Disrupted functional connectivity of default mode and salience networks in chronic pancreatitis patients

Recent advances in the understanding and management of chronic pancreatitis pain

Abdominal pain is the most common symptom of chronic pancreatitis (CP) and is often debilitating for patients and very difficult to treat. To date, there exists no cure for the disease. Treatment strategies focus on symptom management and on mitigation of disease progression by reducing toxin exposure and avoiding recurrent inflammatory events. Traditional treatment protocols start with medical management followed by consideration of procedural or surgical intervention on selected patients with severe and persistent pain. The incorporation of adjuvant therapies to treat comorbidities including psychiatric disorders, exocrine pancreatic insufficiency, mineral bone disease, frailty, and malnutrition, are in its early stages. Recent clinical studies and animal models have been designed to improve investigation into the pathophysiology of CP pain, as well as to improve pain management. Despite the array of tools available, many therapeutic options for the management of CP pain provide incomplete relief. There still remains much to discover about the neural regulation of pancreas-related pain. In this review, we will discuss research from the last 5 years that has provided new insights into novel methods of pain phenotyping and the pathophysiology of CP pain. These discoveries have led to improvements in patient selection for optimization of outcomes for both medical and procedural management, and identification of potential future therapies.

White matter brain changes in chronic pancreatitis: A 7-year longitudinal follow-up study

BACKGROUND/OBJECTIVES

The progression of cerebral white matter changes over time has not been explored in chronic pancreatitis (CP). We aimed to characterize such alterations in individuals with CP at baseline and after 7-years as compared with controls and to explore associations to risk factors and clinical parameters.

METHODS

Diffusion tensor imaging was used to evaluate 20 individuals with CP and 13 healthy controls at baseline and after 7-years (CP: n = 9, controls: n = 11). Tract-based spatial statistics were used to assess whole-brain white matter structure, extract significant fractional anisotropy (FA) clusters between groups, mean FA skeleton, mean FA and mean diffusivity (MD). FA of the extracted significant clusters between groups were used for regression analyses with risk factors and clinical parameters, including duration of CP, smoking, and diabetes.

RESULTS

At baseline, widespread reductions in FA were found in CP compared to controls involving corpus callosum, the anterior, posterior thalamic radiation, and superior and posterior corona radiata (cluster volume: 49,431 mm³, all P < 0.05). At baseline, also the mean FA (P = 0.004) and FA skeleton (P = 0.002) were reduced in CP compared to controls. FA of the extracted significant cluster was associated with the daily tobacco use (P = 0.001) and duration of CP (P = 0.010). At follow-up, the whole-brain FA skeleton was reduced by 1.7% for both CP individuals and controls (P = 0.878).

CONCLUSION

Individuals with CP had widespread cerebral white matter alterations at baseline that can likely be explained by the CP disease and exposure to toxic substances. Otherwise, further progression resembles that in healthy controls.

Pain, depression, and poor quality of life in chronic pancreatitis: Relationship with altered brain metabolites

BACKGROUND

To evaluate if altered brain metabolites are connected to pain, depression and affective responses in CP.

METHODS

In this prospective study we evaluated pain characteristics, QOL (EORTC QLQc30+PAN28), depression (Beck depression inventory [BDI] II) in 558 patients with CP and 67 healthy controls. Brain metabolites were evaluated using magnetic resonance spectroscopy (MRS) in 49 patients and 5 healthy controls. We measured plasma metabolites using gas chromatography-mass spectrometry (GC-MS/MS). Relationship between metabolomic alterations, pain, depression and QOL components were assessed using statistical/bioinformatics methods. Benjamini-Hochberg FDR correction was applied for multiple testing.

RESULTS

261 (46.8%) patients had depression compared to 5 (7.5%) among healthy controls [n = 67](p < 0.0001). Risk [OR (95% CI) of developing depression in the presence of pain was 1.9 (1.33-1.68); p = 0.0004. The depression scores correlated negatively with functional components and positively with symptom components of EORTC QLQ30. Significant negative correlation, though based on a small sample size, was observed between N-acetyl aspartate in the left hippocampus and choline in the left prefrontal cortex with emotional and cognitive functions. PLS-DA modelling revealed significant alteration in the plasma metabolomic profile among patients with CP who had depression. Six metabolites were significantly different between CP with depression and healthy controls, of which glycine contributed most significantly to the PLS-DA model (VIP score of 3.5).

CONCLUSIONS

A significant proportion of patients with CP develops depression that correlate with poor QOL functions. Pain, depression, and emotional components of QOL in patients with CP correlated with N-acetyl aspartate and choline in the left hippocampus and left prefrontal cortex of the brain.

Two-Week Cervical Vagus Nerve Stimulation in Chronic Pancreatitis Patients Induces Functional Connectivity Changes of Limbic Structures

OBJECTIVES

Noninvasive vagus nerve stimulation (nVNS) has not only shown antinociceptive effects, but also demonstrated anti-inflammatory and antidepressant effects. These effects could be beneficial in chronic pancreatitis (CP) patients suffering from chronic abdominal pain, even though the underlying central mechanisms remain unclear. The aim was to investigate the effect of cervical nVNS in patients with painful CP on brain functional connectivity and cerebral metabolites.

MATERIALS AND METHODS

In a randomized double-blind, sham-controlled crossover trial, we used resting-state functional magnetic resonance imaging to investigate functional connectivity changes of limbic structures (seed-based analysis) after two weeks cervical nVNS treatment (GammaCore) as compared with two weeks sham treatment. Similarly, magnetic resonance spectroscopy was performed in the anterior cingulate cortex (ACC) with assessment of glutamate/creatine (Glu/cre) and N-acetylaspartate/creatine (NAA/cre).

