A Predictive Model for Major Adverse Aortic Events in Patients with Abdominal Aortic Aneurysm Using Clinical and...

Area of Science:

• Cardiovascular Research
• Biomarker Discovery
• Medical Diagnostics

Background:

• Abdominal aortic aneurysm (AAA) management relies on monitoring expansion and intervention risk.
• Individual serum biomarkers for AAA prognosis have shown limited predictive power.
• A multi-protein panel integrated into a predictive algorithm may enhance AAA event forecasting.

Purpose of the Study:

• To develop and validate a predictive model for major adverse aortic events (MAAE) in patients with AAA.
• To assess the incremental value of a four-protein biomarker panel combined with clinical characteristics for AAA prognosis.
• To improve the accuracy of predicting rapid AAA expansion or the need for intervention.

Main Methods:

• A prospective cohort study of 452 patients with and without AAA was conducted.
• Serum concentrations of seven biomarkers were measured, and patients were followed for 2 years for MAAE.
• A random forest model was trained using 10-fold cross-validation, incorporating clinical data and a four-protein panel (MMP-3, HE4/WFDC2, Chitinase 3-like-1, Kallikrein 6/Neurosin).

Main Results:

• Two-year MAAE occurred in 25% of patients (n=114).
• Four proteins (MMP-3, HE4/WFDC2, Chitinase 3-like-1, Kallikrein 6/Neurosin) were significantly elevated in AAA patients.
• The random forest model achieved an AUROC of 0.64 with clinical features alone, improving to 0.80 with the addition of the four-protein panel.

Conclusions:

• A combination of clinical characteristics and a four-protein biomarker panel accurately predicts 2-year MAAE in patients with AAA.
• This multi-biomarker approach offers improved prognostic accuracy for abdominal aortic aneurysm.
• The developed algorithm holds potential for refining AAA patient management and risk stratification.