Correlation of findings in advanced MR techniques with global severity scales in patients with some grade of cognitive impairment.

INTRODUCTION: Some previous studies in patients with mild cognitive impairment and Alzheimer's disease have probed changes in the results of (1)H magnetic resonance spectroscopy and perfusion- and diffusion-weighted imaging. The purpose of this work was to correlate the results of perfusion- and diffusion-weighted imaging and magnetic resonance spectroscopy with the results of two global severity scales in cognitive impairment: the clinical dementia rating (CDR) and the global deterioration scale (GDS). PATIENTS AND METHODS: We evaluated 87 patients with cognitive impairment of diverse grade (35 men and 52 women; mean age, 70.2 +/- 8.5 years old). All patients were evaluated by a neurological team in our hospital. They applied both global severity scales (CDR and GDS) and referred the patients to our diagnostic imaging department to make a cerebral magnetic resonance imaging study and studies of diffusion- and perfusion-weighted imaging and magnetic resonance spectroscopy. We excluded patients with history of Parkinson's disease, frontotemporal dementia, cerebrovascular disease, intracranial tumors, hydrocephaly, epilepsy, alcoholism and psychiatric disorders. Magnetic resonance spectroscopy was carried out in the left occipital cortex and in the posterior cingulate gyrus. The evaluated metabolites were N-acetylaspartate (NAA), choline (Cho), creatine (Cr) and myo-inositol (mI). After diffusion-weighted imaging, we calculated apparent diffusion coefficient values in the region of interest located in hippocampi, white matter of temporal lobes, occipital lobes, parietal lobes, frontal lobes and posterior cingulate gyrus of both hemispheres. In perfusion-weighted imaging, we calculated the relative cerebral blood volume in hippocampi, gray matter of frontal lobes, occipital lobes, temporoparietal regions, posterior cingulate gyri and somatic-sensorial cortex. We used Spearman coefficient to analyse the correlation among the different factors. Statistical analysis was made with SPSS 14 software. RESULTS: We found 33 patients with Alzheimer's disease and 54 with mild cognitive impairment. The Spearman coefficient had statistical significance in the correlation of CDR and GDS (R(2)=0.596, p<0.001). Magnetic resonance spectroscopy showed a good correlation between ratios of NAA/Cr and NAA/mI with CDR and GDS in both evaluated regions and a weak correlation between Cho/Cr in the left occipital lobe and GDS. In diffusion-weighted imaging, we found a weak correlation between GDS and apparent diffusion coefficient values in hippocampi, temporal lobes, left frontal lobe and left occipital lobe. Finally, perfusion showed a weak correlation between GDS and relative cerebral blood volume in occipital lobes and posterior cingulate gyrus. CONCLUSION: In patients with cognitive impairment, there is a good correlation between CDR and GDS. The tool that showed the closest correlation with the clinical scales (CDR and GDS) was magnetic resonance spectroscopy in the left occipital cortex and posterior cingulate gyrus. Perfusion- and diffusion-weighted imaging are tools with a weak correlation with clinical scales, GDS being unique that gave us significant statistical results; this could be explained by the major number of items considered for cognitive impairment (GDS 2 and 3) compared with CDR (CDR 0.5). Magnetic resonance spectroscopy can be used in the diagnostic, following and evaluation of the response to the treatment in patients with cognitive impairment (mild cognitive impairment and Alzheimer's disease), complementing the information obtained in the clinical evaluation.

Utility of different MR modalities in mild cognitive impairment and its use as a predictor of conversion to probable dementia.

RATIONALE AND OBJECTIVES: Mild cognitive impairment has been regarded as a pre-Alzheimer condition, but some patients do not develop dementia. The authors' objective was to determine whether findings from a combined use of H1 magnetic resonance spectroscopy (MRS), perfusion imaging (PI), and diffusion-weighted imaging (DWI) would predict conversion from amnesic mild cognitive impairment to dementia and to compare the diagnostic accuracy in discriminating patients with probable Alzheimer disease (AD), mixed dementia (MD), Lewy body dementia (LBD), pre-Alzheimer disease mild cognitive impairment (MCI), vascular MCI (VaMCI), and anxious or depression patients with cognitive impairment (DeMCI). MATERIALS AND METHODS: A longitudinal cohort of 119 consecutive and incident subjects (73 women, 46 men; age 70+/-9.5 years) who fulfilled the criteria of amnesic MCI was followed for a mean period of 29 months. At baseline, a neuropsychological examination and standard blood test were performed, and different areas were examined by proton MRS, PI, and DWI. Among the group of patients considered to have AD, we also included patients with MD because these patients have a neurodegenerative component. RESULTS: After the follow-up period, 54 patients were considered as converted to dementia (49 with AD; 5 with LBD), 28 patients as MCI, 22 patients as DeMCI, and 15 patients as VaMCI. We found that N-acetylaspartate (NAA)/creatine (Cr) ratios in posterior cingulated gyri (PCG) predict the conversion to probable AD with a sensitivity of 82% and specificity of 72%, and NAA/Cr ratios in the left occipital cortex (LOC) had a sensitivity of 78% and specificity of 69%. When we used spectroscopy in the PCG and LOC to differentiate the types of MCI and dementias, we found significance differences in NAA/Cr, NAA/myoinositol (mI), NAA/choline (Cho), mI/NAA, and Cho/Cr ratios. The apparent diffusion coefficient (ADC) values in the right hippocampus showed differences in patients with LBD and DeMCI (P=.003), LBD with MCI (P=0.48), and LBD and VaMCI (P=.009). CONCLUSIONS: NAA/Cr ratios in PCG and LOC can predict the conversion from MCI to dementia with high sensitivity and specificity. MRS can differentiate AD from MCI, but cannot differentiate the types of MCI. DWI in the right hippocampus presents higher values of ADC in LBD and allows differentiating it from MCI.