Schultz, Kris Ann P. MD, MS*; Schneider, Dominik T. MD†; Pashankar, Farzana MD‡; Ross, Jonathan MD§; Frazier, Lindsay MD∥
OVARIAN SEX CORD-STROMAL TUMORS IN CHILDREN
Ovarian malignancies comprise approximately 1% of childhood cancer.1 Ovarian sex cord-stromal tumors (SCST), which are distinct from germ cell tumors and epithelial ovarian tumors, comprise approximately 7% of ovarian tumors overall and approximately 15% of ovarian tumors in children.
Early in embryonic life, the sex cords arise from the primitive genital ridge or coelomic epithelium. In females, the sex cords develop into the cortical cords and later into ovarian follicles. SCSTs arise from the sex cords or from ovarian stroma/mesenchyme of the developing gonad and include both juvenile-type and adult-type granulosa cell tumors, Sertoli-Leydig cell tumors, Sertoli cell and Leydig cell tumors, as well as theca and granulosa-theca tumors, sclerosing stromal tumors, SCSTs with annular tubules, and gynandroblastomas with simultaneous Sertoli and granulosa cell differentiation (Table 1, Fig. 1). In children, juvenile granulosa cell tumors are most common followed by Sertoli-Leydig cell tumors. In adults, adult granulosa cell tumors are most common. The recently described microcystic stromal tumor of the ovary is also likely within this category, although to date, the youngest patient reported with this diagnosis was 26 years of age.2
SCSTs may present in young girls with signs and symptoms similar to those of other ovarian tumors such as abdominal pain or distention, gastrointestinal symptoms, or abdominal mass. These patients often have clinical signs of sex hormone production and may present with isosexual precocity including breast swelling and vaginal bleeding, primary or secondary amenorrhea, and/or virilization.
Granulosa cell tumors are of 2 distinct subtypes, juvenile and adult. Juvenile granulosa cell tumors are seen in young children and may present with precocious puberty. The vast majority of juvenile granulosa cell tumors are localized (ie, stage I) at diagnosis and associated with a favorable prognosis. Juvenile granulosa cell tumors rarely recur and when this occurs, it is generally within the first 2 to 3 years after diagnosis.3 Adult granulosa cell tumors are very rare in children. Adult granulosa cell tumors are more commonly seen in the early menopausal years and have a propensity for late relapse, often later than 10 years after diagnosis.4–7 Granulosa cell tumors of both types may secrete inhibin.
Sertoli-Leydig cell tumors generally occur in adolescents and young women. In rare instances, pure Sertoli cell tumors occur and may be predominantly estrogenic and may produce renin leading to hypertension. Pure Leydig cell tumors have been rarely reported, and when they occur, are predominantly androgen-secreting.
Extragonadal SCSTs also occur, albeit rarely.8 When they arise, they are usually in adult women and usually within the pelvis, most commonly in the broad ligament, although retroperitoneal, fallopian tube, umbilical hernia sac, adrenal gland, and mesenteric sites have been described. These cases have more commonly been of granulosa cell histology, or sex cord-stromal tumor with annular tubules (SCST-AT), although a mesenteric Sertoli-Leydig cell tumor has also been reported.
Recurrence of stage I ovarian SCST in children is overall uncommon. When it occurs, recurrence is within a few years of diagnosis (median 2.8 y), most often within the abdominopelvic area or in regional lymph nodes.9,10 Hematogenous spread to the chest, liver, or bones has been described only rarely, most commonly in those with positive lymph nodes at the time of initial surgery and in those with adult granulosa cell tumor histology.9 In children with localized disease at presentation, later development of metastatic disease outside the abdomen/pelvis is considered very unlikely.
Ovarian SCST are associated with multiple enchondromatosis (Ollier disease) and with Peutz-Jeghers syndrome (PJS). Patients with enchondromatosis may develop granulosa cell tumors. Ovarian tumors in PJS are generally SCST-AT; one third of patients with SCST-AT have PJS. Tumors in patients with PJS tend to present at a younger age and may be bilateral. Patients with PJS may also develop granulosa cell tumors. PJS is associated with germ-line mutations in STK11/LKB1 tumor suppressor gene located on chromosome 19p13.3. Loss of heterozygosity at this location was seen in 41% of presumed sporadic SCST in 1 series, however, no mutations or promoter methylations of STK11 were seen in these tumors.11
Ovarian SCSTs, particularly Sertoli-Leydig cell tumors, but also juvenile granulosa cell tumors and gynandroblastomas, are also seen in families with a history of pleuropulmonary blastoma.12 DICER1 mutations are common in children and families with a history of pleuropulmonary blastoma and have been observed in children and adolescents with ovarian Sertoli-Leydig cell tumor, juvenile granulosa cell tumors, and/or gynandroblastoma) and a personal or family history of pleuropulmonary blastoma or other conditions seen in this familial syndrome such as renal tumors, lung cysts, additional ovarian tumors, sarcomas, or nodular thyroid disease.13,14
Ovarian Sertoli-Leydig tumors are also known to be associated with thyroid disease and this association has been reported in constellation with DICER1 mutations.15,16 All children and adolescents with ovarian tumors should be carefully screened for a personal or family history of dysplastic or neoplastic conditions. Clinical testing for DICER1 is available.
In contrast to the epithelial ovarian cancers, there is no known association between BRCA1 or BRCA2 germ-line mutations and ovarian SCST.
Ovarian SCSTs usually have a balanced karyotype, however, 25% of juvenile granulosa cell tumors in 1 series were found to have a chromosomal imbalance, usually gain of chromosome 12.17 The prognostic significance of this is unknown and warrants future study.
