle, malignant melanomas, cancer of the thyroid gland). Andrologia 1988;20:75-82. 6. Berthelsen JG, Skakkebaek NE. Gonadal function in men with testis cancer. Fertil SteriI1983;39:68-72. 7. Ho GT, Gardner H, DeWolf WC, Loughlin KR, Morgentaler A. Influence of testicular carcinoma on ipsilateral spermatogenesis. J Urol 1992;148:821-5. 8. Johnsen SG. Testicular biopsy score count-a method for registration of spermatogenesis in human testis: normal values and results in 335 hypo gonadal males. Hormones (Basel) 1970;1:2-25. 9. Azer PC, Braumstein GD. Malignant Leydig cell tumor. Cancer 1981;47:1251-4. 10. Warren DW, Pasupuleti V, Lu Y, Platler BW, Horton R. Tumor necrosis factor and interleukin-1 stimulate testosterone secretion in adult male rat Leydig cell in vitro. J Androl 1990;11:353-5.
Fertility and Sterility
FERTILITY AND STERILITY
Printed on acid-free paper in U. S. A.
Copyright" 1994 The American Fertility Society
The effect of testicular nongerm cell tumors on local spermatogenesis
George T. Ho, M.D.* Humphrey Gardner, M.D.t Kash Mostofi, M.D.:\:
William C. DeWolf, M.D.§ Kevin R. Loughlin, M.D.II Abraham Morgentaler, M.D.§1f
Beth Israel Hospital, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, and Armed Forces Institute of Pathology, Washington, D.C.
Objective: To assess ipsilateral spermatogenesis in men with malignant and benign nongerm cell tumors of the testis. Design, Patients: Histologic review of radical orchiectomy specimens performed for 20 men with malignant nongerm cell tumors and 15 with benign testicular lesions, including five Leydig cell tumors with benign clinical features. Main Outcome Measures: Degree of spermatogenesis was determined on a 1 to 10 scale, with 10 representing mature sperm within a seminiferous tubule. For each patient "near" and "far" scores were determined by obtaining the mean score of 50 tubules adjacent «3 mm) to the tumor and 50 tubules distant (>3 mm) from the tumor, respectively. Results: Total, near, and far scores were all lower for malignant tumors than for benign lesions. Scores for Leydig cell tumors were similar to benign lesions. Malignant tumors demonstrated a gradient effect, with greatest impairment of spermatogenesis occurring adjacent to tumor. In contrast, a distinction between near and far scores was not observed for benign lesions or Leydig cell tumors. Conclusions: Malignant nongerm cell tumors of the testis were associated with significant impairment of ipsilateral spermatogenesis, particularly in areas adjacent to tumor. These findings are similar to those observed for testicular germ cell tumors, suggesting a generalized negative Fertil Steril 1994;62:162-6 influence on ipsilateral spermatogenesis by malignant tumors. Key Words: Spermatogenesis, testis, cancer, infertility, tumor
Clinical subfertility has been reported among men with a variety of malignancies, most notably testicular carcinoma (1-6). The cause of this subfertility, however, is not known. In a histologic
Received October 27, 1993: revised and accepted February 21, 1994. * Harvard Program in Urology (Longwood Area), Harvard Medical School. t Department of Pathology, Beth Israel Hospital. :j: Armed Forces Institute of Pathology. § Division of Urology, Beth Israel Hospital. 1\ Division of Urological Surgery, Brigham and Women's Hospital. 1f Reprint requests: Abraham Morgentaler, M.D., 330 Brookline Avenue, Boston, Massachusetts 02215 (FAX: 617-2787292).
162
Ho et al.
