Safety of endoscopy after recent bevacizumab use in metastatic colorectal cancer.

BACKGROUND AND AIMS: Bevacizumab improves metastatic colorectal cancer (mCRC) survival but increases risks including GI perforation. This study compared 30-day postendoscopy (EGD/colonoscopy) perforation and bleeding risks between mCRC patients receiving recent (≤1 month) bevacizumab versus other active chemotherapy.

METHODS

TriNetX, a multinational federated electronic health network database, was used to create and analyze 1:1 propensity score-matched cohorts of patients who underwent endoscopy for either diagnostic purposes (n = 396 total postmatch) or therapeutic purposes (n = 1306) as well as combined (n = 2100). Time-to-event analysis on 30-day perforation risk and 30-day postprocedural bleeding was conducted using Kaplan-Meier analysis on matched cohorts as well as adjusted Cox proportional hazards models on unmatched cohorts for sensitivity analysis.

RESULTS

Baseline covariates were balanced post matching. No significant difference in 30-day perforation risk was found (Combined: hazard ratio [HR], 0.93; P = .791; Diagnostic: HR, 2.33; P = .143; Therapeutic: HR, 0.96; P = .925). Bevacizumab was associated with a significantly higher 30-day postprocedural bleeding risk in the Combined (matched HR, 1.28; P = .006) and Therapeutic (matched HR, 1.45; P = .001; adjusted HR, 1.46; P < .001) analyses but not in the Diagnostic subgroup (matched HR, 1.03; P = .894; adjusted HR, 0.98; P = .921).

CONCLUSIONS

Recent bevacizumab exposure was not associated with increased 30-day postendoscopy perforation risk compared with other active chemotherapy, but was associated with higher bleeding risk, primarily in therapeutic (specifically EGD) procedures.