EPD511 - Perioperative Non-Steroidal Anti-Inflammatory Drug Use and Risk of Colorectal Anastomotic Leak: An Updated Systematic Review and Meta-Analysis

Purpose/Background: Non-steroidal anti-inflammatory drugs (NSAIDs) are integral to multimodal, opioid-sparing analgesia in enhanced-recovery programs for colorectal surgery. Despite their analgesic benefits, concerns persist that NSAIDs may impair anastomotic healing and increase the risk of anastomotic leak (AL), which has limited their use in clinical practice. Previous meta-analyses have yielded conflicting results. Since the latest reviews, several large registry and multicentre cohort studies have been published, contributing over 10,000 additional patients. Therefore, this systematic review and meta-analysis aimed to reassess the impact of perioperative NSAID use and AL after colorectal resections, incorporating updated evidence.

Methods/Interventions: EMBASE, Medline, and Central databases were searched from inception to September 2025 for studies comparing postoperative NSAID exposure versus no NSAID exposure in adults undergoing colorectal resection with an anastomosis. Risk ratios (RRs) with 95% confidence intervals (CIs) were pooled using a random-effects model (Mantel–Haenszel method). Heterogeneity was assessed using the I² statistic, and prespecified subgroup analyses were performed for study design (randomized vs. non-randomized) and NSAID class (non-selective vs. COX-2 selective). Risk of bias was assessed using Cochrane RoB 2 for randomized trials and ROBINS-I for observational studies.

Results/Outcomes: Twenty-two studies including 3 randomized controlled trials and 19 observational studies involving 43,226 patients (14,186 NSAID-exposed; 29,080 NSAID-unexposed) were identified. In the overall pooled analysis, NSAID use was not associated with increased risk of AL (RR 1.16, 95% CI 0.93–1.45; p = 0.20; I² = 75%). Among subgroup analyses of randomized (RR 2.12, 95% CI 0.65–6.97; p = 0.21; I² = 0%) and non-randomized (RR 1.14, 95% CI 0.91–1.43; p = 0.27; I² = 78%) studies, the pooled estimates remained nonsignificant. Subgroup analyses by drug class showed no differences in the NSAID-exposed and -unexposed groups: non-selective NSAIDs (9 studies, 7,748 patients) RR 1.21 (95% CI 0.95–1.55; p = 0.13; I² = 0%), and COX-2 selective agents (6 studies, 3,863 patients) RR 0.76 (95% CI 0.40–1.44; p = 0.40; I² = 40%). Most registry-based studies were judged moderate risk of bias, single-centre retrospective cohorts serious risk, and all RCTs some concerns due to small size and limited reporting of allocation concealment. Sensitivity analysis restricted to low- or moderate-risk studies produced directionally unchanged results.

Conclusion/Discussion: Across more than 43,000 patients, perioperative NSAID use was not associated with an increased risk of colorectal anastomotic leak. The overall certainty of evidence was low, limited by the small number of randomized trials, predominance of observational data, and high heterogeneity. Although not statistically significant, the divergence in effect direction between non-selective and COX-2 selective NSAIDs suggests potential class-specific differences. Considering their well-established opioid-sparing benefits, these findings support the use of COX-2 selective NSAIDs for perioperative analgesia in colorectal surgery. However, a definitive, adequately powered randomized controlled trial is warranted to confirm these class-specific effects and establish the safety of perioperative NSAID use with greater certainty.