Functional validation of somatic variability in TP53 and KRAS for prediction of platinum sensitivity and prognosis in epithelial ovarian carcinoma patients.

Abstract

Concerning the dismal prognosis of chemoresistant patients with epithelial ovarian carcinoma (EOC), we aimed to follow up the findings of a previous whole-exome sequencing study using an orthogonal Sanger sequencing on the same patients and a separate set of 127 EOC patients (N = 177, all fresh frozen tumor samples). We focused on TP53 as a frequently mutated gene relevant for chemosensitivity, included KRAS as an additional therapeutically relevant target, complemented the study with transcript levels of both genes, and compared results with clinical parameters. All variants in TP53 and KRAS detected by exome sequencing were confirmed. KRAS mutated patients had significantly more frequent FIGO stages I or II (p = .002) and other than high-grade serous tumor subtypes (nonHGSCs) (p < .001), which was connected with lower KRAS transcript levels (p = .004). Patients with nonHGSC subtypes had less frequent TP53 mutations (p = .002). Carriers of TP53 variants disrupting the DNA binding loop had significantly longer platinum-free intervals than the rest (p = .037). Tumors bearing nonsense, frameshift, or splice site TP53 variants had a significantly lower TP53 transcript level, while those with missense variants had significantly higher levels than wild types (p < .001). The normalized intratumoral TP53 and KRAS transcript levels were correlated, and patients with co-mutated genes had poorer overall survival than others (p = .015). Protein levels of both genes significantly correlated with their respective transcripts (p = .028 and p = .001, respectively). Our study points to KRAS as a target for future therapy of nonHGSCs and reveals the prognostic value of TP53 variants in the DNA binding loop.