| Part one:A comparative study of cognitive function between patients with HBV-related cirrhosis without overt hepatic encephalopathy and healthy subjects.Objective:1. To obtain the5formulas used to calculate predicted values in psychometric hepatic encephalopathy score （PHES） by using PHES normative data from healthy subjects.2. To explore differences in cognitive function between patients with HBV-related cirrhosis without overt hepatic encephalopathy （OHE） and healthy subjects by using PHES battery of tests.3. To explore the relationship between impaired cognitive of patients with function HBV-related cirrhosis without OHE and abnormal liver function.Materials and Methods:One hundred and thirty-three healthy subjects and34patients with HBV-related cirrhosis without OHE and were investigated in this study. Standard group included133healthy subjects （81male,52female）, mean age45.06±10.79years （range22～65）, years of education10.51±3.70（range6～19）. Cirrhotic group was comprised of34patients （29male,5female）, mean age45.09±9.88years （range27～67）, years of education10.18±3.51（range6-19）; and a total of34matched healthy subjects （age:well matched; sex and education level:within3years） were selected from standard group as a control group （29male,5female）, mean age44.62±9.70years （range28～63）, years of education11.56±3.09（range6-16）.Biochemical tests were performed in all patients with HBV-related cirrhosis, and the liver function status of each patient was assessed using the Child-Pugh score. All167subjects underwent the PHES battery of tests. This paper-pencil test battery consisted of the number connection test A （NCT-A） and B （NCT-B）, the digit symbol test （DST）, the serial dotting test （SDT）, and the line tracing test （LTT-time and errors）. At the same time, the test results were recorded.Analyses were conducted using software （SPSS, version13.0; Chicago, Ⅲ）. Five psychometric test results are expressed as means±SD. In the first step, valid regression models were obtained with Pearson’s correlations between the psychometric test results of and age （years）, gender （male=l; female=2） and education level （years） in standard group. In the second step, the unstandardized beta coefficients of these analyses were used in the final formulas to correct for these factors. These formulas were then used to predict values （Z values） for cirrhotic patients, and the difference between the predicted and observed results for each test was divided by the corresponding SD for the reference population to obtain the deviation from’normal’as a multiple of the SD. Finally, differences for each test in multiples of the SD were summed as follows:a result≥1SD above the predicted was scored as+1, results-1SD,-2SDs and-3SDs below the predicted were scored as-1,-2and-3, respectively. Two independent sample t-tests were performed to assess the differences in raw results of five psychometric tests between two groups. Mann-Whitney U test was performed to assess the differences in the performance level on the PHES between two groups. A spearman correlation analysis was performed between their performance level on the PHES and Child-pugh score in cirrhotic patients. A two-side P value less than0.05was considered as statistically significant. Results:1. In the multivariate analysis using multiple linear regressions, both age and education were found to be independent variables related to all the tests. Five formulas used to calculate predicted values were obtained.2. Compared with the control subjects, the completion time for NCT-A、NCT-B、 SDT、LTT was increased significantly （P<0.01） and completion numbers for DST （P=0.007） were decreased significantly in cirrhotic patients. Cirrhotic patients had significantly worse performance of PHES （P<0.001） than healthy subjects.3. A significant negative correlation was observed between the performance of PHES and Child-pugh score in cirrhotic patients （rs=-0.367, P=0.033）.Conclusions:1. PHES battery of tests can be used to assess comprehensively the alterations of cognitive function in patients with HBV-related cirrhosis without OHE, but it should noted that there still exist factors influencing the results in PHES battery of tests.2. Patients with HBV-related cirrhosis without OHE had significantly worse performance of all the five psychometric test （NCT-A, NCT-B, DST, SDT, LTT） results and PHES than healthy subjects. Those findings reveal the existence of deficits in motor performance, visual perception visuoconstructive abilities, concentration and attention, and memory in those patients.3. The poorer the liver function in patients with HBV-related cirrhosis without OHE is, the more serious impaired cognitive function tends to be. Part two:Regional homogeneity abnormalities in patients with hepatitis B virus-related cirrhosis without OHE:a resting-state fMRI studyObjectives:1.To investigate regional activity abnormalities of patients with HBV-related cirrhosis without OHE by using resting-state fMRI and Regional homogeneity （ReHo） method in the