AbstractBACKGROUND & AIMS
There is controversy over the optimal management for T1 colorectal cancer (T1 CRC). This study compared initial endoscopic resection with or without additional surgery, or initial surgery for T1 CRC, and assessed risk factors for lymph node metastases (LNMs) and long-term recurrence.METHODS
We performed a registration study that included all patients diagnosed with T1 CRC from 1995 through 2011 in the southeast area of The Netherlands (n = 1315). High-risk histology (with regard to LNM) was defined as the presence of poor differentiation, lymphangio-invasion, and/or deep submucosal invasion. The primary outcome measure was the combined rate of local and distant CRC recurrence during a mean follow-up period of 6.6 years. Logistic regression and Cox proportional hazards regression analyses were performed to evaluate independent risk factors for LNM and CRC recurrence, respectively.RESULTS
Endoscopic resection was performed in 590 patients (44.9%); of these, 220 (16.7%) underwent additional surgery. Initial surgery was performed in 725 patients (55.1%). The risk of LNM was higher in T1 CRC with histologic risk factors (15.5% vs 7.1% without histologic risk factors; odds ratio, 2.21; 95% confidence interval, 1.33–3.70). Thirty-day mortality did not differ between patients who received additional surgery (0.9%) and those who underwent only endoscopic resection (1.4%; P = .631). Rates of CRC recurrence were 6.2% (9.8/1000 patient-years) after only endoscopic resection vs 6.4% (9.4/1000 patient-years) after additional surgery (P = .912), and 3.4% (5.2/1000 patient-years) after initial surgery (P = .031). In multivariate analysis, this difference was not significant. The only independent risk factor for long-term recurrence was a positive resection margin (hazard ratio, 6.88; 95% confidence interval, 2.27–20.87).CONCLUSIONS
Based on a population analysis of patients diagnosed with T1 CRC, additional surgery after endoscopic resection should be considered only for patients with high-risk histology or a positive resection margin.