Predictors and Quantitative Assessment of Incomplete Response After Radiofrequency Ablation for Dysplastic Barrett's Esophagus: Analysis of Randomized Sham-Controlled Clinical Trial (The AIM Dysplasia Trial)

Abstract

Background: Radiofrequency ablation (RFA) has been shown to completely eradicate dysplastic intestinal metaplasia (IM) in most patients, yet residual IM may persist in some. Aims: The primary endpoint for RFA therapy is complete response-IM (CR-IM, no histological evidence of IM). We sought to describe pt characteristics related to incomplete response-IM (IR-IM, any residual IM). We also assessed dysplasia grade, and extent/location of any residual IM. Methods: We enrolled 127 pts with dysplastic BE (63 HGD, 64 LGD) in a multi-center trial of RFA. Pts were randomized 2:1 (RFA vs. sham) then biopsied q 3 or 6 mo, with centralized path review. RFA was performed until CR-IM or max 4 sessions. Results: 52 pts (35 RFA, 17 sham) have evaluable 12 mo histology. This sub-analysis of the RF group compares CR-IM to IR-IM at 12 mo. The groups had similar hiatal hernia size. IR-IM had a longer pre-treatment period with dysplasia (p < 0.05). They were also older and had higher BMI, more years with BE, longer BE cm, and more multi-focal dysplasia, but given the small sample size of IR-IM, none of these was significant (table). All IR-IM pts had downgrading of dysplasia. For the 3 IR-IM pts with baseline HGD, the worst grade of residual IM was non-dysplastic (1), indefinite (1), or LGD (1). For the 3 IR-IM pts with baseline LGD, all were downgraded to non-dysplastic IM. Of the 6 IR-IM pts at 12 mo, 4 had a single-level IM focus, while 2 had multi-level disease. Five of 6 IR-IM pts had IM only within 1 cm of the top of gastric folds (TGF), while 1 pt had more proximal IM (4-5 cm from the TGF). One IR-IM pt had persistent GERD esophagitis, 1 had ibuprofen-induced ulceration, and 1 had a baseline stricture preventing focal balloon contact. Conclusions: All IR-IM had downgrading of dysplasia and substantial reduction of IM burden. IR-IM pts had a longer pre-treatment period with dysplasia than CR-IM. IR-IM also had insig increases in age, baseline BE length, BMI and % multi-focal. Follow-up RFA is planned for these pts, with the goal to eliminate residual disease. Tabled 1 CR-IM IR-IM Patients (n, %) 29 (83%) 6 (17%) Male (n, %) 24 (83%) 5 (83%) Age years (median, IQR) 64 (59-69) 70.5 (64-74) BMI (median, IQR) 26.5 (25-31) 28.9 (27-32) Years with BE / dysplasia (median) 3.4 / 0.6 6.0 / 2.2 ∗ p < 0.05 Baseline Prague (CM) median 1.5 / 4.0 4.5 / 6.5 Baseline multifocal dysplasia (n, %) 21 (72%) 6 (100%) Baseline HGD (n, %) 9 (31%) 3 (50%) RFA procedures (median) 4 4 12 mo fragments (n) 806 266 12 mo fragments with IM (n, % total) 0 (0%) 25 (9.3%) ∗ p < 0.05 Open table in a new tab