Construction of a risk prediction model for early postoperative recurrence in stage II/III colorectal cancer.

Background: Colorectal cancer (CRC) recurrence within a year post-surgery poses significant challenges for stage II/III patients. Few models currently predict this early recurrence with multi-dimensional considerations for risk stratification.

Aim: To develop a model for predicting the risk of recurrence within one year after surgery in patients with stage II/III CRC.

Methods: We conducted a retrospective cohort study at Zhejiang Provincial Hospital of Chinese Medicine, including 349 stage II/III CRC patients. Clinical data were collected, and the dataset was randomly divided into training (n = 244) and testing (n = 105) sets. Univariate and multivariate logistic regression analyses identified risk factors for postoperative recurrence. Then a nomogram model was constructed and evaluated via receiver operating characteristic curves, calibration curves and decision curve analysis.

Results: During the one-year follow-up, 10.9% (38/349) of patients experienced recurrence. Univariate analysis identified tumor size, lymph node metastasis (N2 stage), neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, fatigue, and appetite loss as significant correlates of recurrence. Multivariate logistic regression confirmed N2 stage, appetite loss, tumor size, and neutrophil-to-lymphocyte ratio as independent risk factors. The nomogram model showed excellent performance. The area under the receiver operating characteristic was 0.98 (95% confidence interval: 0.97-1.00) in training set and 0.91 (95% confidence interval: 0.84-0.97) in testing set. The decision curve analysis curves showed strong concordance between predicted and observed recurrence probabilities.

Conclusion: The model effectively predicts early postoperative recurrence in stage II/III CRC, integrating clinical, inflammatory, and symptomatic factors.