LAUSR

Editorial for “Progressive Deterioration of Dynamic Functional Network Connectivity in Patients With HBV-Related Cirrhosis” Hepatitis B virus-related cirrhosis (HBV-RC) patients with no minimal hepatic encephalopathy (NMHE) may develop minimal hepatic encephalopathy (MHE)—a neurocognitive dysfunction characterized by a deficit in attention, working memory, and executive functions. The disease progression from NMHE to MHE can be characterized by the alteration of functional brain connectivity. Functional connectivity (FC) is defined as the temporal statistical dependencies between different brain regions and is a method that is widely used in most functional brain imaging studies to analyze brain network interactions. Several studies have demonstrated that MHE patients had weaker FC in multiple brain networks, including the visual network (VIS), default mode network (DMN), salience network, central executive network, and thalamocortical circuits. In contrast to FC, dynamic functional network connectivity (dFNC) offers more temporal information on dynamics indices and provides a more accurate representation of functional brain networks. The dFNC has recently been used to study atypical dynamic network connectivity alteration in cirrhotic patients without overt hepatic encephalopathy. Nevertheless, the dFNC alteration underlying the progression of NHME to MHE remains unelucidated. In this issue of JMRI, Lin et al. further examined the deterioration of dFNC in HBV-RC patients from NMHE to MHE using resting-state functional magnetic resonance imaging. The authors expanded their investigation beyond FC as a standard method for analyzing the functional interaction between distinct networks by incorporating dFNC in their analyses. They used the Child-Pugh score to determine the severity of the patients’ liver disease, whereas the psychometric hepatic encephalopathy score (PHES) was used to determine the patients’ neurocognitive functions. Structural and functional images were acquired using a 1.5-T magnetic resonance (MR) scanner. Using the Statistical Parametric Mapping (SPM) and Graph Theoretical Network Analysis (GRETNA) toolbox, functional data were pre-processed before parcellated using spatial group independent component analysis (GICA). A total of 30 independent components were identified and categorized into seven intrinsic brain networks: auditory network (AUD), VIS, somatomotor network (SMN), cognitive control network (CCN), DMN, subcortical (SC) network, and cerebellar (CB) network. Finally, dFNC indices were compared between 30 HBV-RC patients (17 NMHE and 13 MHE) and 38 healthy controls. These indices were mean dwell time (DT), fraction time (FT), number of transitions (NT), and number of states (NS). This well-designed and carefully executed study reveal that the NMHE and MHE groups had significantly lower PHES scores than their age-, gender-, and education-matched healthy controls. More interestingly, the authors discovered that the dFNC deteriorated progressively along with the disease progression in patients with HBV-RC, thus supporting the study’s hypothesis. They also noted deterioration of the dFNC within the CCN-VIS and DMN-VIS networks as NMHE progressed to MHE. Comparisons between patient groups revealed that the two brain networks’ functional integrity was weaker in the MHE than NMHE patients. Aside from that, the authors found a significant positive correlation between dFNC indices (DT and FT) and PHES performance, suggesting that the deterioration of dynamic FC in the HBV-RC patients leads to most of the neurocognitive impairments in MHE. This finding is not surprising, as the CCN is one of the network hubs that orchestrate information across cortical regions, and therefore, deterioration of these networks may lead to poor cognitive performance in HBV-RC patients. These findings collectively suggest that the deterioration of network connectivity between the CCN and VIS could serve as a biomarker for MHE. As the quest to improve our understanding of how the progression of the disease alters functional network interactions continues, dFNC indices that can robustly correlate with cognitive behaviors may also valuably add to the development of MHE treatment. However, the authors highlighted that the employed dFNC method is a new technique for investigating dynamic brain network states and may lack a gold standard in terms of parameters. Other than that, it is not justified to extend the conclusions to the general population given the study’s small sample size. Therefore, future study is warranted to verify this preliminary evidence