Testis tumors and local spermatogenesis
review of radical orchiectomy specimens for testicular germ cell tumors, Ho et al. (7) observed widespread impairment of spermatogenesis throughout the ipsilateral testis. This impairment was particularly severe in regions adjacent to the tumor, with the degree of impairment dependent on tumor type. Since nongerm cell tumors are also associated with subfertility, the possibility that malignant tumors exert a generalized negative influence on spermatogenesis warrants consideration. To assess the effect of nongerm cell tumors on spermatogenesis, we performed a histologic review of orchiectomy specimens from men with malignant and benign nongerm cell tumors of the testis. Spermatogenesis adjacent to and distant from the primary tumor was examined to determine whether Fertility and Sterility
a gradient effect might be present, as noted with germ cell tumors ofthe testis (7). Clinical parameters such as patient age, tumor type, and tumor size were also evaluated for their influence on spermatogenesis. MATERIALS AND METHODS
All radical orchiectomy specimens evaluated at the Beth Israel Hospital of Boston, the Brigham and Women's Hospital Boston, Massachusetts, and the Armed Forces Institute of Pathology Washington, D.C., from 1980 to 1991 were reviewed. Patients with nongerm cell testicular tumors without prior history of chemotherapy or radiation therapy and with adequate medical history and histologic material were included in this study. Patients were divided into three groups as follows: [1] 20 men with malignant testicular nongerm cell tumors (12 large cell lymphomas, 3 Burkitt's lymphoma, 2 rhabdomyosarcoma, 2 gastric carcinoma, 1 pancreatic carcinoma); [2] 10 men with benign testicular tumors (5 cysts, 3 fibrosis, 1 sarcoidosis, 1 splenic rest); and [3] 5 men with Leydig cell tumors. None of the last group manifested clinical evidence of malignancy. Because a majority of the malignant tumors were reviewed as consult slides at the Armed Forces Institute of Pathology, only limited information was available on the general health of these patients. The quality of spermatogenesis was determined by scoring seminiferous tubules according to the method of Johnsen (8). Tubules were scored on a scale of 1 to 10, with 10 indicating more than five mature spermatozoa within the lumen. Two groups of 50 seminiferous tubules were scored for each case: "near" scores were obtained from tubules within 3 mm from the tumor margin, and "far" scores from beyond 3 mm. All tubules in a series of contiguous fields were scored. Scoring for the far group was performed in areas demonstrating the highest degree of local spermatogenesis. Whenever possible, the near and far groups of tubules were scored on the same slide. In a few cases, no slide was available that showed tumor contiguous with uninvolved tubules, and in these cases a near score could not be obtained. The mean far Johnsen scores of patients with malignant nongerm cell testicular tumors were stratified on the basis of patient age and tumor size to determine the possible contribution of these parameters on the degree of spermatogenesis impairment. Testis and tumor sizes were determined from the original pathology reports. Patients were stratiVol. 62, No.1, July 1994
fied into two subgroups on the basis of age with one group consisting of patients 55 years old or younger and a second subgroup consisting of patients older than 55. Patients were also stratified on the basis of absolute tumor size by dividing the patients into two subgroups, one consisting of patients whose tumors were <20 cc in size and a second consisting of patients with tumors ~20 cc in size. Last, the contribution of relative tumor size was evaluated by dividing the patients into two subgroups consisting of one group with tumor involving <50% of the testis and a second subgroup composed of men whose tumors comprised >50% of the testis. Statistical analysis was performed using the mean Johnsen scores for each of 50 tubules, consisting of a mean near and a mean far score for each patient. Comparison of mean Johnsen scores for tumor types was conducted using a Mann-Whitney test. Other parameters were analyzed by t-test (two-tail). Relative tumor size was calculated by dividing the absolute tumor size by the size of the testicle.
