A nomogram for predicting nutritional risk before gastric cancer surgery

Area of Science:

• Oncology
• Medical Informatics
• Radiology

Background:

• Gastric cancer (GC) is a leading cause of cancer mortality globally.
• Patients with GC face elevated nutritional risks.
• Precise preoperative nutritional risk assessment is crucial for GC management.

Purpose of the Study:

• To develop and validate a nomogram model for predicting preoperative nutritional risk in gastric cancer patients.
• To enhance the accuracy and efficiency of preoperative nutritional risk assessment in GC.
• To aid physicians in early identification and diagnosis of nutritional risk in GC.

Main Methods:

• Utilized clinical, laboratory, and radiomics data from GC patients undergoing surgery.
• Employed Least Absolute Shrinkage Selection Operator (LASSO) and multivariable logistic regression.
• Validated the nomogram's predictive performance using ROC curves, calibration curves, and Decision Curve Analysis (DCA).

Main Results:

• Identified Body Mass Index (BMI), hemoglobin (Hb), and radiomics score (Radscore) as key predictors.
• The nomogram demonstrated good predictive ability with AUCs of 0.895 (training) and 0.880 (validation).
• DCA confirmed the clinical utility and benefit of the developed nomogram model.

Conclusions:

• A simple and reliable nomogram model for clinical prediction of preoperative nutritional risk in GC patients has been established.
• This tool facilitates time-efficient and accurate assessment, aiding early diagnosis and intervention.
• The nomogram supports physicians in managing nutritional status for improved gastric cancer patient outcomes.