Introduction: Total brachial plexus injury not only significantly affects the motor and sensory function of the affected upper limbs but also causes further physical and mental damage to patients with long-term intractable pain. Previous studies mainly focused on the surgical treatment, while only a few paid attention to the intractable neuropathic pain caused by this injury. Changes in the volume of gray matter in the brain are thought to be associated with chronic neuropathic pain. Methods: Voxel-based morphometry analysis was used to compare the difference in cerebral gray matter volume between total brachial plexus injury patients with neuropathic pain and healthy controls. Correlations between pain duration, pain severity, and GM changes were analyzed. Results: The volume of cerebral gray matter in the patient group was decreased significantly in multiple regions, including the parahippocampal gyrus, paracentric lobule, inferior frontal gyrus, auxiliary motor cortex, middle occipital gyrus, right middle temporal gyrus, while it was increased in the insular, pons, middle frontal gyrus, cingulate gyrus, inferior parietal lobule, bilateral thalamus, and globus pallidus. There were no significant correlations between pain duration and rGMV changes, while a positive correlation was observed between pain severity and rGMV changes in one specific region, involving the anterior cingulate cortex. Conclusion: Total brachial plexus injury patients with chronic pain have widespread regions of gray matter atrophy and hypertrophy. The only positive correlation was observed between pain severity and rGMV changes in one specific region, suggesting that nociceptive stimuli trigger a variety of nonpain-specific processes, which confirms the multidimensional nature of pain.

Neuropathic pain is caused by peripheral nerve lesions that progressively alter neurotransmission and sensory processing within the spinal cord and brain nociceptive pathways, resulting in a complex chronic condition with sensory, emotional, and cognitive components. The functional pain network in humans includes the somatosensory cortex, anterior cingulate cortex (ACC), thalamus, insular lobe, basal ganglia, hippocampus, and the temporal and parietal cortex [1‒4]. During large-scale neuroimaging studies of neuropathic pain, patients have detected widespread morphological brain changes specific to the disorder.

Voxel-based morphometry (VBM), an image processing method used to identify brain structural abnormalities by comparing images of the left and right hemisphere or images from an afflicted population to healthy controls, is widely used to reveal structural changes associated with neurological and mental diseases [5‒7]. A VBM analysis of cerebral changes in neuropathic pain has shown significant volume loss in the gray matter areas of the insula, left middle frontal gyrus, and right subgenual ACC, while increased volumes were not observed in any regions [8]. Using the same technique, Sugimine et al. [9] showed significant positive correlations between the neuropathic pain characteristics of patients with chronic pain and changes in brain regions associated with pain modulation. The results of a large quantitative meta-analysis that included 240 patients with neuropathic pain and 263 healthy controls indicated decrease and increase in gray matter regions known for their involvement in pain processing [10‒16]. These findings suggest that neuropathic pain can cause widespread changes in the brain morphology independent of initial etiology.

Brachial plexus injury is a relatively common cause of neuropathic pain and, similar to other etiologies, is associated with multiple structural and functional changes in the brain [17‒19]. To date, most studies on brachial plexus injury have focused on functional changes by applying resting-state functional connectivity magnetic resonance imaging (MRI) [20‒22], while only a few have examined structural changes. One such study reported loss of gray matter in the postcentral gyrus, a region including the primary somatosensory cortex (S1), after a single peripheral nerve transection [23]. However, a study of patients with brachial plexus injury found that most structural and functional changes were not only in brain areas involving the sensorimotor network but also included higher cognitive networks, such as the salience network and default mode network, indicating compensatory processes had taken place [24]. To the best of our knowledge, there is only one study in which VBM has been used to study structural modifications in brachial plexus injury patients [25]. Lu et al. [25] included 9 patients and 10 healthy volunteers who underwent functional MRI, and VBM was used to analyze the T1 images. Moreover, they also used the visual analog scale (VAS) to assess pain in patients. Their results revealed significant gray matter volume (GMV) atrophy in regions primarily associated with motor function. This is the first study that has provided insight into structural changes in brachial plexus injury patients; however, it has some critical limitations. First, the number of participants is relatively small. Second, they only focused on cerebral regions with atrophy and did not report hypertrophy. Third, their study lacks an analysis between the observed structural changes and clinical parameters, such as VAS scores.

Therefore, the scope of relative gray matter volume (rGMV) changes associated with brachial plexus injury is uncertain, especially in relation to clinical parameters. These changes may provide clues to the underlying pathobiology as well as biomarkers for diagnosis and treatment evaluation. The current study used VBM analysis to identify rGMV changes in patients with chronic pain following total brachial plexus injury. Moreover, to evaluate the potential relationship between clinical parameters of neuropathic pain and changes in GMV, correlation analyses were conducted with the VAS scores and pain duration.

