Somatic KRAS Mutation Found in Sporadic Arteriovenous Malformations

Cerebral arteriovenous malformations (AVMs) are estimated to occur in 15 per 100 000 persons, with an annual risk of hemorrhage that is quite variable.1,2 Knowledge regarding AVM formation in patients with no relevant past familial history (sporadic AVMs) is sparse. Patients with hereditary hemorrhagic telangiectasia (HHT) have a higher incidence of AVM formation. The most common HHT genotype involves germline mutations affecting the TGF-β signaling pathway.2 These findings suggest that sporadic AVMs may originate from subsequent somatic mutations. Nikolaev et al3 investigated human AVM tissue and blood samples to determine the role of genetic mutation in sporadic AVM formation. Between 2013 and 2017, investigators collected AVM tissue and blood samples. Patients with digital subtraction angiogram-confirmed unifocal sporadic AVM with no past family history of AVMs or vascular disease were included. Whole-exome sequencing (WES) with confirmatory polymerase chain reaction (PCR) analysis was performed on the collected tissue. Results were compared with control samples comprising of the same patients’ blood sample and cerebral arterial tissue from circle of Willis, cavernous malformations, and dural arteriovenous fistulas taken from separate patients. Results were validated with parallel genomic analysis performed on independent samples of 33 sporadic Finnish patient AVMs, which were previously collected at a separate institution. Furthermore, endothelial cells taken from patient AVMs and control tissue were cultured and studied with PCR and various molecular biological tests. Migration, proliferation, apoptosis, and additional angiogenic factors were investigated in the cell cultures. A total of 39 patients had AVM tissue included in the study with 21 matched blood samples. WES was performed on the AVM tissue to detect genetic abnormalities, revealing 6 somatic missense variants in 3 genes. While 2 of these variants had unknown clinical significance, the other 4 coded for 2 activating KRAS mutations: p.Gly12Asp and p.Gly12Val. Further analysis of WES data found that 12 of the 26 samples contained at least 1 of these KRAS mutations in 0.9% to 4% of the sequence reads. PCR identified 6 previously WES-negative samples with KRAS mutations, in addition to verifying the mutations in the 12 previously identified samples. Furthermore, 11 of 13 samples excluded from WES analysis demonstrated 1 of the mutant variants with PCR analysis; in total 74% (29 of 39) of samples showed KRAS mutation with PCR analysis. No KRAS variants were detected following WES or PCR analysis of the 21 paired blood samples or the control tissue. Their results were further verified when WES-PCR analysis found similar KRAS mutant frequency in the separate Finnish AVM samples. Increased KRAS activity can potentially affect multiple downstream signal pathways: RAF-MEK, MAP-ERK, PI3K, and AKT-MTOR. Nikolaev et al3 investigated these molecular pathways in vitro. Vascular and endothelial cells were taken from 6 patient AVMs; 5 of the 6 endothelial cell samples were positive for 1 of the previously identified KRAS variants. These variants were not observed in the vascular smooth muscle cells. Compared with normal endothelial cells (EC-Norm), KRAS variant cells (EC-Var) demonstrated increased MAP-ERK pathway activation and normal PISK and AKT-MTOR expression. These findings were verified when increased florescence was observed in AVM tissue immunohistochemically stained for phosphorylated-ERK. RNA sequencing of KRAS-Var cells revealed increased expression of genes responsible for angiogenesis, proliferation, migration, and endothelial to mesenchymal cell transition. Additionally, EC-Norm treated with angiogenic factors, vascular endothelial growth factor and fibroblast growth factor, expressed genes similar to those expressed by EC-Var. Compared with EC-Norm, EC-Var demonstrated faster migration and cell turn-over on scratch assay, were phenotypically larger and elongated, had more cytoskeletal actin projections (lamellipodia), and less vascular endothelial cadherin junctions. Furthermore, EC-Var treated with a MEK inhibitor, decreasing ERK phosphorylation, demonstrated increased cadherin expression, and reduced lamellipodia formation and angiogenesis gene expression. The authors demonstrated that sporadic cerebral AVMs often contain endothelial cells that express somatic KRAS variants that solely activate the MAPK-ERK pathway, leading to increased angiogenesis, cell migration, and proliferation. Since mutant KRAS expression was only observed in AVM endothelial cells, it suggests that endothelial dysfunction may be the primary driver of AVM pathogenesis. However, there were many AVM samples that lacked significant KRAS somatic mutations suggesting that there are additional drivers of AVM formation that have yet to be elucidated. Investigators found that inhibition of the MAPK-ERK pathway with MEK inhibitors dampened the observed AVM-endothelial cell dysfunction in cell culture. Common MEK inhibitors are currently utilized for the treatment of melanoma and include Trametinib and Cobimetinib.4 MEK inhibitors may be a promising new treatment modality for AVMs that are challenging to treat by other measures. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.