Background: Heller myotomy (HM) is one of the most effective treatments for esophageal achalasia and has a five-year symptom free rate of 77-90%. However, failures do exist and the success rate tends to decrease with longer follow-up. The efficacy of traditional rescue treatments is limited. The advent of peroral endoscopic myotomy (POEM) brings new hope to these patients. Several studies have reported outcomes of POEM in these patients, but are mostly limited by small sample sizes and/or short follow-up. We present here outcomes from the largest Western series of salvage POEM in patients with prior Heller with 107 patients followed for a median of 5.8 years and comparison of outcomes to POEMs in treatment-naïve patients by the same operator. Methods: All POEM data is entered into a prospectively maintained database. Consecutive POEM patients enrolled between 10/2009 and 11/2021 were analyzed. Baseline patient characteristics, procedure details, hospital course, follow up and adverse events (AEs) were recorded and compared between post-HM and treatment-naïve patients. Results: 107 post-HM and 421 treatment-naïve patients received POEM during the study period. Baseline characteristics were comparable (see Table 1) except that post-HM patients had longer disease duration (144 vs 24 months, p<0.001), HRM type 1 vs 2 predominance (<0.001), more end-stage/sigmoid disease (42% vs 11% p<0.001), slightly lower pre-POEM Eckardt score (7 vs 8 p<0.001) and, as expected, higher pre and post POEM EndoFLIP diameter and distensibility (p<0.001). Procedure duration was longer in post-HM patients (60 vs 46 min, p<0.001). No difference in hospital LOS (median 2 days) or adverse events (<2% overall; no aborted POEMs/surgical conversions, no surgical, IR or other interventions or drains) (Table 1). Similar overall success of 96.0% vs 94% at median follow-up of 69 vs 60 months. We analyzed predictors of salvage POEM failure (Table 2). The analysis was limited by the low number of failures (6/107). On univariate analysis, plausible factors such as Heller-POEM interval, disease duration, GERD metrics and disease stage proved unhelpful. The most significant predictor was a high pre-POEM EndoFLIP distensibility at 50 ml, median of 6.9 [5.9-7.8] in failures vs 3.6 [2.3-5.5] in responders (p=0.01), followed by type 3 achalasia (50% in failures vs 12% in responders, p=0.044). Patients with EndoFLIP Distensibility ≥6 appeared unlikely to benefit from POEM even after the careful vetting based on our extensive experience and comprehensive evaluation including Barium and GERD studies and detailed endoscopic assessment. Conclusion: Salvage POEM after failed Heller myotomy is safe, highly effective and durable. These excellent outcomes, however, may be predicated on a high level of POEM expertise and careful patient selection for which EndoFLIP appears particularly relevant.

Table 1

Table 2

Collin Brathwaite

Winthrop University Hospital

Appearances