High prevalence of KRAS/BRAF somatic mutations in brain and spinal cord arteriovenous malformations

Hong, T; Yan, YP; Li, JW; Radovanovic, I; Ma, XY; Shao, YW; Yu, JX; Ma, YJ; Zhang, P; Ling, F; Huang, SC; Zhang, HQ; Wang, YB

Wang, YB (reprint author), Peking Union Med Coll, 167 Beilishi Rd, Beijing 100037, Peoples R China.; Wang, YB (reprint author), Chinese Acad Med Sci, Fuwai Hosp, 167 Beilishi Rd, Beijing 100037, Peoples R China.

BRAIN, 2019; 142 (): 23


The genetic basis of many brain and spinal arteriovenous malformations is unclear. Hong et al. reveal a causative role for somatic tumour-related mutations in KRAS/BRAF in the majority of patients tested. This homogeneity supports therapeutic targeting of the RAS/RAF/MAPK pathway without the need for tissue genetic diagnosis.Brain and spinal arteriovenous malformations are congenital lesions causing intracranial haemorrhage or permanent disability especially in young people. We investigated whether the vast majority or all brain and spinal arteriovenous malformations are associated with detectable tumour-related somatic mutations. In a cohort of 31 patients (21 with brain and 10 with spinal arteriovenous malformations), tissue and paired blood samples were analysed with ultradeep next generation sequencing of a panel of 422 common tumour genes to identify the somatic mutations. We used droplet digital polymerase chain reaction to confirm the panel sequenced mutations and identify the additional low variant frequency mutations. The association of mutation variant frequencies and clinical features were analysed. The average sequencing depth was 1077 298. High prevalence (87.1%) of KRAS/BRAF somatic mutations was found in brain and spinal arteriovenous malformations with no other replicated tumour-related mutations. The prevalence of KRAS/BRAF mutation was 81.0% (17 of 21) in brain and 100% (10 of 10) in spinal arteriovenous malformations. We detected activating BRAF mutations and two novel mutations in KRAS (p.G12A and p.S65_A66insDS) in CNS arteriovenous malformations for the first time. The mutation variant frequencies were negatively correlated with nidus volumes of brain (P = 0.038) and spinal (P = 0.028) arteriovenous malformations but not ages. Our findings support a causative role of somatic tumour-related mutations of KRAS/BRAF in the overwhelming majority of brain and spinal arteriovenous malformations. This pathway homogeneity and high prevalence implies the development of targeted therapies with RAS/RAF pathway inhibitors without the necessity of tissue genetic diagnosis.

Download PDF

Full Text Link