Introduction

Sensory plasticity encodes environmental experiences through functional and structural reorganizational processes that shape memory, perceptual sensitivity and behavior [1]. However, pathological alterations in sensory system’s nociceptive pathways can cause corresponding plasticity to become chronically maladaptive, in response to particularly intense or repetitive noxious triggers, and mediate pain chronification as a form of memory [2]. This is especially evident for the somatosensory system in specific types of neuropathic pain through potentiation of nociception, central sensitization and altered brain connectivity [3, 4]. Since plasticity represents an intrinsic activity-dependent neuronal ability, this suggests that it is essentially reversible or at least modifiable and that sensory experience, learning mechanisms and psychology are critical treatment factors affecting pain perception [2, 5]. Accordingly, a promising therapeutic approach would be to interfere with these alterations or induce the reversal process as “physiologically” as possible; that is, accelerating recovery by resetting the system to the original baseline state via activity or use-dependent mechanisms, reflecting sensory experience, such as neurostimulation [6, 7]. Certain non-pharmacological interventions including neurostimulation and other modalities show significant potentials in the management of maladaptive plasticity of chronic pain and the prevention of acute-to-chronic pain transition. In this review we discuss novel research findings on the efficacy of these treatment modalities in nociceptive desensitization and associated neural plasticity mechanisms.

Nociceptive pathways

Noxious mechanical, thermal and chemical stimuli of cutaneous and visceral tissues are detected by afferent sensory neurons known as nociceptors, which are generally classified into fast Aδ and slow C-type fibers. Acute tissue injury triggers the local release of various inflammatory mediators including adenosine triphosphate (ATP), bradykinins, histamine, prostaglandins, neurotrophic factors and cytokines [8]. These mediators activate nociceptors to generate action potential impulses to transmit detected inputs to central neurons. In addition, activation of C-fibers initiates neurogenic inflammation characterized by the upregulated production and release of neuropeptides such as neurokinin A, substance P and calcitonin gene-related peptide (CGRP) in a retrograde manner [9]. Together, these alterations lead to reversible sensitization of peripheral nociceptors with reduction in activation threshold and increased stimulus-induced and spontaneous discharge [10]. Within the laminae of the spinal dorsal horn the primary afferents, except sensory cranial nerves, synapse with second-order afferents, which include three groups: proprioceptive, nociceptive and wide dynamic range (WDR) neurons. The nociceptive signals are transmitted from primary to secondary afferents via the release of different neurotransmitters, particularly glutamate and substance P. However, dorsal horn nociceptive transmission is also regulated by inhibitory mechanisms involving endogenous opioids, descending inhibitory pathways and inhibitory interneurons releasing γ-amino butyric acid (GABA) and glycine. The secondary afferents form ascending pathways that project to the brain stem and medulla and terminate in the thalamus and cerebral cortex. Lastly, an extensive cortical network commonly referred to as the pain matrix processes nociceptive and other salient sensory inputs [11].

Nociceptive sensitization

Sensitization of nociception is a temporary adaptive process that occurs following inflammatory and noxious tissue insults and involves lowered pain threshold and amplified responses due to nociceptive neuronal hyperexcitability [12]. However, maladaptation in this protective response can arise from central or peripheral pathologies; for instance neuropathy, and lead to chronic pain characterized by allodynia and hyperalgesia [13]. The enhanced responsiveness of central nociceptive neurons is known as central sensitization, which is hypothesized to result in chronic amplification of pain associated with ongoing tissue inflammation, following neuronal injury or even in the absence of peripheral pathology such as in migraine, fibromyalgia and irritable bowel syndrome [14]. Central sensitization is mediated by neural plasticity mechanisms that involve increased neuronal activity, potentiated synaptic efficacy, enlarged receptive fields and reduced inhibition [15]. Therefore, pain “perception” would no longer be coupled to the presence, intensity or duration of noxious inputs, rendering localized peripheral treatments less effective; thus, central sensitization is thought to account, at least partly, for unexplained chronic pain [16]. Peripheral sensitization processes on the level of primary afferents and free nerve endings have also been identified and involve plasticity changes of nociceptors leading to primary hyperalgesia [17]. The maladaptive plasticity processes outlasting tissue healing are, in certain subgroups of patients, associated with various forms of intractable chronic pain; accordingly, central sensitization and associated psychocognitive factors should be taken into account for developing individualized treatments [18].