resting state.2. To investigate whether these ReHo abnormalities of neural activity in patients with HBV-related cirrhosis without OHE can be related to their impaired cognitive function and abnormal liver function.Materials and Methods:Sixty-four subjects were investigated in this study. Cirrhotic group was comprised of32patients with HBV-related cirrhosis without OHE （27male,5female）, mean age44.69±9.86years （range27～67）, years of education10.34±3.62（range6～19）. Control group included32matched healthy subjects （age:well matched; sex and education level:within3years）,27male and5female, mean age44.22±9.67years （range28～63）, years of education11.84±3.20（range6-16）. The liver function status of all patients was assessed using the Child-Pugh score. All subjects underwent the PHES battery of tests （the same as the first part）.MR imaging data were obtained with a1.5-T MR imager with a16channel neurovascular coil to receive the signal. A gradient-echo echo-planar （GRE-EPI） sequence was used to acquire functional images. Scan parameters:TR/TE=3,000/50ms, flip angle=90°, thickness/gap=4.5/0mm, matrix=64×64, field of view （FOV）=230×230mm, total volumes=160.The imaging data were mainly preprocessed with a MATLAB toolbox called DPARSF （http://restfmri.net/forum/DPARSF） for "pipeline" data analysis of resting-state fMRI.1. The first ten time points were discarded to avoid transient signal changes before magnetization reached steady-state.2. Slice timing.3. Realignment:the raw data were corrected for the head motion. Subjects with head motion exceeding1.5mm in any dimension through the resting-state run will be discarded from further analysis.4. Normalization:all data were spatial normalized to the Montreal Neurological Institute （MNI） template.5. Removal of linear trends.6. Temporally filtered （band pass,0.01-0.08Hz）:to remove the effects of very low-frequency drift and physiological high frequency respiratory and cardiac noise. REST software was used to calculate the ReHo value of each subject. This is accomplished on a voxel-by-voxel basis by calculating Kendall’s coefficient of concordance （KCC） of time series of a given voxel with those of its nearest26neighbors. The KCC value was calculated to every voxel, and an individual KCC map was obtained for each subject. To reduce the influence of individual variations in the KCC value, normalization of ReHo maps was preformed through dividing the KCC among each voxel by averaged KCC of the whole brain. The resulting data were then spatially smoothed with an8-mm full-width at half-maximum （FWHM） Gaussian kernel.By using SPM8and REST software, a second-level random-effect two-sample t-test was performed on the individual normalized ReHo maps in a voxel-by-voxel manner by taking years of age and education, cardiac rates and respiratory rates as confounding covariates. Significant differences were set at the threshold of a corrected cluster level of P less than0.05（AlphaSim corrected, a combined threshold of P<0.01, and a minimum cluster size of74voxels）. Spearman correlation analysis of the mean ReHo values in significant different areas with performance level on the PHES and Child-pugh score in patients were performed, respectively. A two-side P value less than0.05was considered as statistically significant.Results:1. Compared with the control group, the cirrhotic patients group showed significant ReHo decreases in the bilateral precuneus/cuneus （PCu/Cu）, precentral gyrus （PCG） and paracentral lobule （PCL）, left lingual gyrus （LG） and middle temporal gyrus （MTG） and right middle occipital gyrus （MOG）. A significant ReHo increase was found in the bilateral inferior/medial frontal gyrus （IFG/MFG）（P<0.05; AlphaSim corrected, a combined threshold of P<0.01, and a minimum cluster size of74voxels）.2. Correlation analysis of the mean ReHo values in significant different brain areas against performance level on the PHES in patients revealed significantly positive correlation in the left LG （rs=0.369; P=0.037）, right MOG （rs=0.38; P=0.012） and bilateral PCu/Cu （rs=0.468; P=0.007）, PCG （left:rs=0.442, P=0.011; right: rs=0.575, P=0.001）, PCL （rs=0.475; P=0.006）. 3. Correlation analysis of the mean ReHo values in all significant different brain areas against Child-pugh score in patients revealed no statistical significant correlation （P>0.05）.Conclusions:1. Patients with HBV-related cirrhosis without OHE showed decreased ReHo most lay in motor cortex （left LG and MTG, and right MOG）, visual cortex （bilateral PCG and PCL） and default mode network （bilateral PCu/Cu）, while showed increased ReHo in the bilateral IFG/MFG. Those abnormalities reflect the destruction of local synchronization of spontaneous low-frequency BOLD in those regions.2. The mean ReHo values in the identified regions, mainly including visual （left LG and right MOG）, motor （bilateral PCG and PCL） association cortex and default mode network （bilateral PCu/Cu） were significantly positively correlated with the PHES in patients with HBV-related cirrhosis without OHE. These