PD-018Feasibility of next-generation sequencing of circulating tumor cells for predicting relapse after liver resection in metastatic colorectal cancer

Introduction: Colorectal cancer is a major healthcare burden globally. Patients with metastatic disease will eventually die of the disease but liver resection offers the potential for cure in those who present with liver-only metastasis. However, early relapse after liver resection still occurs in up to 30% of cases. The current prospective pilot study evaluated the feasibility of preop next-generation sequencing (NGS) of circulating tumor cell (CTC) in predicting relapse after curative liver resection in metastatic colorectal cancer (mCRC). Methods: Patients, who were planned for resection of liver metastasis with or without the primary tumour, were screened for eligibility. The decision to liver resection was made in the multidisciplinary team comprising oncologists, surgeons, and diagnostic radiologists. Blood samples for CTC were collected within 7 days prior to the scheduled liver resection. CTCs were enriched by a white blood cell depletion strategy using a microfluidic isolation machine with a CTChip that separates cells by size. For enumeration, cell images were captured by fluorescence microscopy and CTCs (CD45-ve; CK/ EpCAM +ve) were counted by imaging software. For NGS, DNA was prepared from CTCs isolated from blood using the ClearCell® FX1 System (Biolidics, Singapore). Whole genome amplification (WGA) was subsequently performed using REPLI-g Single Cell Kit (QIAGEN). DNA concentration was determined by Qubit (Life Technologies). The library preparation was performed using the HEAT-Seq Target Enrichment system (Roche). In this study, the HEAT-Seq Oncology panel, having a capture target of 245 kb across 60 cancer-associated genes, was used. All study procedures were initiated after approval from the Institutional Review Board (UW16-235) and according to the Declaration of Helsinki. Results: Twenty consecutive patients with liver-only mCRC were recruited from May 2017 to Jan 2018. High-purity run with whole genome amplification and subsequent next-generation sequencing with HEAT-Seq Oncology panel were successful in all samples. CTC (CD45-ve; CK/ EpCAM +ve) counts ranged from 0-129. Mutations were detected in all samples and the number of mutations ranged from 2- 45 per sample. Potentially druggable mutations were found in 17 patients (85%). While the reported frequency of mutant SMO was 2.7% among patients with metastatic colorectal cancer, 45% of patients in the current study harbored SMO mutations in their CTCs. Fifteen variants of SMO mutations were detected and 7 of these were predicted to be driver mutations. Overall, 6 of the 20 patients harbored at least one driver mutation of SMO in their CTCs. Among patients with or without disease relapse, the rate of SMO mutations was 60% and 20%, respectively. Upon logistic regression analysis, patients with SMO mutation in their CTCs had a significantly higher risk of relapse (OD 1.23; 95% CI, 1.03-1.47, P = .035). Conclusion: The current study demonstrated the feasibility of genomic profiling of CTCs from patients with metastatic colorectal cancer before potentially curative liver resection. Mutations were detected in all CTC samples and the majority of them had druggable potential. SMO mutation, a druggable target in the Hedgehog signaling pathway, was associated with disease relapse in the current study and further validation of this finding in a larger cohort is warranted.