Comprehensive genomic profiling by liquid biopsy captures tumor heterogeneity and identifies cancer vulnerabilities in patients with RAS/BRAFV600E wild-type metastatic colorectal cancer in the CAPRI 2-GOIM trial

Ann Oncol. 2024 Aug 29:S0923-7534(24)03914-0. doi: 10.1016/j.annonc.2024.08.2334. Online ahead of print.

Abstract

Background: Emerging evidence supports tumor tissue-based comprehensive genomic profiling (CGP) in metastatic colorectal cancer (mCRC). Data on liquid biopsy-based circulating tumor DNA (ctDNA) CGP are scarce and mainly retrospective. Prospective comparison between the two tests is not currently available.

Materials and methods: The CAPRI 2-GOIM trial investigates efficacy and safety of ctDNA-driven, cetuximab-based sequence of three treatment lines in patients with RAS/BRAFV600E wild-type (WT) mCRC, as determined by the local laboratory. Before first-line therapy, CGP is carried out with FoundationOne (F1) CDx and F1 Liquid (F1L) CDx (324 genes) on tumor tissue DNA and plasma ctDNA, respectively.

Results: For 2/207 (0.96%) patients, no ctDNA was detected by F1L CDx. No patient displayed tumor fraction (TF) below 1%, whereas elevated ctDNA (TF ≥ 10%) was detected among 140/205 (68.3%) patients. One thousand and thirteen genomic variants were identified. F1L CDx found KRAS, NRAS, or BRAFV600E alterations in 19 patients, whose tumors were classified as RAS/BRAFV600E WT by the local laboratory. Both F1 CDx and F1L CDx were available for 164/205 (80%) patients. A concordance of 61.4% between the two tests was observed. The concordance increased to 72.7% for F1L CDx with TF ≥ 10%. Concordance for genes potentially involved in anti-epidermal growth factor receptor resistance was found in 137/164 (83%) patients, increasing to 91.5% for F1L CDx with TF ≥ 10%. A higher number of genomic alterations was detected by F1L CDx compared with F1 CDx, including six cases with KRAS and NRAS alterations. Overall, 109/205 (53.2%) patients displayed at least one actionable genomic alteration (I to IIIB), according to the European Society for Medical Oncology Scale for Clinical Actionability of Molecular Targets (ESCAT).

Conclusion: Baseline liquid biopsy-based CGP is feasible, has high concordance with tumor tissue-based CGP, could better recapitulate tumor heterogeneity, and is clinically informative by identifying additional actionable genomic alterations in approximately half of RAS/BRAFV600E WT mCRC patients.

Keywords: anti-epidermal growth factor receptor (EGFR) drugs; circulating tumor DNA (ctDNA); colorectal cancer (CRC); comprehensive genomic profiling (CGP); liquid biopsy.