S409 The Prevalence of Dysplasia in Ultrashort Barrett's Esophagus in a United States Veteran Population

Abstract

Introduction: Surveillance of Barrett’s esophagus (BE), the only known precursor to esophageal adenocarcinoma (EA), is the basis for secondary prevention of EA. Length of BE has been associated with EA risk, and current guidelines recommend against biopsies of ultrashort BE (i.e., BE length < 1cm), however this is based on limited evidence. Our study aims to characterize the prevalence of dysplasia in ultrashort BE. Methods: We conducted a retrospective cohort study at the Houston VA of consecutive patients with new BE diagnosis from 11/1990-1/2019 with follow-up through 12/2021. We classified ultrashort BE as length < 1cm at initial BE-diagnosing (index) EGD. We defined persistent ultrashort BE as those with ultrashort BE on index EGD and all follow-up EGDs through the entire study period, requiring 2 or more EGDs. We calculated the prevalence of any dysplasia (indefinite [IND], low-grade [LGD], and high-grade [HGD]) and definite dysplasia (LGD and HGD). We examined the association of age, race, gender, body mass index (BMI), smoking history, and alcohol history with dysplasia in ultrashort BE and BE ≥1cm with chi-square test. Results: Our cohort had 741 patients with BE. The mean follow-up time was 4.6 years (standard deviation, 5.5 years). Ultrashort BE was present in 20.2% (n=150) of BE patients. Of the 150 with ultrashort BE, 131 underwent at least two EGDs. Persistent ultrashort BE was present in 92 patients with BE (12.4%). The prevalence of any dysplasia and definite dysplasia were 24.0% (n=36, 29 IND, 2 LGD, 5 HGD) and 4.7% (n=7) among those with ultrashort BE, and 52.6% and 23.8% respectively among those with BE ≥1cm. Of those with ultrashort BE, 4 (2.7%) developed any dysplasia after the index EGD. Of those with persistent ultrashort BE, the prevalence of any dysplasia was 21.7% (n=20, 16 IND, 2 LGD, 2 HGD), while the prevalence of definite dysplasia was 4.4% (n=4). All cases of dysplasia in persistent ultrashort BE were diagnosed at the index EGD. Normal BMI was associated with any dysplasia among those with ultrashort BE (p-value 0.003), while non-Hispanic White race/ethnicity was associated with any dysplasia among those without ultrashort BE (p-value 0.009; Table). Conclusion: While the prevalence of dysplasia is lower than that for BE ≥1cm, the prevalence of dysplasia in ultrashort BE in this study was as high as 24.0%. These patients may not be safely excluded from surveillance. Further studies should be conducted to confirm these findings. Table 1. - Sociodemographic and clinical characteristics of any dysplasia (indefinite for dysplasia, low-grade dysplasia, and high-grade dysplasia) among those with and without ultrashort Barrett’s esophagus Ultrashort BE < 1cm at Index EGD BE ≥1cm at Index EGD No DysplasiaN=114N (%) Any dysplasiaN = 36N (%) p-value No DysplasiaN = 280N (%) Any DysplasiaN = 311N (%) p-value Age < 60 years 45 (39.5%) 9 (25.0%) 0.115 94 (33.6%) 104 (33.4%) 0.973 60+ years 69 (60.5%) 27 (75.0%) 186 (66.4%) 207 (66.6%) Sex Male 106 (93.0%) 34 (94.4%) 0.759 269 (96.1%) 304 (98.1%) 0.147 Female 8 (7.0%) 2 (5.6%) 11 (3.9%) 6 (1.9%) Race Non-Hispanic White 86 (75.4%) 26 (72.2%) 0.47 216 (77.1%) 264 (85.9%) 0.009 African American 17 (14.9%) 4 (11.1%) 36 (12.9%) 16 (5.1%) Hispanic 11 (9.7%) 6 (16.7%) 26 (9.3%) 30 (9.7%) Other/Missing 0 (0.0%) 0 (0.0%) 2 (0.7%) 1 (0.3%) BMI Categories Normal (< 25) 18 (15.8%) 11 (30.6%) 0.003 52 (18.6%) 51 (16.4%) 0.878 Overweight (25-29.9) 38 (33.3%) 18 (32.1%) 106 (37.9%) 126 (40.5%) Obese (30+) 58 (50.9%) 7 (19.4%) 120 (42.9%) 132 (42.4%) Missing 0 (0.0%) 0 (0.0%) 2 (0.7%) 2 (0.6%) Smoking Never Smoker 29 (25.4%) 8 (22.2%) 0.813 78 (27.9%) 83 (26.7%) 0.429 Former Smoker 59 (51.8%) 18 (50.0%) 121 (43.2%) 150 (48.2%) Current Smoker 26 (22.8%) 10 (27.8%) 81 (28.9%) 78 (25.1%) Alcohol Use Never Alcohol use 36 (31.6%) 12 (33.3%) 0.862 128 (45.7%) 115 (37.0%) 0.088 Former Alcohol Use 28 (24.6%) 10 (27.8%) 58 (20.7%) 70 (22.5%) Current Alcohol Use 50 (43.9%) 14 (38.9%) 94 (33.6%) 126 (40.5%) Abbreviations: BE (Barrett's esophagus), cm (centimeter), BMI (body mass index)