In any patient with a suspected ovarian tumor, serum tumor markers [α-fetoprotein (AFP), β-human chorionic gonadotropin (hCG), carcinoembryonic antigen, cancer antigen 125 (CA-125), inhibin A and B, and lactate dehydrogenase (LDH)] and serum calcium should be obtained. Serum calcium may help to distinguish small cell carcinoma of the hypercalcemic type, a nonstromal ovarian carcinoma that may present in childhood or adolescence.18 Ovarian granulosa cell tumors generally produce inhibin A and inhibin B both of which may be useful in diagnosis and in follow-up, however, the role of these measurements has yet to be fully elucidated. Measurement of inhibin B is more specific and may be more readily available than inhibin A. The wide range of normal in prepubertal children may lead to some uncertainty, however, when elevated, inhibin A and B may be a useful marker for follow-up. Ovarian SCSTs may also produce Mullerian inhibitory substance.19 Measurement of testosterone and androsteindione should be performed when virilizing signs or symptoms are present and should be considered even in the absence of obvious clinical symptomatology.
Sertoli-Leydig cell tumors may be associated with AFP elevation, although the level is usually <500 ng/mL.20 When AFP is elevated, the histopathology may show poor differentiation and/or retiform pattern with hepatoid differentiation and heterologous elements.
The differential diagnosis in childhood and adolescence includes the more common ovarian germ cell tumors as well as epithelial ovarian carcinoma and small cell ovarian carcinoma. Differentiating between juvenile granulosa cell tumor and small cell ovarian carcinoma may be challenging as small cell carcinoma may mimic the pseudofollicular growth of juvenile granulosa cell tumors. Serum inhibin and calcium measurements as well as immunohistochemistry for inhibin may be helpful. Ovarian SCST tumors must also be distinguished from tumors of nongonadal origin that are metastatic to the ovary.
Histopathologic diagnosis of ovarian SCST may be challenging and consultation is recommended. Most (approximately 95%) of ovarian SCST stain focally positive for inhibin although some granulosa cell tumors and the recently described microcystic stromal tumors may be inhibin negative.2 This finding of inhibin positivity may help to differentiate SCSTs from small cell carcinomas of the hypercalcemic type and other epithelial neoplasms. As the mitotic activity may have a prognostic impact the mitotic count should be evaluated. This should be supplemented by the immunohistochemical evaluation of proliferation markers such as Ki67, which still require prospective evaluation.
SCSTs generally have coexpression of cytokeratin and vimentin. They are usually epithelial membrane antigen negative and calretinin positive. In addition, nearly all sex cord tumors are CD56 positive and the staining is usually diffuse and predominantly membranous.21 Charcot-Bottcher filaments are a distinguishing feature of Sertoli cells and may be present in Sertoli cell tumors or Sertoli-Leydig cell tumors. Histopathologic assessment of Sertoli-Leydig cell tumors should include description of the grade of differentiation, presence of a retiform subtype, and documentation of heterologous elements, all of which are considered prognostically unfavorable.
Retiform Sertoli-Leydig cell tumors may be confused with yolk sac tumors, serous carcinomas, or carcinosarcomas, suggesting that the diagnosis of a serous carcinoma in a child or young woman requires prior exclusion of a retiform Sertoli-Leydig cell tumor.22
As in other solid tumors, stage at diagnosis has been shown in several analyses to be the most important prognostic factor. Histologic subtype within SCST is also important.
Schneider et al23 showed patients with juvenile granulosa cell tumors (n=34) had higher 10-year event-free survival (EFS) than those with Sertoli-Leydig cell tumors (n=6) (0.88±0.05 vs. 0.67±0.19). Within the category of Sertoli-Leydig cell tumors, higher stage or poorly differentiated tumors are associated with a worse prognosis.24
Schneider et al25 also showed that patients with stage Ic tumors because of intraoperative rupture (n=12) had significantly better EFS compared with those patients with stage Ic tumors due to preoperative rupture or those with malignant ascites (n=9). Of 12 patients with intraoperative rupture, 9 received no adjuvant therapy and all survived without relapse. Of 9 patients with preoperative rupture or malignant ascites, 4 received chemotherapy and only 4 (3 who received adjuvant chemotherapy and 1 who did not), experienced EFS (1.0 vs. 0.44±0.17, log-rank 0.003).
Outcome also correlated with mitotic activity with those patients with <20 mitoses/10 HPF (n=35/35) having improved 10-year EFS compared with those with ≥20 mitoses/10 HPF (n=7/13) (10-y EFS 100 vs. 53%, log-rank test P=0.0001). This difference remained significant within the group with stage Ic disease. Those with <20 mitoses/10 HPF (n=14) had significantly better EFS than those with ≥20 mitoses/10 HPF (n=6) (100% vs. 48%, log-rank 0.006).23
Although poor differentiation is a clear poor prognostic factor in Sertoli-Leydig cell tumors of the ovary as discussed above, the same is not necessarily true for patients with ovarian granulosa cell tumors. For stage I patients with juvenile granulosa cell tumors, marked nuclear pleomorphism is common (approximately 15% of cases) and has not been shown to be associated with adverse prognosis.22
Ovarian sex cord-stromal tumor should be staged according to the International Federation of Gynecology and Obstetrics (FIGO) classification for ovarian carcinoma (Table 2). Careful attention should be given to description of stage Ic patients. Although preoperative or intraoperative spillage may both fit the stage Ic description, the details of the spillage and the cytology results may have therapeutic and prognostic implications given the data described above.