Participants

This case-control study was conducted at the Department of Hand Surgery, Department of Trauma Orthopedics, First Affiliated Hospital of Guangxi Medical University, from 2019 to 2020. Inclusion criteria for the patient group were (1) age 16–65 years, (2) confirmed total brachial plexus injury (see section 2.3 Diagnosis of Total Brachial Plexus Injury), (3) right handed, (4) VAS pain score ≥4, (5) normal neurodevelopment, and (6) no history of traumatic brain injury. No painkillers were used during the study period. Exclusion criteria were (1) mental retardation, poor compliance, or otherwise unable to complete the tasks required; (2) excessive head movement during neuroimaging; (3) other forms of chronic pain before or after injury; (4) other diseases that may cause neuropathic pain in the upper limb such as diabetes, multiple sclerosis, and stroke; (5) amputation; (6) contraindications to MRI; and (7) asked to be excluded for any reason.

A healthy control group was selected according to the following inclusion criteria: (1) adult, (2) right handed, (3) normal brain structure with no history of brain injury, (4) volunteered to participate, and (5) without systemic diseases or current pain. Twenty-eight patients met inclusion criteria (26 males and 2 females, 14 with left and 14 with right side injuries, age range 15–60 years, mean 30.60 years). The duration of the disease (from onset of pain to first visit) ranged from 1 week to 21 years (mean 32 months). The VAS score for pain ranged from 4 to 10 (moderate to severe). In addition, 29 healthy adults met the inclusion criteria for the control group (27 males and 2 females, age range 21–40 years, mean 29.55 years, all right handed) (see Table 1 for details).

Table 1.

Baseline patient characteristics

 Baseline patient characteristics
 Baseline patient characteristics

All procedures were conducted in compliance with the ethical standards of Guangxi Medical University and the National Research Council (China), as well as with the 1964 Helsinki Declaration and subsequent amendments. The study was approved by the Ethics Committee of Guangxi Medical University, and all participants provided informed consent.

Clinical Evaluation

We used a VAS to evaluate pain as follows: 0 (no pain), 1–3 (mild pain), 4–7 (moderate pain), and 8–10 (severe pain).

Diagnosis of Total Brachial Plexus Injury

Total brachial plexus injury was diagnosed based on the following evidence: (1) clear history of trauma, (2) sensory and motor disorders in C5-T1 dominant area, (3) preoperative electromyography showing damage to the five major nerve branches of the brachial plexus, (4) MRI findings of C5-T1 avulsion of all or part of the nerve roots from the intervertebral foramen, and (5) no nerve root in the scalene space or ganglion avulsion outside the intervertebral foramen.

Image Acquisition

All images were acquired using a Siemens PRISMA 3T instrument and 64-channel head coil. A high-resolution three-dimensional structural T1-weighted image of the whole brain was obtained in the axial plane using a magnetization-prepared rapid gradient echo imaging (MPRAGE) sequence with the following parameters: repetition time = 2,300 ms, echo time = 2.98 ms, flip angle = 9°, field of view = 256 × 256 mm, slice thickness = 1 mm, voxel size = 1.0 × 1.0 × 1.0 mm3, and the total number of scanning layers = 176.

Image Data Processing

The raw MRI images were preprocessed as described previously by Ashburner and Friston [26] by using Statistical Parametric Mapping 12 (SPM12; Wellcome Department of Cognitive Neurology, University of London) and the CAT12 toolbox running in MATLAB (version 2013b; The MathWorks, Natick, MA, USA). The quality and homogeneity of raw images were first evaluated by a senior radiologist to filter out unsuitable series due to low contrast, image artifacts, or excessive head motion. Subsequently, the different types of tissues, including the gray matter, white matter, and cerebrospinal fluid were segmented in SPM12. Following segmentation, gray matter templates were normalized by using a DARTEL approach. Then, the tissue data were converted and transferred to the standard template of the Montreal Neurological Institute (MNI) with a cubicle voxel size of 1.5  mm3. The resulting images were modulated by using DARTEL to retain the relative volumes of the gray matter during spatial normalization. As a last step, following segmentation, normalization, and modulation, smoothing was applied to the final GMV images by using an 8-mm full-width at half-maximum Gaussian kernel.

Statistical Analysis

All group comparisons were conducted using the SPSS 23.0. Demographic variables were compared between groups by independent-samples t-test or two test as indicated. A p < 0.05 (two-tailed) was considered significant. rGMV values were also compared by independent-samples t-test, with voxel-wise threshold of p < 0.001 (uncorrected), and cluster-level false discovery rate (FDR) of p < 0.05 (corrected) considered as significant. Relationships between rGMV and VAS scores and rGMV and pain duration were assessed by Pearson’s correlation coefficient.

Data Availability Statement

All data, models, and code generated or used during the study appear in the submitted article.

There were no significant differences in mean age or sex ratio between patient and healthy control groups. Compared with the healthy control group, patients demonstrated significantly reduced rGMV in the bilateral paracentral gyrus, superior temporal gyrus, paracentral lobule, postcentral gyrus, precentral gyrus, middle frontal gyrus, inferior frontal gyrus, auxiliary motor cortex, middle occipital gyrus, superior parietal gyrus, and cuneus, as well as right middle temporal gyrus, right inferior temporal gyrus, and left precuneus. Conversely, patients exhibited significantly greater rGMV in the bilateral insula, pons, middle frontal gyrus, cingulate gyrus, inferior parietal lobule, thalamus, and globus pallidus (see Fig. 1; Table 2, 3).

