Aromatase inhibitors in the treatment of recurrent ovarian granulosa cell tumors: Brief report and review of the literature


Dr Mariam M. AlHilli, Department of Obstetrics and Gynecology, 200 First Street, SW, Rochester, MN 55905, USA. Email:


Granulosa cell tumors are ovarian sex cord stromal cancers that are unique for their indolent nature and late recurrence. Standard treatment regimens utilized include surgery, chemotherapy, and radiotherapy. However, hormonal suppression with aromatase inhibitors, which have shown promising results, may be a viable alternative to these modalities. We present a heavily pre-treated, multi-operated patient who experienced significant tumor shrinkage following treatment with an aromatase inhibitor for her recurrent granulosa cell tumor.


Granulosa cell tumors (GCT) represent 2–5% of all ovarian malignancies.1,2 These hormonally active tumors are derived from the granulosa cells of the ovarian stroma.3 They frequently secrete estrogens among other hormones.1,2 Histologically classified as sex cord stromal tumors, they account for 70–85% of tumors arising from that portion of the ovary.2,3

GCT are characterized by an indolent clinical course, a low malignant potential, and a tendency to recur up to 20 years following the initial diagnosis.1 The majority of cases are diagnosed with stage I disease (78–91%), and have an excellent prognosis, with a 10-year survival rate of 85–95%.3,4 Relapses up to 37 years after initial diagnosis have been reported.3 Hence, the necessity for continuous life-long disease surveillance.5 At least in part, the indolent nature of these tumors accounts for the lack of standard treatment guidelines for recurrent disease.6

As with all ovarian malignancies, surgery with total abdominal hysterectomy and bilateral salpingo-oophorectomy, bilateral pelvic and para-aortic lymphadenectomy, omentectomy, and cytoreduction is the mainstay of primary treatment for all stages of the disease.2 Conservative, fertility-sparing surgery in the treatment of GCT is an option for women of reproductive age who have not completed child bearing.7 Platinum-based combination chemotherapy with bleomycin, etoposide, and cisplatin (BEP) is the standard adjuvant chemotherapy regimen for advanced disease and recurrence not previously treated with chemotherapy.2,4,7,8 Adjuvant treatment with radiotherapy has been considered for both early stage disease and select advanced or residual disease with varying results.9

It has been suggested that the suppression of endogenous estrogens may provide antiproliferative benefits that may enhance the treatment effects of GCT.10 Aromatase inhibitors (AI), such as anastrozole and letrozole, have been used in the management of six recurrent GCT cases reported in the literature with promising results.4,10 Here, we report a heavily pretreated patient with recurrent GCT responding to letrozole. A search of PubMed for the terms ‘granulosa cell tumor’, ‘treatment’, ‘ovary’, and ‘aromatase inhibitor’ was performed to identify all published reports of ovarian GCT treated with AI.

Case Report

A 76-year-old woman with a history of seven previous recurrences of adult-type GCT presented to her primary care physician with chronic low back pain. Computed tomography (CT) of the abdomen and pelvis revealed evidence of intraperitoneal disease involving the tip of the left hepatic lobe and right pelvic sidewall consistent with recurrent GCT. She was referred to gynecologic oncology for surgical consultation.

She was initially diagnosed with a stage IA GCT in 1979 and underwent a total abdominal hysterectomy with bilateral salpingo-oophorectomy followed by 6 months of adjuvant chemotherapy with melphalan (L-PAM). Her first recurrence, a 5-cm right lower quadrant mass, occurred in 1991 and was managed with secondary surgical cytoreduction. She was again recurrence-free for more than a decade but developed asymptomatic CT-detected solitary masses in February 2002 and again in June 2004, both surgically resected. In January 2005, she again underwent an exploratory laparotomy with resection of multiple metastatic lesions involving the small bowel, sigmoid colon, and left pelvic sidewall. Resection margins were positive with microscopic evidence of residual disease along the pelvic sidewall.

