Date of Award

January 2011

Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)



First Advisor

Joanne B. Weidhaas

Subject Area(s)



Purpose: Ovarian cancer has a poor prognosis, yet pathologic and clinical data do not accurately predict which patients will ultimately succumb to their disease. We previously reported an association between rs61764370, a germline functional variant in the 3'UTR of the KRAS oncogene, and epithelial ovarian cancer (EOC) risk. Here we evaluate this variant as a biomarker of clinical outcome and chemotherapy resistance in EOC.

Patients and Methods: Four groups of EOC patients with complete clinical data were genotyped for the KRAS-variant and analyzed: Sporadic EOC patients (n=451); BRCA mutant EOC patients (n=79); EOC patients treated with neoadjuvant chemotherapy (n=122), and; EOC patients treated adjuvantly with platinum-based chemotherapy after cytoreductive surgery (n=292).

Results: The KRAS-variant predicts for significantly worse survival for EOC patients over 55 years-old by multivariate Cox regression analysis (HR=1.71, 95% CI=1.09 - 2.69, p = 0.02). However, for the subgroup of EOC patients with known BRCA mutations, the KRAS-variant did not predict altered outcome (HR=0.994, CI=0.28-3.56, p=0.99). KRAS-variant positive EOC patients respond poorly to neoadjuvant carboplatin and paclitaxel chemotherapy, having significantly more residual disease remaining after surgery (OR=26.27, CI=1.56- 441.83, p=0.0232). In addition, EOC patients that harbor the KRAS-variant are more likely to be resistant to adjuvant platinum chemotherapy (OR=2.86, CI=1.13-7.23, p=0.026).

Conclusions: These findings expand the potential importance of the KRAS-variant in EOC, from acting as a marker of risk to being a biomarker that predicts worse outcome, perhaps due to its association with platinum resistance. These data may ultimately help lead to treatment optimization and improved outcome for KRAS-variant positive EOC patients.