Speakers - 2026

Title: Genetic analysis of neurological deficit development in patients with arteriovenous malformations

Abstract

Cerebrovascular diseases account for approximately 7 million deaths each year, remaining the second leading cause of mortality worldwide and the third leading cause of long-term disability. Cerebral arteriovenous malformations (AVMs) represent severe vascular anomalies characterized by a complex network of abnormal vessels that form direct connections between arteries and veins. In recent years, substantial global progress has been achieved in genomic research focused on uncovering the molecular mechanisms underlying complex multifactorial diseases.

Relevance: Vascular malformations are benign, non-neoplastic lesions that are present from birth, although they may not become apparent until weeks, months, or even years later. AVM is classified as a multifactorial disease.

Materials and Methods. The study enrolled 94 patients with a verified diagnosis. The control group included 60 individuals with no history of cerebrovascular disorders. Neuroimaging assessments were carried out using magnetic resonance imaging, multispiral computed tomography, and cerebral angiography to assess AVM structure, location, and related vascular characteristics. Molecular genetic testing was performed to analyze polymorphic variants of the BDNF and CDKN2B genes.

The results of the study: In our study, a multigene prognostic model was developed to estimate the probability of developing a torpid clinical form based on the genetic predictors CDKN2B and BDNF. The analysis demonstrated that carriage of the T allele of the BDNF gene was associated with a reduced likelihood of remaining healthy (adjusted odds ratio, AOR = 0.332; 95% CI: 0.129–0.856; p = 0.022), indicating a pathogenetic role of BDNF in the development of neuronal hyperexcitability. In contrast, the T allele of the CDKN2B gene was found to significantly reduce the risk of developing the torpid form of the disease (AOR = 3.417; p = 0.005).

Conclusions. The model demonstrated acceptable discriminatory performance (AUC = 0.645; p = 0.002), with a sensitivity of 68.3% and a specificity of 60%. These findings confirm the relevance of the identified genetic markers for the torpid form of the disease, and the model may be applied within a multiparametric approach to predicting disease course.