| Objective To prospectively study the changes of the absolute concentrations estimated with LCModel and the metabolite ratios from single-voxel proton magnetic resonance spectroscopy (MRS) in patients with different stages of cerebral infarction;to determine the correlation between metabolite changes and clinical measures of neurological disability and functional outcome in the setting of early cerebral infarction;and to investigate the difference and significance of the two methods in the evaluation of cerebral infarction.Materials and Methods According to the time of attack, 41 patients with cerebral infarction were divided into five groups: (1)hyperacute group(﹤12h), n=5; (2)acute group(12h~72h), n=15; (3)subacute group(72h~10d), n=11; (4)early chronic group(11d~1m), n=4; (5) late chronic group(>1m), n=6. The conventional MR imaging, diffusion weighted imaging(DWI) and 1H-MRS were performed in all patients with GE Signa Hi-Speed 1.5T MRI scanner. T1-Flair, T2-FRFSE sequence and SE-EPI sequence were used for conventional MRI and DWI. Single-voxel spectra were obtained with a point-resolved spectroscopy sequence. Neurological functional deficit scales and the functional prognosis scales were performed in patients with an infarction of the middle cerebral artery territory within 18h~5d after stroke onset(n=17, respectively). Chinese scales of clinical neurologic functional deficit of stroke patients were performed within two hours after the MR scan. The clinical functional outcomes were assessed by the Barthel Index at the time of three months after stroke onset. The metabolic changes of the apparently normal brain tissue contralateral to the lesion were studied in fourteen patients with high risk factors of stroke such as hypertension and diabetes. A control population of fourteen age-matched normal healthy volunteers were studied with identical method for comparison. NAA, Lac, Cr, Cho, mI and Lip were the main metabolites to be analyzed. The absolute concentrations of metabolites were obtained by the LCModel method. The metabolite ratios(relative quantification) were the peak height ratios. Statistical analysis was performed by SPSS 10.0 software. The Statistical tests included paired t test, One-Way ANOVA,and so on. Results1. Changes of metabolite ratios: At the hyperacute stage, both NAA/Cr and Cho/Cr ratios showed no significant difference compared with the contralateral hemisphere(P﹥0.05), but mI/Cr ratio increased significantly (P﹤0.05). At the acute stage, the NAA/Cr ratio decreased significantly (P﹤0.001), but the Cho/Cr ratio had no obvious changes too(P﹥0.05) and mI/Cr ratio was further increased (P﹤0.001). At the subacute stage, the NAA/Cr ratio was significantly reduced too(P﹤0.001), but the Cho/Cr and mI/Cr ratios were both significantly increased(P﹤0.01).2. Changes of absolute metabolite concentrations(analyzed by LCModel): NAA concentration reduced significantly (P=0.007) and decreased to 64.1% of the contralateral hemisphere at the hyperacute stage. Further marked reductions were seen at the acute stage. During this period, NAA concentration decreased to 14.8% of the contralateral hemisphere. At the subacute stage, NAA concentration was at very low level and decreased to 13.2% of the contralateral hemisphere. At the chronic stage, NAA concentration kept very low level or declined to zero. The Cr concentration also reduced significantly (P=0.031) and decreased to 82.5% of the contralateral hemisphere at the hyperacute stage. At acute and subacute stage, Cr concentration decreased continuously and decreased to 48.5% and 22.1% of the contralateral hemisphere, respectively. At the chronic stage, Cr concentration kept very low level or declined to zero too. At the hyperacute stage, Cho concentration decreased slightly but showed no significant difference compared with the contralateral hemisphere(P﹥0.05), and mI concentration also showed no obvious change (P﹥0.05). At acute and subacute stage, the concentrations of Cho and mI were both significantly reduced(P﹤0.01), but the extent of reduction was lower than NAA and Cr. At the chronic stage, Cho and mI were at lower levels. Lactate doublet was detected in all hyperacute patients. Its concentration increased significantly and reached the highest level at the acute stage, and then gradually decreased at the subacute stage. In the chronic stage, no obvious lactate doublet could be seen.3. Comparision of two methods of quantitative analysis of metabolite concentrations : At the hyperacute stage, comparision of the NAA/Cr ratios between the central region of the lesion and contralateral hemisphere which showed no significant