Patients with T1 colorectal cancer might not benefit from additional surgery after endoscopic resection unless they have positive or indeterminate resection margins or high-risk histology, according to a retrospective, population-based study of 1,315 patients.

After a median follow-up of 6.6 years, the rates of colorectal cancer (CRC) recurrence were 6.2% in patients who underwent endoscopic resection only and 6.4% in patients who also had additional surgery (P = .9), reported Tim D.G. Belderbos, MD, of University Medical Center Utrecht (the Netherlands). Rates of local recurrence also were similar between these groups (4.1% and 3.7%, P = .3), he and his associates reported in the March issue of Clinical Gastroenterology and Hepatology (doi: 10.1016/j.cgh.2016.08.041).

Among high-risk patients, however, the rates of overall recurrence were 14% with endoscopic resection only and 7% with endoscopic resection plus additional surgery (P = .06), and the rates of local recurrence were 12% and 1%, respectively (P = .004). “Based on our study, we recommend performing additional surgery after initial endoscopic resection in cases of high-risk T1 CRC, determined by high-risk histology and/or positive resection margins,” the researchers concluded. Invasive CRCs confined to the colonic submucosa (T1 CRC) present a treatment dilemma – they are usually cured by complete endoscopic resection, but up to 13% involve lymph node metastases and need additional surgery, the investigators noted. To identify predictors of recurrence and metastasis, they studied all patients diagnosed with T1 CRC in the Southeast Netherlands from 1995 through 2011. A total of 370 patients (28%) underwent endoscopic resection only, 220 (17%) underwent endoscopic resection with additional surgery, and 725 (55%) had an initial surgical resection.

Surgery after endoscopic resection was more likely when patients had positive or doubtful resection margins (P less than .001), and this link remained significant after high-risk histology, tumor location, time period, age, sex, and comorbidities were controlled for. Endoscopic resection plus surgery did not reduce the risk of recurrence, compared with endoscopic resection only (P = .3), after the investigators accounted for age, sex, year of procedure, tumor location, and margin characteristics. Initial surgery was associated with significantly lower rates of overall and local recurrence, compared with endoscopic resection only, but the differences also lost significance in the multivariable analysis (P = .2).

Only the presence of positive resection margins significantly predicted recurrence among patients undergoing endoscopic resection (hazard ratio, 6.9; 95% confidence interval, 2.3-20.9). Positive or doubtful resection margins also predicted recurrence after initial surgery, with hazard ratios of 13.2 and 3.4, respectively. High-risk histology – that is, poor differentiation, deep submucosal invasion, or lymphangioinvasion – was significantly associated with lymph node metastasis (OR, 2.2; 95% CI, 1.3-3.7; P less than .002), but not with recurrence after resection margins were accounted for. This might result from missing histology data or the fact that patients with high-risk histology tended to undergo surgical rather than endoscopic resection, the researchers said.

They noted several other study limitations, including a lack of details about lesions and procedures. Also, endoscopic submucosal resection was not practiced in the Netherlands during the study period, they said.

The investigators did not report funding sources and had no disclosures.