Abstract 2498: Assessing the clinical value of ctDNA sequencing for initial tumor profiling in metastatic colorectal cancer patients with sufficient tumor tissue
Yongjun Cha,Bun Kim,Dong Woon Lee,Hyun Yang Yeo,Chang Won Hong,Kyung Su Han,Byung Chang Kim,Hee Jin Chang
DOI: https://doi.org/10.1158/1538-7445.am2024-2498
IF: 11.2
2024-03-22
Cancer Research
Abstract:Abstract Introduction Tumor profiling including RAS, BRAF, HER2, and MSI/MMR status, is required to determine the treatment for patients with metastatic colorectal cancer (mCRC) at the time of diagnosis. While comprehensive tumor profiling using tissue-based next-generation sequencing (NGS) is increasingly adopted for initial profiling in mCRC, the role of circulating tumor DNA (ctDNA) NGS as initial testing in patients with sufficient tumor tissue is not clearly understood. We assessed the clinical value of ctDNA sequencing compared to tumor NGS in patients with newly diagnosed mCRC who have sufficient tumor specimens. Methods We prospectively enrolled consecutive patients with newly diagnosed mCRC at the National Cancer Center Korea. As per the institutional protocol for mCRC, initial tumor profiling was performed on primary tumor tissue using an in-house NGS panel (NCC PCP ver.3), which included 525 genes. For ctDNA sequencing, patients were evaluated using the AlphaLiquid®100 comprehensive cancer panel (IMBdx, Inc.), which included 118 genes, before the initiation of chemotherapy. Additionally, immunohistochemical (IHC) testing for HER2 and polymerase chain reaction (PCR)/IHC testing for MSI and/or MMR were performed to assess the accuracy of HER2 and MSI/MMR status. Results A total of 188 patients were enrolled. In 139 eligible patients, 275 potentially actionable mutations were found in 12 selected CRC-related genes (APC, TP53, KRAS, NRAS, BRAF, FBXW7, PIK3CA, ERBB2, SMAD4, NF1, EGFR, MET). Of these, 32% were found both in ctDNA and tissue; 54% were found in ctDNA only; 12% in tissue only, and 2% were discordant. For RAS/BRAF mutations, which are required for anti-EGFR treatment decisions, the concordance rate between ctDNA and tissue NGS was 83.1%, and the concordance was higher in patients with higher ctDNA concentrations. For 9 patients with potentially actionable copy number variations (CNV) in EGFR, HER2, MET, and FGFR1, 3 cases were found by both assays; 4 were found by ctDNA only, and 2 were found by tissue only. ctDNA NGS correctly predicted MSI/MMR status in 2 out of 4 patients with MSI-H/dMMR; in the 2 other patients, the MSI and MMR statuses were different, suggesting potential false positivity. In addition, the fold changes in ctDNA dynamics during treatment significantly correlated with changes in tumor size and CEA levels, as well as with droplet digital PCR copy number fold changes. In a patient with MET amplification, ctDNA NGS identified MET Y1230H as a potential acquired resistance mutation after crizotinib treatment, which responded to cabozantinib but not to capmatinib. Conclusions Initial tumor profiling using ctDNA NGS yielded outcomes comparable to those of tumor tissue NGS in guiding treatment for patients with newly diagnosed mCRC, thereby suggesting its utility as an initial profiling method in mCRC. Citation Format: Yongjun Cha, Bun Kim, Dong Woon Lee, Hyun Yang Yeo, Chang Won Hong, Kyung Su Han, Byung Chang Kim, Hee Jin Chang. Assessing the clinical value of ctDNA sequencing for initial tumor profiling in metastatic colorectal cancer patients with sufficient tumor tissue [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2498.
oncology