April 24, 2009
2 min read
Save

MRI identified EGFRvIII mutation in glioblastomas

AACR 100th Annual Meeting

Researchers were able to detect EGFRvIII mutations — the hallmark of aggressive glioblastoma — using magnetic resonance perfusion-weighted imaging to measure cerebral blood volume levels, according to a late-breaking abstract presented at the AACR 100th Annual Meeting in Denver.

“We believe that the cerebral blood volume value as determined by MRI can be used to predict mutation status, which then has implication for not only diagnoses and treatment selection, but for evaluation of treatment response,” Donald M. O’Rourke, MD, an associate professor of neurosurgery at University of Pennsylvania, said during his presentation. “We can easily visualize a situation in which noninvasive analysis allows us to not in all cases obtain tissue and then evaluate, more importantly, tumor progression and treatment response over time.”

O’Rourke and colleagues used MRI to analyze glioblastoma multiforme tissue samples from 97 participants and used real-time polymerase chain reaction to determine VEGF expression.

The researchers found that patients with the EGFRvIII mutation had higher maximum relative cerebral blood volume compared with those who were EGFRvIII-negative (P=.0017). They also found that VEGF expression was higher in EGFRvIII-positive tumors compared with EGFRvIII-negative tumors (P=.101).

Multivariate logistic regression showed that relative cerebral blood volume assessment was a “significant independent predictor” of EGFRvIII mutation. Relative cerebral blood volume had excellent accuracy for predicting the presence of the mutation and had a better predictive value than VEGF. – by Jason Harris

PERSPECTIVE

I don't fully agree with several of the assumptions made in the introduction. First, it is not clear that presence of the vIII mutant confers a worse prognosis. Second, recent clinical trials of erlotinib in glioblastoma multiforme did not confirm that the presence of the vIII mutant and intact PTEN predicted for favorable tumor outcome with erlotinib. Consequently, the authors' statement that noninvasive identification of vIII mutational status would be important in choosing therapies for individual patients is not currently well-supported in clinical practice. Their methods seem appropriate and their conclusions that vIII leads to upregulated VEGF and increased rCBF is of interest. Other MRI studies have shown some utility in predicting EGFR overexpression in glioblastoma multiforme, so this is not the first study on the subject, though it may be the first to look at EGFRvIII status as opposed to EGFR overexpression.

David Schiff, MD

Co-director
University of Virginia Health System Neuro-Oncology Center, Charlottesville

For more information: