Autor:innen:
Dr. Robin Ristl | aCenter for Medical Statistics, Informatics, and Intelligent Systems, Medical University of Vienna, Austria | Austria
Dr. med. Johannes Klopf | Medical University of Vienna, Department of Surgery, Division of Vascular Surgery | Austria
Andreas Scheuba | bDepartment of General Surgery, Division of Vascular Surgery, Medical University of Vienna, Austria | Austria
Prof. Florian Wolf | cDepartment of Biomedical Imaging and Image Guided Therapy: Division of Cardiovascular and Interventional Radiology | Austria
Prof. Dr. Martin Funovics | cDepartment of Biomedical Imaging and Image Guided Therapy: Division of Cardiovascular and Interventional Radiology | Austria
Prof. Dr. Bernd Gollackner | bDepartment of General Surgery, Division of Vascular Surgery, Medical University of Vienna, Austria | Austria
Prof. Dr. Anders Wanhainen | dDepartment of Surgical Sciences, Uppsala University, Uppsala Sweden, and Department of Surgical and Perioperative Sciences, Umeå University, Umeå, Sweden | Sweden
Prof. Dr. Christoph Neumayer | bDepartment of General Surgery, Division of Vascular Surgery, Medical University of Vienna, Austria | Austria
Prof. Dr. Christine Brostjan | bDepartment of General Surgery, Division of Vascular Surgery, Medical University of Vienna, Austria | Austria
Dr. Wolf Eilenberg | bDepartment of General Surgery, Division of Vascular Surgery, Medical University of Vienna, Austria | Austria
Objective: The most relevant determinant to schedule monitoring intervals of abdominal aortic aneurysms (AAA) is aneurysm maximum diameter and expected aneurysm growth, as clinical events are associated with aneurysm size. AAA growth rates, however, exhibit considerable between-patient variability as well as within-patient variability across time, making predictions for individual patients difficult. The aim of the study was to develop astatistical model that accounts for specific characteristics of AAA growth distributions and allows for probabilistic statements on the expected AAA growth.
Design, materials and methods: A total of 363 computed tomographic angiographies (CTAs) obtained from a sample of 87 patients whose AAA maximum diameter was monitored every 6 months with a mean follow-up time of 1.9 years were analyzed. By extending the model of geometric Brownian motion with a log-normal random effect, a stochastic growth model for AAA that accounts for the heavily right-skewed growth distribution and within-patient variability of growth rates was developed. For further validation, an external data set of US-based growth data (390 patients) was utilized.
Results: The AAA maximum diameter at baseline was between 29 and 86 mm (median 52 mm). Estimated from the growth model, the median growth in maximum diameter within one year was 1.2, 1.7 and 2.1 mm for patients with initial maximum diameter of 30, 40 and 50 mm respectively. The respective 95% quantiles maximum diameter increase within one year were 4.0, 5.3 and 6.6 mm. Internal cross-validation and external validation using an additional data set suggested that the stochastic growth model gives an accurate estimation of the distribution of aneurysm growth. An online calculator base on the fitted model was made available.
Conclusion: The stochastic growth model was found to be valid and can provide a reliable tool for predicting AAA growth.