Margaret C. Liu, MD1, Joseph C. Anderson, MD2, Lynn F. Butterly, MD3, Christina Robinson, MS2, Amitabh Strivastava, MD4; 1Dartmouth-Hitchcock Medical Center, Lebanon, NH; 2Dartmouth College Geisel School of Medicine, Hanover, NH; 3Dartmouth College Geisel School of Medicine, Lebanon, NH; 4Brigham & Women's Hospital, Boston, MA
Introduction: Distinguishing HPs from SSPs can be challenging for pathologists. Our goal was to use New Hampshire Colonoscopy Registry (NHCR) data to determine if diagnoses of HP versus SSP as reported by pathologists have captured distinctions of HPs (especially large (> 1 cm) HPs) vs. SSPs so that they define 2 clinically distinct groups with different colorectal neoplasia risk profiles and risks for metachronous large (> 1 cm) SPs. Methods: We divided NHCR patients (2004-18) with only SPs into 4 groups by most advanced SP: non-significant HPs (small (< 1 cm) rectosigmoid), small HPs (< 1 cm proximal to sigmoid), large ( > 1 cm) HPs, and SSPs. We compared associations for colorectal neoplasia risk factors (age, sex, family history, BMI, and smoking) for adults with HPs versus SSPs. We excluded exams that were incomplete or with poor bowel prep, IBD, and polyposis syndromes. We examined metachronous risk on a subset of patients with a 2nd exam > 12 months after index– 1st examining absolute risk in the 4 groups and then using 2 logistic models to predict metachronous risk of large SPs; first in the 4 groups and then across polyp factors; histology (SSP vs HP), size (< 1 vs >1 cm), and location (proximal vs distal to splenic flexure). Results: We included 22129 patients (average age 54.4 (+ 9.9) yrs, 50.0 % male). Smoking (22.9% vs 13.2%; p < 0.001) and synchronous advanced neoplasia (12.9% vs 7.1%; p < 0.001) were associated with large HPs vs SSPs and female sex (54.5% vs 48.3%; p < 0.001) was associated with SSPs (Table 1). There were no significant differences for metachronous advanced neoplasia risk. SSPs were associated with a higher risk for future large SPs in the subset (n=3493) with a 2nd exam as compared to non-significant HPs (OR=8.29(95%CI:5.25-13.09)) (Table 2). Large HPs also had an increased risk but the OR was lower than SSPs(OR=4.23(95% CI:2.15-8.33)). In the 2nd model, which adjusted for histology, size > 1 cm (OR 1.71, CI 1.16-2.53) and proximal location (OR 2.36, CI 1.61-3.46) were predictors of future large SPs. Discussion: Our data suggest that patients with polyps histologically identified by pathologists as SSPs have a higher risk for future large SPs, suggesting that histologic diagnosis in practice risk stratifies adults with SPs. However, our second logistic model suggests that proximal location and size > 1 cm are additional important features, even after adjusting for histology. Thus, our data suggest that there may still be overlap between these HPs and SSPs.
Table 1 showing baseline risk factors for the 4 SP groups as well as absolute risk for metachronous lesions in subset of patients with second exam.
Table 2 showing increased risk for metachronous large SPs in SSP group and the importance of size and proximal location of SPs even after adjusting for histology.
Disclosures: Margaret Liu indicated no relevant financial relationships. Joseph Anderson indicated no relevant financial relationships. Lynn Butterly indicated no relevant financial relationships. Christina Robinson indicated no relevant financial relationships. Amitabh Strivastava indicated no relevant financial relationships.