Review of Previous Treatment Strategies
Gersheson et al26 reported the results of an adult phase II study where 8 adult patients, age 24 to 65 years (median 43 y) received cisplatin, doxorubicin, and cyclophosphamide. Two patients had stage II disease, 1 had stage III, and 5 had recurrent disease. The overall (Complete response+Partial response) response rate was 63%. Four patients [previously stage Ic recurrent, stage IIa recurrent, and stage IIc (n=2)] remained disease free at 13 to 48 months from the start of chemotherapy. Gershenson et al27 also reported the results of the administration of cisplatin, etoposide, and bleomycin to 9 patients with metastatic SCST or poorly differentiated Sertoli-Leydig cell tumors (median age 37 y, range 12 to 69 y). Of 6 patients with measurable disease, the overall response rate was 83%. Median progression-free survival was 28 months with 2 patients alive and disease free at the time of the report. This group later compared the results in patients treated with taxanes to those treated with the previous regimen28 and found a higher but not statistically significant difference in response rates (71% for cisplatin, etoposide and bleomycin (PEB) vs. 37% for taxanes). Progression-free survival was not statistically different.
Colombo et al29 reported the results of prospective treatment of 11 patients with metastatic adult granulosa cell tumor of the ovary (median age 53 y, range 7 to 69 y). All underwent staging laparotomy and attempted total resection. Each received combination chemotherapy with vinblastine, cisplatin, and bleomycin. Six patients had a surgically verified complete response to therapy. Three died of progressive disease and there were 2 toxic deaths, 1 due to sepsis and 1 due to respiratory failure.
Zambetti et al30 reported the results of 7 patients of age 15 to 57 with advanced or recurrent granulosa cell tumor treated with cisplatin, vinblastine, and bleomycin. Three complete and 1 partial response occurred, resulting in an overall response rate of 66%. Pautier et al31 reported the use of PEB in 20 patients with metastatic or recurrent ovarian granulosa cell tumors. Overall response rate was 90%; 9 patients demonstrated a complete clinical response. Of the initial 20 patients, 12 are alive and 9 are without disease. Seven patients remained free of disease at a 4 to 84 months after therapy. Two patients were without disease in complete response 2 at the time of the report. Homesley et al32 reported a group of primarily adults with granulosa cell tumors treated with bleomycin, etoposide, and cisplatin. Overall, the therapy was well tolerated with 37% (14/38) having a negative second-look surgery.
Management Outline for Ovarian SCSTs
* Careful history and physical exam to determine presence of gastrointestinal or systemic symptoms, isosexual precocity or virilization (hirsutism, acne), or menstrual irregularities.
* Careful family history with attention to colon polyps, pleuropulmonary blastoma/lung cysts, sarcomas, thyroid disease, lung surgeries in early childhood and renal tumors, particularly cystic nephroma.
* Physical exam with attention to any stigmata of PJS including melanocytic macules, or signs of enchondromatosis.
* Laboratory evaluation: serum AFP, β-hCG, CA-125, LDH, and inhibin B and serum calcium. Careful attention to serum calcium is warranted given the possibility of small cell carcinoma of the hypercalcemic type.
* Surgery for diagnosis and therapy: Surgical approach requires attention to FIGO staging criteria and sampling of peritoneal fluid. The operative report should clearly state whether there was any violation of the tumor capsule and whether any rupture was felt to have occurred preoperatively versus intraoperatively.
* Fertility-sparing surgery for localized disease: For stage I tumors, unilateral salpingo-oophorectomy with examination of the contralateral ovary is likely sufficient. There is no need for hysterectomy in stage I tumors in childhood and adolescence or in adult women when fertility preservation is desired. It should be noted that in some centers, preservation of the uterus and contralateral tube and ovary is also attempted for children with stage II and III disease, unless there is significant infiltration of these structures (D. Schneider, personal communication, April 16, 2009. Oral communication).
* Removal of visible tumors should be performed when considered “reasonable and safe.”
* Stage Ia: Generally careful follow-up.
* Juvenile granulosa cell tumors: Adjuvant platinum-based chemotherapy for stage Ic tumors with preoperative rupture and/or malignant ascites or stages II, III, IV. Adjuvant chemotherapy should also be considered for juvenile granulosa cell tumors with stage Ic due to intraoperative rupture with ≥20 mitoses/10 HPF.
* Sertoli-Leydig cell tumors: Adjuvant platinum-based chemotherapy for stage Ic (due to preoperative or intraoperative rupture or with malignant ascites) or >Ic disease.
* Gynandroblastoma: Very few cases of gynandroblastoma have been described. Most have been localized at presentation and most have been associated with a benign course. A management strategy similar to that of juvenile granulosa cell and Sertoli-Leydig cell tumors seems prudent.
Adjuvant chemotherapy utilized in past studies of ovarian SCST has generally included platinum-based regimens: cisplatin/etoposide/ifosfamide (PEI), cisplatin/etoposide/bleomycin (PEB), etoposide/cisplatin, paclitaxel/carboplatin, vincristine/adriamycin/cytoxan, cisplatin/vinblastine/bleomycin. Regional, deep hyperthermia has been used for patients with recurrent or refractory ovarian tumors.33 Some investigators have also considered the role of ovarian suppression in these malignancies (D. Schneider, personal communication, April 16, 2009. Oral communication).
Given the rarity of these tumors, no prospective randomized studies have been performed comparing these regimens. A prospective German MAKEI protocol (principal investigator DT Schneider) offers the following treatment strategy with comparisons to historical controls:
* FIGO Ia: surgery, no adjuvant therapy.
