Liquid biopsies from circulating tumor DNA (ctDNA) have been employed recently as a non-invasive diagnostic tool for detecting cancer-specific gene mutations. Here, we show the comprehensive gene mutation profiles of ctDNA in 51 patients with different histological subtypes of stage I-IV ovarian cancer, and their association with clinical outcomes. The ctDNA extracted from pre-treatment patients' plasma were analyzed using Cancer Personalized Profiling by Deep Sequencing targeting 197 genes. Of 51 patients, 48 (94%) showed one or more non-synonymous somatic mutations, including TP53 (37.3%), APC (17.6%), KRAS (15.7%), EGFR (13.7%), MET (11.8%), PIK3CA (11.8%), NPAP1 (11.8%), and ALK (9.8%). The most frequently mutated genes were as follows: TP53 in high-grade serous carcinoma (66.7%), APC in clear cell carcinoma (30.8%), PIK3CA in endometrioid carcinoma (40%), and KRAS in mucinous carcinoma (66.7%). Higher cell-free (cf)DNA concentration significantly correlated with worse progression-free survival (PFS) in all patients as well as stage III-IV patients (p = 0.01 and 0.005, respectively). Further, patients with any pathogenic mutations showed significantly worse PFS (p = 0.048). Blood tumor mutational burden detected from ctDNA did not significantly correlate with the histological subtypes or survival. Collectively, clinico-genomic profiles of individual ovarian cancer patients could be identified using ctDNA and may serve as a useful prognostic indicator. These findings suggest that ctDNA-based gene profiling might help in establishing personalized therapeutic strategies.
Keywords: circulating tumor DNA (ctDNA); gene mutation profile; liquid biopsy; ovarian cancer; prognosis.