Reevaluating the Evidence for Intensive Postoperative Extracolonic Surveillance for Nonmetastatic Colorectal Cancer

•A total of 3 recent trials (FACS, GILDA, and COLOFOL) failed to demonstrate a statistically significant survival advantage from intensive extracolonic surveillance for resected nonmetastatic colorectal cancer (CRC). This contrasts sharply with previous meta-analyses that found a moderate-to-large benefit from intensive follow-up. The recent evidence has been interpreted as demonstrating that extracolonic surveillance does not improve survival and, consequently, as indirect evidence that the increasingly common practice of metastasectomy may be less beneficial than previously thought.•This article leverages a modeling analysis to argue that the results of the trials have been misinterpreted. Given the large reduction in CRC recurrence rates over the last 30 years, the average 5-year survival benefit from intensive extracolonic surveillance would likely now be much smaller even if the benefit to patients who experience a recurrence was large. The trials were terribly underpowered to detect such a benefit. Overall, although very imprecise, the results of the 3 trials are consistent with a realistically sized survival benefit and thus should not motivate de novo skepticism about the efficacy of metastasectomy nor even risk-stratified extracolonic surveillance. Moreover, although the 5-year average benefit would be small, the predicted life-expectancy gain from the latter is comparable with CRC screening and thus would be clinically significant.•In the absence of experimental evidence, the true benefit of metastasectomy and consequently extracolonic surveillance remains unknown. Power analyses show that any future trial designed to compare surveillance regimens with respect to survival is practically guaranteed to be uninformative and thus would be ethically suspect. A moderate-to-large (n = 200-300) trial of salvage surgery (or of metastasectomy in particular) is needed to inform clinical practice in both cases.