Long-term outcomes of surveillance or endoscopic therapy for low-grade dysplastic Barrett's according to a selective...

Current European and American guidelines conflict in their recommendations for surveillance versus endoscopic therapy for low-grade dysplastic Barrett's (LGD). We aimed to evaluate the performance of a selective management algorithm and provide real-world outcomes.

Data on 497 patients with dysplastic Barrett's were collected prospectively between 2008 and 2022 at a Barrett's referral unit. LGD was defined as confirmation of LGD by an expert gastrointestinal pathologist. Persistent unifocal LGD or multifocal LGD were considered high-risk features for progression and patients underwent endoscopic eradication therapy (EET). Patients with non-persistent unifocal LGD were deemed low-risk and were surveilled. Primary outcome was progression rate to high grade dysplasia or neoplasia.

A total of 135 patients had LGD (median [interquartile range] follow up: 4.8 years [1.0-7.1]): 22 patients met low-risk criteria and were surveilled (LR-S), eight patients met high-risk criteria and were surveilled (HR-S; patient preference n = 4, medical comorbidities n = 4), and 105 patients met high-risk criteria and underwent EET (HR-EET). Progression rates were similar between the LR-S and HR-EET cohorts (4.5% [n = 1/22] vs. 6.7% [n = 8/105]; P = 0.43). The HR-S group had a significantly higher progression rate (25% [n = 2/8]; P = 0.04). Univariable analysis showed reflux esophagitis (sub-distribution hazard ratio 3.21, 95% confidence interval 1.02-10.1, P = 0.04) was associated with progression risk in the high-risk LGD cohort only.

This selective management algorithm for LGD is safe. Surveillance is appropriate in low-risk LGD patients. Patients with high-risk features who are surveilled and/or have reflux esophagitis may have an increased progression risk and should undergo EET with optimized acid suppression therapy.