Researchers found that 8.9% of patients had harmful gene fusions—higher than previously reported. Five genes accounted for most of these fusions: FGFR3 (37%), MET (21%), EGFR (20%), NTRK2, and PDGFRA. The most common fusion types included FGFR3:TACC3, PTPRZ1:MET, and EGFR:SEPT14.
Tumors with gene fusions had different molecular features compared with typical glioblastoma. They more often showed amplifications of MET, FGFR3, CDK4, and MDM2, but had fewer EGFR and TP53 mutations. EGFR fusions were almost always found together with EGFR amplification. Patients with fusion-positive tumors had a slightly longer median survival (16.6 months) than those without fusions (15.5 months). However, treatment with tyrosine kinase inhibitors (TKIs) has not yet shown a clear survival benefit, possibly because of small patient numbers or difficulty delivering drugs across the blood-brain barrier.
The findings support the growing role of precision oncology in brain cancer. Although each individual fusion is rare, researchers suggest “umbrella” clinical trials that screen patients for specific fusions and match them with targeted therapies. This approach could eventually lead to routine fusion testing in glioblastoma, offering additional treatment options for patients who do not respond to standard radiation and chemotherapy.