Prevalence of advanced histological features in diminutive and small colon polyps

Background: Investigators have proposed "predict, resect, and discard" strategies for diminutive (≤ 5 mm) or small (6-9 mm) polyps to reduce screening colonoscopy costs. Advanced histological features such as villous histology, high-grade dysplasia, and/or cancer in these polyps could deter adoption of these strategies.

Objective: Determine the prevalence of advanced histological features in diminutive and small colon polyps.

Design: Retrospective analysis of data from 3 prospective clinical trials.

Setting: Two tertiary-care referral centers.

Patients: This study involved patients undergoing screening or surveillance colonoscopy.

Intervention: The location, size, and morphology of each polyp detected was documented. Each polyp was then resected, placed in a unique specimen jar, and sent for histopathological evaluation.

Main outcome measurements: Rates of advanced histological features (villous histology, high-grade dysplasia, and cancer).

Results: A total of 2361 polyps were detected, removed, and retrieved. Both diminutive and small polyps had a lower frequency of any advanced histological features compared with large polyps (0.5% and 1.5%, respectively vs 15.0%; P < .001 for both comparisons). Polyps <10 mm in size had a lower frequency of advanced histology compared with polyps ≥ 10 mm (0.8% vs 15.0%; P < .001). During sensitivity analysis, the frequency of advanced histological features varied from 0.2% to 0.7% within diminutive polyps, 1.5% to 3.6% within small polyps, and 0.8% to 1.2% within polyps <10 mm.

Limitations: Retrospective analysis from tertiary-care referral centers; predominantly white, male, veteran patient population resulting in limited generalizability of results.

Conclusion: The prevalence of advanced histological features in colon polyps ≤ 5 mm is very low (0.5%). This has important implications for the potential practice of "predicting, resecting, and discarding" diminutive colon polyps.