* FIGO Ic: intraoperative rupture: surgery without adjuvant therapy for juvenile granulosa cell tumor. Surgery followed by adjuvant chemotherapy for Sertoli-Leydig cell tumor.
* FIGO Ic: ruptured preoperatively or positive cytology: surgery and cisplatin-based chemotherapy (PEI).
* FIGO II, III: surgery followed by cisplatin-based chemotherapy.
* FIGO IV or stage II or III with high mitotic index: cisplatin-based chemotherapy and regional deep hyperthermia.
Despite the lack of prospective evidence, certain general principles are suggested by retrospective reviews. Early, aggressive adjuvant therapy in tumors other than stage I appears important; nearly all who relapsed died. Recent evidence9 suggests that given the rarity of lymph node spread at diagnosis, routine staging lymphadenectomy be omitted from surgical management of ovarian SCST. Lymph nodes that are enlarged by imaging or are felt to be suspicious intraoperatively should be removed.
Overall, there is no strong evidence to recommend PEB vs. PEI adjuvant therapy. Many institutions in the United States are familiar with the PEB regimen utilized for germ cell tumors, thus this regimen is available at most institutions and a reasonable option. There has generally been little pulmonary toxicity associated with this regimen when the germ cell protocol schedule has been given. On the basis of individual patient circumstances such as preexisting pulmonary dysfunction or presence of sarcomatous heterologous elements, individual health care providers may choose alternate chemotherapeutic regimens including PEI.
Stage Ia: Complete resection. No further adjuvant therapy recommended. Monitor clinically and radiographically as well as with tumor markers. We suggest imaging and laboratory studies approximately every 3 months for the first 2 years with more frequent monitoring if clinical concerns arise. In patients suspicious of a genetic predisposition, follow-up should also consider the development of tumors at other sites.
Stage Ic (preoperative rupture or malignant ascites), II, and III: resection followed by 4 or 6 cycles of PEB therapy (Table 3). If there is response but concern for persistent disease by imaging or tumor marker elevation after 4 cycles, consider second-look surgery. If persistent disease is documented, consider second-line chemotherapy.
Management of patients with stage Ic disease due to intraoperative rupture is controversial and histology, histologic differentiation, and number of mitoses may be important. For those patients with juvenile granulosa cell tumor with intraoperative rupture and low numbers of mitoses, consideration should be given to observation with clinical follow-up, imaging, and serial tumor markers. Patients with Sertoli-Leydig cell tumors appear to be at higher risk of relapse and tumor-related death than patients with juvenile granulosa cell tumor, thus for patients with Sertoli-Leydig cell tumor and preoperative or intraoperative rupture or malignant ascites, adjuvant chemotherapy is recommended.34
For those patients with preoperative rupture and malignant ascites or metastases, along with the additional high-risk features of poorly differentiated histology and high number of mitoses, particularly those with Sertoli-Leydig cell tumors, consideration could be given to higher intensity chemotherapy regimens.
Stage IV tumors are quite rare and carry a poor prognosis. Given the rarity of these tumors, very little evidence is available to guide therapy. The German cooperative group has utilized locoregional hyperthermic chemotherapy for the treatment of recurrent or refractory tumors with locoregional extension,33 however, this approach is limited by availability and lack of prospective data. High-dose chemotherapy with stem cell rescue has also been utilized in some patients, however, there is little data thus far on outcomes using this approach. At least 2 patients with recurrent, metastatic Sertoli-Leydig cell tumor are alive without evidence of disease using PEI combination chemotherapy followed later by high-dose chemotherapy (carboplatin/etoposide) with autologous stem cell rescue (D. Schneider, personal communication, April 16, 2009. Oral communication).
There are anecdotal reports of responses to radiation therapy,35,36 however, there are no published studies to support its routine use in ovarian SCST. Also, there are no published recommendations as to radiation dose or regimen.
Computed tomography (CT), ultrasound, and magnetic resonance imaging scans have been used to follow patients with gonadal stromal cell tumors both during and after therapy. Under most clinical circumstances, abdominopelvic magnetic resonance imaging or ultrasound is preferred due to concerns for radiation exposure with CT imaging. We recommend imaging at approximately 3-month intervals over the first 3 years after diagnosis with increasing intervals thereafter. In children and adolescents with history of localized abdominopelvic disease, intrathoracic relapse is unlikely in absence of recurrent abdominopelvic disease. Most patients who relapse with Sertoli-Leydig cell tumors will do so within the first few years after diagnosis.10 Juvenile granulosa cell tumors most frequently recur in the same time period, but adult granulosa cell tumors may be associated with late relapse, even after >10 years.4–7
Tumor markers should also be followed at routine intervals, particularly if unavailable or elevated at diagnosis. We recommend following inhibin B, hormonal markers (if elevated at diagnosis), and CA-125 at regular intervals, approximately every 3 months for at least the first 3 years after diagnosis and at increasing intervals thereafter. If elevated at diagnosis, AFP should also be routinely monitored. Relapse, when it occurs, is associated with a poor prognosis; no clear evidence regarding specific chemotherapeutic regimens for relapse is available.
TESTICULAR STROMAL CELL TUMORS IN CHILDREN
Prepubertal testis tumors account for approximately 1% of pediatric solid tumors (Fig. 1). Of these, 11% are SCSTs.37
Review of data from the Prepubertal Testis Tumor Registry,37 a previously open registry study through the American Academy of Pediatrics Urology Section, suggests that most testicular tumors in children behave in a benign manner with the notable exceptions of yolk sac tumors and undifferentiated stromal tumors. Occasionally, older children may also have Sertoli cell tumors that behave in a malignant manner.
