Rectal cancer : when is the local recurrence risk low enough to refrain from the aim to prevent it?

Recently, preliminary results of the OCUM study (optimized surgery and MRI-based multimodal therapy of rectal cancer) were published and raised concern in the scientific community. In this observational study, the circumferential resection margin status assessed in preoperative MRI (mrCRM) was used to decide for either total mesorectal excision (TME) alone or neoadjuvant radiochemotherapy (nRCT). In contrast to current guidelines, neither T3 stage (with negative CRM) nor clinically positive lymph nodes were an indication for nRCT. Pathologically node-positive patients received chemotherapy (ChT). Overall, 230 patients were included, of whom 96 CRM-positive patients received nRCT. The CRM was accurately predicted in MRI, the rate of mesorectal plane resection was high. Recurrence rates have not yet been reported, but an impressive rate of down-staging for both T and N stage after nRCT was observed, while acute side effects were minimal. Nonetheless, the authors conclude that a substantial number of patients could be "spared severe radiation toxicity" and propagate their concept for prospectively replacing current guidelines. This is based on the hypothesis that CRM is a valid surrogate parameter for the risk of local recurrence and in case of a negative CRM, nRCT becomes dispensable. Moreover, it is assumed that lymph node status is no more relevant. Both assumptions are a contradiction to recent data from randomized studies as specified below. As 5-year locoregional recurrence rate (LRR) of only of 5-8% and < 5% in low risk rectal cancer can be achieved by the addition of RT, the noninferiority of surgery alone can not be presumed unless the expected 5-year LRR is ≤ 5-8%, whereas any excess of this range renders the study design inacceptable. Unless a publication explicitly specifies 5-year LRR, results are not exploitable for clinical decisions.