Does retroperitoneal lymph node dissection have a curative role for patients with sex cord–stromal testicular tumors?

Abstract

BACKGROUND

Sex cord–stromal tumors account for < 5% of all adult testicular tumors, and 10% are malignant. Due to the limited reported experience, there is no agreement on the best management, especially in patients who have tumors with malignant pathologic features or who present with metastatic disease. The authors attempt to evaluate the role of retroperitoneal lymph node dissection (RPLND) in the management of patients with these malignant sex cord–stromal tumors.

METHODS

Reviewing the Indiana University testis cancer registry revealed 17 patients who underwent RPLND for sex cord–stromal tumors. Pathology was reviewed for features suggestive of malignancy. The data examined included clinical and pathologic stage, surgical procedure, additional therapy received, and outcome.

RESULTS

Pathology included Leydig tumors in six patients, Sertoli tumors in four patients, sex cord–stromal tumors in five patients, a granulosa cell tumor in one patient, and a poorly differentiated non–germ cell tumor in one patient. Nine patients had histologic features suggestive of malignancy. Clinical stage at surgery was Stage I in nine patients and Stage IIA–IIIA in eight patients. Patients underwent modified or bilateral RPLND. Nine patients had pathologic Stage I tumors, and the remaining eight patients and had pathologic Stage IIB–IIIA tumors. Follow-up ranged from 8 months to 11 years. Of the eight patients with Stage II–III disease, six patients eventually died of metastatic disease despite additional radiotherapy and/or chemotherapy.

CONCLUSIONS

Sex cord–stromal tumors have a potentially aggressive malignant behavior that is difficult to predict based on clinical and pathologic features. Although the therapeutic role of RPLND in patients with small-volume metastatic retroperitoneal tumors is unclear, RPLND remains an option to be performed immediately after orchiectomy, especially in patients who have tumors with malignant features and/or small-volume metastatic disease. Cancer 2003;98:753–7. © 2003 American Cancer Society.

Sex cord–stromal testicular tumors are a group of primary testicular neoplasms that includes tumors of Leydig, Sertoli, and granulosa cell origins in addition to mixed cell types and incompletely differentiated tumors. These tumors, as a group, comprise 4–5% of all testicular tumors.1, 2 Approximately 10–12% of these tumors are malignant. Although a number of studies have analyzed the pathologic features suggestive of malignancy,3–8 no unequivocal histologic criteria for malignancy have been defined. The experience reported with these tumors is limited, and there is controversy regarding the best approach and management. Many centers recommend orchiectomy alone for patients with clinical Stage I tumors because of the uncertain malignant behavior and scant data relating to the value of retroperitoneal lymph node dissection (RPLND) in these tumors. In the current study, we report on our efforts to evaluate the role of RPLND in the management of patients with malignant sex cord–stromal testicular tumors.

MATERIALS AND METHODS

Between 1986 and March 2001, 17 RPLND procedures were performed for patients with sex cord–stromal testicular tumors, as identified by a search of the Indiana University testis cancer registry. Patients ranged in age from 28 years to 54 years (average, 40.9 years). Patients had clinical Stage I disease (T1–T4, N0, M0: no regional lymph node involvement or distant metastasis), Stage IIA disease (any T, N1, M0: metastasis with a lymph node mass measuring ≤ 2 cm in greatest dimension or multiple lymph node involvement with no masses measuring > 2 cm in greatest dimension and no distant metastasis), Stage IIB disease (any T, N2, M0: metastasis with a lymph node mass measuring > 2 cm but not > 5 cm in greatest dimension, or multiple lymph nodes with any 1 mass measuring > 2 cm but not > 5 cm in greatest dimension and no distant metastasis), or Stage IIIA disease (any T, any N, M1a: nonregional lymph node or pulmonary metastasis) at the time of RPLND. If patients underwent orchiectomy elsewhere, then the pathology of the orchiectomy specimen was reviewed at Indiana University to confirm the diagnosis of sex cord–stromal tumor and to assess for features suggestive of malignancy, including size > 5 cm, necrosis, moderate-to-marked nuclear atypia, lymphovascular invasion, invasive margins or extratesticular extension, and > 5 mitotic figures per 10 high-power fields. Additional data examined included presentation, clinical stage at initial presentation and at surgery, surgical procedure, pathologic stage, additional therapy, and outcome. Follow-up ranged from 8 months to 11 years.

