Natural History and Outcomes of Deep-seated Brain Arteriovenous Malformations: a Propensity Score Matched Analysis Using Nationwide Multicenter Prospective Registry Data

Objectives: This study aims to evaluate the natural history of deep-seated brain arteriovenous malformation (AVMs), as well as the risk-benefit outcomes of interventional treatment versus conservative management. Materials & Methods: Patients with deep-seated AVMs were selected from a nationwide prospective multicenter registry study in China (the MATCH study), and univariate and multivariate analyses were conducted to identify factors associated with AVM rupture. In the analysis of outcomes, propensity score matching (PSM) was performed between the interventional and conservative treatment groups, adjusting for baseline differences. The primary outcomes were hemorrhagic stroke or death, while the secondary outcomes focused on obliteration rates and neurological status. Subgroup and sensitivity analyses were conducted, incorporating various study designs to assess the robustness and consistency of the results. Results: Among 4,286 consecutive AVM cases registered from August 2011 to December 2021, 1,057 (24.7%) were classified as deep-seated AVMs. The natural annualized rupture risk before the treatment decision is 5.58%. The independent risk factors for rupture included diffuse lesions (aOR: 1.79 [1.29–2.49]), single drainage (aOR: 1.88 [1.20–2.93]), and drainage stenosis (aOR: 2.33 [1.44–3.75]). In the analysis of outcomes, 883 cases maintained continuous follow-up (128 conservative management, 755 intervention). After PSM, there were 119 cases in each group. After a median follow-up duration of 4.34 (1.72, 7.23) years in the intervention group, 47.93% achieved complete obliteration, with an annualized rupture risk of 4.82%. Compared to conservative management, intervention was associated with a higher rate of hemorrhagic stroke or death (AR: 3.85 [1.84–5.86] per 100 person-year, P<0.001; HR: 4.862 [1.869-12.651], P<0.001) and higher obliteration rates (OR: 108.56 [14.57–809.01], P<0.001). No significant differences were observed in terms of neurological functional outcomes. In a further analysis stratified by interventional strategies, embolization and multimodality treatment significantly increased the risk of hemorrhagic stroke or death compared with conservative treatment (embolization: HR: 4.414 [95%CI, 1.642-11.867]; multimodality treatment: HR, 6.238 [95%CI, 2.146-18.136]), while microsurgical resection and stereotactic radiosurgery did not. Subgroup and sensitivity analyses showed consistent trends, though with slight differences in statistical power. Conclusion: This study indicates that in deep-seated AVMs, interventional treatment is associated with an increased risk of hemorrhagic stroke or death. However, the negative effect may result from the adverse effects of embolization and multimodality treatment, whereas microsurgical resection and stereotactic radiosurgery did not.