Expression Profiling of Ovarian Cancer: markers and targets for therapy

The full text of this item cannot yet be made available, due to a publisher's embargo

Ovarian cancer is the leading cause of death from gynecological cancer in the Western world. The initial response of the primary tumor to taxane and platinum-based chemotherapy is high, however 20% of patients never achieve a clinical response and the majority of the patients will relapse and eventually die of drug-resistant disease. Chapter 1 includes a general overview of ovarian cancer, its epidemiology, histology, typing and the different therapies. The major drawback in the treatment of ovarian cancer is late detection and therapy failure due to intrinsic and acquired chemotherapy resistance and several mechanisms involved in the platinum-based chemotherapy resistance are described. Furthermore, the importance of expression profiling (mRNA or protein) in the search for tumor markers suitable for early detection of ovarian cancer, response prediction, progression monitoring and identification of targets for therapy is discussed. Chapter 2A The expression profiling of 24 ovarian carcinomas led to the discovery of a discriminating 69-gene signature from which a predictive nine-gene set was extracted. The nine-gene set predicted the resistance in an independent validation set (n=72) with a sensitivity of 89% (95% CI: 0.68-1.09) and a specificity of 59% (95% CI: 0.47-0.71)(OR=0.09, p=0.026). The predictive nine-gene set consists of the following genes, FN1, TOP2A, LBR, ASS, COL3A1, STK6, SGPP1, ITGAE and PCNA. Interestingly, three of these nine genes are already direct or indirect targets for therapy, i.e. topoisomerase 2A (TOP2A), serine/threonine kinase 6 (STK6) and argininosuccinate synthetase (ASS). The predictive power of the nine-gene set needs to be further validated in larger independent multicenter study before this model can be implemented in the clinical practice. Chapter 2B In their â?~letter to the editorâ?T, Gevaert et al. suggest that in clinical practice, a higher specificity would have been more successful assuming that patients predicted not to respond are given a different treatment not containing platinum drugs. We agree that the predictive gene signature needs further validation before implementation in the clinical practice can be advised. However, it is was not our intention to withhold platinum treatment from patients predicted not to respond, but to tailor the treatment based on the expression profile. An overexpression of TOP2A indicates that adding a TOP2A inhibitor, like etoposide, to the conventional platinum treatment, might proof to be beneficial for the patient. Chapter 2C Underexpression of one of the nine genes from the predictive gene set, i.e. Argininosuccinate synthetase (ASS) was associated with platinum-based chemotherapy resistance. To determine if this observed association was functional, ASS was downregulated with siRNA in three ovarian cancer cell lines that were relatively sensitive to cisplatin. For all three cell lines, this did not result in a reduced response to cisplatin measured with an MTT assay. However, due to differences between cell lines and carcinomas, we cannot exclude that ASS might still play a role in platinum-based chemotherapy resistance in ovarian cancer patients. Chapter 3 One of the nine genes of the predictive gene set i.e. proliferating cell nuclear antigen (PCNA), is involved in the DNA mismatch repair (MMR). In vitro, a relationship between MMR deficiency and platinum-drug resistance was suggested. However, no microsatellite instability (which is a marker for MMR deficiency) was detected in 75 primary ovarian carcinomas of which 46 received platinum-based chemotherapy. Thus resistance seen in 11 patients was not associated with an deficient MMR. An overview of published data revealed an overall frequency of MMR inactivation of 13% (165/1315, range: 0-39%). Interestingly, this was higher in mucinous and endometrioid when compared to clear cell and serous ovarian carcinomas. Chapter 4 Gene expression profiling studies leading to prediction of treatment resistance in breast and ovarian cancer, show only a few overlapping genes. This could be explained by heterogeneity of the disease, patient populations studied, different platforms and approach of statistical analysis. Moreover, overlapping pathways may be present but overlooked since each gene signature could contain different genes from the same pathway. One gene, fibronectin (FN1), is overlapping in our two profiling studies in breast and ovarian cancer. FN1 is involved in cell adhesion mediated drug resistance (CAM-DR) and down regulation was associated with paclitaxel resistance in breast and ovarian epithelial cancer cell lines. Therefore, FN1 and its relation with the response to different therapeutics might be interesting for future investigations. Chapter 5 Little overlap was seen between seven published gene signatures associated with platinum-based chemotherapy response in ovarian cancer. Therefore, pathway analysis was done to determine if there are common pathways. Gene Ontology analysis revealed six functional processes that were represented in all seven gene signatures. Each process was further analyzed in Ingenuity Pathway Analisys and remarkably TNF, P53 and TGFB were identified as common key genes. In the future, besides using the identified signatures for prediction, pathway analysis could lead to a better insight in the functional relationship between specific pathways and ovarian cancer and chemotherapy resistance. This could lead to a more educated selection of targets for therapy. Chapter 6 SELDI-TOF MS was used to discover ovarian cancer biomarkers by comparing serum protein profiles of ovarian cancer patients at diagnosis (n=35) or at progression (n=43), with that of healthy individuals (n=31). This resulted in eight primary and eleven progression ovarian cancer biomarkers, with four overlapping biomarkers. In addition, we compared the sera profiles from ovarian cancer patients after platinum-based chemotherapy (n=12) with that of ovarian cancer patients at progression (n=24) and discovered ten potential progression monitoring markers of which one was identified as serum amyloid A1 and validated with ELISA. Further validation would reveal the use of these biomarkers in screening and monitoring progression of ovarian cancer. Chapter 7 The generation and characterization of an A2780 mutant ovarian cancer cell line that is resistant to and shows a reduced accumulation of cisplatin, carboplatin, oxaliplatin and tetraplatin but not transplatin. This impaired accumulation could not be attributed to the known cisplatin transporters MRP2, CTR1, ATP7A or ATP7B and the platinum efflux rate was similar to that of A2780. The resistance could be overcome by circumvention of the natural influx mechanisms with cisplatin nanocapsules (nanoprecipitates of cisplatin coated with a lipid bilayer), indicating a defect in influx mechanism. The striking specificity of this influx mechanism to platinum compounds in the cis-configuration suggests that not passive diffusion but a platinum specific transporter is involved in the impaired platinum influx and resistance in A2780-Pt.