Neural plasticity of pain

Plasticity changes occur along pain pathways throughout the neuroaxis and mediate peripheral nociceptor sensitization, central (spinal) sensitization and brain remodeling (Fig. 1). Peripherally, inflammatory mediators and retrograde neuropeptides “sensitize” nociceptors leading to upregulation of substance P, transient receptor potential vanilloid (TRPV) and purinergic receptors; in addition to altered membrane ion channels, protein kinase activity and growth factor expression resulting in hypersensitivity and primary hyperalgesia [17]. The main excitatory transmitter of nociceptive neurons is glutamate, which acts upon 3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), N-methyl-d-aspartate (NMDA), kainic acid (KA) and metabotropic glutamate (mGlu) receptors. Under baseline conditions (resting membrane potential), NMDAR channel pore is blocked by Mg2+ ions [19]. During intense, repeated or sustained nociceptor activation, as with neuroinflammation and nerve injury, the continued release of neuropeptides such as substance P and CGRP from primary afferents on dorsal horn neurons provides enough postsynaptic depolarization to expel Mg2+ ions and relieve NMDAR channel blockade. Subsequently, the binding of glutamate to NMDA receptors generates a strong inflow of Ca2+ ions and mediates long-term potentiation (LTP) of dorsal horn excitatory transmission [20]. Other sources for intracellular Ca2+ in dorsal horn neurons include the Ca2+-permeable GluA1 subunit-lacking AMPARs and the mobilization of Ca2+ from intracellular stores, the latter of which is mediated by the activation of group-I mGluRs. The generated Ca2+ signals and the activity of various peptides including substance P, CGRP and brain-derived neurotrophic factor (BDNF) activate protein kinases such as calcium/calmodulin-dependent protein kinase II (CAMKII), protein kinase A (PKA) and protein kinase C (PKC) [15]. These kinases are found to mediate the induction and early-phase of LTP through AMPA receptor phosphorylation and synaptic insertion following C-fiber tetanization, while the maintenance or late-phase of LTP requires de novo protein synthesis [21, 22]. Other mechanisms of functional plasticity in dorsal horn neurons include disinhibition, glial activation and nitric oxide (NO)-dependent retrograde signaling, which leads to increased neurotransmitter release probability from C-fiber terminals [15]. In addition, delayed structural plasticity changes are observed on the level of dendritic spine size and density and involve alterations in gene expression and connectivity [23]. Within the brain, similar functional and structural plasticity alterations associated with multiple chronic pain syndromes have been documented in various regions including the brain stem, thalamus, insular cortex, cingulate cortex and primary somatosensory cortex [24]. While it remains unclear if brain remodeling is a cause or consequence of chronic pain, it was proposed that pain-associated plasticity changes within brain circuits resemble associative learning and memory trace formation; thus, rendering pain perception more affective than somatic in nature [25]. Accordingly, structural spinal and supraspinal remodeling are hypothesized to mediate pain chronification with representational shifting towards emotional than nociceptive circuits [26]. It should be noted; however, that not all forms of chronic pain, and not all patients with a specific type of pain, show these reorganizations. As previously shown, patients with orofacial neuropathic pain exhibit cortical somatosensory remodeling; however, patients with chronic non-neuropathic orofacial pain do not [27]. Therefore, individualized targeting of central sensitization and brain plasticity may provide the means for preventing acute-to-chronic pain transitioning and the reversal of pathologic plasticity in specific forms of chronic pain. Various modalities have been investigated for potential effectiveness in nociceptive desensitization as modulators of neural activity and plasticity such as neurostimulation, virtual reality, cognitive therapy and rehabilitation.

