The application of dynamic monitoring of circulating tumor DNA for detecting minimal residual disease and predicting recurrence in colorectal cancer patients.

Abstract

e15531 Background: Although the prognosis of colorectal cancer (CRC) has improved in the past decade, a subset of CRC patients may still suffer from relapse due to the progression from minimal residual disease (MRD) after surgical resection. A sensitive and non-invasive method to detect MRD and early diagnosis of recurrence disease is warranted to start therapeutic interventions at an earlier timing and improving the overall survival rate. In this study, we have evaluated the feasibility of circulating tumor DNA (ctDNA) analysis to detect MRD and early detection of recurrence in CRC patients. Methods: Plasma samples were collected prospectively from 38 CRC patients (stage I to IV), who underwent surgical resection. Preoperative blood samples were obtained just before surgery and post-operative samples were collected at multiple time-points to monitor the changes of tumor mutation profiles. Tumor-derived mutations were detected in preoperative blood samples as well as surgically resected-tumor tissues using ultradeep targeted next generation sequencing. Patient-paired peripheral blood cells (PBCs) were sequenced concurrently to exclude clonal hematopoiesis-related mutations. Results: Among the 38 patients, 74 non-synonymous mutations were identified in tumor tissues and 64 mutations in the preoperative plasma samples. Paired PBCs sequencing identified 11 mutations in plasma samples to be clonal hematopoiesis-related mutations. After the exclusion of clonal hematopoiesis-related mutations, 34 (89.5%) of the 38 patients harbors at least one somatic mutation either from tumor tissues or plasma samples to be monitored longitudinally. ctDNA was detectable in 5 of 14 (36%) post-surgical samples of patients who did not receive adjuvant chemotherapy and in 9 of 18 (50%) post-chemotherapy samples. Up to date, 6 patients have been detected with clinical recurrence and ctDNA analysis identified all 6 recurrences before imaging. Serial ctDNA analyses were able to detect disease recurrence up to 6 months before imaging tests. Furthermore, all patients that were ctDNA negative post-operative or post-chemotherapy showed no signs of clinical relapse. Conclusions: Our current results indicate that ctDNA analysis allows the detection of MRD in CRC patients. The integration of ctDNA analysis with current standard monitoring guidelines holds great promise in early detection of recurrence to allow clinical intervention to be applied promptly.