RESULTS
Scores for malignant tumors were lower than benign tumors in all categories: near (4.4 versus 8.5, P < 0.0001); far (5.7 versus 8.8, P < 0.0001); and overall score (5.1 versus 8.7, P < 0.0001). Scores for Leydig cell tumors were similar to other nonma1ignant tumors (Table 1) and also differed significantly from the malignant group (near 8.1, P < 0.001; far 8.1, P < 0.01). Score distributions for individual tubules are presented in Figure lA, revealing minimal overlap of scores between benign and malignant tumors. Almost 90% of tubules in the malignant group failed to demonstrate spermatids or
Table 1
Patient Demographics and Tumor Characteristics Group 1 malignant
No. of patients Mean age (y) Mean tumor size (cc) Overall score Near score Far score
64.9 36.3 5.1 4.4 5.7
20 19.8* 11.6 1.5 1.3 1.4
± ± ± ± ±
Group 2 benign 10 25.6 ± 15.8 ± 8.7 ± 8.5 ± 8.8 ±
Group 3 leydig cell 5
4.7 6.6 1.0 0.8 1.2
62.4 25.3 8.1 8.1 8.1
± ± ± ± ±
9.2t 8.8t 1.1:1: 1.1§ 1.4:1:
* Values are means ± SD. t Group 2 differs significantly from groups 1 and 3 (P < 0.05). :I: Group 1 differs significantly from groups 2 and 3 (P < 0.01). § Group 1 differs significantly from groups 2 and 3 (P < 0.001).
Ho et al.
Testis tumors and local spermatogenesis
163
A
% OF TOTAL TUBULES
50 40
30 20
2
3
4 5 6 7 8 JOHNSEN SCORE
9
10
B
% OF TOTAL TUBULES
2
3
4 5 6 7 JOHNSEN SCORE
8
9
10
Figure 1 (A), Distribution of individual tubule scores for malignant and benign tumors in the near category «3 mm from tumor). ~, malignant tumors; 0, benign tumors. (B), Comparison of near and far tubule scores for malignant tumors. ~, tubules within 3 mm from tumor (near); 0, tubules >3 mm from tumor (far).
spermatozoa, suggesting a maturation arrest at the spermatocyte level of differentiation. In contrast, spermatids or spermatozoa were present in 90% of tubules in the benign group. Among men with malignant tumors, the near score was significantly lower than the far score (4.4 versus 5.7, P < 0.0001). A histogram of tubule scores reveals similar curves with a shift toward lower scores for the near category (Fig. IB). Spermatogenesis was consistently poor in regions adjacent to these tumors regardless of tumor type. In contrast, spermatogenesis distant from the tumor appeared to vary with tumor type, with far scores ranging from 5.5 in men with large cell lymphomas to 6.9 in men with non lymphomatous metastases to the testis, such as gastric carcinoma. Marked intrapatient variation of score counts was also observed among tubules in these distant regions. For example, small islands of tubules with high scores could be found amid tubules with lower scores. In addition, moderate lymphocytic infiltrates were noted routinely in areas adjacent to malignant tumors. 164
Ho et al.
Testis tumors and local spermatogenesis
Among the benign and Leydig cell groups, there was little variation in scores, and lymphocytic infiltrates were not observed. Two subgroups merit special mention. The first consists of two patients with large cell lymphomas who exhibited completely normal spermatogenesis throughout the specimen. The second subgroup is comprised ofthe three patients with Burkitt's lymphoma. Although they were among the youngest men in this study, they demonstrated a diffuse, homogeneous impairment of spermatogenesis throughout the testis, with similar near (mean, 4.9) and far (mean, 5.1) scores. Statistical analysis of tumor size and patient age failed to identify features contributing significantly to score variation. Although larger malignant tumors demonstrated lower scores than smaller tumors (5.7 versus 6.5, P = 0.07), this difference did not reach statistical significance. No statistical difference was noted between tumors occupying more than half of the testicular volume compared with those involving less than half the testis. Patient age did not appear to be a factor in Johnsen score. The age of the Leydig cell group was similar to that of the malignant group (62.4 years versus 64.9 years), yet their overall scores differed considerably (8.1 versus 5.1, P < 0.0001). Further, despite a large difference in age between the Leydig cell group and the benign group (62.4 years versus 25.6 years), comparison of Johnsen scores revealed no significant difference (8.1 versus 8.7, P> 0.05). Overall, men :2:55 years of age scored 5.7, and men <55 scored 5.8. DISCUSSION