Ciftci and colleagues reported 51 patients with prepubertal testicular tumors, 3 of whom had a stromal neoplasm; all 3 were Leydig cell tumors. All presented with precocious puberty and elevated serum testosterone and androsteindione levels and underwent radical inguinal orchiectomy for stage I disease. No recurrences were noted.38 Harms and Kock39 described 29 patients with testicular SCSTs. The most frequent histologic type was the large cell calcifying Sertoli cell tumor, followed by juvenile granulosa cell tumors. In this group no relapses were observed. Several smaller case series of infantile juvenile granulosa cell tumors have recently been published.40–42
Clinical Presentation and Diagnosis
Nearly all testicular stromal tumors present with a painless mass. The mass is usually easily palpable and visible on ultrasound and may have a heterogenous appearance. Patients may have gynecomastia or signs of precocious puberty. Patients with testicular juvenile granulosa cell tumor often present in the first 6 months of life and usually lack hormone-related symptoms. Children with Leydig cell tumors may present with precocious puberty or gynecomastia and increased 17-ketosteroids.
When a testicular tumor is suspected, AFP measurement is needed and inhibin measurement should be considered. In patients near or after puberty, a β-hCG level should also be obtained. Testicular SCST frequently show immunohistochemical staining for inhibin. Testicular granulosa cell tumors may secrete inhibin, however, the use of inhibin as a tumor marker has not yet been established. Further study may elucidate this issue.
Immunohistochemical staining of testicular stromal tumors has been well described.43 Inhibin A staining has been found to distinguish testicular stromal cell tumors from germ cell tumor. Estrogen and progesterone receptors may also be present in up to 39% of malignant Leydig cell tumors.
Testicular juvenile granulosa cell tumors may be seen in association with structural abnormalities of the Y chromosome or in patients with ambiguous genitalia.44
Briefly, Carney syndrome may present with cutaneous lentiginosis, atrial myxoma, Cushing syndrome, acromegaly, and/or papillary and follicular thyroid cancers in addition to gonadal stromal tumors. PJS is characterized by cutaneous lentiginosis, intestinal polyps, and a predisposition to gastrointestinal and extraintestinal malignancies. Both may be associated with large cell calcifying Sertoli cell tumors. Carney syndrome has also been associated with unilateral or bilateral Leydig cell tumors. In both PJS and Carney syndrome, the lesions are often hormone-producing and may produce gynecomastia in children or adults. Testicular SCST-AT seen in association with PJS are associated with progesterone production and are more often bilateral than those tumors not associated with PJS.
PJS has also recently been reported to be associated with a unique testicular lesion, multifocal intratubular neoplasia of large Sertoli cells, which is distinct from large cell calcifying Sertoli cell tumor and SCST-AT. This lesion has been reported to follow an indolent course. Readers are referred to the original report for more information and pathology images.45
Multifocal testicular Leydig cell tumors may be associated with adrenogenital syndrome.46
On the basis of their individual management strategy, each testicular tumor will be considered separately.
Testicular juvenile granulosa cell tumors present in infancy, usually in children under the age of 6 months. By immunohistochemistry, they are characteristically positive for inhibin. These tumors are considered benign, with little to no evidence of recurrence or metastases. Testis-sparing surgery is generally reasonable if a reasonable amount of normal testicular tissue remains although some authors advocate orchiectomy. No adjuvant therapy is recommended for patients with stage I testicular juvenile granulosa cell tumors. In these tumors, serum AFP is expected to be within the normal range for age, although there is a wide physiological range in infancy that may make this determination challenging. We refer readers to the reference published by Blohm et al47 regarding normal AFP ranges. Surgical resection is curative for testicular juvenile granulosa cell tumor.
Sertoli cell tumors are generally benign in children less than age 5 and radical orchiectomy alone is considered appropriate treatment. In older children, consideration should be given to a metastatic evaluation including CT of the chest, abdomen, and pelvis. If found, metastatic disease requires aggressive surgical and adjuvant therapy.46,48 In addition, available evidence suggests that when there are a large number of mitotic figures or local invasion or poor differentiation, metastatic evaluation and follow-up are warranted. The role of chemotherapy in those patients who show histologic features concerning for malignant behavior is undefined. Small, benign appearing lesions may carry a lower likelihood of malignant behavior.49
Large cell calcifying Sertoli tumors tend to follow a more benign course than other Sertoli cell tumors in older children and adolescents although a malignant large cell calcifying tumor in a 28-year-old patient has recently been described. That tumor had histopathologic findings associated with aggressive malignancy and the patient died 4 years after diagnosis.50
Large cell calcifying Sertoli tumors may be associated with PJS and Carney syndrome. The large cell, calcifying Sertoli tumor seen in Carney syndrome is unlikely to metastasize and orchiectomy is likely to be curative.51 Testis-sparing surgery has also been described in this group.52
Leydig cell tumors are seen in prepubertal males and carry a favorable prognosis. Orchiectomy or testis-sparing surgery alone is likely to be curative. Hormonal symptoms may be associated with elevation of 17-ketosteroids and may require the consultation of a pediatric endocrinologist. Testicular Sertoli-Leydig cell tumors are quite rare.
Undifferentiated stromal tumors harbor malignant potential and prepubertal and postpubertal males with these tumors should undergo a metastatic evaluation.