Table 2.

Peak MNI coordinates and number of voxels of brain regions in which GMV was significantly reduced in neuropathic pain patients compared to healthy controls (FDR-corrected p < 0.001)

 Peak MNI coordinates and number of voxels of brain regions in which GMV was significantly reduced in neuropathic pain patients compared to healthy controls (FDR-corrected p < 0.001)
 Peak MNI coordinates and number of voxels of brain regions in which GMV was significantly reduced in neuropathic pain patients compared to healthy controls (FDR-corrected p < 0.001)
Table 3.

Peak MNI coordinates and number of voxels of brain regions in which GMV was significantly greater in neuropathic pain patients compared to healthy controls (FDR-corrected p < 0.001)

 Peak MNI coordinates and number of voxels of brain regions in which GMV was significantly greater in neuropathic pain patients compared to healthy controls (FDR-corrected p < 0.001)
 Peak MNI coordinates and number of voxels of brain regions in which GMV was significantly greater in neuropathic pain patients compared to healthy controls (FDR-corrected p < 0.001)
Fig. 1.

Differences in regional GMV between healthy controls and patients with neuropathic pain following total brachial plexus injury. Differences between groups are shown as a color-coded t-value map overlaid on structural images in the axial plane. Red areas indicate regions with reduced GMV in patients compared to healthy controls, while blue areas indicate regions with greater GMV in patients than controls (FDR-corrected p< 0.001).

Fig. 1.

Differences in regional GMV between healthy controls and patients with neuropathic pain following total brachial plexus injury. Differences between groups are shown as a color-coded t-value map overlaid on structural images in the axial plane. Red areas indicate regions with reduced GMV in patients compared to healthy controls, while blue areas indicate regions with greater GMV in patients than controls (FDR-corrected p< 0.001).

Close modal

To assess the potential relationship between the duration and severity of neuropathic pain and changes in rGMV, Pearson’s correlation analysis was applied. The results indicated no statistically significant correlations between rGMV and pain duration. The correlation analysis of pain severity as represented by the VAS scores and rGMV only showed a statistically significant positive correlation in cluster 11 (r(28) = 0.71, p < 0.001), which comprises the middle frontal gyrus, cingulate gyrus, and inferior parietal lobule (see Fig. 2).

Fig. 2.

Correlation coefficient between the rGMV of cluster 11 and VAS scores (p< 0.001). rGMV, relative gray matter volume; VAS, visual analog scale; cluster 11, middle frontal gyrus, cingulate gyrus, and inferior parietal lobule.

Fig. 2.

Correlation coefficient between the rGMV of cluster 11 and VAS scores (p< 0.001). rGMV, relative gray matter volume; VAS, visual analog scale; cluster 11, middle frontal gyrus, cingulate gyrus, and inferior parietal lobule.

Close modal

High-resolution T1-weighted structural brain imaging combined with VBM analysis identified 10 voxel clusters in which rGMV was significantly reduced in patients with neuropathic pain following total brachial plexus injury compared to health-matched controls as well as five clusters in which rGMV was higher in patients than controls. The correlation analyses between clinical parameters and rGMV showed that only VAS scores and rGMV had a statistically significant positive correlation in cluster 11.

Most studies on neuropathic pain mechanisms have focused on the spinal cord as both experimental results and theory suggest that neuroplastic changes within the spinal dorsal horn circuitry can induce chronic pain in the absence of noxious inputs. Indeed, abnormal afferent nerve impulses in the spinal dorsal horn, activation of glial cells and the ensuing production of factors promoting hyperalgesia, central sensitization, overexcitation of dorsal horn neurons, and reduction of intrinsic and descending inhibitory signals have been documented following somatosensory nerve injury [1‒3].

Pain is a sensory experience as well as a multidimensional emotional experience involving disgust, anxiety, fear, and motivation to terminate the source. These negative emotions result from the integration of pain-related information in several higher brain centers. Functional imaging studies have shown that numerous cortical and subcortical structures and networks are activated by pain signals, including motor and sensory cortices, anterior cingulate gyrus, the midbrain dopaminergic network, thalamus, periaqueductal gray, raphe nucleus, and locus cerulean, which collectively constitute a complex pain matrix that regulates the development and maintenance of chronic pain.