RESULTS

Pathology of testicular tumors included Leydig cell tumor in six patients, Sertoli cell tumor in four patients, sex cord–stromal tumor (type not otherwise specified) in five patients, granulosa cell tumor in one patient, and poorly differentiated non–germ cell tumor in one patient. The presence and frequency of features suggestive of malignancy are illustrated in Table 1. Nine patients (five patients with clinical Stage I disease and four patients with clinical Stage II disease) exhibited two or more of these features in the orchiectomy specimen.

Table 1. Pathologic Features Suggestive of Malignancy

Pathologic feature	No. of patients
HPF: high power field. a With a × 40 objective.
Tumor size > 5 cm	2
Necrosis	7
Moderate or marked nuclear atypia	6
Angiolymphatic invasion	2
Infiltrating margin/paratesticular invasion	4
Greater than 5 mitotic figures per 10 HPFa	10

The clinical stage at RPLND was Stage I in nine patients and Stage II–III in eight patients (IIA–IIB in seven patients and Stage IIIA in one patient). Four of the eight patients with clinical Stage II–III disease presented first with as clinical Stage I disease. They initially were managed by surveillance, and they developed evidence of metastatic disease 14 months, 30 months, 37 months, and 58 months, respectively, after undergoing orchiectomy. All patients had normal tumor markers (α-fetoprotein; human chorionic gonadotropin;, and, if measured, lactate dehydrogenase).

Patients underwent modified unilateral RPLND (7 patients) or full bilateral RPLND (10 patients). One patient underwent an additional unilateral suprahilar dissection; and one patient underwent pelvic dissection. The patient who had Stage IIIA disease, in addition to bilateral RPLND, underwent a right thoracotomy for excision of multiple pulmonary nodules. Three patients underwent nerve-sparing surgery. Additional procedures that were required intraoperatively included an arterial graft and nephrectomy in one patient and nephrectomy in another patient. Postoperative complications occurred in three patients (all with pathologic Stage II–III tumors) and included pancreatitis and pulmonary atelectasis in one patient, ascites and pleural effusion in another patient, and ascites in the third patient. The pathologic stage was Stage I in all nine patients who had clinical Stage I disease and Stage IIB–IIIA in the remaining eight patients.

The characteristics and outcomes of the 17 patients studied are detailed in Table 2. Follow-up ranged from 8 months to 11 years (average, 3.3 years). All patients with Stage I disease were alive with no evidence of disease at last follow up (9 months to 11 years; average, 4.5 years). Six of 8 patients with Stage II–III tumors died of disease within 9 months to 6 years after RPLND (median survival, 1.2 years; average, 2.4 years). Of the 2 remaining patients, 1 patient was alive with metastatic disease at 13 months, and 1 patient had no evidence of disease at last follow-up (8 months).

Table 2. Characteristics and Outcome of 17 Patients with Sex Cord–Stromal Tumors Who Underwent Retroperitoneal Lymph Node Dissection

