A real-world study confirms that circulating tumor DNA (ctDNA)–defined minimal residual disease (MRD) is highly prognostic for recurrence in patients with colorectal cancer (CRC) after curative intent, even after adjusting for current clinicopathologic features. The INTERCEPT program enrolled patients undergoing curative intent surgery for stages II-IV CRC at MD Anderson Cancer Center. After a median follow-up of 13 months, the 12-month disease-free survival (DFS) was 86.8% in the ctDNA-negative cohort compared with 15.9% for patients with at least 1 ctDNA-positive test.
The data offer support for the feasibility of incorporating ctDNA testing as part of routine surveillance, according to Giulia Maddalena, MD, and colleagues, who reported their findings in a poster presented at the 2024 American Society of Clinical Oncology (ASCO) Gastrointestinal Cancers Symposium.
In patients with a positive ctDNA test, the median lead time to relapse was 3.0 months, and if patients with concomitant relapse were excluded, the median lead time was extended to 5.6 months. Dr Maddalena, from The University of Texas MD Anderson Cancer Center, Houston, said, “Our study adds to the body of literature that helps improve stratification of colorectal cancer patients after surgery based on risk of recurrence that will eventually aid in better approaches for adjuvant therapies and improve staging. When MRD is detected during surveillance, there is a window of opportunity during the lead time to prevent eventual recurrence. Clinical trials are ongoing testing this hypothesis.” Only in a few cases did the test fail in relapse detection, the investigators noted.
The prognostic implication of MRD had been reported in observational cohorts, but little information had existed on the impact of returning and acting on ctDNA results on the subsequent ability to detect radiographic disease.
A total of 1140 patients were included in the INTERCEPT study. Patients had MRD testing integrated into routine colorectal cancer clinical care. Routine surveillance visits were performed according to National Comprehensive Cancer Network Guidelines, with the addition of ctDNA testing to each surveillance visit. A tumor-informed MRD assay was drawn postoperatively and every 3 months, and the results were returned to patients and providers, offering, according to the researchers, “opportunities to understand the prognostic performance of ctDNA in practice, including the rate of false negatives defined as radiographic recurrence without detectable ctDNA.”
At least one ctDNA-positive result was obtained in 338 patients (29.6%). Patients with advanced disease were more likely to have ctDNA detected postoperativly compared with patients with localized disease (odds ratio, 5.2; P<.0001).
With a median follow-up time of 13.0 months, the median DFS was not reached in the ctDNA-negative cohort compared with patients with at least 1 ctDNA-positive test (hazard ratio, 12.0; P<.0001).
When segregated by stage, ctDNA remained the most important predictor of outcomes. Patients with stage IV disease who were negative for ctDNA had better 12-month DFS rates than those with stage II disease who were ctDNA-positive. Among stage II patients, the 12-month DFS was 16.7% in those with ctDNA detected postoperatively compared with 94.9% in those with no detectable postoperative ctDNA. In the patients with stage III colorectal cancer, 12-month DFS was 34.9% in those with ctDNA detected postoperatively versus 93.3% when ctDNA was not detected, and in stage IV patients, the 12-month DFS rates were 12.4% and 74.1%, respectively.
The DFS was not reached for ctDNA-negative patients with localized disease; DFS was 21.7 months for those who were ctDNA-negative with an advanced tumor, was 7.7 months for those with localized tumors who were ctDNA-positive, and was 4.3 months for those with advanced disease who were ctDNA-positive.
A false negative ctDNA (negative test but macroscopic disease detection within 4 months) was detected in 17 patients, representing a rate of 1.49%. For 9 patients, relapse was detected on a single lesion: subcentimeter lung node in 6, to the liver in 2, and local relapse in 1.
“Based on these findings, ctDNA may be considered a standard of care for all stages of colon cancer and is prognostic of DFS,” said Cathy Eng, MD, from Vanderbilt-Ingram Cancer Center, Nashville, who was not involved in the study.