PROTON PUMP INHIBITOR USAGE AND BARRETT’S ESOPHAGUS AFTER PER-ORAL ENDOSCOPIC MYOTOMY AND HELLER MYOTOMY: PROPENSITY MATCHED ANALYSIS OF A LARGE MULTI-INSTITUTIONAL COHORT

Background:
Per-oral endoscopic myotomy (POEM) and Heller myotomy (HM) are endoscopic and surgical treatment options for achalasia. Post procedure gastroesophageal reflux disease (GERD) is common; however, the indications, role and rate of proton pump inhibitor (PPI) use in this population is not well defined. Furthermore, evidence suggests that endoscopic esophagitis rates are higher than symptomatic reflux in this population, hence the incidence of late GERD complications including Barrett’s esophagus (BE) needs to be studied, as these may be silent in many patients.
Methods:
The multi-institutional research network TriNetX was used to identify patients who underwent POEM or Heller myotomy procedures. PPI use and BE diagnosis were compared in the two cohorts after a lag time of 6 months and one year including only those patients who had documented follow up for each duration. After crude analysis, a propensity score matched (PSM) analysis was performed using demographic variables, history of GERD, esophagitis, obesity, and comorbidities as covariates to control for confounders. Standardized mean difference (SMD) >0.1 was considered positive for residual confounding. Aggregate counts <10 are not reported when utilizing this database. Cox proportional hazard analysis was performed to calculate Hazard ratios for BE to account for difference in follow up duration in cohorts.
Results:
A total of 1911 patients with POEM and 4824 with HM were identified. At baseline, patients with POEM were older, had higher rates of CVA, HTN, DM, and prior GERD. Hiatal hernia was more common in HM cohort. All studied characteristics were well balanced in the propensity matched cohorts (SMD <0.1 for all covariates)(Table 1).
High rates of documented PPI usage were observed in both cohorts. Higher PPI use was noted after POEM compared to HM after 1 year on crude (45.66% vs 32.27%; RR:1.41 95% CI:1.29-1.56) and PSM analysis (45.55% vs 30.01% RR:1.52, 95% CI:1.32-1.74). There was no difference in rates of BE in the two cohorts in matched (HR:0.55; 95% CI 0.14-2.22) or unmatched (HR:0.42; 95% CI 0.13-1.35) analysis after one year of procedure (Table 2).
Among patients who underwent EGD at least 6 months after procedure, </=10 patients (</=2.38%) in POEM cohort and 4.1% in HM cohort were diagnosed with BE. Patients with post-myotomy BE were older (p<0.01), had higher rates of pre procedure hiatal hernia (p<0.01), GERD (p<0.01) and esophagitis (p<0.01); no difference in smoking or obesity was noted.
Conclusions:
A sizable cohort of patients with documented PPI use after POEM as well as HM was found in this analysis of a multi-institutional sample, with higher rates of PPI usage noted in the POEM cohort. Rate of BE in patients undergoing any myotomy was higher than accepted thresholds for BE screening; EGD for BE surveillance should be considered in the post-myotomy population.