Overall,
liver cirrhosis patients presented significantly impaired executive functions and attention (Fig.
1) independent from the etiology of liver cirrhosis (means: z= -1.49 (SD 1.19)).
Table 2 shows the neuropsychological performance of HE subgroups.
Compared to Non-HE patients,
patients with HE,
regardless of the HE-state (MHE or HE1) performed worse in the TMT-B (MHE: p =0.003,
HE1: p=0.000),
logical memory (MHE: p=0.026,
HE1: p=0.011) and TAP divided attention tests (MHE: p=0.049,
0.000,
0.001,
HE1: p=0.029,
0.048,
0.004).
Importantly,
HE-states themselves (MHE and HE1) did not differ significantly in cognitive performance (p>0.05).
In comparison to healthy controls,
cirrhosis patients showed reduced volume of GM (p<0.05,
FWE corrected) in frontal and temporal cortices,
putamen and cerebellum (Fig.
2).
By comparison of subgroups,
Non-HE patients did not differ from MHE patients significantly.
The HE grade I-subgroup showed GM volume loss of the left motor respectively supplementary motor area in the temporal pole in contrast to Non-HE patients (Table 3 and Fig.
3a).
Furthermore,
the subgroup of HE grade I-patients revealed bilateral GM volume loss in the inferior temporal and temporal fusiform cortex when comparing to MHE-patients (Table 3 and Fig.
3b).
We did not observe significant differences comparing controls to Non-HE subgroup in respect to Fa and MD values.
In contrast,
the comparison of controls to patients with HE (MHE or HE1) revealed substantial white matter alterations showing significantly decreased fractional anisotropy (FA) and increased mean diffusivity (MD values) in bilateral white matter long association,
projection and commissural fibers along with the internal capsule.
Figure 4 shows the group comparisons between different HE stages.
Differences of white matter tracts were most pronounced when comparing the Non-HE to MHE and the Non-HE to HE1 subgroup,
demonstrating decreased FA and increased MD values bilaterally in the long association,
projection and commissural fibers as well as in the internal capsule,
cerebellum and corticospinal tract.
The maximum effect sizes were identified for frontal and parieto-occipital association fibers.
Comparing MHE to HE1 subgroups,
we did not find significant differences for FA values,
but in HE1 significant increased MD values of long association fibers as the superior and inferior fronto-occipital fasciculus and inferior longitudinal fasciculus.
Considering the d-maps of FA values when comparing MHE to HE1 subgroup,
we found topologically similar effects to MD values.
Correlation analyses to test the relationship between grey and white matter variables and neuropsychological test scores displayed significant correlations between FA/MD values and neuropsychological test results (p<0.05).
Especially,
scores of the TMT-A (r= -0.523; p=0.001) and the visual TAP divided attention test (r= -0.631; p=0.000) displayed correlations with FA and cingulate gyrus of statistically higher significance (p<0.01,
multiple comparisons corrected) (Fig.
5).
Results showed no significant correlation between grey matter volume and neuropsychological test scores.