Proton MR spectroscopy of delayed cerebral radiation in monkeys and humans after brachytherapy.

PURPOSE To determine whether radiation necrosis can be differentiated from residual/recurrent tumor by proton MR spectroscopy. METHODS We studied the effects of interstitial brachytherapy on the brains of healthy monkeys and in humans with glioblastoma multiforme. The effects of radiation therapy on normal brain tissue in monkeys were assessed with sequential proton MR spectroscopic studies 1 week to 6 months after brachytherapy. Proton MR spectroscopy was also performed in five patients with residual/recurrent glioblastoma multiforme (three of whom had radiation necrosis after brachytherapy), seven patients with newly diagnosed untreated glioblastoma multiforme, and 16 healthy volunteers, who served as a control group. RESULTS In monkeys, the ratio of N-acetylaspartate (NAA) to creatine-phosphocreatine (Cr) and the ratio of choline-containing compounds (Cho) to Cr of the reference point were significantly lower 1 week after brachytherapy than before treatment. The ratio of NAA to Cho of the irradiated area tended to be higher 1 week after brachytherapy than before irradiation. These peak metabolic ratios showed characteristic changes 6 months after treatment. In two of three monkeys, lipid signal was elevated 6 months after irradiation. In the clinical study, the ratio of NAA to Cho in the area of radiation necrosis was significantly different from that in glioblastoma multiforme when compared with the contralateral hemisphere after irradiation. In addition, lipid signal was detected in all patients with radiation necrosis. CONCLUSION It might be possible to use proton MR spectroscopy to differentiate radiation necrosis from residual/recurrent glioblastoma multiforme on the basis of comparisons with the contralateral hemisphere after radiation therapy.