Sa1067 Rates and Predictors of Progression in Barrett's Esophagus With Low Grade Dysplasia: Results From a Prospective U.S. Registry

BACKGROUND Estimates of progression in BE-LGD are limited by heterogeneity in pathological diagnoses. While recent data show that ablation in BE-LGD reduces risk of progression to HGD/EAC, they are limited by high rates of progression not seen in most studies. Identification of risk factors for progression in BE-LGD may help in selecting subjects for endoscopic therapy. We aimed to assess the rates and predictors of progression in a cohort of BE-LGD subjects who were part of a large prospective BE-EAC registry in a tertiary care center. METHODS Subjects with a histologic diagnosis of LGD (made by expert GI pathologists) were identified from a prospective BE registry at our institution. The registry database has records of demographic details, endoscopic findings and histologic data from esophageal biopsies. Index date was defined as the first date of LGD diagnosis. Progressors were defined as BE-LGD subjects who developed HGD / EAC more than 12 months after the index date. Risk factors assessed included demographic variables, family history of BE/ EAC, GERD symptoms, duration of BE, medication use, endoscopic findings and histological findings (multifocal vs unifocal LGD, H pylori status in gastric biopsy). Univariate and multivariable analyses were performed to identify predictors of progression using Cox Proportional Hazards models. RESULTS 353 BE-LGD subjects were identified, of which 337 were included after excluding those with missing data. The mean (SD) age of these subjects was 63.2 (11.2) years. 283 (84 %) were males. Baseline characteristics are summarized in table 1. Of the 337 BE-LGD patients, 21 (6.2 %) subjects progressed to HGD / EAC. The mean (SD) follow up of subjects was 7 (5) years. The annual incidence of HGD/EAC in subjects with BE-LGD was 0.8%. Univariate and multivariable predictors of progression to HGD/ EAC are presented in table 2. On univariate analysis, longer BE segment length and presence of nodularity increased risk of progression, while a longer follow up duration after LGD diagnosis reduced the risk of progression. Multifocal dysplasia (dysplasia at multiple levels) approached statistical significance as a predictor of progression. On multivariable analysis, increased BE segment length remained a predictor of progression while older age and longer LGD follow up (likely reflecting persistent LGD) reduced the risk of progression. CONCLUSIONS: In this well-defined cohort of BE-LGD patients with all histology confirmed by expert GI pathologists, progression rates for LGD subjects were substantially lower than those reported in some European studies. Longer BE segment length and younger age at LGD diagnosis predicted a higher risk of progression. These subjects may be candidates for intensive surveillance or endoscopic therapy.