findings shed light on the pathophysiological mechanisms underlying cognitive alterations of cirrhotic patients and demonstrate the feasibility of using ReHo as a research and clinical tool to monitor the progression of cognitive impairment of cirrhotic patients without OHE.3. There is no significant correlation between ReHo abnormalities in patients with HBV-related cirrhosis without OHE and their abnormal liver function. Part three:Research on the gray matter volume of patients with hepatitis B virus-related cirrhosis without OHE:a voxel-based morphometry studyObjective:1. To investigate the gray matter （GM） volume abnormalities in patients with HBV-related cirrhosis without OHE by using high resolution MRI and voxel-based morphometry （VBM） method.2. To investigate whether these GM volume abnormalities in patients with HBV-related cirrhosis without OHE can be related to their impaired cognitive function and abnormal liver function.Materials and Methods:Fifty-four subjects were investigated in this study. Cirrhotic group was comprised of27patients with HBV-related cirrhosis without OHE （24male,3female）, mean age45.15±9.53years （range27～67）, years of education10.48±3.72（range6～19）. Control group included27matched healthy subjects （age:well matched; sex and education level:within3years）,24male and3female, mean age45.04±9.56years （range28～63）, years of education12.19±2.99（range6～16）. The liver function status of patients was assessed using the Child-Pugh score. All subjects underwent the PHES battery of tests （the same as the first part）.MRI data were obtained on a Philips Achieva1.5T Nova Dual MR scanner. A three-dimensional fast field echo （FFE） pulse sequence was used to produce contiguous sagittal images. Scan parameters:TR/TE=25/4.1ms, thickness/gap=1/0mm, matrix=232×231, field of view （FOV）=230×230mm, flip angle=30°.Images analysis was performed using the VBM8tool, an extension tool of SPM. The main procedures include:1. Spatial normalization: High-dimensional spatial normalization was chosen and the T1images were normalized to an already existing Dartel-template in MNI space.2. Segment: International Consortium for Brain Mapping （ICBM） template was used to remove non-brain tissue from the images. Then, maximum a posterior （MAP） and partial volume estimation （PVE） approaches were used to segment the images into GM, white matter and cerebrospinal fluid.3. Modulation: non-linear was introduced to modulate the resulting GM images so that the voxels’values indicate the absolute amount of tissue corrected for individual brain sizes.4. The resulting images were smoothed with an8-mm full width at half maximum （FWHM） isotropic Gaussian kernel.By using SPSS soft （version13.0）, two independent sample t-tests were performed to assess the differences in the whole brain volume and total GM volume between two groups. A two-side P value less than0.05was considered as statistically significant. A voxel based comparison was used to identify the differences between two groups. Statistical maps were set at a cluster-level threshold of p<0.001, uncorrected with extended threshold of200contiguous voxels. Partial correlation analysis was performed with age and year of education as covariates to assess the relation of total GM volume and the GM volume in significant different brain areas with PHES and Child-pugh score in cirrhotic patients, respectively. A two-side P value less than0.05was considered as statistically significant.Results:1. There was no significant difference in the whole brain volume between two groups （P=0.504）. Compared with the controls, the total GM volume was significantly increased （P<0.001） in cirrhotic patients.2. Compared with the controls, GM volume increased significantly in bilateral cerebellar hemisphere, fusiform gyrus （extend to bilatera thalamus/caudate/precuneus/cuneus/middle occipital gyrus/middle temporal gyrus/inferior frontal gyrus/insula/lingual gyrus）, bilateral orbitofrontal cortex, bilateral middle frontal gyrus, bilateral inferior frontal gyrus, right middle temporal gyrus, right inferior temporal gyrus, bilateral precentral gyrus, bilateral paracentral lobule, middle cingulate cortex and pons （P<0.001,uncorrected, clusters>200mm3） in cirrhotic patients.3. Significant negative correlation was observed between the total GM volume and PHES in cirrhotic patients （r=-0.631, P=0.001）. Partial correlation analysis of the GM volume in significant different brain areas against PHES in patients revealed significantly negative correlation in bilateral fusiform gyrus （r=-0.709, P<0.001）, left middle frontal gyrus （r=-0.546, P=0.005）, bilateral inferior frontal gyrus （left: r=-0.446, P=0.026; right:r=-0.406, P=0.044）, right middle temporal gyrus （r=-0.736, P<0.001）, right inferior temporal gyrus （r=-0.545, P=0.005）, bilateral precentral gyrus （left:r=-0.635, P=0.001; right:r=-0.655, P<0.001）, bilateral paracentral lobule （r=-0.594, P=0.002）, middle cingulate cortex （r=-0.524, P=0.007） and pons （r=-0.569, P=0.003）.4. Correlation analysis of the