Predicting the unpredictable: a robust nomogram for predicting recurrence in patients with ampullary carcinoma

Area of Science:

• Oncology
• Surgical Oncology
• Biostatistics

Background:

• Ampullary carcinoma (AC) is a rare malignancy requiring effective recurrence risk assessment post-resection.
• Identifying predictive factors is crucial for personalized patient management and improved outcomes.

Purpose of the Study:

• To identify risk factors influencing recurrence in ampullary carcinoma patients after radical resection.
• To develop and validate a predictive model for AC recurrence using Lasso-Cox regression.

Main Methods:

• Retrospective analysis of 162 ampullary carcinoma patients undergoing pancreaticoduodenectomy.
• Lasso regression for risk factor screening, followed by Lasso-Cox regression and Random Survival Forest (RSF) model comparison.
• Model validation using time-dependent ROC curves, calibration plots, and decision curve analysis (DCA).

Main Results:

• Lasso regression identified CA19-9/GGT, TNM staging, lymph node invasion, differentiation, tumor size, CA19-9, gender, GPR, PLR, drinking history, and complications as significant factors.
• The Lasso-Cox regression model demonstrated superior predictive performance (C-index=0.845) compared to RSF (C-index=0.719).
• The validated model showed high accuracy across 1, 3, and 5-year prediction intervals, with excellent calibration and broad clinical utility via DCA.

Conclusions:

• A robust nomogram based on Lasso-Cox regression effectively predicts recurrence in ampullary carcinoma patients.
• This developed model offers enhanced accuracy and reliability for clinical risk stratification compared to other methods.