Vascular biomarkers derived from dynamic contrast-enhanced MRI predict response of vestibular schwannoma to antiangiogenic therapy in type 2 neurofibromatosis

BACKGROUND Antiangiogenic therapy of vestibular schwannoma (VS) in type 2 neurofibromatosis can produce tumor shrinkage with response rates of 40%-60%. This study examines the predictive value of parameter-derived MRI in this setting. METHODS Twelve patients with 20 VSs were recruited. Each had at least one rapidly growing tumor. Patients were treated with bevacizumab, 5 mg/kg every 2 weeks. Patients with stable or reduced VS volume were maintained at 2.5-5 mg every 4 weeks after 6 months. Those who failed treatment had their bevacizumab discontinued. Dynamic contrast-enhanced (DCE) MRI performed prior to treatment using a high temporal resolution technique, and data were analyzed to allow measurement of contrast transfer coefficient (K(trans)), vascular fraction (v(p)), extravascular-extracellular fraction (v(e)). Relaxation rate (R1(N)) was measured using a variable flip angle technique. Apparent diffusional coefficient (ADC) was calculated from diffusion-weighted imaging. The predictive power of microvascular parameters and ADC were examined using logistic regression modeling. RESULTS Responding tumors were larger (P < .001), had lower R1(N) (P < .001), and higher K(trans) (P < .05) and ADC (P < .01). They showed increases in R1(N) (P < .01) and reduction of K(trans) (P < .01) and ADC (P < .01). Modeling to predict response demonstrated significant independent predictive power for R1(N) (Β = - 0.327, P < .001), and K(trans) (Β = 0.156, P < .05). Modeling to predict percentage change in tumor volume at 90 days identified baseline tumor volume (Β = 5.503, P < .05), R1(N) (Β = - 5.844, P < .05), and K(trans) (Β = 5.622, P < .05) as independent significant predictors. CONCLUSIONS In patients with type 2 neurofibromatosis, biomarkers from DCE-MRI are predictive of VS volume response to inhibition of vascular endothelial growth factor inhibition.