Areas of Reduced Gray Matter Volume in Chronic Pain following Total Brachial Plexus Injury

The largest cluster of reduced GMV after total brachial plexus injury spanned the limbic system, prefrontal lobe, and temporal lobe, while the second largest cluster was located in the bilateral temporal lobe. The limbic system, prefrontal lobe, and temporal lobe all regulate emotion, and the volumes of various structures within this general region are altered in patients with emotional disorders. For instance, patients with schizophrenia or severe depression have demonstrated reductions in GMV correlated with disease severity [4]. The main symptom after brachial plexus injury in this study was stimulus-independent pain, and many participants had experienced this symptom for many years. A variety of emotional, cognitive, and motivational changes may occur over this period, including maladaptive neuroplastic responses in brain regions and networks associated with emotional, cognitive, and motivational processes. For instance, the parahippocampal gyrus in cluster 1 forms a loop through the hippocampal formation, papillary body, anterior thalamic nucleus, and cingulate gyrus associated with emotion, learning, and memory. This hyperconnectivity can disrupt the normal hippocampal network, in turn impairing learning, memory, and emotional regulation [27]. We speculate that the decreased rGMV in clusters 1 and 2 may be related to the depression caused by long-term pain and poor functional recovery of the affected limb after brachial plexus injury.

Clusters 3, 5, 7, and 10 showing reduced GMV in neuropathic pain patients were located primarily within sensorimotor cortices. Atrophy in these regions may result from reduced neural activity and a concomitant decrease in trophic support, especially in patients showing poor recovery of sensorimotor function of the affected limb. The S1 receives ascending pain signals from the periphery via the thalamus and provides descending feedback that modulates subsequent input. After brachial plexus injury, the sensory cortex of the corresponding limb will lose input, resulting in gray matter atrophy. Further, this area will no longer respond to the signals transmitted by the thalamus, thus causing abnormal sensitization of the thalamus and ensuing stimulus-independent pain. The primary motor cortex (M1) also exhibited significant atrophy. The M1 is an effective target of repetitive transcranial magnetic stimulation to treat neuropathic pain, strongly suggesting involvement in pain control [28]. In addition, phantom limb pain after amputation is associated with reorganization of primary somatosensory and motor areas [14, 29], and the degree of cortical reorganization is closely related to the intensity of pain. Alternatively, even if the decrease in sensorimotor GMV is not directly related to pain, it may indirectly reflect the maladaptive reorganization of the brain that causes neuropathic pain after total brachial plexus injury.

Clusters 4 and 8 contain areas of the prefrontal lobe, the most highly developed area of the brain. The prefrontal lobe can be divided into medial prefrontal cortex (MPFC), preorbital cortex, ventrolateral frontal cortex, dorsolateral prefrontal cortex (DLPFC), and frontal pole. The MPFC and DLPFC are involved in cognitive and emotional regulation of pain. The DLPFC receives pain signals directly from the thalamus, while the MPFC projects to the midbrain aqueduct, allowing this structure to suppress ascending pain transmission. Patients with chronic complex regional pain syndrome demonstrated MPFC atrophy accompanied by decreased white matter integrity and reduced connectivity to the basal ganglia [15]. Pain in these patients was positively correlated with left posterior hippocampal and left amygdalar volumes but negatively correlated with bilateral DLPFC volume [16]. Compared to a healthy control group, fibromyalgia patients exhibited volume decreases in the insula, cingulate gyrus, medial frontal cortex, and parahippocampal gyrus [17]. Further, patients with chronic myofascial pain showed reduced gray matter density in right DLPFC, and greater atrophy was associated with lower pain threshold. Moreover, transcranial magnetic stimulation of the DLPFC can be used to suppress chronic pain. Atrophied clusters 4, 5, 8, and 10 also included regions of the prefrontal lobe. We suggest that a major outcome of this atrophy is reduced connectivity to the midbrain aqueduct and loss of ascending pain suppression.

Clusters 6 and 9 included regions of the anterior cuneate. Experimental pain can lead to inactivation of the precuneus [19] in the anterior cuneate lobe, and an association has been found between gray matter density of the anterior cuneate lobe and pain sensitivity [23]. After total brachial plexus injury, GMV was reduced in the superior parietal lobule, paracentral lobule, precentral gyrus, and extensive areas of prefrontal cortex. Reduced precuneus GMV may be related to the decreased functional connectivity among these cortices previously reported in chronic neuropathic pain patients [30].

Areas of Increased Gray Matter Volume in Chronic Pain after Total Brachial Plexus Injury

Cluster 11, the largest region of increased GMV, included parts of the middle frontal gyrus, cingulate gyrus, and inferior parietal lobule, all structures belonging to the default mode (DMN). The DMN is most active in the resting state and in implicated in monitoring of the internal and external environments, maintenance of consciousness, emotional processing, introspection, and situational memory extraction among other functions. The DMN is most strongly activated when the individual is awake but not focused on the outside world, while routine tasks with external stimuli inhibit DMN activity. Given this function in emotion and introspection, DMN dysfunction may contribute to neuropsychiatric diseases such as Alzheimer’s disease, autism, and schizophrenia [31]. Dysfunctional self-referenced cognitive processing, self-awareness, and self-monitoring by the DMN has also been implicated in other pathologies related to chronic pain. Regardless of the cause of the disease, it is reasonable to assume that long-term invasive diseases such as chronic pain itself alter endosensory function [32, 33]. The chronic spontaneous pain experienced by patients with total brachial plexus injury occurs mostly in the resting state, which may be related to DMN overactivation. The increased rGMV in cluster 11 may be a reflection of this overactivation.