total GM volume, GM volume in all significant different brain areas against Child-pugh score in patients revealed no statistical significant correlation （P>0.05）.Conclusion:1. Patients with HBV-related cirrhosis without OHE have increased GM volume in many brain regions.2. Most of increased GM volume in patients with HBV-related cirrhosis without OHE was negative correlated with the degree of impaired cognitive function. It suggests that the increased GM volume may be underlying basis of impaired cognitive function. It olso demonstrates that the incresed GM volume in those brain regions can be used as a morphology indictor to monitor the progression of cognitive impairment of cirrhotic patients without OHE.3. There is no significant correlation between GM volume abnormalities in patients with HBV-related cirrhosis without OHE and their abnormal liver function. Part four:White Matter changes in patients with HBV-related cirrhosis without OHE:A DTI and TBSS studyObjective:1. To assess the white matter （WM） changes by using diffusion tensor imaging （DTI） and tract based spatial statistic （TBSS） in patients with HBV-related cirrhosis without OHE.2. To explore the relationship between WM changes in patients with HBV-related cirrhosis without OHE and impaired cognitive function, and abnormal liver function.Materials and Methods:Sixty subjects were investigated in this study. Cirrhotic group was comprised of30patients with HBV-related cirrhosis without OHE （25male,5female）, mean age46.43±9.49years （range32～67）, years of education9.97±3.22（range6～19）. Control group included30matched healthy subjects （age:well matched; sex and education level:within3years）,25male and5female, mean age45.83±8.67years （range28～63）, years of education11.43±3.00（range6-16）. The liver function status of patients was assessed using the Child-Pugh score. All subjects underwent the PHES battery of tests （the same as the first part）.MRI data were obtained on a Philips Achieva1.5T Nova Dual MR scanner. A spin-echo echo-planar imaging （SE-EPI） sequence was used to acquire DTI data. Scan parameters:acquired in33noncollinear diffusion gradient directions, b values=800and0s/mm2, T R/TE=19837/62ms, thickness/gap=2/0mm, flip angle=90°.DTI data was analysis by using FSL tools. First, diffusion-tensor images were corrected for head movements and eddy current distortion. Then, brain extraction tool （BET） was using to delete non-brain tissue from an image of the whole head and output a binary brain mask image. The diffusion tensor was estimated on a voxel-by-voxel basis by using dtifit （a part of the FDT）. Maps of mean diffusivity （MD）, fractional anisotropy （FA） were obtained. Running TBSS first involves running next steps:tbss1preproc, tbss2reg, tbss3postreg, tbss4prestats. Subsequently, using the threshold-free cluster enhancement （TFCE） option in randomise to perform statistics for FA and MD maps between two groups [Permutation-based correction for multiple comparisons （FWE） at P<0.05].By using SPSS soft （version13.0）, two independent sample t-tests were performed to assess the differences in the whole-brain mean FA and MD values, mean FA and MD values of the abnormal regions. Partial correlation analysis was performed with age and year of education as covariates to assess the relation of those indexes in TBSS with PHES and Child-pugh score in cirrhotic patients, respectively. A two-side P value less than0.05was considered as statistically significant.Results:1. Compared with controls, MD values of right frontal lobe, parietal lobe, temporal lobe, bilateral internal and external capsule, bilateral superior cerebellar peduncle, middle cerebellar peduncle, Cerebellar Vermis, cerebelar hemisphere and brain stem （P<0.05, FWE corrected）. There was no difference in FA values between two groups.2. There was no difference in whole-brain mean FA between two groups （P=0.716）. Compared with the controls, the whole-brain mean MD values （P=0.044） and mean MD values in the abnormal regions （P<0.001） significantly increased in cirrhotic patients.3. No significant correlation were observed between PHES and the whole-brain mean FA （r=0.180, P=0.358） or MD values （r=-0.310, P=0.108）, between PHES and mean MD values （r=-0.287, P=0.138）, and between Child-pugh score and the whole-brain mean FA values （r=-0.197, P=0.315） in cirrhotic patients. Significant positive correlation were observed between Child-pugh score and whole-brain mean MD values （r=0.686, P<0.001） and mean MD values in the abnormal regions （r=0.620, P<0.001）.Conclusion:1. The increased MD values with no concomitant changes in FA in many WM regions in patients with HBV-related cirrhosis without OHE indicates the presence of extracellular brain edema, while without damage of WM microstructure.2. Increased MD value in patients with HBV-related cirrhosis without OHE was positive correlated with the degree of impaired liver function. It suggests that the worse the liver function, the more serious the extracellular WM edema will be. |
PDF Full Text Request