Diagnostic study and meta-analysis of C-reactive protein as a predictor of postoperative inflammatory complications after gastroesophageal cancer surgery

Abstract

This study assessed the diagnostic accuracy of C-reactive protein (CRP) after gastroesophageal cancer resection for postoperative inflammatory complications (PIC). The clinical data and CRP values of patients operated on for gastroesophageal cancer surgery between 1997 and 2009 were retrospectively analyzed. The results of this study were compared with published data using a meta-analytic approach for diagnostic outcomes. Of 210 patients included in the study, 59 developed PIC (28.1 %; 95 % CI: 22.5-34.5 %). On the postoperative day (POD) 4 and 7, CRP had the best diagnostic accuracy for PIC (AUC 0.77; 95 % CI, 0.64-0.91, AUC 0.81; 95 % CI, 0.71-0.91). Using a cut-off value of 141 mg/L (95 % CI, 131-278 mg/L) for CRP on POD 4, the sensitivity was 0.78 (95 % CI, 0.55-0.91), the specificity was 0.70 (95 % CI, 0.53-0.83) and the NPV was 0.89 (95 % CI, 0.77-0.95). The in-hospital mortality rate was 3.3 % (95 % CI, 1.5-6.9 %). In a diagnostic meta-analysis that included two additional studies, CRP had a significant predictive value after POD 3. There is limited evidence for the diagnostic accuracy of CRP levels for PIC after gastroesophageal cancer surgery. CRP levels on POD 4 might be useful to rule out PIC, but its diagnostic accuracy is moderate at best. For clinical routine use CRP levels are clearly not sufficient to predict PIC and have to be interpreted in the context of the whole clinical picture.