Association Between Length of Barrett's Esophagus and Risk of High-grade Dysplasia or Adenocarcinoma in Patients Without Dysplasia

BACKGROUND & AIMS: It is not clear whether length of Barrett's esophagus (BE) is a risk factor for high-grade dysplasia (HGD) or esophageal adenocarcinoma (EAC) in patients with nondysplastic BE. We studied the risk of progression to HGD or EAC in patients with nondysplastic BE, based on segment length. METHODS: We analyzed data from a large cohort of patients participating in the BE Study-a multicenter outcomes project comprising 5 US tertiary care referral centers. Histologic changes were graded as low-grade dysplasia, HGD, or EAC. The study included patients with BE of documented length without dysplasia and at least 1 year of follow-up evaluation (n = 1175; 88% male), and excluded patients who developed HGD or EAC within 1 year of their BE diagnosis. The mean follow-up period was 5.5 y (6463 patient-years). The annual risk of HGD and EAC was plotted in 3-cm increments (<= 3 cm, 4-6 cm, 7-9 cm, 10-12 cm, and >= 13 cm). We calculated the association between time to progression and length of BE. RESULTS: The mean BE length was 3.6 cm; 44 patients developed HGD or EAC, with an annual incidence rate of 0.67%/y. Compared with nonprogressors, patients who developed HGD or EAC had longer BE segments (6.1 vs 3.5 cm; P < .001). Logistic regression analysis showed a 28% increase in risk of HGD or EAC for every 1-cm increase in BE length (P = .01). Patients with BE segment lengths of 3 cm or shorter took longer to develop HGD or EAC than those with lengths longer than 4 cm (6 vs 4 y; P = nonsignificant). CONCLUSIONS: In patients with BE without dysplasia, length of BE was associated with progression to HGD or EAC. The results support the development of a risk stratification scheme for these patients based on length of BE segment.