The Prepubertal Testis Tumor Registry described 43 patients registered with stromal tumors.53 Ten of these had an unspecified stromal tumor and one of these patients developed metastases. The remaining 32 included 10 Sertoli cell tumors, 5 Leydig cell tumors, 9 juvenile granulosa cell tumor, and 8 mixed/undifferentiated stromal tumors. The single patient diagnosed with metastatic mixed/undifferentiated stromal tumor died. None of the patients with Sertoli, Leydig cell, or juvenile granulosa cell tumor had metastases at diagnosis, however, 4 cases of Sertoli cell tumors with malignant behavior in prepubertal males have been reported in the literature.
A current analysis of 40 testicular SCSTs of the MAKEI Registry confirms these data. In this cohort, no metastases were observed; all patients are remaining in continuous remission (unpublished data).
Testicular SCSTs in childhood are often benign and generally confined to the testis at presentation. Given this, there is no evidence-based, pediatric staging system. On the basis of available data from patients with germ cell tumors, it would seem reasonable to stage patients with malignant tumors using the staging system used by Children’s Oncology Group for germ cell tumors (Tables 4 and 5). Regardless of the formal staging system, staging of potentially malignant tumors should include assessment of local control (including surgical margins and involvement of the spermatic cord), assessment of retroperitoneal regional lymph nodes, and determination of the presence of distant metastases, most often in the lungs.
Principles of Surgical Management
Surgical approach is a critical aspect of patient management. The inguinal approach (not transcrotal approach) is favored, when the presurgical diagnosis includes testicular mass.
The Prepubertal Testis Tumor Registry investigators37 suggest testis-sparing surgery for most potentially benign tumors including those in patients with a normal preoperative AFP. Serum AFP interpretation is fraught with difficulty during infancy because of physiologically high levels, thus the authors suggest that testis-sparing surgery be excluded as a management option in those children aging 6 to 12 months with a preoperative AFP>100 ng/ml or those older than 1 year with an elevated AFP.
Shukla et al55 concurred with these recommendations, describing the first reported case of testis-sparing surgery in a patient with juvenile granulosa cell tumor of the testis. Other authors suggest a similar management strategy.56 Given the aforementioned experience, testis-sparing surgery is considered a reasonable approach, however, better long-term data are needed to further delineate its role in management.
In the intraoperative period, the authors suggest that the spermatic cord be temporarily occluded and excisional biopsy performed. If frozen section confirmed a benign histology, they suggest that the remainder of the testis be spared. If a malignant diagnosis is suspected based on frozen section, they suggest complete orchiectomy. This strategy is felt to be reasonable for patients who are clearly prepubertal. It should be noted, however, that during and after puberty, these tumors may follow a more malignant course. Thus, for patient near the age of puberty, an area of normal testis should be assessed on frozen histology and if pubertal, orchiectomy should be performed for patients with teratoma.
It should be noted that the above recommendations spring largely from the Prepubertal Testis Tumor Registry that included only patients younger than age 12 and surveyed health care providers only at the time of registration. Thus, follow up information on the patients on whom these recommendations are made, are limited.
For extremely rare testicular SCST with documented metastases, we recommend surgical resection followed by adjuvant chemotherapy. There is little evidence on which to base the chemotherapeutic regimen. On the basis of evidence in ovarian SCST and adult testicular germ cell tumors, a platinum-based chemotherapy regimen seems prudent. Thus, we recommend 4 cycles of PEB. After the fourth cycle, if imaging suggests persistent disease, consideration should be given to second-look surgery and up to 2 additional cycles of chemotherapy.
Patients who have postpubertal or histologically concerning Sertoli cell tumors or any child or adolescent male with an undifferentiated stromal tumor should undergo CT imaging. If tumor or hormonal markers are elevated at diagnosis, these should be serially followed. Metastatic Sertoli cell tumors or undifferentiated stromal cell tumors should be treated aggressively and carry a poor prognosis, however, limited data are available. Given the lack of available data, it would be reasonable to treat with a cisplatin-containing regimen similar to the regimen described for ovarian SCSTs.
In patients in whom there is concern for the development of metastatic disease, serial imaging is recommended. We recommend a follow up schedule of imaging every 3 months for 2 years. If tumor markers are elevated at diagnosis, these also should be followed. In general, for patients with testicular juvenile granulosa cell tumor, serial imaging and laboratory studies are not required. If AFP is believed to be elevated because of age alone, we recommend confirmation of an age-appropriate decline in AFP.
Both ovarian and testicular SCSTs are rare in children. They often, although not always follow a benign course. Histopathologic and clinical assessment may be challenging. Given the rarity of these tumors, expert consultation may be helpful. Data collection will be helpful in the management of future patients with these rare conditions and thus registry and clinical trial participation and international collaboration are encouraged.
The authors thank Children’s Oncology Group, the MAKEI group, and Pine Tree Apple Tennis Classic for their support of clinical research in the field of rare tumors. DTS is supported by the Barbara and Hubertus Trettner Fund. All authors warmly thank the children and families participating in the registries and clinical studies on whom this guideline is based.
