Abstract

High-resolution magic angle spinning proton (HRMAS 1H) magnetic resonance spectroscopy produces well-resolved spectra of metabolites from intact tissue specimens. Here we report the results of a preliminary study of 19 human brain tumors obtained by applying this method. Among these 19 cases were 2 low-grade astrocytomas, 1 anaplastic astrocytoma, 8 glioblastomas, 6 meningiomas, and 2 schwannomas. In addition, autopsy human brain tissues from two subjects without any known neurological diseases were used as normal controls. The HRMAS 1H magnetic resonance spectroscopy measurements were performed at 2 degrees C on a 400-MHz NMR spectrometer with a HRMAS speed of 2.5 kHz. From these HRMAS 1H MR spectra, we measured the concentrations of 11 metabolites, the ratios of 15 metabolites (resonances) to creatine (at 3.03 ppm), and the spin-spin relaxation time for these metabolites (resonances). Our results indicate that these parameters have the potential to characterize tumor types and grades with statistical significance as well as identify tumor biochemical characteristics. In particular, we found that compared with metabolite concentrations and metabolite spin-spin relaxation time, the metabolic ratios presented the highest sensitivity in differentiating normal tissue from tumors, as well as in distinguishing between tumor groups. Of 15 analyzed metabolic ratios, 12 showed statistical significance in differentiating normal tissue from low-grade and anaplastic astrocytomas, 13 showed statistical significance in differentiating normal tissue from glioblastomas, 14 showed statistical significance in differentiating normal tissue from schwannomas, and 9 showed statistical significance in differentiating normal tissue from meningiomas. Moreover, our results strongly indicate that the resonance ratio of inositol (at 4.05 ppm) to creatine may help distinguish tumor type. Our results suggest that the HRMAS method for intact tissue measurement may function as an adjunct to histopathology and contribute to improved accuracy for brain tumor diagnoses.