New predictors of recurrence in post-polypectomy surveillance: a retrospective analysis of risk stratification for advanced adenomas and polyps.

Abstract

Background and aims: Recent guidelines emphasized that not all patients with recurrent polyps benefit from post-polypectomy surveillance, except for advanced adenomas recurrence. Our study aimed to analyze the recurrence risk factors for advanced adenomas and investigating the difference from any polyp recurrence.

Methods: This retrospective observational study included patients who underwent colonoscopy and at least once post-polypectomy surveillance. Multivariate regression models identified risk factors for the recurrence of polyps and advanced adenomas. The Youden index determined the optimal cut-off value for risk factors. Kaplan-Meier curve estimated the cumulative recurrence rates, and the log-rank tests were compared the differences between these curves.

Results: A total of 1818 patients were detected polyps at baseline examination. During post-polypectomy surveillance, 1063 patients had recurrent polyps, and 64 patients experienced recurrent advanced adenomas. Multivariate logistic analysis identified age as an independent risk factor for both advanced adenoma [OR(95% CI): 1.028 (1.003-1.056)] and polyps [OR(95% CI): 1.019 (1.009-1.028)], with cut-off values of 57 years and 53 years, respectively. Cut-off value of 40 years can increase the predictive sensitivity to 95%. Polyps size [OR(95% CI):1.070 (1.014-1.147)] and high-risk pathology [OR(95% CI): 6.339 (2.057-23.919)] were significantly associated with the recurrence of advanced adenomas, but not with any polyps, with a size cut-off value of 7.5 mm. Except for high-risk pathology, neither hyperplastic polyps nor tubular adenomas with low-grade dysplasia increased the recurrence risk of advanced adenomas.

Conclusions: 40 years can increase the predictive sensitivity to 95% for both advanced adenomas and any polyps. Polyps size and high-risk pathology were associated with the recurrence of advanced adenoma, but not with any polyps.