Patient	Clinical stage at presentation	Pathology	No. of malignant pathologic features	Clinical stage at RPLND	Pathologic stage	Additional treatment	Outcome	Follow-up (mos)
RPLND: retroperitoneal lymph node dissection; LC: Leydig cell tumor; SC-ST: sex cord–stromal tumor; SCT: Sertoli cell tumor; GrCT: granulosa cell tumor; Non-GCT: non–germ cell tumor; Chemo: chemotherapy; RT: radiotherapy; NED: no evidence of disease; DOD: died of disease; AWD: alive with disease. a Patient was managed initially with surveillance.
1	I	LCT	3	I	I	—	NED	135
2	I	SC-ST	2	I	I	—	NED	80
3	IIIA	SCT	2	IIIA	IIIA	Chemo, surgery	DOD	69
4a	I	LCT	1	IIB	IIB	Surgery	DOD	54
5	I	GrCT	1	I	I	—	NED	80
6	I	LCT	3	I	I	—	NED	25
7	I	SCT	0	I	I	—	NED	63
8	I	LCT	4	I	I	—	NED	38
9	I	SC-ST	3	I	I	—	NED	35
10	I	SCT	0	I	I	—	NED	9
11a	I	SC-ST	5	IIB	IIC	—	DOD	15
12	IIB	SCT	3	IIB	IIB	Chemo, surgery	DOD	9
13a	I	LCT	0	IIB	IIC	Surgery	DOD	11
14	IIA	Non-GCT	1	IIA	IIB	Chemo, surgery	NED	8
15a	I	LCT	3	IIB	IIB	Chemo, RT	DOD	11
16	I	SC-ST	0	I	I	—	NED	20
17	IIB	SC-ST	1	IIB	IIC	Chemo, surgery	AWD	13

Additional therapy included one or more of various treatment modalities. Five patients with Stage II–III disease received chemotherapy (as adjuvant treatment or for tumor recurrence). Various regimens (most commonly, bleomycin, etoposide and cisplatin [BEP]; etoposide and cisplatin [EP]; and vinblastine, ifosfamide, and cisplatin [VIP]) and schedules were used according to the individual patient situation. The responses varied but were mostly partial and short term. A number of patients underwent additional surgeries for recurrent disease, including abdominal and pelvic mass excisions, bowel resection, hepatic lobectomy, and splenectomy. One patient received palliative radiotherapy for thoracic spine disease with spinal cord compression.

DISCUSSION

Sex cord–stromal tumors are rare, accounting for 4–5% of all primary testicular neoplasms.1, 2 According to the literature, approximately 10% of these tumors have metastatic potential. A number of studies suggest even greater malignant potential and a higher metastatic rate. Bertram et al. reported a 20% occurrence rate of metastatic disease in patients with Leydig cell tumors,3 and Farkas et al. reported metastases leading to death in 4 of 7 patients with Leydig cell tumors.4 Dilworth et al., in their review, suggested a malignant potential in up to 20% of Sertoli cell tumors.1

A number of studies have evaluated various pathologic features in orchiectomy specimens that may correlate with the malignant potential of these tumors.1, 3–8 The features associated most strongly with malignant behavior were large primary tumor size (> 5 cm), necrosis, moderate-to-severe nuclear atypia, lymphovascular invasion, local invasive qualities, and high mitotic rate (> 5 mitotic figures per 10 high-power fields). No absolute morphologic criteria for malignancy for sex cord–stromal tumors have been established; however; in the current group of patients, at least two of those features were present in five of nine patients with Stage I tumors who did well after orchiectomy and in only four of eight patients who had documented metastatic tumors. We realize that the current series represents a group of patients who were offered RPLND or were referred for RPLND; therefore, a selection bias may have contributed to the apparent low ability of the pathologic features to predict clinical metastatic behavior compared with other studies.

The first and most common site of metastatic disease in our patients was the retroperitoneal lymph nodes. This is in agreement with most of the published literature.1, 4, 9 Other sites of metastases included the lungs, mediastinal and cervical lymph nodes, pelvis, liver, and bone.

Initial treatment in all reported patients with sex cord–stromal testis tumors has been inguinal orchiectomy. Additional treatment, especially in patients who have tumors with pathologic features suggestive of malignancy or in patients with metastatic disease, remains controversial. Due to the rarity of these tumors, the natural history has not been described well, and treatment decisions have to be based on a limited reported experience.

