Models Based on Dynamic Clinicopathological Indices for Predicting Prognosis During the Perioperative Period for Patients with Colorectal Cancer

Recent studies have found that clinicopathological indices, such as inflammatory and biochemical indices, play a significant role in the prognosis of colorectal cancer (CRC) patients. However, few studies have focused on the effect of dynamic changes in these indicators. In our study, we studied the influence of dynamic changes in inflammatory and biochemical indices on patient outcomes during the perioperative period.

We enrolled 551 patients from Hubei Cancer Hospital who had undergone radical resection of CRC and collected the results of laboratory examinations performed within 1 week before surgery and at the first admission after surgery. The whole population was randomly divided into the training (386) and testing (185) cohorts. We used postoperative inflammatory and biochemical indices/preoperative inflammatory and biochemical indices (ΔX) to reflect the dynamic changes. Chi-square tests, Kaplan-Meier survival analyses, and univariate and multivariate Cox regression analyses were used to evaluate the prognosis. The prediction accuracies of models for overall survival (OS) and disease-free survival (DFS) were estimated through Harrell's concordance index (the C-index) and Brier scores. Nomograms of the prognostic models were plotted for evaluations of individualized outcomes.

The median follow-up time of the 551 patients was 35.6 (range: 1.1-73.8) months. Ultimately, the prognostic models based on age, sex, TNM stage, pathological conditions, inflammatory and biochemical indices, CEA, and CA199 were found to have exceptional performance for OS and DFS. The C-index of the nomogram for OS was 0.806 (95% CI, 0.75-0.86) in the training cohort and 0.921 (95% CI, 0.87-0.96) in the testing cohort. The C-index of the nomogram for DFS was 0.781 (95% CI, 0.74-0.82) in the training cohort and 0.835 (95% CI, 0.78-0.88) in the testing cohort.

We successfully established a novel model based on inflammatory and biochemical indices to guide clinical decision-making for CRC.