Comprehensive genomic profiling (CGP) via peripheral blood liquid biopsies identifies therapeutically relevant genomic alterations in tubo-ovarian and peritoneal cancer

Abstract

Objectives: Analysis of circulating tumor DNA (ctDNA) extracted from peripheral blood liquid biopsies is a less-invasive alternative to traditional solid tumor tissue testing. In this study, we analyzed a large cohort of liquid biopsies collected from advanced tubo-ovarian and peritoneal cancer patients to determine the clinical utility and landscape of potentially actionable genomic alterations. Methods: CGP by a hybrid-capture based sequencing assay was performed on blood samples from 574 patients with primary ovarian, fallopian tube, or peritoneal cancer. Short variant alterations, rearrangements, and copy number gains were interrogated across exons from 70 genes and introns from 7 genes; gene deletions were not evaluated. Microsatellite instability (MSI) was also assessed by the assay. An additional cohort assayed by an expanded gene panel and an analysis of paired liquid and tissue samples will be presented. Results: A total of 527 ovarian primary, 36 fallopian tube primary, and 11 peritoneal primary cancers were analyzed. Patients had a median age of 68 years (range 22-89+). There were 202 serous carcinomas, 7 carcinosarcomas, 7 clear cell carcinomas, 6 endometrioid carcinomas, 4 mucinous carcinomas, 1 carcinoma of mixed histology, and the remainder of the cases (n=347) had no specified histologic subtype. A total of 508 (89%) patients had at least 1 genomic alteration detected by the assay. The most frequently altered genes were TP53 (74%) consistent with high-grade carcinomas, CHEK2 (24%), ATM (14%), BRCA1 (7%), NF1 (7%), PIK3CA (6%), BRCA2 (5%), and KRAS (5%). Patients often harbored more than one alteration in TP53, CHEK2, and ATM (48%, 22%, and 16% of altered cases respectively). Alterations in BRCA1/2 were detected in a total of 71 patients (12.3%). 8/71 cases (11.3%) had multiple alterations in either BRCA1 or BRCA2, and these alterations were consistent with known reversion mechanisms that restored BRCA1/2 function and conferred resistance to PARP inhibitors (PARPi). Follow-up was available for one of these patients, who had BRCA2 alterations and had received 10 months of PARPi therapy prior to the collection of their liquid biopsy. A second patient showed acquisition of a second, likely reversion BRCA1 alteration on their liquid biopsy from a solid-tissue CGP test performed 22 months prior. Rearrangements in genes associated with approved targeted therapies were detected in FGFR2 (n=3), RET (n=1), ROS1 (n=1), and ALK (n=1). Other potentially targetable genes altered in this cohort included ERBB2 (1%), BRAF (1%), and PALB2 (1%). Additionally, one case was found to be MSI-high. Conclusions: CGP via liquid biopsies identified actionable alterations in metastatic or recurrent ovarian, fallopian tube, and peritoneal carcinomas across various histologic subtypes. The mutational landscape included alterations that supported the tissue diagnosis as well as potentially targetable alterations by both FDA-approved and investigational agents in clinical trials, inlcuding BRCA1/2 mutations that conferred sensitivity and/or explained resistance to PARPi therapy. These results support the clinical utility and integration of liquid biopsy in guiding the treatment of advanced tubo-ovarian and peritoneal cancers. Analysis of circulating tumor DNA (ctDNA) extracted from peripheral blood liquid biopsies is a less-invasive alternative to traditional solid tumor tissue testing. In this study, we analyzed a large cohort of liquid biopsies collected from advanced tubo-ovarian and peritoneal cancer patients to determine the clinical utility and landscape of potentially actionable genomic alterations. CGP by a hybrid-capture based sequencing assay was performed on blood samples from 574 patients with primary ovarian, fallopian tube, or peritoneal cancer. Short variant alterations, rearrangements, and copy number gains were interrogated across exons from 70 genes and introns from 7 genes; gene deletions were not evaluated. Microsatellite instability (MSI) was also assessed by the assay. An additional cohort assayed by an expanded gene panel and an analysis of paired liquid and tissue samples will be presented. A total of 527 ovarian primary, 36 fallopian tube primary, and 11 peritoneal primary cancers were analyzed. Patients had a median age of 68 years (range 22-89+). There were 202 serous carcinomas, 7 carcinosarcomas, 7 clear cell carcinomas, 6 endometrioid carcinomas, 4 mucinous carcinomas, 1 carcinoma of mixed histology, and the remainder of the cases (n=347) had no specified histologic subtype. A total of 508 (89%) patients had at least 1 genomic alteration detected by the assay. The most frequently altered genes were TP53 (74%) consistent with high-grade carcinomas, CHEK2 (24%), ATM (14%), BRCA1 (7%), NF1 (7%), PIK3CA (6%), BRCA2 (5%), and KRAS (5%). Patients often harbored more than one alteration in TP53, CHEK2, and ATM (48%, 22%, and 16% of altered cases respectively). Alterations in BRCA1/2 were detected in a total of 71 patients (12.3%). 8/71 cases (11.3%) had multiple alterations in either BRCA1 or BRCA2, and these alterations were consistent with known reversion mechanisms that restored BRCA1/2 function and conferred resistance to PARP inhibitors (PARPi). Follow-up was available for one of these patients, who had BRCA2 alterations and had received 10 months of PARPi therapy prior to the collection of their liquid biopsy. A second patient showed acquisition of a second, likely reversion BRCA1 alteration on their liquid biopsy from a solid-tissue CGP test performed 22 months prior. Rearrangements in genes associated with approved targeted therapies were detected in FGFR2 (n=3), RET (n=1), ROS1 (n=1), and ALK (n=1). Other potentially targetable genes altered in this cohort included ERBB2 (1%), BRAF (1%), and PALB2 (1%). Additionally, one case was found to be MSI-high. CGP via liquid biopsies identified actionable alterations in metastatic or recurrent ovarian, fallopian tube, and peritoneal carcinomas across various histologic subtypes. The mutational landscape included alterations that supported the tissue diagnosis as well as potentially targetable alterations by both FDA-approved and investigational agents in clinical trials, inlcuding BRCA1/2 mutations that conferred sensitivity and/or explained resistance to PARPi therapy. These results support the clinical utility and integration of liquid biopsy in guiding the treatment of advanced tubo-ovarian and peritoneal cancers.