Fig. 1
figure 1

Nociceptive pathways and the main loci of maladaptive plasticity. Activation of peripheral nociceptors generates action potentials that are transmitted to secondary afferents in the dorsal horn, which then project to the brain for processing. Particularly intense tissue insults result in maladaptive plasticity including peripheral sensitization in primary afferents, central sensitization in dorsal horn secondary afferents and brain remodeling. AP action potential, AMPAR 3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor, DRG dorsal root ganglion, Glu glutamate, mGluR metabotropic glutamate receptor, NK1 neurokinin receptor, NMDAR N-methy-d-aspartate receptor, PKA/PKC protein kinase A/C, SP substance-P, VGCC voltage-gated calcium channel

Full size image

Modulating maladaptive plasticity

Neurostimulation

Neurostimulation is a neuromodulatory method based on the delivery of electrical impulses to stimulate specific neurological sites within the body intended for various diagnostic and treatment purposes [28]. Multiple invasive and non-invasive neurostimulation techniques have been developed for the management of pain, especially chronic forms, which can be used to stimulate peripheral nerves, the spinal cord or specific brain regions. Peripherally, the techniques include transcutaneous electrical nerve stimulation (TENS), peripheral nerve stimulation (PNS) and peripheral nerve field stimulation (PNFS) [29]. Central neurostimulation techniques include spinal cord stimulation (SCS), non-invasive brain stimulation (NIBS) and invasive brain stimulation techniques such as deep brain stimulation (DBS). NIBS techniques include transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS) [30]. In this review, we will focus on SCS and NIBS techniques in relation to maladaptive plasticity of central sensitization and brain remodeling in chronic pain.

