Abstract PR08: Longitudinal sampling of ctDNA reveals actionable mutations to optimize treatment of patients with high-grade serous ovarian cancer

Abstract

Abstract Tumor cells leak their DNA into the blood stream, which allows us to use plasma sample from a patient to measure biomarker levels. In particular, circulating tumor DNA (ctDNA) sampling together with next-generation sequencing technologies provide a rapid and noninvasive test for quantifying tumor response to a treatment as well as finding potentially actionable mutations. The main objective of our study was to use ctDNA to establish a therapeutic window for targeting potentially actionable mutations to optimize the treatment of high-grade serous ovarian cancer (HGSOC) patients. We employed a targeted sequencing panel of 508 clinically annotated cancer genes to screen for actionable copy-number variations (CNV) and single-nucleotide variants (SNV) in 56 plasma and 20 fresh tumor samples from 13 patients with stage IIB-IVB HGSOC treated at the Turku University Hospital, Finland. All patients were surgically debulked and had received standard first-line carboplatin and paclitaxel chemotherapy. We analyzed longitudinal samples (from 3-6 time points per patient) from 11 patients during primary therapy, and from two patients with multiple (two and three samples per patient) relapsed disease. DNA isolated from tumor tissue and plasma was analyzed for genetic alterations by targeted deep-sequencing. White blood cell DNA was used as germline controls. Pathogenic germline mutations in 14 known ovarian cancer susceptibility genes were evaluated with CADD annotation framework and ClinVar database. Minimum requirements for passed somatic SNVs were CADD phred score 10, sequencing depth 100 (ctDNA) or 30 reads (tumor and blood), variant read count 4 or frequency 0.001. Variants detected in blood with > 2 reads were excluded. Additionally, a variant that passed filtering in at least one sample of a patient was included, regardless of threshold, if present in any other specimen. Thresholds of < 0.7 and > 4 copies were used as functionally relevant losses and gains, respectively. Clinically relevant germline pathogenic mutations in BRCA2, ATR, APC, and RB1 genes were identified from blood samples of five (38%) patients. The preoperative samples were also characterized by the highest somatic mutation burden, with a median total variant allele frequency (VAF) of 0.16 (range 0.02-0.77). Importantly, clinically relevant CNVs were detected in preoperative ctDNA of 6 (46%) patients, in potentially targetable driver genes such as ERBB2, PIK3CA, and CDK12. In six patients undergoing neoadjuvant chemotherapy (NACT), we noted a significant decrease in total ctDNA mutation burden during treatment, consistent with a drop in CA-125 levels. Interestingly, the levels of CNV counts increased in ctDNA during NACT in two patients, which associated with improved response to chemotherapy. During adjuvant treatment after debulking surgery, potentially actionable mutations were detected in all patients, with median total VAF of 0.03 (range 0.006-0.43). Importantly, in patients with no macroscopic residual disease, we detected a median of 2.5 (range 1-21) mutations in adjuvant treatment samples, indicating a molecularly persistent disease. Further, a subset of these alterations in, e.g., FANCA and PRCKB persisted during adjuvant treatment, revealing a potential therapeutic window for targeted treatments. Finally, two patients with progressive disease were characterized by predominantly either ctDNA gains or losses. Further, potentially actionable mutations in, e.g., TNXRD1, which sensitizes to AKT inhibitors, could be identified and tracked during disease progression. Taken together, longitudinal analysis of a panel-based ctDNA targeted sequencing can reliably detect very low VAF alterations, which can be used to identify and track actionable genomic alterations. Thus, ctDNA analysis can open a therapeutic window to target molecular residual disease in patients with newly diagnosed and advanced HGSOC. This abstract is also being presented as Poster B06. Citation Format: Kaiyang Zhang, Liina Salminen, Jaana Oikkonen, Kaisa Huhtinen, Johanna Hynninen, Seija Grénman, Sakari Hietanen, Rainer Lehtonen, Anniina Färkkilä, Sampsa Hautaniemi. Longitudinal sampling of ctDNA reveals actionable mutations to optimize treatment of patients with high-grade serous ovarian cancer. [abstract]. In: Proceedings of the AACR Conference: Addressing Critical Questions in Ovarian Cancer Research and Treatment; Oct 1-4, 2017; Pittsburgh, PA. Philadelphia (PA): AACR; Clin Cancer Res 2018;24(15_Suppl):Abstract nr PR08.