1. Breen JL, Maxson WS. Ovarian tumors in children and adolescents. Clin Obstet Gynecol. 1977;20:607–623
2. Irving JA, Young RH. Microcystic stromal tumor of the ovary: report of 16 cases of a hitherto uncharacterized distinctive ovarian neoplasm. Am J Surg Pathol. 2009;33:367–375
3. Leyva-Carmona M, Vazquez-Lopez MA, Lendinez-Molinos F. Ovarian juvenile granulosa cell tumors in infants. J Pediatr Hematol Oncol. 2009;31:304–306
4. McCann EC, Zerner J. Granulosa cell tumor of the ovary in a 7-year-old female with late (12-year) lung metastasis. A case report and discussion. J Maine Med Assoc. 1973;64:201–203
5. Merras-Salmio L, Vettenranta K, Mottonen M, et al. Ovarian granulosa cell tumors in childhood. Pediatr Hematol Oncol. 2002;19:145–156
6. Schumer ST, Cannistra SA. Granulosa cell tumor of the ovary. J Clin Oncol. 2003;21:1180–1189
7. Hines JF, Khalifa MA, Moore JL, et al. Recurrent granulosa cell tumor of the ovary 37 years after initial diagnosis: a case report and review of the literature. Gynecol Oncol. 1996;60:484–488
8. Trabelsi AAS, Hadfi M, Fatnaci R. Primary mesenteric Sertoli-Leydig cell tumor: a case report and review of the literature. J Oncol. 2008
9. Brown J, Sood AK, Deavers MT, et al. Patterns of metastasis in sex cord-stromal tumors of the ovary: can routine staging lymphadenectomy be omitted? Gynecol Oncol. 2009;113:86–90
10. Metzinger DWM. Surgical management of Sertoli-Leydig cell tumors of the ovary. CME J Gynecol Oncol. 2002;7:140–142
11. Kato N, Romero M, Catasus L, et al. The STK11/LKB1 Peutz-Jegher gene is not involved in the pathogenesis of sporadic sex cord-stromal tumors, although loss of heterozygosity at 19p13.3 indicates other gene alteration in these tumors. Hum Pathol. 2004;35:1101–1104
12. Schultz KA, Pacheco MC, Yang J, et al. Ovarian sex cord-stromal tumors, pleuropulmonary blastoma and DICER1 mutations: a report from the International Pleuropulmonary Blastoma Registry. Gynecol Oncol. 2011;122:246–250
13. Hill DA, Ivanovich J, Priest JR, et al. DICER1 mutations in familial pleuropulmonary blastoma. Science. 2009;325:965
14. Heravi-Moussavi A, Anglesio MS, Cheng SW, et al. Recurrent somatic DICER1 mutations in nonepithelial ovarian cancers. N Engl J Med. 2012;366:234–242
15. Slade I, Bacchelli C, Davies H, et al. DICER1 syndrome: clarifying the diagnosis, clinical features and management implications of a pleiotropic tumour predisposition syndrome. J Med Genet. 2011;48:273–278
16. Rio Frio T, Bahubeshi A, Kanellopoulou C, et al. DICER1 mutations in familial multinodular goiter with and without ovarian Sertoli-Leydig cell tumors. JAMA J Am Med Assoc. 2011;305:68–77
17. Schneider DT, Calaminus G, Harms D, et al. Ovarian sex cord-stromal tumors in children and adolescents. J Reprod Med. 2005;50:439–446
18. Distelmaier F, Calaminus G, Harms D, et al. Ovarian small cell carcinoma of the hypercalcemic type in children and adolescents: a prognostically unfavorable but curable disease. Cancer. 2006;107:2298–2306
19. Gustafson ML, Lee MM, Scully RE, et al. Mullerian inhibiting substance as a marker for ovarian sex-cord tumor. N Engl J Med. 1992;326:466–471
20. Young RH, Perez-Atayde AR, Scully RE. Ovarian Sertoli-Leydig cell tumor with retiform and heterologous components. Report of a case with hepatocytic differentiation and elevated serum alpha-fetoprotein. Am J Surg Pathol. 1984;8:709–718
21. McCluggage WG, McKenna M, McBride HA. CD56 is a sensitive and diagnostically useful immunohistochemical marker of ovarian sex cord-stromal tumors. Int J Gynecol Pathol. 2007;26:322–327
22. Young RH. Sex cord-stromal tumors of the ovary and testis: their similarities and differences with consideration of selected problems. Mod Pathol. 2005;18(suppl 2):S81–S98
23. Schneider DT, Calaminus G, Wessalowski R, et al. Ovarian sex cord-stromal tumors in children and adolescents. J Clin Oncol. 2003;21:2357–2363
24. Young RH, Scully RE. Ovarian Sertoli-Leydig cell tumors. A clinicopathological analysis of 207 cases. Am J Surg Pathol. 1985;9:543–569
25. Schneider DT, Calaminus G, Göbel U. In Reply J Clin Oncol. 2004;22:2033–2035
26. Gershenson DM, Copeland LJ, Kavanagh JJ, et al. Treatment of metastatic stromal tumors of the ovary with cisplatin, doxorubicin, and cyclophosphamide. Obstet Gynecol. 1987;70:765–769
27. Gershenson DM, Morris M, Burke TW, et al. Treatment of poor-prognosis sex cord-stromal tumors of the ovary with the combination of bleomycin, etoposide, and cisplatin. Obstet Gynecol. 1996;87:527–531
28. Brown J, Shvartsman HS, Deavers MT, et al. The activity of taxanes compared with bleomycin, etoposide, and cisplatin in the treatment of sex cord-stromal ovarian tumors. Gynecol Oncol. 2005;97:489–496
29. Colombo N, Sessa C, Landoni F, et al. Cisplatin, vinblastine, and bleomycin combination chemotherapy in metastatic granulosa cell tumor of the ovary. Obstet Gynecol. 1986;67:265–268