In December 2006, she was found to have radiologic evidence of disease along the left pelvic sidewall and between the bladder and sigmoid colon. Following cytoreduction of disease in the central pelvis and obturator fossa, intraoperative radiation therapy (IORT) was administered to the left pelvic sidewall (1250 cGy) followed by pelvic floor reconstruction with an omental pedicle. Postoperatively, she underwent external beam radiotherapy (EBRT) to the left iliac lymph-node-bearing area (4500 cGy). In 2008, she developed a sixth recurrence around the left common iliac vessels. She underwent EBRT with shrinkage of the tumor, followed by resection of the residual tumor as well as a left para-aortic lymphadenectomy and IORT to the left common iliac area.

The patient presented for further evaluation of chronic low back pain in April 2010. She was known to have indeterminate sclerotic spine lesions, multi-level disc disease, and lumbar plexopathy secondary to radiation. CT of the abdomen and pelvis showed a 3.0 × 2.6 × 2.0-cm mixed density mass along the lateral left hepatic lobe and a 1.7 × 2.0 × 2.7-cm mass in the right pelvis compressing the bladder. Both lesions were suspicious for peritoneal metastases and had doubled in size compared to previous CT imaging 6 months prior. The patient was asymptomatic from her peritoneal disease with normal bowel and bladder function, good appetite, and no weight fluctuation. Surgical resection of both peritoneal lesions with IORT to the right pelvis was initially recommended. An alternative approach using hormonal suppressive therapy with the aromatase inhibitor letrozole was discussed and the patient chose a trial of letrozole (2.5 mg daily) for a period of 3 months.

In follow up, the patient remained asymptomatic from her persistent disease and a repeat CT scan of the abdomen and pelvis in August 2010 and October 2010 revealed progressive decrease in the size of the tumor along the left lobe of the liver from 3.1 cm to 2 cm in greatest dimension. Likewise, the lesion on the right pelvic sidewall had decreased in size from 3.0 cm to 1.5 cm (Figs 1,2). Given the clinical response, the patient continued letrozole and an eighth tumor debulking and third IORT were avoided.

Figure 1.

Computed tomography of abdomen and pelvis in April 2010 showing right pelvic sidewall metastasis.

Figure 2.

Computed tomography of abdomen and pelvis in October 2010 showing a decrease in size of right pelvic sidewall metastasis.


GCT typically present as large, occasionally asymptomatic, masses. Approximately two-thirds of patients present with manifestations secondary to endocrine excesses, such as hirsutism and virilization, or with abnormal vaginal bleeding.3 Approximately 25–50% of GCT are associated with endometrial hyperplasia and 5–13% with endometrial carcinoma.2 In addition to estrogen, GCT often secrete inhibin, a glycoprotein produced by granulosa cells. Patients with GCT frequently have elevated serum inhibin A and B levels that fall precipitously following tumor removal.2

Aromatase inhibitors (AI) are a family of oral non-steroidal (letrozole and anastrozole) and steroidal (exemestane) medications that bind to aromatase, an enzyme involved in the conversion of androstenedione to estriol E1 and testosterone to estradiol (E2).11 In postmenopausal women, AI exhibit a strong anti-estrogen effect and may result in a reduction of up to 90% of the aromatization of androgens with low adverse effect profiles.11 AI have been widely utilized in the treatment of advanced breast cancer and are the current first-line adjuvant hormonal therapy for estrogen-receptor-positive postmenopausal breast cancer.12

In gynecologic oncology, AI have been used in the treatment of endometrial cancer with modest efficacy.11 Epithelial ovarian cancers have minimal aromatase activity, which limits their applicability in treatment.11 Ovarian GCT, on the other hand, are theoretically excellent targets for AI due to their high degree of estrogen activity.10

