Is C-reactive protein a prognostic factor of colorectal cancer?

PURPOSE: C-reactive protein, a commonly used inflammation marker, has been reported to be a prognostic factor of colorectal cancer. This prospective study was designed to confirm the prognostic value of its preoperative levels and to observe their perioperative change. METHODS: Between January 2001 and September 2005, preoperative C-reactive protein levels were obtained for 212 consecutive patients (140 males) receiving elective open resection of colorectal cancer. A level higher than 0.5 mg/dl was defined as positive. They were analyzed against clinicopathologic factors. The survival of 158 curative resections was analyzed. Postoperative levels (at months 1, 3, and 6) were collected for analysis of changing trend, from the patients receiving curative surgeries. RESULTS: Median value of preoperative C-reactive protein was 0.54 mg/dl (48.6 percent positive). Positive rate was significantly correlated with ulcerative type, larger size, higher stage, and positive carcinoembryonic antigen (>5 ng/ml). In both univariate log-rank test and multiple Cox proportional hazards regression, stage (univariate P=0.011, and multivariate P=0.016; hazard ratio, 6.23; 95 percent confidence interval, 1.41-27.54), C-reactive protein (0.5 mg/dl; P=0.005, and P=0.016; hazard ratio: 6.51; 95 percent confidence interval: 1.41-30.05), and differentiation (P=0.006, and P=0.043; hazard ratio, 3.53; 95 percent confidence interval, 1.04-11.98) were significant factors. Analysis of disease-free interval showed C-reactive protein was significant (P=0.03): as level rose, prognosis worsened. The quiescent inflammation-response group (≤0.1 mg/dl) had excellent outcomes. Postoperatively, the C-reactive protein levels declined at the third postoperative month. CONCLUSIONS: Preoperative C-reactive protein is an independent prognostic factor. The levels declined postoperatively, although with a lag. These findings seem to support the response hypothesis regarding C-reactive protein.