Prevalence of Serrated Polyp Subtypes and Predictors of Advanced Serrated Histology in a Large Colorectal Cancer Screening Cohort
JT Hetzel, S Cerda, J Coukos, FA Farraye, CS Huang, K Omstead, S Hammond, S Yang, MJ O'Brien. Boston University Medical Center, Boston, MA
Background: The aim of the present study was to determine the prevalence of serrated polyps in a large average risk screening cohort using prospectively defined classification criteria and to clarify demographic risk factors associated with more advanced pathology.
Design: Pathology reports were reviewed for all screening colonoscopies at Boston Medical Center from 2006 to 2008. Histological diagnoses were abstracted and classified as adenomas, serrated polyps and other. All cases classified as serrated polyps were reviewed jointly by 2 GI pathologists using prospectively defined classification criteria (Torlakovic et al) and reclassified: Microvesicular Serrated polyp (MVSP), Goblet Cell Serrated Polyp (GCSP), Sessile Serrated Adenoma (SSA), or Serrated Adenoma nos (dysplastic serrated polyp). Prevalence was summarized as patients with at least one polyp per 100 patients. Predictors of histological diagnosis among serrated polyps were modeled with multivariate logistic regression.
Results: The study included 7116 average risk screening colonoscopies. Specimens of 4272 polyps were received from 2365 colonoscopies (33.2%). Prevalence of all polyps is shown in table 1. Among serrated polyps the distribution of subtypes was as follows: GCSP 51.8% (n=575); MVSP 40.6% (n=451); SSA 6.4% (n=71); SA 1.2% (n=13). In the multivariate analysis predictors of SSA histology among serrated polyps were proximal location (Adj OR 13.7; CI 7.8 - 24.1) and larger size (Adj OR 6.2; CI 2.2 - 17.7); predictors of dysplasia were age in years (adj OR 1.1; CI 1.1 - 1.2), larger size (OR 8.7; CI 1.7 - 44.6) and female gender (marginally significant).
|Proximal||Distal||Proximal and distal|
|% Prevalence||% Prevalence||N||% of Serrated||% Prevalence|