Only six patients with recurrent GCT treated with AI have been reported in the literature.4,10Table 1 summarizes each of the six cases described in the literature. All six had durable clinical responses exceeding 12 months, including a case with a reported disease-free interval of 54 months.10 To our knowledge, Freeman et al. described the first two cases in the literature.4 Both patients received treatment with anastrozole following multiple treatment modalities for recurrent GCT, including surgery, chemotherapy (carboplatin/paclitaxel and paclitaxel alone), radiotherapy, and gonadotrophin-releasing hormone (GnRH) agonists. These patients experienced normalization of their inhibin levels, which remained normal for 14 and 18 months. Likewise, Korach et al. reported treating four patients with recurrent GCT with AI (two treated with anastrozole and two with letrozole) and showed clinical responses with both AI.10 In fact, three of the four patients experienced complete responses. We describe the seventh case in the literature of recurrent GCT successfully treated with an AI.

Table 1.  Summary of reported granulosa cell tumor cases treated with an aromatase inhibitor
AuthorCaseAge (years)PresentationStageInitial treatmentOutcomeRemission following aromatase inhibitor (months)
  1. BEP, bleomycin, etoposide, cisplatin; NA, not available/described.

Freeman et al., 20064148Chronic pelvic painICSurgery alone– Recurrence 4 years following surgery (pelvic mass) treated with surgery and chemotherapy (Paclitaxel/carboplatin)18
– Leuprolide started for persistent disease
– Radiation therapy × 6 weeks; mass persisted
– Anastrozole started
Freeman et al., 20064252Pleural effusion and adnexal massIIBSurgery alone– Recurrence 2 years following surgery (liver and peritoneum)14
– Leuprolide × 18 months without improvement)
– Chemotherapy (Paclitaxel) started × 10 cycles with partial response
– Anastrozole started
Korach et al., 200910342Ovarian massIIBSurgery and chemotherapy– Etoposide/cisplatin × 4 cycles54
– Pelvic recurrence at 32 months treated with surgery and radiotherapy
– Abdomen/pelvic recurrence 23 months later treated with surgery and letrozole
Korach et al., 200910479Endometrial cancerICSurgery alone– Recurrence at 11 months (left diaphragm and pelvis)12
– Etoposide and cisplatin initiated and stopped due to side-effects
– Anastrozole started
Korach et al., 200910549NAICSurgery alone– Recurrence at 39 months (abdomen/ pelvis) following surgery treated surgically12
– BEP × 4 cycles with resultant grade 3 neuropathy
– Recurrence after 8 months in bladder treated surgically only
– Recurrence after 4 months treated surgically and letrozole started
Korach et al., 200910657NAICSurgery alone– Recurrence 33 months after surgery treated surgically and with BEP × 4 cycles15
– Recurrence after 8 months (mediastinal/pleural nodules)
– Anastrozole started

Due to the low incidence of GCT and their indolent nature, prospective trials investigating the optimal systemic therapy for recurrent disease are difficult to conduct. In 1999, the Gynecologic Oncology Group published their phase II study of BEP as first-line treatment for incompletely resected stage II–IV and recurrent sex cord-stromal tumors, most of which (48/57; 84%) were GCT.8 While 69% of patients with advanced disease and 51% of patients with recurrent disease had clinical responses, grade 4 myelotoxicity occurred in 61% and grade 4 gastrointestinal toxicity occurred in 14%.8 Although this study took 6 complete years to accrue 57 eligible patients, it has shaped our evidence-based approach to GCT requiring treatment beyond surgery alone.

Current first-line chemotherapy for advanced-stage or recurrent GCT is BEP.10 However, reports on the activity of taxanes, GnRH agonists,1 tamoxifen, progestins and, more recently, AI have been published.1,2 Certainly the responses to AI in the limited reports published to date are indeed promising and the side-effects of AI are considerably more favorable compared to BEP. Further exploration of the efficacy of AI in GCT is warranted in the form of a clinical trial.