523 VETERANS WITH CONFIRMED PERSISTENT BARRETT’S-ASSOCIATED LOW-GRADE DYSPLASIA ARE AT INCREASED RISK OF PROGRESSION TO HIGH GRADE DYSPLASIA AND ADENOCARCINOMA

The incidence and prevalence of certain diseases and neoplasms among veterans are known to deviate from the general population; perhaps because of service associated exposures or lifestyle risk factors. Studies in the general population have shown that the number of pathologists confirming Barrett's esophagus (BE) associated low grade dysplasia (LGD) is strongly associated with the risk of progression. However, the risk of progression of persistent confirmed LGD (cLGD) has not been well studied in the US population. To characterize the incidence of progression in confirmed persistent BE-associated LGD in US veterans. All patients with BE and cLGD were identified at Minneapolis Veterans Affairs Health Care System between January 2006 and December 2016. A diagnosis of cLGD was established by consensus of 2 to 7 pathologists including an expert GI pathologist. Information was collected on demographics, past medical history, EGD findings, histopathology and lifestyle risk factors. Evidence of persistent cLGD was defined as cLGD on subsequent endoscopy greater than 3 months after initial diagnosis. The primary endpoint of the study was the development of high-grade dysplasia (HGD) or esophageal adenocarcinoma (EAC). Univariate logistic regression analysis was used to assess the association between outcomes and risk factors for progression. A Kaplan-Meier curve was used to evaluate progression-free survival then compared using Log-rank test. Among 72 patients with cLGD, 18 (25%) developed HGD/EAC during a median follow up of 3.17 years (interquartile range, 1.15-5.24). The incidence of HGD/EAC in non-persistent cLGD was 6.96 cases per 100 patient-years (95% confidence interval [CI], 4.12-10.99). In patients with persistent cLGD, the incidence of HGD/EAC was 11.59 cases per 100 patient-years (95% CI, 5.99-20.24), and there was a 5-fold increased risk of malignant progression (odds ratio 5.20; 95% CI, 1.62-16.70). Figure 1 shows the statistical difference in progression-free survival between persistent and non-persistent cLGD (P < 0.001). All patients with neoplastic progression had hiatal hernia (n = 18). Other traditional risk factors including obesity, length of BE, and tobacco use did not show a significant association with progression (Table 1). Among veterans, the risk of progression in patient with BE containing persistent cLGD was 5-fold higher than in those without persistent cLGD. Progression rates for persistent and non-persistent cLGD were 12% and 7% per year, respectively. Regardless of persistence, cLGD progression rates in US veterans are substantially higher than reported rates in unconfirmed LGD.Table 1Demographic and clinicopathological factors associated with the progression of cLGDParaemterProgressors (n = 18)Non-progressors (n = 54)P-valueAge (years)65.5665.200.88Body Mass Index (kg/m2)29.8129.530.81Persistence12/1815/540.006Hiatal Hernia18/1843/540.055Alcohol Use3/1814/540.53Tobacco Use11/1837/530.56Diabetes Mellitus6/1814/540.56Cirrhosis1/173/541.00Gastroesophageal Reflux Disease14/1845/540.72Long Segment BE10/1826/451.00 Open table in a new tab