Non-steroidal Anti-inflammatory Drugs and Timing-specific Anastomotic Leak Risk Following Minimally Invasive Proctectomy: A Retrospective Cohort Study.

Abstract

Reports regarding postoperative administration of non-steroidal anti-inflammatory drugs (NSAIDs) and anastomotic leak (AL) after rectal surgery are conflicting. Distinctions between early and late AL have often been overlooked, leaving it uncertain whether the NSAID-related effects differ according to the timing of leak presentation. This retrospective cohort study included 1663 patients undergoing minimally invasive rectal resection (2010-2021). NSAID exposure was defined as one or more dose within the first postoperative day. AL was classified as early (≤ 6 days) or late (> 6 days). Propensity score matching (PSM) and multivariable regression were used to assess associations between NSAIDs and AL. Subgroup analyses explored NSAID type/selectivity and dosing effects. The overall AL rate was 8.7% (n = 144). Early NSAID administration was associated with a higher risk of early AL in both unadjusted (7.7% vs. 4.7%, P = 0.030) and PSM-adjusted (8.4% vs. 4.6%, P = 0.041) analyses but not with late AL. This effect was mainly attributable to non-selective NSAIDs (odds ratio [OR] 1.717; 95% confidence interval [CI] 1.128-2.615, P = 0.012) and multiple NSAID doses (OR 1.687; 95% CI 1.104-2.576, P = 0.016). Perioperative bleeding was also more common in patients using NSAIDs (4.0% vs. 1.2%, P = 0.003). Subgroups identified a heightened NSAID-associated risk of AL in male patients (OR 1.836; 95% CI 1.190-2.832, P = 0.006) and patients without diverting stomas (OR 1.689; 95% CI 1.086-2.627, P = 0.020). Early postoperative NSAIDs, particularly non-selective agents and repeated dosing, were associated with early AL after minimally invasive rectal surgery but not late AL. This study highlights the necessity of classifying AL by timing, providing a crucial new perspective for future research on NSAID effects.