Magnetic Resonance Spectroscopy

The Parkinson's-Reversing Breakthrough

Parkinson Disease Treatment Options

Get Instant Access

MRS provides chemical information on tissue metabolites (3,4). The molecules that can be studied by MRS in human brain tissue are hydrogen 1 (:H) and phosphorus 31 (31P). Magnetic resonance sensitivity is far greater for protons than it is for phosphorus (3). Therefore, most commercial MR scanners are capable of only proton MRS. Spectra are usually obtained from localized brain regions. The brain region is defined on a single slice by placing a small voxel, on the order of 1 or 2 cm2, in the area of interest. The compounds that can be observed in proton spectra are primarily identified by their frequency (i.e., their position in the spectrum), expressed as the shift in frequency in pars per million (ppm) relative to a standard. A normal spectrum shows peaks from N-acetyl groups, especially N-acetylaspartate (NAA) at 2.0 ppm, creatine (Cr), and phosphocreatine at 3.0 ppm, and cho-line-containing phospholipids (Ch) at 3.2 ppm (3). An additional peak at 1.3 ppm arises from the methyl resonance of lactate and is normally barely visible above the baseline noise (3). Pathologic conditions that involve regional neuronal loss lead to region-specific decreases in the relative NAA

concentrations (3,4). Since NAA is a marker of neuronal integrity, proton MRS provides a noninvasive means of quantifying neuronal loss or damage in vivo. Ch and other lipids are markers of altered neuronal membrane synthesis. Cr is a possible marker of defective energy metabolism. Typically, individual metabolic ratios obtained from peak areas of the spectrum are used as input to the statistical analysis. Because total Cr concentration is relatively resistant to change, Cr is often used as an internal standard to which the concentrations of other metabolites are normalized.

Proton MRS has been used by several groups to study brain metabolism in nondemented PD yielding conflicting results (15-27). In general, the majority of these studies have failed to detect consistent abnormalities for NAA, Ch, or Cr in the basal ganglia (15-19,21,24,25; but see refs. 20,26,27) or the cerebral cortex (19,21,25; but see refs. 22,23). A multicenter study that included 151 patients with PD and 97 age-matched controls found no overall differences in the striatal proton spectra between groups (16). However, there was a decrease in the NAA/Cho ratio in a subgroup of elderly patients (aged 51-70 yr) and patients that were not treated with levodopa (16). These findings suggest that proton MRS may reveal subtle metabolic abnormalities in patients with PD depending on age and medication. Therefore, heterogeneities of clinical features across patient groups may at least in part account for discrepancies across studies (15-27).

Three out of four proton MRS studies have reported abnormal proton spectra in patients with atypical parkinsonian disorders (15,18,19; but see ref. 26). Davie et al. (15) performed proton MRS of the lentiform nucleus in seven patients with MSA-P (predominantly striatonigral variant) and five patients with MSA-C (olivocerebellar variant). Compared with healthy controls, the NAA/Cr ratio was significantly decreased in patients with MSA (15). Tedeschi et al. (19) found a reduced NAA/Cre ratio in the brainstem, centrum semiovale, and frontal and precentral cortex, and a reduced NAA/Ch ratio in the lentiform nucleus in 12 patients with PSP compared with healthy controls (Fig. 1). The same study also included nine patients with corticobasal degeneration (CBD). CBD patients also showed an abnormal metabolic pattern with a reduced NAA/Cre ratio in the centrum semiovale and a reduced NAA/Ch ratio in the lentiform nucleus and the parietal cortex contralateral to the most affected side (Fig. 1). A reduction in NAA/Ch and NAA/Cr ratios in the lentiform nucleus was also observed by Federico et al. (18) in seven patients with MSA and seven patients with PSP. All three studies additionally examined the proton spectra in a group of PD patients and found no abnormalities in the selected region of interest (15,18,19). It should be noted, however, that Clarke et al. failed to find any abnormality for NAA, Ch, and Cr in six patients with probable MSA (26).

In summary, the published data suggest that proton MRS may be of potential diagnostic value in atypical parkinsonian disorders. However, the demonstration of significant group differences in patients with well-defined clinical features does not imply that proton MRS can reliably discriminate between PD and atypical parkinsonian disorders in individual patients. A recent study by Alexon et al. used pattern recognition techniques (i.e., neural network and related data analyses) to analyze proton spectra in 15 patients with probable PD, 11 patients with possible PD, 5 patients with atypical PD, and 14 healthy age-matched controls (27). In contrast to conventional analyses, all information within the proton spectrum can be entered in the statistical analysis simultaneously. The neuronal networks approach allowed them to distinguish between the four groups with considerable accuracy; approximately 88% of the predictions were correct. By contrast, conventional analysis revealed no significant differences in metabolite ratios among groups.

Proton MRS of the basal ganglia in conjunction with statistical analyses that take into account the pattern of the entire proton spectrum may help to distinguish PD from atypical parkinsonian disorders at initial presentation. Discriminative power may be further increased by using a multimodal imaging approach that combines MRS with other imaging techniques such as structural MRI or 18F-6-fluorodopa PET.

Spectroscopy Mri Parkinson

Fig 1. H-MRS findings in control subjects and patients with parkinsonian disorders.PSP, progressive supranuclear palsy; PD, Parkinson's disease; CBD, corticobasal degeneration. Columns and error bars present means ± SD. Asterisks indicate significant differences between controls and PSP patients and between controls and CBD (*p = 0.05, **p < 0.01, ***p < 0.001). The data were taken from Table 2 published in ref. 19.

Fig 1. H-MRS findings in control subjects and patients with parkinsonian disorders.PSP, progressive supranuclear palsy; PD, Parkinson's disease; CBD, corticobasal degeneration. Columns and error bars present means ± SD. Asterisks indicate significant differences between controls and PSP patients and between controls and CBD (*p = 0.05, **p < 0.01, ***p < 0.001). The data were taken from Table 2 published in ref. 19.

Phosphorus MRS provides information on levels of cerebral phospholipids and high-energy phosphates. The cerebral phosphorus spectra contains at least seven resonances that can be attributed to phosphomonoesters (PME), inorganic phosphate (Pi), phosphodiesters (PDE), phosphocreatine (PCr), and adenosine triphosphate (ATP). So far, only one group has used MRS to investigate the phosphorus spectra in 13 patients with PD and 15 patients with MSA (28). Assuming a constant cytosolic concentration of ATP, Barbiroli et al. measured PCr and Pi and calculated cytosolic pH and free Mg2+ (28). Compared with an age-matched control group, patients with PD showed a significant increase in Pi and a decrease in cytosolic free Mg2+, whereas PCr was reduced and Pi was increased in patients with MSA. In the MSA group, there was no difference in the phosphorus spectra between the MSA-P and the MSA-C variants of MSA. A discriminant analysis that considered only the concentrations of PCr and Mg2+ provided a correct classification of MSA and PD patients in 93% of the cases. These preliminary data suggest that, in addition to proton MRS, phosphorus MRS may be useful to probe cerebral phosphate metabolism and ion contents in patients with atypical parkinso-nian disorders.

Was this article helpful?

0 0

Post a comment