Abstract GMM-043: CTDNA PROFILING TO PREDICT PROGNOSIS AND OPTIMIZE TREATMENT IN HIGH-GRADE SEROUS OVARIAN CANCER

Tumor cells leak their DNA into the blood stream, which allows detection of tumor mutations and copy number variants from circulating tumor DNA (ctDNA) from plasma samples. These variants offer a dynamic “molecular snapshot” to the changing landscape of genomic events occurring during tumor treatment and progression. Our aim is to translate ctDNA profiling into clinical benefit in patients with high-grade serous ovarian cancer (HGSOC). We used a comprehensive cancer-specific sequencing panel of over 500 genes to identify mutations and copy-number alterations (CNA) in ctDNA from plasma. We collected 78 plasma samples in 12 patients: at pre-treatment, primary treatment, follow-up and possible progression. For each patient, we also collected tissue and ascites samples (from 1-4 different time points per patient, totally 21 samples) and a white blood cell sample for germline variant detection. From variation detected in ctDNA, we identified clinically relevant information to predict prognosis and detect actionable genomic alterations that could be used to target treatment. After extensive filtering of non-somatic mutations, we detected high concordance between ctDNA and tumor tissue samples: 77% of the mutations detected in ctDNA were also detected in tumor tissue samples. Mean correlation between CNAs detected in plasma versus tissue was also high, 0.7. We identified several actionable mutations and CNAs. For example, we identified ERBB2 amplification in pre-treatment ctDNA in a poor-responding patient (platinum free interval (PFI) 5 months). The HER2 over-expression was validated in interval tumor tissue sample with immunohistochemistry and in-situ hybridization. Based on these results, the patient was treated with trastuzumab combined with reduce-dose carboplatin and dose-dense paclitaxel during disease progression, which yielded promising clinical response. In two other patients, mTOR pathway activation was predicted based on mutations detected in ctDNA. In both patients, the activation was validated with immunohistochemistry. These patients could benefit from mTOR inhibitors in case of disease progression. These identified clinically relevant variants illustrate the clinical value of ctDNA in the treatment of HGSOC patients. Overall, patients with longer PFIs showed fast response to chemotherapy: ctDNA level was considerably reduced and mutational composition changed after first cycles of chemotherapy. Contrary, the poor-responding patients with PFI less than 12 months showed failure to drop ctDNA level after start of chemotherapy, smaller changes in mutational composition during primary treatment and higher number of detected mutations. The early prognosis prediction in combination with identification of clinically relevant variants can allow window of opportunity to treat poor-prognosis patients even before relapse. Additionally, ctDNA allows detection of changes in mutational composition during treatment that can reveal subclonal selection which cannot be covered by single biopsies. Citation Format: Jaana Oikkonen, Kaiyang Zhang, Liina Salminen, Kaisa Huhtinen, Ingrid Schulman, Noora Andersson, Olli Carpen, Sakari Hietanen, Seija Grenman, Rainer Lehtonen, Johanna Hynninen, Anniina Farkkila, Sampsa Hautaniemi. CTDNA PROFILING TO PREDICT PROGNOSIS AND OPTIMIZE TREATMENT IN HIGH-GRADE SEROUS OVARIAN CANCER [abstract]. In: Proceedings of the 12th Biennial Ovarian Cancer Research Symposium; Sep 13-15, 2018; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2019;25(22 Suppl):Abstract nr GMM-043.