Clinical feasibility of gastric endoscopic submucosal dissection in patients on glucocorticoids or immunomodulators: Propensity-score-matched study.

Background and study aims: Evidence on gastric endoscopic submucosal dissection (ESD) under glucocorticoids or immunomodulators (GC/IM) is limited. We evaluated whether GC/IM use affects gastric ESD outcomes. Patients and methods: We retrospectively analyzed 411 consecutive ESDs (April 2017-April 2022). GC/IM users (n = 32) were compared with controls (n = 379); 1:3 propensity-score matching yielded 27 vs 81 patients. The primary outcome was overall complications, defined as a composite of pain, fever, delayed bleeding, and delayed perforation. Secondary outcomes were each component, intra-procedure perforation, hospital stay, and use of symptom-directed treatments (analgesics, antipyretics, antibiotics). Results: Overall complications were more frequent with GC/IM than controls (44.4% vs 21.0%; P = 0.024; relative risk [RR] 2.11, 95% confidence interval 1.16-3.84), driven by pain (40.7% vs 18.5%; P = 0.035) and fever (11.1% vs 1.2%; P = 0.047). Rates of major complications did not differ (delayed bleeding 3.7% vs 2.5%; delayed perforation 0% vs 0%). Intra-procedure perforation was numerically higher (7.4% vs 3.7%) without significance. Median (interquartile range) hospital stay showed a small, non-significant difference (9 [7-12] vs 8 [7-9] days; P = 0.064). Symptom management was used more often with GC/IM (analgesics 25.9% vs 3.7%, P = 0.002; antipyretics 7.4% vs 2.5%, P = 0.270; antibiotics 7.4% vs 3.7%, P = 0.597). Conclusions: In patients receiving GC/IM, gastric ESD was associated with a higher incidence of minor, clinically managed events-chiefly pain and transient fever-whereas major complications remained uncommon. With close monitoring and prompt symptom-directed care, gastric ESD appeared clinically feasible, albeit with slightly greater resource use and observation time.