Human in vivo NMR spectroscopy in diagnostic medicine: clinical tool or research probe?

In this critical review of human in vivo nuclear magnetic resonance (NMR) spectroscopy, the questions of which chemical species can be detected and with what sensitivity, their biochemical significance, and their potential clinical value are addressed. The current in vivo detectability limit is about 10(-6) of that of tissue water protons, necessitating a 1-10 cm3-volume of tissue and approximately 10-minute averaging time. This permits access to fats, membrane lipid metabolism, high-energy phosphate metabolism, glycogen, some neurotransmitters and metabolites in the citric acid cycle, and artificially introduced fluorocompounds. While hydrogen-31, phosphorus-31, carbon-13, sodium-23, and fluorine-19 in vivo results are discussed, the majority of patient studies use P-31 NMR spectroscopy. Here results from metabolic and ischemic disorders substantiate a case for spectroscopy as a diagnostic modality. The use of a broad range of spatial localization strategies is justifiable on the basis of the location and size of the pathologic condition and of NMR sensitivity. Abnormalities in spectra from many other disorders, most notably cancer, and improvements are often observed early in the course of successful therapy. Yet the potential impact of these results on clinical diagnosis and therapeutic monitoring is not always well understood, and many questions remain. Neurotransmitters and citric acid cycle metabolites exhibit high H-1 NMR sensitivities and represent major untapped potential for human clinical spectroscopy research. Studies evaluating spectroscopy in the context of existing modalities are needed. The unique ability of spectroscopy to provide noninvasive information about tissue chemistry in patients bodes well for its impact on clinical research and disease diagnosis.