Association Between Index Adenomas and Advanced Findings at Third Colonoscopy: Data From the New Hampshire...

The United States Multi-Society Task Force for Colorectal Cancer (CRC) provides recommendations for the second surveillance colonoscopy for patients with adenomas on index exam. However, evidence which informs risk at second surveillance is limited due to lack of data on quality measures and known risk factors. We evaluated the risk for an advanced outcome on second surveillance colonoscopy for patients with index adenomas and a first surveillance colonoscopy without polyps, while accounting for risk factors and adenoma detection rates.

Our primary outcome was the detection of advanced outcomes at second surveillance colonoscopy as defined by large (>1 cm)/dysplastic serrated polyps, advanced adenoma (AA), or CRC. Another outcome was AA or CRC. We used prospectively collected New Hampshire Colonoscopy Registry (NHCR) data. Patients had 2 surveillance colonoscopies >12 months apart and a first surveillance colonoscopy without polyps. Patients had no previous colorectal neoplasia, genetic syndromes, or inflammatory bowel disease. Patients were risk-stratified based on index colonoscopy findings; no adenomas; 1 to 2 nonadvanced adenomas; 3 to 4 nonadvanced adenomas; AA /5+ nonadvanced adenomas. The reference group had no adenomas on index and first surveillance. Covariates in the Poisson model included family history of CRC, smoking, adenoma detection rates, and serrated polyps on index exam.

We included 3171 patients from the NHCR, 3.7% (95% confidence interval [CI], 3.0%-4.6%) of patients with no index adenomas, 6.3% (95% CI, 4.9%-8.2%) with 1 to 2 index nonadvanced adenomas, 4.3% (95% CI, 1.5%-12.0%) with 3 to 4 index nonadvanced adenomas, and 6.2% (95% CI, 4.0%-9.4%) with 5+ nonadvanced adenomas and/or AAs developed advanced outcomes on second surveillance colonoscopy. The adjusted rate ratio for advanced outcomes on second surveillance exam was 2.12 (95% CI, 1.06-4.25) in patients with AA/5+ nonadvanced adenomas, 1.23 (95% CI, 0.26-5.74) in patients with 3 to 4 nonadvanced adenomas, and 1.67 (95% CI, 1.01-2.74) in patients with 1 to 2 nonadvanced adenomas. The adjusted rate ratio for AA or CRC on second surveillance exam was 3.47 (95% CI, 1.46-8.24) in patients with AA/5+ nonadvanced adenomas, 2.26 (95% CI, 0.48-10.67) in patients with 3 to 4 nonadvanced adenomas, and 1.84 (95% CI, 0.94-3.59) in patients with 1 to 2 nonadvanced adenomas.

Our analysis demonstrated that patients with AAs or 5+ adenomas on index exam and a normal first surveillance exam have increased risk for advanced findings on second surveillance colonoscopy. This association is independent of quality measures including adenoma detection rate and known risk factors such as smoking and body mass index.