Using machine learning to preoperatively stratify prognosis among patients with gallbladder cancer:...

Area of Science:

• Oncology
• Machine Learning
• Surgical Oncology

Background:

• Gallbladder cancer (GBC) is an aggressive malignancy with high recurrence and mortality rates.
• Accurate prognostic stratification is crucial for effective patient management.

Purpose of the Study:

• To develop a machine learning model for stratifying GBC patients into distinct prognostic groups.
• To identify preoperative clinical factors predictive of overall survival (OS) in GBC.

Main Methods:

• A classification and regression tree (CART) model was employed.
• Preoperative variables were analyzed using a multi-institutional database of patients undergoing curative-intent GBC resection.

Main Results:

• CART analysis identified tumor size, biliary drainage, carbohydrate antigen 19-9 (CA19-9), and neutrophil-lymphocyte ratio (NLR) as key prognostic factors.
• Four prognostic groups were defined, showing a significant incremental decrease in median OS from 59.5 months (Group 1) to 12.1 months (Group 4) (p < 0.0001).

Conclusions:

• A machine-based model successfully stratified GBC patients into four prognostic groups using only preoperative characteristics.
• This predictive model can enhance patient-centered care by providing more accurate prognosis assessments.