Ovarian granulosa cell tumours (GCT) are rare gynaecological malignancies with recurrence possible decades after initial treatment. Although mostly treated surgically, often chemotherapy and aromatase inhibitors (AIs) are used with little supporting data. Despite evidence in breast and epithelial ovarian cancer that hormone receptor expression predicts response to AIs, there is no such evidence in GCTs. Oestrogen/progesterone receptor (ER/PR) status poorly reported in case series. Study aim: evaluate clinical efficacy of hormonal manipulation in GCT, correlation of efficacy with immunohistochemical (IHC) marker presence.
Clinical details, demographics, survival data collected- 15 patients with recurrent GCT treated 2004-2014 in large UK centre. ER/PR IHC performed on tumour samples available. Primary outcome: progression free survival on hormonal manipulation (PFS). Secondary outcomes: ER/PR positivity, correlation with response to treatment.
Median age at diagnosis 54 years (range 29-78years). Median interval from diagnosis to detection of recurrence/advanced disease 5 years (range 0.3-21years). Ten patients (67%) received hormonal manipulation, often previously heavily pretreated. Chemotherapy trialed in 4 patients, multiple lines given prior to commencing hormonal manipulation. AI/GnRH prescribed on 14 occasions, sometimes multiple lines in the same patient. Median PFS 14 months (95%CI 11.04-16.95), 5 patients continue to respond at time of analysis. Tumour analysed in 9 instances of hormonal manipulation. ER+ related to longer PFS in those treated with AIs, median 20.5 months (range 9-32); 14 months (range 3-24) in ER- cases.
ER/PR status and progression free survival in episodes of management with hormonal manipulation
|Hormonal agent||Line of treatment post recurrence||ER/PR (Q score)||Time to radiological progression from commencement of hormonal agent (months)|
^= same patient
*= censored data, continues to respond
PFS comparable to previous reviews with particularly good response from letrozole. ER+ potentially correlates with prolonged response with AIs,but also seen in ER-. Prospective studies would further clarify usefulness of AIs as standard of care and utility of ER IHC to predict response.
Clinical trial identification
Legal entity responsible for the study
Charitable funding donated to Dr Sarah Williams
All authors have declared no conflicts of interest.