Clusters 12 to 15 showing increased rGMV included areas of the insula, pons, thalamus, and globus pallidus, important components of the brain pain matrix. The pain matrix includes cortical and subcortical areas that interact to form a complex neural network involved in sensory perception, cognition, emotion, and memory. For example, the medial pain system transmits emotional aspects of pain from the thalamus to the insula, while the thalamic projection via the globus pallidus to the ventrolateral dorsal medulla nucleus constitutes a pain suppression system. Similarly, the pons is part of a spinal-thalamus-pons-cerebral cortex pain regulation loop.

Functional overactivation of these pain networks may promote the restructuring of network areas and increased rGMV [34, 35]. The bilateral insula and anterior cingulate gyrus make up the brain’s salience network, which is also associated with emotional control. Activation of the anterior insular lobe is related to the emotional dimension of pain processing and the expectation of pain, while activation of the posterior insular lobe is related to the sensation of pain and body movement. One possible explanation for the increase in insular gray matter volume is chronic input from peripheral pain pathways.

The parabrachial nucleus located in the pons transmits sensory information on visceral discomfort, temperature, and pain to forebrain structures such as thalamus, hypothalamus, and amygdala. The PB is also strongly connected to the amygdala and anterior cingulate gyrus, both of which are implicated in the development of chronic pain. The locus coeruleus (LC), also located in the pons, promotes arousal and vigilance, including under stress. Painful emotional stimulation can also activate neurons in the LC. In addition, stress activates LC neurons and enhances NE synthesis and secretion, pain perception in the prefrontal lobe, and sympathetic activity [36]. Experimental studies have also found that peripheral nerve injury induces neuronal plasticity of inhibitory neurons releasing γ-aminobutyric acid (GABA) in the LC and spinal dorsal horn, resulting in induction of anxiety after pain [37].

The thalamus is another core region for morphological, functional, and biochemical abnormalities associated with neuropathic pain. Neuropathic pain leads to neuroplasticity in the thalamus, including reduced inhibitory input, sensitization, and overexcitation of thalamic neurons, activation of glial cells, and excessive production of inflammatory chemokines. In addition, long-term potentiation of excitatory spinal cord neurons causes overexcitation of nociceptive neurons in the thalamus. A magnetic resonance spectroscopy study found that γ-aminobutyric acid content in the thalamus was significantly reduced in patients with chronic neuropathic pain, which may in turn increase the functional connectivity between the thalamus and S1, secondary somatosensory cortex, and anterior insular [38]. The resulting increase in thalamocortical activity may lead to elevated insular activity and continuous perception of pain. In addition, these rGMV changes may lead to thalamic-cortical dysrhythmia, another contributing factor to the chronic pain state [39].

We speculate that brachial plexus avulsion causes long-term abnormal discharges in the dorsal horn of the spinal cord, and that the upstream signals excite the pons and thalamus, which are involved in the transmission of ascending pain signals and in descending pain control, resulting in increased pontine and thalamic GMV. Overexcitation of these structures could influence basal ganglia-thalamic loop inhibitory transmission and the GMV of brain areas (such as the globus pallidus) involved in pain inhibition.

A meta-analysis by Cauda et al. [32] concluded that chronic pain is associated with reduced GMV in the bilateral medial frontal gyrus, bilateral superior frontal gyrus, right precentral gyrus, and postcentral gyrus (including primary somatosensory and M1), bilateral anterior insular lobe, right dorsal cingulate gyrus, basal ganglia, thalamus, and periaqueductal gray but increased GMV in bilateral postcentral gyrus, left inferior parietal lobule, right precentral gyrus (M1), right postcentral gyrus, dorsal frontal lobe, caudate nucleus, thalamus, cerebellum, and pons. Another meta-analysis found a progressive GMV decrease in the insula, thalamus, superior frontal gyrus, and postcentral gyrus of neuropathic pain patients but greater GMV in the medial frontal gyrus and posterior insular [10]. While these findings are not in complete accord with the current study, no meta-analysis included patients with pain after brachial plexus injury. We suggest that the structural changes in the cerebral gray matter following total brachial plexus injury are very similar to those caused by other forms of chronic pain, albeit with certain unique characteristics.

Gray Matter Volume and Clinical Parameters

Patients with a broad range of different pain durations (from 1 month to 21 years) were enrolled in this study. To evaluate whether the duration of pain or severity could affect changes in rGMV, correlation analysis was conducted. Statistically significant correlations were not observed between rGMV and the duration of pain, while a positive correlation was found between the rGMV increase in cluster 11 and VAS scores. Sugimine et al. [9] and Schmidt-Wilcke et al. [40] both reported comparable findings. Cluster 11 comprises different regions among which the ACC is part of the cingulate gyrus. Studies have suggested that structural, functional, and biochemical changes in the spinal dorsal horn can enhance action potential generation in nociceptive pathways and subsequent activation of higher pain-associated centers in the central nervous system that are vital in chronic pain, such as the amygdala, insular cortex, prefrontal cortex, and ACC [41, 42]. These findings may explain the positive correlation observed in this study.

