Impact of a diagnosis of ‘low-grade dysplasia’ in patients with Barrett’s esophagus

Summary Barrett’s esophagus (BE) is the precursor to esophageal adenocarcinoma (EAC). Progression to high-grade dysplasia (HGD) /EAC can be directly from non-dysplastic BE or via low-grade dysplasia (LGD). There is a lack of consensus about the implications of LGD diagnosis on progression risk and whether modifying risk progression with endoscopic interventions such as radiofrequency ablation (RFA) is appropriate. The aim of this study was to determine the clinical implications of LGD and the cost-effectiveness of different management strategies. Outcomes from a large single-center prospective BE surveillance database were retrospectively analyzed. All patients with BE in a structured surveillance program were included. The cohort was divided into three groups: non-dysplastic BE throughout, LGD at surveillance entry, and LGD developing during surveillance. Each group’s annual incidence of progression to HGD/EAC was calculated per 100 person-years. Outcomes were then applied within a health economic model for health economic analysis of ongoing endoscopic surveillance versus RFA of LGD to identify the most cost-effective management strategy. Nine hundred fourteen patients were included; 727 had non-dysplastic BE, 97 had LGD at entry, and 90 developed LGD during surveillance. Total surveillance time was 5212 person-years. Forty-six (5. 0%) patients progressed to HGD/EAC, at an annual progression rate of 0. 9 per 100 person-years. For subgroups, the progression rates to HGD/EAC were 0. 6 per 100 person-years for non-dysplastic BE, 1. 0 for LGD at entry, and 2. 2 for LGD during surveillance (P 0. 0001). The most cost-effective management strategy was RFA if LGD was identified under surveillance, with an estimated cost per quality-adjusted life year gained being AU26, 763. Overall progression from LGD to HGD/EAC was comparable to most previous studies. RFA is a cost-effective management strategy for BE once LGD arises during surveillance.