Spinal cord stimulation

The basic principle behind SCS development is the stimulation of ascending non-nociceptive Aβ fibers to block or close the gate for C-fiber nociceptive transmission, through activating inhibitory interneurons, based on the Gate Control Theory of pain [31]. However, the analgesic mechanism of SCS is complex and affects various aspects of pain through spinal and supraspinal mechanisms, which together with the analgesic efficacy can vary across different stimulation protocols [32]. These include tonic (conventional), high-frequency (paresthesia-free or high-dose stimulation), burst and closed-loop SCS waveforms [33]. In relation to clinical efficacy, high-frequency (10 kHz) and burst SCS are shown to be effective and superior to tonic or conventional SCS in chronic back and leg pain, failed back surgery syndrome (FBSS) and intractable and diabetic neuropathy [34,35,36,37,38]. In addition, various clinical studies demonstrate effectiveness of high-frequency SCS in relieving other forms of pain including chronic neck and upper limb pain [39], thoracic back pain [40], chronic pelvic pain [41], chronic post-surgical pain [42] and chronic widespread pain / fibromyalgia [43]. In relation to central maladaptive plasticity, high-frequency SCS is clinically found, at 3 months of application, to enhance the functional connectivity between the insula, frontoparietal and central executive networks in patients with FBSS, suggesting potential influence on affective saliency and thus emotional awareness of pain [44]. Furthermore, magnetic resonance imaging (MRI) in FBSS patients who received high-frequency SCS for 3 months revealed significant volumetric alterations of white and grey matter in various brain regions, which correlated with pain relief [45]. These studies confirm the supraspinal modulatory effects of SCS and further support the reversibility of chronic pain-induced alterations of brain connectivity. Regarding central sensitization of secondary spinal afferents, multiple in vitro studies investigated the effects of SCS using rat models. Early findings using rodent models showed that tetanization, nerve injury and acute noxious stimuli (chemical, mechanical and thermal) induce C-fiber synaptic LTP on dorsal horn neurons [46] including WDR neurons [47] with subsequent hyperexcitability of WDR neurons [48]. Appropriately, the application of SCS is found to block these effects by inhibiting dorsal horn C-fiber LTP on WDR neurons with no effect on A-fiber responses [49], decreasing spinal excitatory amino acid release via a GABAergic mechanism in neuropathic rats experiencing allodynia [50] and attenuating the increased WDR neuronal excitability without affecting induced or spontaneous discharge in control non-allodynic rats [51]. To unravel the underlying mechanisms, extensive research has been recently conducted providing novel insights into the molecular basis corresponding to SCS excitability normalization and reversal of central sensitization. A study by Tilley and colleagues (2021) showed, through proteomic analysis, that conventional SCS influenced the expression of over 150 proteins, many of which are involved in stress, nociception and neuroglial interactions [52]. Accordingly, the results not only show the reversal of pain-associated proteomic profiles but further indicate that the mechanism of SCS is not solely dependent on the interruption of electrical transmission. Another study by Liao and colleagues (2020) revealed that spared nerve injury (SNI) model of neuropathic pain results in mechanical hyperalgesia and increased expression and phosphorylation of extracellular signal-regulated kinases (ERKs), c-Jun N-terminal kinases (JNKs), and p38 mitogen-activated protein kinase (p38-MAPK), which are important regulators of neuronal activity and plasticity. Importantly, early application of high-frequency SCS was able to prevent these alterations in dorsal root ganglia (DRG) and spinal dorsal horn, which associated with the attenuation of hyperalgesia [53]. Furthermore, Shinoda and colleagues (2020) investigated the effects of SCS on SNI, applied at 60 Hz for 6 h on the third day of SNI, as well. The findings show that SCS suppressed mechanical hypersensitivity, microglial activation and dorsal horn nociceptive hyperexcitability; additionally, SCS reduced somatosensory neuronal activity [54]. Interestingly, it is also reported that conventional SCS can activate microglia and thus compromise its own analgesic efficacy as preventing microglial activation prolonged the pain inhibition induced by SCS [55]. As microglia are important modulators of neurotransmission, neuroglial interactions and pain signaling with hypothesized roles in the pathogenesis of neuropathic pain [56], these results highlight the regulation of microglial activity by SCS as a potentially essential analgesic mechanism in neuropathic pain. Lastly, the application of high-frequency SCS in rats with SNI-induced neuropathic pain was observed to restore, at least partially, spinal glutamate uptake activity, spinal glutamate levels and miniature excitatory postsynaptic current (mEPSC) frequency [57]. Therefore, SCS induces various spinal and supraspinal modifications that reverse pain-associated maladaptive plasticity in relation to brain connectivity and central sensitization.