30. Zambetti M, Escobedo A, Pilotti S, et al. Cis-platinum/vinblastine/bleomycin combination chemotherapy in advanced or recurrent granulosa cell tumors of the ovary. Gynecol Oncol. 1990;36:317–320
31. Pautier P, Gutierrez-Bonnaire M, Rey A, et al. Combination of bleomycin, etoposide, and cisplatin for the treatment of advanced ovarian granulosa cell tumors. Int J Gynecol Cancer. 2008;18:446–452
32. Homesley HD, Bundy BN, Hurteau JA, et al. Bleomycin, etoposide, and cisplatin combination therapy of ovarian granulosa cell tumors and other stromal malignancies: A Gynecologic Oncology Group study. Gynecol Oncol. 1999;72:131–137
33. Wessalowski R, Schneider DT, Mils O, et al. An approach for cure: PEI-chemotherapy and regional deep hyperthermia in children and adolescents with unresectable malignant tumors. Klin Padiatr. 2003;215:303–309
34. Schneider DT. Abstracts from the SIOP. Pediatr Blood Cancer. 2011;56:1155–1164
35. Wolf JK, Mullen J, Eifel PJ, et al. Radiation treatment of advanced or recurrent granulosa cell tumor of the ovary. Gynecol Oncol. 1999;73:35–41
36. Schneider DT, Calaminus G, Wessalowski R, et al. Therapy of advanced ovarian juvenile granulosa cell tumors. Klin Padiatr. 2002;214:173–178
37. Ross JH, Rybicki L, Kay R. Clinical behavior and a contemporary management algorithm for prepubertal testis tumors: a summary of the Prepubertal Testis Tumor Registry. J Urol. 2002;168(pt 2):1675–1678 discussion 1678–1679
38. Ciftci AO, Bingol-Kologlu M, Senocak ME, et al. Testicular tumors in children. J Pediatr Surg. 2001;36:1796–1801
39. Harms D, Kock LR. Testicular juvenile granulosa cell and Sertoli cell tumours: a clinicopathological study of 29 cases from the Kiel Paediatric Tumour Registry. Virchows Arch. 1997;430:301–309
40. Zugor V, Labanaris AP, Witt J, et al. Congenital juvenile granulosa cell tumor of the testis in newborns. Anticancer Res. 2010;30:1731–1734
41. Fagin R, Berbescu E, Landis S, et al. Juvenile granulosa cell tumor of the testis. Urology. 2003;62:351
42. Gun F, Erginel B, Klcaslan I, et al. A rare neonatal testicular tumor: juvenile granulosa cell tumor of infant testis: a report of 3 cases. J Pediatr Hematol Oncol. 2010;32:e158–e159
43. Iczkowski KA, Bostwick DG, Roche PC, et al. Inhibin A is a sensitive and specific marker for testicular sex cord-stromal tumors. Mod Pathol. 1998;11:774–779
44. Cortez JC, Kaplan GW. Gonadal stromal tumors, gonadoblastomas, epidermoid cysts, and secondary tumors of the testis in children. Urol Clin North Am. 1993;20:15–26
45. Ulbright TM, Amin MB, Young RH. Intratubular large cell hyalinizing sertoli cell neoplasia of the testis: a report of 8 cases of a distinctive lesion of the Peutz-Jeghers syndrome. Am J Surg Pathol. 2007;31:827–835
46. Ahmed HU, Arya M, Muneer A, et al. Testicular and paratesticular tumours in the prepubertal population. Lancet Oncol. 2010;11:476–483
47. Blohm ME, Vesterling-Horner D, Calaminus G, et al. Alpha 1-fetoprotein (AFP) reference values in infants up to 2 years of age. Pediatr Hematol Oncol. 1998;15:135–142
48. Agarwal PK, Palmer JS. Testicular and paratesticular neoplasms in prepubertal males. J Urol. 2006;176:875–881
49. Kaplan GW, Cromie WJ, Kelalis PP, et al. Gonadal stromal tumors: a report of the Prepubertal Testicular Tumor Registry. J Urol. 1986;136(pt 2):300–302
50. Petersson F, Bulimbasic S, Sima R, et al. Large cell calcifying Sertoli cell tumor: a clinicopathologic study of 1 malignant and 3 benign tumors using histomorphology, immunohistochemistry, ultrastructure, comparative genomic hybridization, and polymerase chain reaction analysis of the PRKAR1A gene. Hum Pathol. 2010;41:552–559
51. Ross JH, Kay R. Prepubertal testis tumors. Rev Urol. 2004;6:11–18
52. Nonomura K, Koyama T, Kakizaki H, et al. Testicular-sparing surgery for the prepubertal testicular tumor. Experience of two cases with large cell calcifying Sertoli cell tumors. Eur Urol. 2001;40:699–704
53. Thomas JC, Ross JH, Kay R. Stromal testis tumors in children: a report from the prepubertal testis tumor registry. J Urol. 2001;166:2338–2340
54. Benedet JL, Bender H, Ngan HYS, et al. FIGO Committee on Gynecologic Oncology. FIGO staging classifications and clinical practice guidelines in the management of gynecologic cancers. Int J Gynecol Obstet. 2000;70:209–262
55. Shukla AR, Huff DS, Canning DA, et al. Juvenile granulosa cell tumor of the testis: contemporary clinical management and pathological diagnosis. J Urol. 2004;171:1900–1902
56. Trobs RB, Krauss M, Geyer C, et al. Surgery in infants and children with testicular and paratesticular tumours: a single centre experience over a 25-year-period. Klin Padiatr. 2007;219:146–151
© 2012 Lippincott Williams & Wilkins, Inc.