RESUMO
BACKGROUND: Serum C-reactive protein (CRP) is commonly used by surgeons to raise suspicion of anastomotic leakage and other infectious complications, but most studies on optimal cut-off values are retrospective with a small sample of patients. The aim of this study was to determine the accuracy and optimal cut-off value of CRP for anastomotic leakage in patients following esophagectomy for cancer. MATERIALS AND METHODS: Consecutive minimally invasive esophagectomy for esophageal cancer patients was included in this prospective study. Anastomotic leakage was confirmed if a defect or leakage of oral contrast was seen on a CT scan, by endoscopy or if saliva was draining from the neck incision. Diagnostic accuracy of CRP was assessed by receiver operator curve (ROC) analysis. Youden's index was adopted to determine the cut-off value. RESULTS: A total of 200 patients were included between 2016 and 2018. Postoperative day 5 showed the highest area under the ROC (0.825) and optimal cut-off value of 120 mg/L. This resulted in a sensitivity of 75%, specificity of 82%, negative predicting value of 97%, and positive predicting value of 32%. CONCLUSIONS: CRP on postoperative day 5 can be used as a negative predictor for and can be used as a marker to raise suspicion of anastomotic leakage following esophagectomy for esophageal cancer. When CRP exceeds 120 mg/L on postoperative day 5, additional investigations should be considered.