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

Serum biomarkers associated with abdominal aortic aneurysm (AAA) have been studied individually; however, an algorithm that considers panel of proteins to inform AAA prognosis may improve predictive accuracy. We conducted a prognostic study using a prospectively recruited cohort of patients with and without AAA (n = 452). Serum concentrations of seven biomarkers were measured at baseline, and the cohort was followed for 2 years. The primary outcome was major adverse aortic event (MAAE; composite of rapid AAA expansion [>0.5 cm/6 months or >1 cm/12 months] or AAA intervention). Using 10-fold cross-validation, we trained a random forest model to predict 2-year MAAE using: (1) clinical characteristics, (2) biomarkers, and (3) clinical characteristics and biomarkers. Two-year MAAE occurred in 114 (25%) patients. Four proteins were significantly elevated in patients with AAA compared to those without AAA (matrix metalloproteinase 3 [MMP-3], human epididymal secretory protein 4 [HE4/WFDC2], Chitinase 3-like-1, and Kallikrein 6/Neurosin), composing the protein panel. For predicting 2-year MAAE, our random forest model achieved an area under the receiver operating characteristic curve (AUROC) of 0.64 using clinical features alone and the addition of the four-protein panel improved performance to an AUROC of 0.80. Using a combination of clinical and biomarker data, we developed a model that accurately predicts 2-year MAAE.