Does postoperative plasma IL-6 improve early prediction of infection after pulmonary cancer surgery? A two-centre prospective study.

Abstract

Postoperative hyperinflammation increases infection risk. We hypothesized that interleukin-6 (IL-6) is an early predictor of infection after pulmonary cancer surgery. A two-centre prospective cohort study, including consecutive elective pulmonary cancer surgery patients. The primary outcome was any postoperative infection within 30 days. Multivariable logistic regression was used to create a core model (age, sex, surgery duration and Charlson comorbidity index) to which maximum IL-6, C-reactive protein (CRP), procalcitonin (PCT) concentrations and white blood cell count (WBC) between start of anaesthesia and 24 hours were added. The predictive performance of the models was assessed. 170 patients were analysed, of whom 38 (22%) developed a postoperative infection. IL-6 concentrations peaked 6 hours postoperatively, whereas CRP had not yet reached peak levels at 24 hours (time of prediction). Maximum IL-6 concentrations were associated with postoperative infection (adjusted odds ratio (aOR) 1.04 per 10 pg/ml, 95% confidence interval (CI) 1.00-1.09, p = 0.047) as was CRP (aOR 1.01 per mg/L, 1.00-1.03, p = 0.032). WBC and PCT were not associated with postoperative infection. The c-statistic of the prediction models that included IL-6 or CRP concentrations were 0.67 (95%CI: 0.56-0.77) and 0.68 (0.57-0.77), respectively, compared to 0.67 (0.56-0.76) for the core model. IL-6 and CRP slightly improved calibration by broadening the range of predicted probabilities. Reclassification did not improve. Plasma IL-6 and CRP levels observed within 24 hours from the start of surgery are associated with postoperative infection risk, yet the added value of these biomarkers to a simple clinical prediction model seems limited.