RESULTS

Sixteen CP patients (mean age 56.6 ± 9.4 years) completed the trial. nVNS induced reduced functional connectivity compared to sham treatment between 1) bilateral thalamus and bilateral superior frontal gyrus, 2) ACC and putamen, and 3) posterior cingulate cortex and right thalamus (all p < 0.05). No changes were observed in Glu/cre (p = 0.96) and NAA/cre (p = 0.43) levels between the nVNS and sham treatments.

CONCLUSION

In our population of CP patients, cervical nVNS compared with sham treatment induced reduced functional connectivity of limbic structures, as also observed in other patient groups. The findings are relevant, since we have previously demonstrated an effect on pain scores in CP patients for both nVNS and sham treatment. Our results elucidate the effects in the central nervous system following nVNS treatment of CP patients, pointing at potential beneficial effects in this patient group.

Audiovisual working memory and association with resting-state regional homogeneity

Multisensory processing is a prevalent research issue. However, multisensory working memory research has received inadequate attention. The present study aimed to investigate the behavioral performance of an audiovisual working memory task and its association with resting-state functional magnetic resonance imaging (fMRI) regional homogeneity (ReHo). A total of 128 healthy participants were enrolled in this study. The participants completed a modified Sternberg working memory task using complex auditory and visual objects as materials involved in different encoding conditions, including semantically congruent audiovisual, semantically incongruent audiovisual, and single modality of auditory or visual object encoding. Two subgroups received resting-state fMRI scans according to their behavioral performances. The results showed that the semantically congruent audiovisual object encoding sped up the later unisensory memory recognition in this task. Moreover, the high behavioral performance (response time, RT) group showed increased ReHo in the executive control network (ECN) and decreased ReHo in the default mode network (DMN) and saline network (SN). In addition, resting-state ReHo values in the ECN nodes (e.g., middle frontal gyrus and superior frontal gyrus) was correlated with RT. These findings suggested that semantically congruent audiovisual processing in working memory was superior to unisensory memory recognition and may be involved in the different functional networks such as ECN.

Altered Cortical Reorganization and Brain Functional Connectivity in Phantom Limb Pain: A Functional MRI Study

OBJECTIVES

Functional neuroimaging studies have shown that amputees have altered cortical reorganization and functional connectivity (FC). This study aimed to investigate whether patients with phantom limb pain (PLP) and PLP-free lower limb amputees exhibit changes in corresponding primary cortical motor area/somatosensory cortex (M1/S1) cortical reorganization and supplementary motor area (SMA) network FC. The association between functional magnetic resonance imaging (fMRI) changes and clinical parameters is also explored.

METHODS

A total of 10 PLP patients were matched with 10 PLP-free amputees and 10 healthy controls (HCs). Before undergoing fMRI, all participants completed questionnaires evaluating pain, anxiety, depression, and health-related quality of life. Task-related activation and regions of interest (ROI)-wise connectivity analysis were applied to differentiate the brain regions of amputees from those of HCs. Linear correlation analysis was used to evaluate the correlation between altered FC and clinical manifestations.

RESULTS

As compared with HCs, PLP patients showed increased cortical activation in M1/S1 when moving the intact foot, imagining phantom big toe movement, or having the corresponding thumb stimulated. The increased FC in the SMA network included the SMA-caudate nucleus, SMA-bilateral insula, and SMA-anterior cingulate cortex. Furthermore, results of the linear correlation analysis demonstrated that this increased FC was positively correlated with VAS scores, negatively correlated with Medical Outcomes Study 36-item Short-Form (SF-36) scores, and not correlated with anxiety or depression scores.

CONCLUSIONS

Phantom limb pain in lower limb amputees is associated with M1/S1 cortical reorganization and altered SMA network FC in different areas of the brain, which could help to support our understanding of the central mechanism of PLP.

Dynamic Properties of Human Default Mode Network in Eyes-Closed and Eyes-Open

The default mode network (DMN) reflects spontaneous activity in the resting human brain. Previous studies examined the difference in static functional connectivity (sFC) of the DMN between eyes-closed (EC) and eyes-open (EO) using the resting-state functional magnetic resonance imaging (rs-fMRI) data. However, it remains unclear about the difference in dynamic FC (dFC) of the DMN between EC and EO. To this end, we acquired rs-fMRI data from 19 subjects in two different statues (EC and EO) and selected a seed region-of-interest (ROI) at the posterior cingulate cortex (PCC) to generate the sFC map. We identified the DMN consisting of ten clusters that were significantly correlated with the PCC. By using a sliding-window approach, we analyzed the dFC of the DMN. Then, the Newman's modularity algorithm was applied to identify dFC states based on nodal total connectivity strength in each sliding-window. In addition, graph-theory based network analysis was applied to detect dynamic topological properties of the DMN. We identified three group-level dFC states (State1, 2 and 3) that reflects the strength of dFC within the DMN between EC and EO in different time. The following results were reached: (1) no significant difference in sFC between EC and EO, (2) dFC was lower in State2 but higher in State3 in EC than in EO, (3) lower clustering coefficient, local efficiency, and global efficiency, but higher characteristic path length in State2 in EC than in EO, and (4) lower nodal strength in the precuneus (PCUN), PCC, angular gyrus (ANG), middle temporal gyrus (MTG) and medial prefrontal cortex (MPFC) in State3 in EC. These results suggested different resting statuses, EC and EO, may induce different time-varying neural activity in the DMN.