Differentiation between glioma recurrence and radiation-induced brain injuries using perfusion-weighted MR imaging

Abstract

Objective To evaluate the feasibility of perfusion weighted imaging (PWI) in the differentiation of recurrent glioma and radiation-induced brain injuries. Methods Fifteen patients with previously resected and irradiated glioma, presenting newly developed abnormal enhancement, were included in the study. The final diagnosis was determined either histologically or clinicoradiologically. PWI was obtained with a gradient echo echo-planar-imaging (GRE-EPI) sequence. The normalized rCBV ratio[CBV(abnormal enhancement)/CBV(contralateral tissue)], rCBF ratio[CBF(abnormal enhancement)/CBF(contralateral tissue)]and rMTT ratio[(MTT abnormal enhancement)/MTT(contralateral tissue)]were calculated, respectively. The regions of interest (ROIs) consisting of 20-40 mm2 were placed in the abnormal enhanced areas on postcontrast T1-weighted images. Ten to fifteen ROIs measurements were performed in each lesion and the mean value was obtained. Mann-Whitney test was used to determine whether there was a difference in the rCBV/rCBF/MTT ratios between glioma recurrence and radiated injuries. Results Nine of the 15 patients were proved recurrent glioma,6 were proved radiation-induced brain injuries. The mean rCBV ratio[2.87(0.70-4.91)]in glioma recurrence was markedly higher than that[0.70(0.12-1.62)]in radiation injuries (Z=-2.55,P＜0.05). The mean rCBF ratio[1.89(0.64-3.96)]in glioma recurrence was markedly higher than that[0.56(0.12-2.08)]in radiation injuries (Z=-2.08,P＜0.05). The areas under rCBV and rCBF ROC curve were 0.893 and 0.821. If the rCBV ratio ≤0.77, the diagnosis sensitivity of radiation-induced brain injuries was 100.0%;If ≥2.44, the diagnosis specificity of recurrent glioma was 100.0%. Conclusion PWI was an effective technique in distinguishing glioma recurrence from radiation injuries and rCBV and rCBF ratios were of great value in the differentiation. Key words: Glioma; Recurrence; Brain injuries; Radioactivity; Magnetic resonance imaging; Perfusion