Targeted approaches to the treatment of ovarian cancer.

Abstract

5694

Background: The majority of patients with advanced stage epithelial ovarian cancer will develop platinum-resistant recurrent disease requiring salvage therapy. Response rates in this setting are generally less than 20%, with the potential for significant cumulative toxicities. Currently, our ability to tailor therapy to patients most likely to respond to individual salvage agents is limited by an incomplete understanding of the biology that underlies chemo-responsiveness. Objectives: To characterize the molecular genetic basis of advanced epithelial ovarian cancer response to topotecan salvage chemotherapy and identify opportunities for novel therapeutic additions to the standard treatment strategy based on genomic data. Methods: Ovarian cancers were obtained at the time of primary surgery and assayed for gene expression in 48 patients who went on to develop persistent or recurrent disease when treated with topotecan salvage therapy. A gene expression based model was developed to predict response to topotecan salvage therapy. In parallel, we further evaluated the role of a series of gene expression signatures of activated oncogenic signaling pathways as indicators of therapeutic opportunities, combined with a prediction of topotecan response, in 48 ovarian cancers, 14 ovarian cancer cell lines and 54 other human cancer cell lines. Results: We have demonstrated the ability to identify ovarian cancer patients unlikely to have a clinical response to topotecan salvage chemotherapy with an overall accuracy of 81% for correctly predicting response to topotecan (80% responders and 83% non-responders) in a leave-one-out cross validation. We also show that the prediction of pathway deregulation in cancer cell lines predicts sensitivity to therapeutic agents that target components of two pathways, src and beta-catenin. We have further investigated the extent to which the effect of combined therapies is additive in relation to pathway status. Specifically, a collection of ovarian cancer cell lines were assayed for sensitivity to topotecan either with or without a src inhibitor (SU6656). The addition of a src inhibitor (SU6656) increased ovarian cancer cell line sensitivity to topotecan. Conclusions: Our data suggest that we can use gene expression profiles from primary ovarian cancers obtained at the time of initial surgical cytoreduction to not only predict response to topotecan, but also to identify alternate therapeutic options for patients predicted to be both topotecan-resistant as well as topotecan-sensitive. Deregulation of the src and beta-catenin oncogenic signaling pathways appears to be associated with relative topotecan resistance in human ovarian cancers and cancer cell lines. This relative resistance to topotecan can be partially reversed in-vitro using a src inhibitor. Our analysis suggests that the src and beta-catenin oncogenic signaling pathways may be important in topotecan responsiveness in many other human cancers.