Subfertility has been noted among young men with malignancies ranging from testicular germ cell tumors to nontesticular tumors such as lymphoma, malignant melanoma, and thyroid cancer (1-6). In studies evaluating the feasibility of semen cryopreservation in patients with malignant disease only, 23% of men with either lymphomas or testicular germ cell tumors demonstrated sperm density and motility comparable to an age-matched control group (1). The etiology or mechanism for this cancer-associated subfertility has not been established. To directly evaluate the influence oftesticular cancer on ipsilateral spermatogenesis, Ho et al. (7) performed a histologic review of radical orchiectomy specimens. This revealed marked impairment of spermatogenesis, particularly in the immediate vicinity of malignant lesions. Tumor type was also associated with different degrees of impairment. Fertility and Sterility
Those results suggested that impairment of spermatogenesis may be mediated by humoral or local factors, either produced directly by malignant germ cell tumors or in response to the presence of malignancy. To determine whether depressed spermatogenesis represents a specific response to germ cell tumors or a general response to malignancy within the testis, a similar histologic review of radical orchiectomy specimens was performed for three groups of men with nongerm cell tumors of the testis: malignant nongerm cell tumors, benign tumors, and Leydig cell tumors. Leydig cell tumors are generally considered benign but may metastasize in up to 10% of affected men (9). The clinical presentation of all men with Leydig cell tumors in this study was consistent with benign pathology. Spermatogenesis was relatively uniform and complete in men with either benign lesions or Leydig cell tumors, and scores in these two groups were similar. In contrast, malignant tumors demonstrated markedly impaired spermatogenesis throughout the testis. This impairment was particularly severe in regions adjacent to tumor, regardless of tumor type. Spermatogenesis was less severely depressed in regions distant from tumor and was influenced by tumor type. For example, men with Burkitt's lymphoma demonstrated particularly low Johnsen scores. These observations regarding malignant nongerm cell tumors are identical to those previously reported for testicular germ cell cancers (7). Although the group with malignant tumors was older than the group with benign lesions, age was not a significant factor in this study. The three individuals with Burkitt's lymphoma had severely depressed spermatogenesis, despite being among the youngest patients in the study. Further, the group with Leydig cell tumors demonstrated Johnsen scores very similar to those with benign lesions, despite a large age difference of 37 years. Finally, the malignant group in this study had a mean age of 64.9 years compared with a mean age of 30.1 years for men with testicular cancer (7), yet Johnsen scores were quite similar. Testicular cancer scores were 4.1 near and 5.9 far, and malignant nongerm cell tumors in this study were 4.4 near and 5.7 far. This indicates that malignancy and not age is the critical determinant of Johnsen score. This histologic review demonstrates that malignancy within the testis is associated with severe impairment of spermatogenesis. The presence of a gradient effect, with the greatest impairment occurring adjacent to malignant tumors and a milder deVol. 62, No.1, July 1994
rangement at a distance, suggests a direct local influence of tumor on spermatogenesis. A mass effect with compression of seminiferous tubules against the relatively inelastic tunica albuginea of the testis seems unlikely because larger tumors did not demonstrate significantly greater impairment of spermatogenesis, and some of the smallest tumors were associated with the lowest Johnsen scores. Further, benign tumors failed to exhibit any difference between near and far scores, suggesting that the mere presence of a mass cannot account for the altered germ cell development seen with malignant lesions. Cachexia associated with malignancy could certainly contribute to impaired spermatogenesis but would not account for the