Oncologic impact of anastomotic leakage after curative rectal cancer resection: early vs. late leakage

To evaluate whether anastomotic leakage and its timing affect oncologic outcomes after curative rectal cancer surgery. We retrospectively analyzed patients with stage 0–III rectal adenocarcinoma undergoing curative resection with anastomosis during 2012–2020. Leakage was classified as early (EL; ≤ 30 days), late (LL; > 30 days), or no leakage (NL). Disease-free survival (DFS) and overall survival (OS) were evaluated using multivariable time-dependent Cox models with direct EL–LL comparisons. Local recurrence (LR) and distant metastasis (DM) were assessed using competing-risk methods. Among 4,338 patients, 140 developed leakages (EL, n = 71; LL, n = 69). Both EL and LL had significantly higher 5-year cumulative incidences of LR and DM than NL. In adjusted models, EL was associated with worse DFS (HR 1.64, 95% CI 1.10–2.43), whereas LL was not (HR 1.29, 95% CI 0.77–2.18). Both EL (HR 1.85, 95% CI 1.16–2.95) and LL (HR 1.99, 95% CI 1.16–3.42) were associated with worse OS. In direct adjusted comparisons using EL as the reference, LL did not differ significantly from EL for DFS (HR 0.79, 95% CI 0.41–1.51) or OS (HR 1.07, 95% CI 0.53–2.18). Anastomotic leakage was associated with adverse long-term oncologic outcomes. EL and LL showed higher cumulative incidences of LR and DM than NL. In time-dependent Cox analyses, both were associated with worse OS, whereas only EL was independently associated with worse DFS. After multivariable adjustment, DFS and OS did not differ significantly between EL and LL.