VBM with stringent FDR correction revealed that patients with chronic pain following total brachial plexus injury exhibit widespread regions of gray matter atrophy as well as gray matter hypertrophy. Although there were no significant correlations between pain duration and rGMV changes, a positive correlation was observed between pain severity and rGMV changes in a specific region. However, no other correlations were observed, possibly because nociceptive stimuli trigger a variety of cognitive, emotional, motivational, autonomic, and motor processes that are not specific to pain but are part of the multidimensional nature of the pain experience. Further study is needed to elucidate the functional implications of these rGMV changes in chronic pain development and disease expression.

The authors would like to thank 51runse (www.51runse.cn) for the English language editing during the preparation of the manuscript.

All procedures were conducted in compliance with the ethical standards of Guangxi Medical University and the National Research Council (China), as well as with the 1964 Helsinki Declaration and subsequent amendments. The study was approved by the Ethics Committee of Guangxi Medical University, approval number (2021[KY-E-328]) and written informed consent was obtained from all participants for participation in this study. For underaged participants, written informed consent was obtained from the participants’ parent/legal guardian/next of kin to participate in the study.

The authors have no conflicts of interest to declare.

This work was supported by the Shanghai Key Laboratory of Peripheral Nerve and Microsurgery (Grant No. 20DZ2270200); the NHC Key Laboratory of Hand Reconstruction (Fudan University), Shanghai, People’s Republic of China; and the Guangxi Science and Technology Base and Talent Special Project (Grant No. GuikeAD19254003).

Jing-wei Wang, Zhu-qing Huang, Wen-mei Li, and Jin-min Zhao contributed to the study concept, design, and data analysis. Jing-wei Wang, Zhu-qing Huang, Yu-jie Lu, Ke Sha, Wen-mei Li, and Jin-min Zhao contributed to the data acquisition and interpretation. Jing-wei Wang and Zhu-qing Huang contributed to the drafting of the manuscript. Jing-wei Wang, Zhu-qing Huang, Yu-jie Lu, Ke Sha, Wen-mei Li, and Jin-min Zhao have read and approved the final draft of the manuscript.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.

Additional Information

Jing-wei Wang and Zhu-qing Huang contributed equally to this work.