differences between near and far tubules noted in this study. It has also been proposed that the subfertility noted among men with testicular cancer may be due to an inherent testicular defect, resulting in both a predisposition to cancer and impaired spermatogenesis. The spermatogenic impairment noted in this study with nongerm cell metastases to the testis renders this argument untenable. The gradient effect observed for malignant tumors, whether of germ cell or nongerm cell origin, suggests that one or more factors may be secreted by these tumors, which then act locally to interfere with normal spermatogenesis. Alternatively, humoral factors may be elicited in response to malignancy. In this regard, it is noteworthy that lymphocytic infiltrates were routinely observed in the vicinity of malignant but not benign tumors, raising the possibility of an immune phenomenon, perhaps mediated by cytokines. Both tumor necrosis factor and interleukin-1 have been demonstrated in vitro to influence testosterone secretion by adult rat Leydig cells (10), providing a rationale for impaired spermatogenesis in the vicinity of malignant tumors. Tumor-secreted factors, such as a-fetoprotein (AFP) and hCG in testicular cancer, have been proposed as candidates for mediating subfertility in affected men (6). However, no correlation was noted between Johnsen scores and serum AFP or hCG levels among men with testicular carcinoma (7). Furthermore, these particular agents are secreted by only a limited number of tumors and could not be expected to account for the impaired spermatogenesis demonstrated by the variety of tumors in this study. However, the possibility remains that some other unidentified tumor product plays a role in this process. Because the histologic appearance of seminiferous tubules in the presence of maligHo et al.
Testis tumors and local spermatogenesis
165
nancy resembles that of maturation arrest, it is possible that the effect of cancer on spermatogenesis is related directly to the differentiation process itself. Elucidation of the mechanism by which malignancy results in impaired spermatogenesis may therefore provide valuable clinical insights into the biological controls of spermatogenesis. In summary, malignant nongerm cell tumors of the testis were associated with marked ipsilateral impairment of spermatogenesis, particularly in areas adjacent to the tumor. In contrast, normal spermatogenesis was observed among specimens with benign tumors. These results are similar to those observed with germ cell tumors and suggest that malignant tumors exert a local negative effect on spermatogenesis as a general phenomenon. The mechanism for this effect is yet to be determined. REFERENCES 1. Sanger WG, Armitage JO, Schmidt MA. Feasibility of se-
men cryopreservation in patients with malignant disease. JAMA 1980;244:789-90.
166
Ho et al.
Testis tumors and local spermatogenesis
2. Thacil JV, Jewett MA, Rider WD. The effects of cancer and cancer therapy on male fertility. J UroI1981;126:141-5. 3. Fossa SD, Aass N, MoIne K. Is routine pre-treatment cryopreservation of semen worthwhile in the management of patients with testicular cancer? Br J Urol 1989;64:524-9. 4. Berthelsen JG. Sperm counts and serum follicle-stimulating hormone levels before and after radiotherapy and chemotherapy in men with testicular germ cell cancer. Fertil Steril 1992;41:281-5. 5. Banth V, Schantelder M. Fertility studies in malignancy (tumors of the testicle, malignant melanomas, cancer of the thyroid gland). Andrologia 1988;20:75-82. 6. Berthelsen JG, Skakkebaek NE. Gonadal function in men with testis cancer. Fertil SteriI1983;39:68-72. 7. Ho GT, Gardner H, DeWolf WC, Loughlin KR, Morgentaler A. Influence of testicular carcinoma on ipsilateral spermatogenesis. J Urol 1992;148:821-5. 8. Johnsen SG. Testicular biopsy score count-a method for registration of spermatogenesis in human testis: normal values and results in 335 hypo gonadal males. Hormones (Basel) 1970;1:2-25. 9. Azer PC, Braumstein GD. Malignant Leydig cell tumor. Cancer 1981;47:1251-4. 10. Warren DW, Pasupuleti V, Lu Y, Platler BW, Horton R. Tumor necrosis factor and interleukin-1 stimulate testosterone secretion in adult male rat Leydig cell in vitro. J Androl 1990;11:353-5.
Fertility and Sterility