脳動静脈奇形における動物モデルのトランスレーショナルな妥当性：システマティックレビュー

We conducted a systematic review to identify and evaluate animal models of true brain arteriovenous malformations (bAVMs), focusing on how they replicate human disease. Outcomes assessed included bAVM growth, rupture, seizures, and survival. The review adhered to PRISMA guidelines and SYRCLE's protocol for animal studies. A search on PubMed and Embase was conducted (April 19, 2023; updated October 24, 2024). Exclusion criteria were; (1) Not an original, peer-reviewed full length research article, (2) not an in vivo study, (3) no bAVM induction, (4) no appropriate control group, (5) no histological or anatomical assessment during the bAVM follow-up, (6) outcome of the bAVM not reported. Meta analyses were planned for all key parameters. Risk of bias and image duplication assessments were conducted. Forty-one studies could be included in this SR. Models primarily involved mice and targeted mutations in MAPK, TGFbeta and Notch related signaling. Study designs varied significantly, limiting meta-analysis and direct comparison. We often noted high risk of bias in studies' reporting. Many studies had high risk of bias and focused on HHT-related mutations, which represent only a minority of clinical bAVM cases. KRAS-based models may offer better clinical relevance, but overall, current bAVM models show substantial variability.