Investigation of PARP inhibitor resistance through the analysis of serially collected circulating tumor DNA (ctDNA) in ovarian cancer patients (027)

Abstract

Objectives: An increasing number of ovarian cancer patients are becoming resistant to PARP inhibitor (PARPi). Aside from BRCA reversion, other patient-specific molecular alterations which underlie the development of acquired resistance are unknown. Moreover, the potential efficacy of serial circulating tumor DNA (ctDNA) in predicting response to PARP is yet to be explored. Thus, we have analyzed serially obtained ctDNA in samples from patients receiving PAPR inhibitors. Methods: Patients receiving PARPi were prospectively enrolled since January 2018. Whole blood samples were obtained prior to receiving PARPi and serially every three months until progression on PARPi. Every patient had tumor next-generation sequencing data. Cell-free DNA was extracted from plasma and germline DNA from buffy coats. Extracted DNA was target enriched with a panel targeting 531 cancer-related genes, and the resultant library was sequenced using the Illumina Novaseq 6000 sequencing system. The sequencing data was analyzed using our custom analysis pipeline. Results: A total of 72 patients have contributed at least one sample, and 29 patients have recurred so far. Preliminary analysis was performed for 17 patients with paired samples. When compared to tumor NGS data, analysis of ctDNA in patient-matched samples showed ctDNA specific variants. Analysis of patient-matched paired samples before and after progression on PARP inhibitor showed an increase in intratumoral heterogeneity in post-progression samples. Genes that were frequently modified after gaining PARPi resistance were BRCA1, TP53, BRCA2, and ATM. BRCA reversion was found in two patients. Moreover, new variants in genes that are linked to DNA repair pathways were found to be associated with increased tumor mutation burden. Conclusions: ctDNA, a less-invasive alternative to serial tumor biopsy, helps capture the changes in mutational status in patients receiving PARPi. In terms of prediction of PARPi response and investigation of resistance mechanism, serial ctDNA samples before and during PARPi may provide useful insights. Updated analysis with additional patients will be shared at the conference. Objectives: An increasing number of ovarian cancer patients are becoming resistant to PARP inhibitor (PARPi). Aside from BRCA reversion, other patient-specific molecular alterations which underlie the development of acquired resistance are unknown. Moreover, the potential efficacy of serial circulating tumor DNA (ctDNA) in predicting response to PARP is yet to be explored. Thus, we have analyzed serially obtained ctDNA in samples from patients receiving PAPR inhibitors. Methods: Patients receiving PARPi were prospectively enrolled since January 2018. Whole blood samples were obtained prior to receiving PARPi and serially every three months until progression on PARPi. Every patient had tumor next-generation sequencing data. Cell-free DNA was extracted from plasma and germline DNA from buffy coats. Extracted DNA was target enriched with a panel targeting 531 cancer-related genes, and the resultant library was sequenced using the Illumina Novaseq 6000 sequencing system. The sequencing data was analyzed using our custom analysis pipeline. Results: A total of 72 patients have contributed at least one sample, and 29 patients have recurred so far. Preliminary analysis was performed for 17 patients with paired samples. When compared to tumor NGS data, analysis of ctDNA in patient-matched samples showed ctDNA specific variants. Analysis of patient-matched paired samples before and after progression on PARP inhibitor showed an increase in intratumoral heterogeneity in post-progression samples. Genes that were frequently modified after gaining PARPi resistance were BRCA1, TP53, BRCA2, and ATM. BRCA reversion was found in two patients. Moreover, new variants in genes that are linked to DNA repair pathways were found to be associated with increased tumor mutation burden. Conclusions: ctDNA, a less-invasive alternative to serial tumor biopsy, helps capture the changes in mutational status in patients receiving PARPi. In terms of prediction of PARPi response and investigation of resistance mechanism, serial ctDNA samples before and during PARPi may provide useful insights. Updated analysis with additional patients will be shared at the conference.