1.
Vranken
JH
.
Mechanisms and treatment of neuropathic pain
.
Cent Nerv Syst Agents Med Chem
.
2009 Mar
;
9
(
1
):
71
8
.
2.
H Vranken
J
.
Elucidation of pathophysiology and treatment of neuropathic pain
.
Cent Nerv Syst Agents Med Chem
.
2012 Dec
;
12
(
4
):
304
14
.
3.
Boadas-Vaello
P
,
Castany
S
,
Homs
J
,
Álvarez-Pérez
B
,
Deulofeu
M
,
Verdú
E
.
Neuroplasticity of ascending and descending pathways after somatosensory system injury: reviewing knowledge to identify neuropathic pain therapeutic targets
.
Spinal Cord
.
2016 May
;
54
(
5
):
330
40
.
4.
Serra-Blasco
M
,
Portella
MJ
,
Gómez-Ansón
B
,
de Diego-Adeliño
J
,
Vives-Gilabert
Y
,
Puigdemont
D
, et al
.
Effects of illness duration and treatment resistance on grey matter abnormalities in major depression
.
Br J Psychiatry
.
2013 Jun
;
202
(
6
):
434
40
.
5.
Kucyi
A
,
Davis
KD
.
The dynamic pain connectome
.
Trends Neurosci
.
2015 Feb
;
38
(
2
):
86
95
.
6.
Kucyi
A
,
Davis
KD
.
The neural code for pain: from single-cell electrophysiology to the dynamic pain connectome
.
Neuroscientist
.
2017 Aug
;
23
(
4
):
397
414
.
7.
del Rey
A
,
Apkarian
AV
,
Martina
M
,
Besedovsky
HO
.
Chronic neuropathic pain-like behavior and brain-borne IL-1β
.
Ann N Y Acad Sci
.
2012 Jul
;
1262
(
1
):
101
7
.
8.
Yoon
EJ
,
Kim
YK
,
Shin
HI
,
Lee
Y
,
Kim
SE
.
Cortical and white matter alterations in patients with neuropathic pain after spinal cord injury
.
Brain Res
.
2013 Dec 2
;
1540
:
64
73
.
9.
Sugimine
S
,
Ogino
Y
,
Kawamichi
H
,
Obata
H
,
Saito
S
.
Brain morphological alternation in chronic pain patients with neuropathic characteristics
.
Mol Pain
.
2016 Jun 8
;
12
:
174480691665240
.
10.
Pan
PL
,
Zhong
JG
,
Shang
HF
,
Zhu
YL
,
Xiao
PR
,
Dai
ZY
, et al
.
Quantitative meta-analysis of grey matter anomalies in neuropathic pain
.
Eur J Pain
.
2015 Oct
;
19
(
9
):
1224
31
.
11.
Eto
K
,
Wake
H
,
Watanabe
M
,
Ishibashi
H
,
Noda
M
,
Yanagawa
Y
, et al
.
Inter-regional contribution of enhanced activity of the primary somatosensory cortex to the anterior cingulate cortex accelerates chronic pain behavior
.
J Neurosci
.
2011 May 25
;
31
(
21
):
7631
6
.
12.
Vierck
CJ
,
Whitsel
BL
,
Favorov
OV
,
Brown
AW
,
Tommerdahl
M
.
Role of primary somatosensory cortex in the coding of pain
.
Pain
.
2013 Mar
;
154
(
3
):
334
44
.
13.
Stadler
JA
3rd
,
Ellens
DJ
,
Rosenow
JM
.
Deep brain stimulation and motor cortical stimulation for neuropathic pain
.
Curr Pain Headache Rep
.
2011 Feb
;
15
(
1
):
8
13
.
14.
Flor
H
,
Denke
C
,
Schaefer
M
,
Grüsser
S
.
Effect of sensory discrimination training on cortical reorganisation and phantom limb pain
.
Lancet
.
2001 Jun 2
;
357
(
9270
):
1763
4
.
15.
Geha
PY
,
Baliki
MN
,
Harden
RN
,
Bauer
WR
,
Parrish
TB
,
Apkarian
AV
.
The brain in chronic CRPS pain: abnormal gray-white matter interactions in emotional and autonomic regions
.
Neuron
.
2008 Nov 26
;
60
(
4
):
570
81
.
16.
Barad
MJ
,
Ueno
T
,
Younger
J
,
Chatterjee
N
,
Mackey
S
.
Complex regional pain syndrome is associated with structural abnormalities in pain-related regions of the human brain
.
J Pain
.
2014 Feb
;
15
(
2
):
197
203
.
17.
Kuchinad
A
,
Schweinhardt
P
,
Seminowicz
DA
,
Wood
PB
,
Chizh
BA
,
Bushnell
MC
.
Accelerated brain gray matter loss in fibromyalgia patients: premature aging of the brain
.
J Neurosci
.
2007 Apr 11
;
27
(
15
):
4004
7
.
18.
Niddam
DM
,
Lee
SH
,
Su
YT
,
Chan
RC
.
Brain structural changes in patients with chronic myofascial pain
.
Eur J Pain
.
2017 Jan
;
21
(
1
):
148
58
.
19.
Seminowicz
DA
,
Davis
KD
.
Pain enhances functional connectivity of a brain network evoked by performance of a cognitive task
.
J Neurophysiol
.
2007 May
;
97
(
5
):
3651
9
.
20.
Liu
B
,
Li
T
,
Tang
WJ
,
Zhang
JH
,
Sun
HP
,
Xu
WD
, et al
.
Changes of inter-hemispheric functional connectivity between motor cortices after brachial plexuses injury: a resting-state fMRI study
.
Neuroscience
.
2013 Jul 23
;
243
:
33
9
.
21.
Lu
Y
,
Liu
H
,
Hua
X
,
Xu
WD
,
Xu
JG
,
Gu
YD
.
Supplementary motor cortical changes explored by resting-state functional connectivity in brachial plexus injury
.
World Neurosurg
.
2016 Apr
;
88
:
300
5
.
22.
Feng
JT
,
Liu
HQ
,
Hua
XY
,
Gu
YD
,
Xu
JG
,
Xu
WD
.
Brain functional network abnormality extends beyond the sensorimotor network in brachial plexus injury patients
