Endoscopic resection-related duodenal strictures: prevalence, risk factors, management, and outcomes

Background Endoscopic resection (ER) of large (≥15 mm) duodenal laterally spreading lesions (D-LSLs) is now standard of care. Data on prevalence, risk factors, and management of strictures after ER of D-LSL are absent; we sought to evaluate this in a large tertiary referral cohort. Methods A prospective cohort of ER-treated D-LSLs in an expert center was retrospectively analyzed. Strictures were considered "severe" if patients experienced obstructive symptoms, "moderate" if a standard gastroscope (diameter 9.9 mm) could not pass the stenosis, or "mild" if there was resistance on successful passage. When necessary, dilation was performed every 2-4 weeks until scope passage without resistance. Primary outcomes included stricture prevalence, risk factors, and management. Results Over 193 months until February 2023, 246 lesions in 239 patients were included (median age 70 years [interquartile range (IQR) 63-77]; 51.5% male; median lesion size 35 mm [IQR 22.5-47.5]). Overall, 30 resections (12.2%) resulted in stricture (14 mild [46.7%], 4 moderate [13.3%], and 12 severe [40%]), and 18 (7.3%) required balloon dilation (median 2 sessions [IQR 0-6]). On multivariable analysis, post-ER defect circumference ≥80% was the strongest independent predictor of stricture formation (OR 60.2, 95%CI 17.5-254.2; P < 0.001). Incidence of stricture formation with ER defects of ≥80%, 60%-79%, and <60% was 72.2% (26/36), 12.5% (4/32), and 0% (0/178), respectively. All severe strictures occurred in ER defects ≥80%. Conclusions ER defect circumference strongly predicted stricture formation following ER of D-LSLs. These findings can be used to guide informed consent and post-procedural care.