Abstract
PURPOSE: Anti-estrogen therapy may be used as a palliative treatment option in high-grade serous ovarian carcinomas (HGSC). However, clinical implementation is limited as the use of estrogen receptor (ER) protein expression by immunohistochemistry remains insufficient in predicting therapy response. To determine the accuracy of ER protein expression as a marker for ER signaling pathway activity, we aimed to correlate ER protein expression to functional ER signaling pathway activity in HGSC. METHODS: Immunohistochemical ER protein expression was visually scored using total percentages of stained tumor cells and histoscores. Subsequently, mRNA was extracted, and RT-qPCR analysis was performed. Functional ER pathway activity was assessed by a computational Bayesian model inferring ER signaling pathway activity from mRNA levels of ER-specific target genes. RESULTS: Our analysis of 29 HGSCs shows that neither total percentage of ER protein expression, nor ER histoscores are significantly correlated to ER signaling pathway activity (respectively, p = 0.473 and p = 0.606). Classification of HGSC into three groups based on ER histoscores 0-100 (n = 6), 101-200 (n = 15) and 201-300 (n = 8) resulted in comparable mean ER signaling pathway activity among the groups (p = 0.356). Several samples in the higher ER histoscore groups had low ER signaling pathway activity, indicating that nuclear ER protein expression is not sufficient to describe transcriptional ER activation. CONCLUSION: Positive immunohistochemical ER staining is not always indicative of an active ER signaling pathway and is, therefore, a poor predictor of anti-estrogen response. Further research is needed to prove the predictive value of ER signaling pathway activity regarding anti-estrogen sensitivity in HGSC patients.