The Prognostic Values Of Serum Enzymological And Inflammatory-Related Indexes In Patients With Resected Colorectal Cancer

Objective: The study was aimed to evaluate the prognostic values of serum enzymological indexes and inflammation-related indexes in patients with resected colorectal cancer.Integrated scores were constructed by combining indexes with good predictive potential of prognosis of colorectal cancer patients.The integrated scores were further combined with clinicopathological factors to provide individualized prognostic evaluation through nomogram.Methods: The present study included 509 patients with resected primary colorectal tumors.Prognostic and clinicopathological data were collected through medical record review and telephone follow-up.Tests of enzymological and inflammation-related indexes were performed before and 1 month after tumor resection.Cox regression was applied to estimate the associations of the aforementioned indexes with progression-free survival(PFS)and cancer-specific survival(CSS).Receiver operating characteristic curve(ROC)was used to evaluate the predictive potential of individual indexes and integrated scores.The prediction accuracy of Cox regression models was estimated through Harrell’s concordance indexes.Nomograms of the combination of integrated scores and clinicopathological factors were plotted for individualized evaluation of cancer progression and mortality.Results: Our study indicated that preoperative and postoperative enzymological and inflammation-related indexes were associated with progression-free survival and cancerspecific survival.We created two integrated scores: cancer-related enzymology and antigen score(CREAS)and cancer-related inflammation,nutrition and antigen score(CRINAS).The areas under ROC curves(AUC)of preoperative and postoperative CREAS associated with PFS were 0.69 and 0.75,respectively.The multivariable-adjusted HRs(95%CIs)for per score increase of preoperative and postoperative CREAS were 1.25(1.18-1.34)and 1.23(1.15-1.36),with Harrell’s concordance indexes of 0.75 and 0.78 for corresponding Cox regression models,respectively.The AUC of preoperative and postoperative CREAS associated with CSS were 0.74 and 0.82,respectively.The multivariable-adjusted HRs and 95% CIs for one score increase of preoperative and postoperative CREAS were 1.19(1.13-1.26)and 1.18(1.12-1.24),with Harrell’s concordance indexes of 0.84 and 0.85 for corresponding multivariable-adjusted models,respectively.The AUC of preoperative and postoperative CRINAS associated with PFS were 0.74 and 0.76,respectively.The multivariable-adjusted HRs(95% CIs)for per score increase of preoperative and postoperative CRINAS were 1.21(1.15-1.29)and 1.23(1.14-1.32),respectively.Harrell’s concordance indexes of corresponding multivariable-adjusted Cox regression models were 0.76 and 0.79,respectively.The AUC of preoperative and postoperative CRINAS associated with CSS were 0.78 and 0.82,respectively.The corresponding multivariable-adjusted HRs(95% CIs)were 1.29(1.20-1.39)and 1.16(1.11-1.21),with both Harrell’s concordance indexes of 0.84.Nomograms of the combination of preoperative CREAS or CRINAS with clinicopathological factors had the Harrell’s concordance indexes of 0.75 for PFS and 0.83 for CSS.Nomograms of the combination of postoperative CREAS or CRINAS with clinicopathological factors harbored the Harrell’s concordance indexes of 0.77 for PFS and 0.83 for CSS.Conclusions: Our study suggests that both CREAS and CRINAS are good predictors for prognosis of colorectal cancer.Nomograms based on the combination of CREAS or CRINAS with clinicopathological factors might act as convenient tools for individualized evaluation of cancer progression and mortality.