Predictors of complications and mortality following left colectomy with primary stapled anastomosis for cancer: results of a multicentric study with 1111 patients

Aim Reports detailing the morbidity–mortality after left colectomy are sparse and do not allow definitive conclusions to be drawn. We aimed to identify risk factors for anastomotic leakage, perioperative mortality and complications following left colectomy for colonic malignancies. Method We undertook a STROBE‐compliant analysis of left colectomies included in a national prospective online database. Forty‐two variables were analysed as potential independent risk factors for anastomotic leakage, postoperative morbidity and mortality. Variables were selected using the ‘least absolute shrinkage and selection operator’ (LASSO) method. Results We analysed 1111 patients. Eight per cent of patients had a leakage and in 80% of them reoperation or surgical drainage was needed. A quarter of patients (24.9%) experienced at least one minor complication. Perioperative mortality was 2%, leakage being responsible for 47.6% of deaths. Obesity (OR 2.8, 95% CI 1.00–7.05, P = 0.04) and total parenteral nutrition (TPN) (OR 3.7, 95% CI 1.58–8.51, P = 0.002) were associated with increased risk of leakage, whereas female patients had a lower risk (OR 0.36, 95% CI 0.18–0.67, P = 0.002). Corticosteroids (P = 0.03) and oral anticoagulants (P = 0.01) doubled the risk of complications, which was lower with hyperlipidaemia (OR 0.3, P = 0.02). Patients on TPN had more complications (OR 4.02, 95% CI 2.03–8.07, P = 0.04) and higher mortality (OR 8.7, 95% CI 1.8–40.9, P = 0.006). Liver disease and advanced age impaired survival, corticosteroids being the strongest predictor of mortality (OR 21.5, P = 0.001). Conclusion Requirement for TPN was associated with more leaks, complications and mortality. Leakage was presumably responsible for almost half of deaths. Hyperlipidaemia and female gender were associated with lower rates of complications. These findings warrant a better understanding of metabolic status on perioperative outcome after left colectomy.