Risk scoring system for the preprocedural prediction of the clinical failure of peroral endoscopic myotomy: a multicenter case–control study

Abstract Background  Peroral endoscopic myotomy (POEM) is effective for the management of achalasia and its variants; however, it can be ineffective in some patients. We aimed to develop and validate a risk scoring system to predict the clinical failure of POEM preoperatively. Methods  Consecutive patients who underwent POEM in 14 high volume centers between 2010 and 2020 were enrolled in this study. Clinical failure was defined as an Eckardt score of ≥ 4 or retreatment. A risk scoring system to predict the short-term clinical failure of POEM was developed using multivariable logistic regression and internally validated using bootstrapping and decision curve analysis. Results  Of the 2740 study patients, 112 (4.1 %) experienced clinical failure 6 months after POEM. Risk scores were assigned for three preoperative factors as follows: preoperative Eckardt score (1 point), manometric diagnosis (−4 points for type II achalasia), and a history of prior treatments (1 point for pneumatic dilation or 12 points for surgical/endoscopic myotomy). The discriminative capacity (concordance statistics 0.68, 95 %CI 0.62–0.72) and calibration (slope 1.15, 95 %CI 0.87–1.40) were shown. Decision curve analysis demonstrated its clinical usefulness. Patients were categorized into low (0–8 points; estimated risk of clinical failure < 5 %) and high risk (9–22 points; ≥ 5 %) groups. The proportions of clinical failure for the categories were stratified according to the mid-term outcomes (log-rank test, P  < 0.001). Conclusions  This risk scoring system can predict the clinical failure of POEM preoperatively and provide useful information when making treatment decisions.