Factors Associated With Classification of Hyperplastic Polyps as Sessile Serrated Adenomas/Polyps on Morphologic Review.

BACKGROUND: Distinguishing sessile serrated adenomas/polyp (SSA/P), a subset of serrated polyps, from hyperplastic polyps (HPs) remains a challenge and has surveillance implications. Our goal was to identify clinical and pathologic factors associated with serrated polyps originally read as HPs being reassessed as SSA/Ps versus confirmed as HPs. METHODS: Data were collected from consecutive patients with a right-sided HP and a corresponding comparison group with conventional adenomas between 1993 and 2003. Two experienced gastrointestinal pathologists, blinded to polyp and clinical factors, reinterpreted the HPs using current SSA/P classification criteria. These HPs were classified as SSA/P when diagnostic histologic feature(s) were present in at least 3 crypts. Analyses, conducted on a per polyp basis, examined the factors associated with risk of individual HPs being reassessed as SSA/Ps as opposed to being confirmed as HPs. RESULTS: Of the 702 HPs (355 adults), 188 (26.8%) were reclassified as SSA/Ps. Predictors of HPs being reinterpreted as SSA/Ps included: size ≥5 mm [odds ratio (OR), 2.09; 95% confidence interval (CI), 1.34-3.26], proximal location (OR, 2.83; 95% CI, 1.69-4.74), synchronous adenomas with advanced pathology (OR, 2.61; 95% CI, 1.22-5.55) and ≥1 synchronous HPs (other than HP being reassessed) reclassified as SSA/Ps (OR, 11.76; 95% CI, 6.75-20.49). CONCLUSIONS: Because HP versus SSP is not very reproducible the predictors of SSA/P that we identified, including size, location, and synchronous lesions, can offer some additional help to endoscopists when determining surveillance intervals in patients with serrated polyps. In addition, observed association between SSA/P with advanced conventional neoplasia (but not low-grade adenomas) suggests 2 distinct groups of patient predisposition, one with both advanced conventional and important serrated precursors (SSA/P) and the other largely restricted to nonadvanced conventional adenomas and HPs only. Whether the association reported here has to do with SSA/P diagnosis per se or generally larger size of SSA/P remains to be determined in future studies.