Abilities of Pre- and Postoperative High-Sensitivity C-Reactive Protein Levels to Predict 90-Day Mortality After Surgery for Abdominal and Thoracic Cancers.

Abstract

Elevated high-sensitivity C-reactive protein (hsCRP) levels are known to be associated with a poor prognosis for cancer patients, but inadequate data exist regarding whether pre- or postoperative hsCRP levels have better predictive value after oncologic surgery. This study evaluated the associations between pre- and postoperative hsCRP levels and 90-day postoperative mortality among patients who underwent surgery for abdominal or thoracic cancers. This retrospective cohort study included 7933 patients who underwent elective surgery between January 2010 and December 2016. Cox regression analysis and receiver operative characteristic curve analyses were used to evaluate the prognostic values of preoperative hsCRP (< 1 month before surgery) and postoperative hsCRP (< 3 days after surgery). For predicting 90-day mortality, the area under the receiver operating characteristic curve was significantly larger for preoperative hsCRP than for postoperative hsCRP [0.76; 95% confidence interval (CI) 0.71-0.81 vs 0.65 95% CI 0.57-0.72; P < 0.001]. The optimal cutoff values were 0.5 mg/dL for preoperative hsCRP and 9.7 mg/dL for postoperative hsCRP. Based on these cutoff values, increased risks of 90-day mortality were significantly associated with preoperative hsCRP levels higher than 0.5 mg/dL [hazard ratio (HR) 7.60; 95% CI 4.43-13.03; P < 0.001] and postoperative hsCRP levels higher than > 9.7 mg/dL (HR 1.83; 95% CI 1.12-2.98; P = 0.016). Both elevated pre- and postoperative hsCRP levels were associated with increased risks of 90-day mortality after surgery for thoracic and abdominal cancer. However, preoperative hsCRP had better prognostic value than postoperative hsCRP.