Background and purpose: The results of the A Randomized Trial of Unruptured Brain Arteriovenous (ARUBA) study, indicating that conservative medical management of unruptured brain arteriovenous malformations (UBAVM) is superior to interventional therapy, have generated debates that have hampered their application into clinical practice. Irrespectively of study conclusions, it seems reasonable to explore how much better interventional therapy would have to be to become competitive with conservative medical management.
Methods: We conducted an exploratory analysis to replicate the original data from ARUBA. The functional form of the replicated ARUBA data, according to their Weibull distribution, allowed estimation of parameters. We carried out Monte Carlo simulations while introducing theoretical reductions of interventional risk, and the results were used to construct theoretical and example Kaplan-Meier curves from simulations.
Results: The "ARUBA Replication" analysis showed results nearly identical to those published in the study, with an estimated hazard ratio of 0.27 (95% CI: 0.14-0.55). At 50% interventional risk reduction, the simulations showed an estimated event rate of 14.9%, and the protective effect of conservative medical management was no longer statistically significant. Greater risk reductions hastened the time to benefit for interventional therapy, and an 80% risk reduction demonstrated superiority of interventional therapy at just over 2 years Hazard Ratio (HR: 1.44, 95% CI: 0.55-4.92).
Conclusions: Reduction in risk of interventional therapy by 50%-80% results in more competitive clinical outcomes, equating or surpassing the benefit of conservative medical management of UBAVM. This conjecture should be taken into consideration in the design of future studies of this patient population, particularly because it is supported by recent observational studies.
Keywords: Intracranial arteriovenous malformations; Monte Carlo method; neurosurgical procedures.
Copyright © 2018. Published by Elsevier Inc.