Development and validation of prognostic nomograms for overall survival in patients with locally advanced colorectal cancer

Abstract

BackgroundMany colorectal cancer (CRC) patients are initially diagnosed at an advanced stage. Some evidence indicating that TNM staging may not be entirely reliable in forecasting survival outcomes. The research aims to develop accurate prediction models that can determine the survival outcomes of patients with clinically locally advanced colorectal cancer (LACRC). The study aims to provide valuable insights for clinical professionals to make right clinical decisions. MethodsWe retrospectively analyzed 636 LACRC patients admitted to the General Hospital of Western Theater Command between Jan 1, 2017, and Dec 31, 2019. Significant factors determining the prognosis of LACRC patients screened by lasso regression and multifactorial cox regression analysis. These factors were utilized to develop a nomogram model, which serves as a valuable predictive tool for estimating the overall survival (OS) of LACRC patients. Next, we validated the nomogram in the validation group. ResultsThe nomogram emphasized that the TNM stage holds the most prognostic value, followed by CA19–9 levels, smoking history, and MLR. Multiple assessments confirmed the superior discriminatory ability of the nomograms. ConclusionsWe have developed and verified nomograms that can predict OS cohort 635 LACRC patients. The survival nomogram may assist clinicians in better strategizing the therapeutic options for patients with LACRC.