Four patients in the current study group initially had clinical Stage I disease at the time they underwent orchiectomy and were managed by surveillance. These patients developed evidence of retroperitoneal disease 14–58 months after orchiectomy and had clinical Stage II–III disease at the time they underwent RPLND. Although most metastases of sex cord–stromal tumors are detected at the time of diagnosis or within 1 year of orchiectomy, a number of studies also report the appearance of metastatic disease as late as 9 years3 and 15 years10 after orchiectomy, suggesting the presence of clinically undetected metastatic disease with variable growth and progression potential. Although the therapeutic impact of RPLND in patients with small-volume metastatic retroperitoneal disease is unclear, it is reasonable to assume that, if RPLND has a therapeutic benefit, then such a benefit would be greater in patients who have small-volume retroperitoneal metastasis (defined as clinical Stage ≤ IIA or pathologic Stage IIA: metastasis with a lymph node mass measuring ≤ 2 cm in greatest dimension or ≤ 5 positive lymph nodes, with none measuring > 2 cm in greatest dimension) compared with patients who have large-volume disease. A study from Japan reported on two patients with malignant gonadal stromal tumors who underwent RPLND with detection of tumor cells in the resected lymph nodes. The patients were alive and free of disease at 6 months and 5 years after surgery, respectively.11 Another report described a patient who, at the time he underwent orchiectomy, had a Leydig cell tumor with features suggestive of malignancy. On lymphangiography, there was a slight abnormality of one of the paraaortic lymph nodes. The patient underwent RPLND, which revealed microscopic metastases. The patient was free of disease at the time of the report, 6 months after undergoing retroperitoneal dissection,12 a follow-up that was too short to predict a cure.

The value of surgery for patients with large-volume or widespread, metastatic disease appears to be limited. In our series as well as in other reports,3, 10 RPLND and surgery for metastatic disease in other sites seemed to provide effective local treatment. Most patients, however, eventually succumbed to widespread recurrence.

Various chemotherapeutic agents and regimens have been used for the treatment of patients with metastatic disease, with limited success. Metastatic Leydig cell tumors appear particularly chemoresistant to all of the agents tried, alone or in combination, including BEP, EP, VIP, dactinomycin, cyclophosphamide, 5-fluorouracil, methotrexate, melphalan, and mitotane.3 There is some evidence that other subtypes of sex cord–stromal tumors, such as Sertoli cell tumors, may respond more favorably to chemotherapy.13 A number of reports have described an objective response of metastatic Sertoli cell tumors and unclassified sex cord–stromal tumors to chemotherapy.1, 10, 13 However, those responses were mostly partial and were short-lived.

Radiotherapy has been of questionable benefit for patients with malignant sex cord–stromal tumors.1 Bertram et al. reviewed 11 patients with metastatic Leydig cell tumors who received radiotherapy. There were two reports of a decrease in pain after radiation treatment, but there were no objective decreases in tumor size.3 In a review of metastasizing Sertoli cell tumors, Madson and Hultberg reported that irradiation of paraaortic lymph nodes may provide effective local treatment. None of 13 patients they so treated developed paraaortic lymph node recurrences.10

The median and average survival rates for our patients with metastatic disease were 1.2 years and 2.4 years, respectively. In the literature, survival after patients were diagnosed with malignant sex cord–stromal tumors ranged from 2 months to 17 years.1 Survival after a diagnosis of metastatic disease ranged from < 1 month to as long as 9 years, with almost two-thirds of patients dying within 2 years.3

In conclusion, sex cord–stromal tumors exhibit potentially aggressive malignant behavior that is difficult to predict based on clinical and pathologic features. The therapeutic role of RPLND in patients with low-volume metastatic disease remains unclear; however, waiting to resect large-volume disease usually is ineffective. These tumors respond poorly to radiotherapy and chemotherapy. RPLND remains an option for patients with sex cord–stromal tumors, and they should undergo the procedure immediately after orchiectomy, especially if they have tumors with malignant features and/or small-volume metastatic disease.