differences (P﹥0.05), but the absolute quantification analyzed by LCModel showed NAA concentration of the lesion was significantly lower than that in contralateral hemisphere (P=0.007). At the acute and subacute stage, the NAA/Cr ratio and NAA concentration of the lesion were both significantly reduced(P﹤0.001). At the hyperacute stage, the Cho/Cr ratio and Cho concentration of the lesion had no obvious changes compared with the contralateral hemisphere(P﹥0.05). At the acute stage, the Cho/Cr ratio of the lesion had no obvious changes too(P﹥0.05), but Cho concentration of the lesion was significantly lower than that in contralateral hemisphere(P=0.001). At the subacute stage, the Cho/Cr ratio of the lesion was significantly higher than that in contralateral hemisphere (P﹤0.01), but Cho concentration of the lesion was significantly lower than that in contralateral hemisphere too(P﹤0.001). At the hyperacute stage, the mI/Cr ratio of the lesion was significantly higher than that in contralateral hemisphere (P﹤0.05), but mI concentration of the lesion had no obvious changes(P﹥0.05). At the acute and subacute stage, the mI/Cr ratio of the lesion was significantly higher than that in contralateral hemisphere too(P﹤0.001), but mI concentration of the lesion was significantly lower than that in contralateral hemisphere (P﹤0.01).4. NAA concentration of the contralateral hemisphere of stroke patients with hypertension and diabetes slightly decreased (7.32±0.94 mmol/kg wet weight) compared with the normal healthy volunteers (7.80±0.96 mmol/kg wet weight). But statistical analysis showed no significant difference. The concentrations of Cr, Cho, mI and Glu+Gln also had no significant difference between the two groups.5. NAA, Lac and other main metabolites of the recent stroke(18h~5d) lesion were not significantly correlated with the clinical neurologic functional deficit scores.6. The lactate concentration of the recent stroke(18h~5d) lesion was strongly correlated with Barthel Index scores (r=-0.713,P=0.001). Lac/Cr ratio(r=-0.582,P=0.014) and Lac/Cho ratio (r=-0.624,P=0.007) correlated with Barthel Index scores. The lactate concentration analyzed by LCModel had greater correlation with clinical outcome measure than Lac/Cr ratio and Lac/Cho ratio.Conclusion1. Absolute quantification was more sensitive than relative quantification to detect NAA reduction in the hyperacute cerebral infarction. But at the acute and subacute stage, the two methods were both sensitive to NAA reduction. Absolute quantification was more sensitive to Cho reduction than relative quantification in the acute cerebral infarction. But at the subacute stage, the results of the two methods to the change of Cho were opposite. Using the Cho/Cr ratio to assess the change of Cho, it would lead to overestimate the Cho level in the subacute cerebral infarction. Likewise, using the mI/Cr ratio to assess the change of mI, it would lead to overestimate the mI level in the hyperacute, acute and subacute cerebral infarction.2. The Cr could not be used as the reference metabolite because it decreased greatly after cerebral infarction. Using the metabolite ratios (their denominator is the creatine) to assess the changes of other metabolites, it would be difficult to reflect the degree of acute ischemia and the true metabolism changes objectively. Relative quantification may lead to misinterpretation of spectral data and to erroneous metabolite values.3. The lactate concentration, Lac/Cr ratio and Lac/Cho ratio of the recent stroke(acute and early subacute stage) lesion were all correlated with Barthel Index scores(at 3 months after stroke onset). They could provide valuable prognostic information for cerebral infarction. The lactate concentration had stronger correlation with clinical outcome measure than the other two ratios. Therefore, absolute quantification may have greater advantage than relative quantification in predicting clinical outcomes.4. NAA concentration of the apparently normal brain tissue(on conventional MRI and DWI ) contralateral to the lesion in stroke patients with hypertension and diabetes slightly decreased compared with the normal healthy volunteers. The result suggests that high risk factors of stroke such as hypertension and diabetes may induce metabolic abnormalities of the brain prior to stroke onset.5. The main metabolites of the recent stroke(acute and early subacute stage) lesion were not significantly correlated with the early clinical neurologic functional deficit scores. So the metabolite changes were difficult to assess the degree of neurologic functional deficit of the patients exactly.
|
PDF Full Text Request