Development of a 3D functional assay and identification of biomarkers, predictive for response of high-grade serous ovarian cancer (HGSOC) patients to poly-ADP ribose polymerase inhibitors (PARPis): targeted therapy

Abstract

Abstract Background Poly(ADP-ribose) polymerase inhibitors (PARPis) specifically target homologous recombination deficiency (HRD) cells and display good therapeutic effect in women with advanced-stage BRCA1/2-mutated breast and epithelial ovarian cancer (EOC). However, about 50% of high grade serous ovarian cancers (HGSOC) present with HRD due to epigenetic BRCA1 inactivation, as well as genetic/epigenetic inactivation(s) of other HR genes, a feature known as “BRCAness”. Therefore, there is a potential for extending the use of PARPis to these patients if HR status can be identified. Methods We have developed a 3D (spheroid) functional assay to assess the sensitivity of two PARPis (niraparib and olaparib) in ascites-derived primary cell cultures (AsPCs) from HGSOC patients. A method for AsPCs preparation was established based on a matrix (agarose), allowing for easy isolation and successive propagation of monolayer and 3D AsPCs. Based on this method, we performed cytotoxicity assays on 42 AsPCs grown both as monolayers and spheroids. Results The response to PARPis treatment in monolayer AsPCs, was significantly higher, compared to 3D AsPCs, as 88% and 52% of the monolayer AsPCs displayed sensitivity to niraparib and olaparib respectively, while 66% of the 3D AsPCs were sensitive to niraparib and 38% to olaparib, the latter being more consistent with previous estimates of HRD (40%–60%) in EOC. Moreover, niraparib displayed a significantly stronger cytotoxic effect in both in 3D and monolayer AsPCs, which was confirmed by consecutive analyses of the HR pathway activity (γH2AX foci formation) in PARPis-sensitive and resistant AsPCs. Global gene expression comparison of 6 PARPi-resistant and 6 PARPi-sensitive 3D AsPCs was indicative for the predominant downregulation of numerous genes and networks with previously demonstrated roles in EOC chemoresistance, suggesting that the PARPis-sensitive AsPCs could display enhanced sensitivity to other chemotherapeutic drugs, commonly applied in cancer management. Microarray data validation identified 24 potential gene biomarkers associated with PARPis sensitivity. The differential expression of 7 selected biomarkers was consecutively confirmed by immunohistochemistry in matched EOC tumor samples. Conclusion The application of this assay and the potential biomarkers with possible predictive significance to PARPis therapy of EOC patients now need testing in the setting of a clinical trial.