.
Brain Imaging Behav
.
2016 Dec
;
10
(
4
):
1198
205
.
23.
Emerson
NM
,
Zeidan
F
,
Lobanov
OV
,
Hadsel
MS
,
Martucci
KT
,
Quevedo
AS
, et al
.
Pain sensitivity is inversely related to regional grey matter density in the brain
.
Pain
.
2014 Mar
;
155
(
3
):
566
73
.
24.
Bhat
DI
,
Indira Devi
B
,
Bharti
K
,
Panda
R
.
Cortical plasticity after brachial plexus injury and repair: a resting-state functional MRI study
.
Neurosurg Focus
.
2017 Mar
;
42
(
3
):
E14
.
25.
Lu
Y
,
Liu
H
,
Hua
X
,
Xu
JG
,
Gu
YD
,
Shen
Y
.
Attenuation of brain grey matter volume in brachial plexus injury patients
.
Neurol Sci
.
2016 Jan
;
37
(
1
):
51
6
.
26.
Ashburner
J
,
Friston
KJ
.
Voxel-based morphometry: the methods
.
Neuroimage
.
2000 Jun
;
11
(
6
):
805
21
.
27.
Mutso
AA
,
Petre
B
,
Huang
L
,
Baliki
MN
,
Torbey
S
,
Herrmann
KM
, et al
.
Reorganization of hippocampal functional connectivity with transition to chronic back pain
.
J Neurophysiol
.
2014 Mar
;
111
(
5
):
1065
76
.
28.
García-Larrea
L
,
Peyron
R
,
Mertens
P
,
Gregoire
CM
,
Lavenne
F
,
Le Bars
D
, et al
.
Electrical stimulation of motor cortex for pain control: a combined PET-scan and electrophysiological study
.
Pain
.
1999 Nov
;
83
(
2
):
259
73
.
29.
MacIver
K
,
Lloyd
DM
,
Kelly
S
,
Roberts
N
,
Nurmikko
T
.
Phantom limb pain, cortical reorganization and the therapeutic effect of mental imagery
.
Brain
.
2008 Aug
;
131
(
8
):
2181
91
.
30.
Alomar
S
,
Bakhaidar
M
.
Neuroimaging of neuropathic pain: review of current status and future directions
.
Neurosurg Rev
.
2018 Jul
;
41
(
3
):
771
7
.
31.
Broyd
SJ
,
Demanuele
C
,
Debener
S
,
Helps
SK
,
James
CJ
,
Sonuga-Barke
EJ
.
Default-mode brain dysfunction in mental disorders: a systematic review
.
Neurosci Biobehav Rev
.
2009 Mar
;
33
(
3
):
279
96
.
32.
Cauda
F
,
Palermo
S
,
Costa
T
,
Torta
R
,
Duca
S
,
Vercelli
U
, et al
.
Gray matter alterations in chronic pain: a network-oriented meta-analytic approach
.
Neuroimage Clin
.
2014 Apr 16
;
4
:
676
86
.
33.
Ostrowsky
K
,
Magnin
M
,
Ryvlin
P
,
Isnard
J
,
Guenot
M
,
Mauguière
F
.
Representation of pain and somatic sensation in the human insula: a study of responses to direct electrical cortical stimulation
.
Cereb Cortex
.
2002 Apr
;
12
(
4
):
376
85
.
34.
Sawamoto
N
,
Honda
M
,
Okada
T
,
Hanakawa
T
,
Kanda
M
,
Fukuyama
H
, et al
.
Expectation of pain enhances responses to nonpainful somatosensory stimulation in the anterior cingulate cortex and parietal operculum/posterior insula: an event-related functional magnetic resonance imaging study
.
J Neurosci
.
2000 Oct 1
;
20
(
19
):
7438
45
.
35.
Zhang
D
,
Raichle
ME
.
Disease and the brain’s dark energy
.
Nat Rev Neurol
.
2010 Jan
;
6
(
1
):
15
28
.
36.
Moazen
P
,
Torabi
M
,
Azizi
H
,
Fathollahi
Y
,
Mirnajafi-Zadeh
J
,
Semnanian
S
.
The locus coeruleus noradrenergic system gates deficits in visual attention induced by chronic pain
.
Behav Brain Res
.
2020 Jun 1
;
387
:
112600
.
37.
Yoshizumi
M
,
Parker
RA
,
Eisenach
JC
,
Hayashida
K
.
Gabapentin inhibits γ-amino butyric acid release in the locus coeruleus but not in the spinal dorsal horn after peripheral nerve injury in rats
.
Anesthesiology
.
2012 Jun
;
116
(
6
):
1347
53
.
38.
Henderson
LA
,
Peck
CC
,
Petersen
ET
,
Rae
CD
,
Youssef
AM
,
Reeves
JM
, et al
.
Chronic pain: lost inhibition
.
J Neurosci
.
2013 Apr 24
;
33
(
17
):
7574
82
.
39.
Llinás
RR
,
Ribary
U
,
Jeanmonod
D
,
Kronberg
E
,
Mitra
PP
.
Thalamocortical dysrhythmia: a neurological and neuropsychiatric syndrome characterized by magnetoencephalography
.
Proc Natl Acad Sci USA
.
1999 Dec 21
;
96
(
26
):
15222
7
.
40.
Schmidt-Wilcke
T
,
Leinisch
E
,
Gänbauer
S
,
Draganski
B
,
Bogdahn
U
,
Altmeppen
J
, et al
.
Affective components and intensity of pain correlate with structural differences in gray matter in chronic back pain patients
.
Pain
.
2006 Nov
;
125
(
1
):
89
97
.
41.
Apkarian
AV
,
Bushnell
MC
,
Treede
RD
,
Zubieta
JK
.
Human brain mechanisms of pain perception and regulation in health and disease
.
Eur J Pain
.
2005 Aug
;
9
(
4
):
463
.
42.
Tsuda
M
,
Koga
K
,
Chen
T
,
Zhuo
M
.
Neuronal and microglial mechanisms for neuropathic pain in the spinal dorsal horn and anterior cingulate cortex
.
J Neurochem
.
2017 May
;
141
(
4
):
486
98
.