Development and Validation of Platelet-Index Based Colorectal Cancer Survival Prognostic Nomogram.

INTRODUCTION

A prognostic model is essential for postoperative treatment planning, follow-up strategies, and aiding physicians in communicating survival rates to patients. Recent studies suggest that platelet indices may serve as valuable prognostic markers. This study aims to develop and validate a nomogram incorporating platelet indices to predict survival outcomes in CRC patients.

METHODS

This prospective cohort study included subjects diagnosed with CRC between 2019 and 2024 at Wahidin Sudirohusodo Hospital, Makassar. Subjects were randomly divided into training and validation sets. Demographic data (age, gender), clinical data (tumor invasion [T], node [N], metastases [M], tumor location, histological grading, history of surgery, chemotherapy, ileus or peritonitis), and platelet indices were collected. Independent prognostic factors were determined using the Cox regression model in SPSS 25.0. Nomogram development and validation were conducted using the "rms" (Regression Modeling Strategies) and "survival" packages in R Studio.

RESULTS

Thirteen prognostic factors, including gender, tumor location, T, N, M, histological grade, history of ileus or peritonitis, history of tumor resection, history of chemotherapy, and platelet indices, significantly impacted CRC survival. A nomogram was constructed with a C-index of 0.9 in the training set and 0.91 in the validation set.

CONCLUSIONS

A prognostic survival nomogram for Indonesian CRC patients was developed and validated, enabling predictions of CRC survival probabilities.