Non-invasive brain stimulation

NIBS techniques have significant potentials in the study and treatment of various psychiatric and neurological disorders including pain [58]. However, it must be stressed that the underlying mechanisms are complex and the clinical effects depend on various factors leading to high degree of variability [59]. Therefore, the best clinical practice for the use NIBS techniques in pain management should be derived from standardized guidelines [60, 61]. The two main studied NIBS techniques are TMS and tDCS. The principle of TMS is the non-invasive application of magnetic field pulses, which carry a current through the skull and excite superficial layers of the cortex. The two main cortical region to which TMS is applied are the primary motor (M1) cortex and the dorsolateral prefrontal cortex (DLPFC), both of which are associated with chronic pain and affect various aspects of pain processing [62, 63]. The repetitive application of TMS (rTMS) produces long-lasting, stimulation frequency-dependent excitability effects [64, 65] and leads to widespread activity changes in connected cortical and subcortical regions, yet the alterations of functional connectivity remain network-specific [66, 67]. The rTMS-induced excitability changes are believed to be mediated through neural plasticity mechanisms [68]. Indeed, evidence shows that rTMS-induced changes are NMDAR-dependent, lead to enhanced BDNF function [69] and able to induce potentiation [70] and depression [71] of excitability. Furthermore, rTMS application in rodents is observed to enhance cognition, facilitate hippocampal plasticity and increase the levels of various plasticity markers [72]. Accordingly, accumulating evidence supports the hypothesis that rTMS accelerates recovery of sensory and motor functions after stroke, incomplete spinal cord injury and nerve injury by promoting synaptic plasticity and thereby reversing maladaptive plasticity [73,74,75]. In addition, the LTP-like plasticity induced by rTMS treatment correlates with cognitive function improvement in Alzheimer’s disease patients [76]. Similarly, early treatment with rTMS is proposed to block pain-associated maladaptive plasticity induced by surgery, spinal injury and brain trauma; thus, preventing acute-to-chronic pain transitioning [77]. In a recent meta-analysis, Che and colleagues (2021) found that rTMS exerts a short-term analgesic effect that is specific to neuropathic pain, a long-term (average of 3 month) analgesic effect across multiple chronic pain conditions and significant analgesia of provoked pain, which could model either acute pain or acute-to-chronic pain transition [78]. These findings support the general consensus that rTMS exerts a multitude of mechanisms that could differentially modulate specific types of pain and indicate an acute analgesic effect that could be independent from the maladaptive plasticity associated with chronic pain. Indeed, rTMS application is shown to activate opioid-mediated analgesia of acute pain in healthy individuals [79], induce dose-dependent immediate analgesia following stimulation in patients with intractable neuropathic pain [80] and elevate electrical pain thresholds up to 40 min following application over the somatosensory cortex of healthy subjects without altering the excitability of the M1 cortex [81]. Therefore, rTMS carries significant potentials in both: prevention of acute-to-chronic pain transition, through acute analgesia and prevention of maladaptive plasticity, and treatment of chronic pain through reversal of maladaptive plasticity. On the other hand, the principle of tDCS is the passage of current between two electrodes; thus, anodal tDCS leads to depolarization while cathodal stimulation causes hyperpolarization. In contrast to TMS, which can stimulate cortical neuronal axons to fire action potentials, the effects of tDCS are more electrically subtle. This is due to weaker current pulses affecting membrane excitability (subthreshold potential alterations); however, depending on the duration and frequency of application it can also induce long-lasting effects mediated by intracortical inhibition and facilitation [82]. In relation to neural plasticity, the application of tDCS is found in rodent models to promote BDNF-dependent synaptic plasticity [83], enhance synaptic plasticity and memory [84] and improve plasticity deficits and cognitive dysfunction associated with diabetes [85]. Clinically, tDCS facilitates the formation of long-term motor memory, reflecting experience-dependent plasticity [86], improves motor performance in the elderly via enhanced facilitation and reduced inhibition [87] and, at short stimulation intervals, leads to LTP-like excitability enhancements in healthy participants [88]. These findings indicate significant facilitatory effects of tDCS on neuronal plasticity, which could thereby accelerate the recovery from, or prevent the development of, maladaptive plasticity similar to rTMS. Various studies support the potential analgesic efficacy of tDCS in multiple chronic neuropathic pain conditions [89] as well as migraine, osteoarthritis and capsaicin-induced mechanical sensitivity [90,91,92]. Furthermore, tDCS has no impact on pain thresholds and mechanical detection in healthy individuals [93]. However, the analgesic response to tDCS depends on many factors; hence, it is not effective in all patients with neuropathic pain [94]. The mechanisms underlying direct tDCS-induced analgesia are not completely understood; however, the effects may not only be related to increased or decreased neuronal firing rates as reports suggest the engagement of endorphins [95, 96] and, in addition to modulation of glutamatergic and GABAergic balance [97], the alteration of certain neuromodulators such as dopamine [98]. In rodent models, other neuromodulators are also found to significantly affect tDCS responses including serotine [99] and norepinephrine [100]. These neuromodulators are known to regulate neuronal activity, synaptic plasticity, input processing and associated neurological functions [101]. Lastly, many other tDCS effects were proposed to mediate analgesia in relation to altered pain processing and modulation of its emotional aspects [102].

Virtual reality

Virtual reality (VR) is a technology that provides an immersive experience in a simulated and interactive environment via multimodal sensory stimuli including visual, auditory and tactile inputs using computer hardware. The potential VR applications in the medical field were recognized over two decades ago, such as education, surgery and rehabilitation [103]. Since then, extensive research has been conducted to evaluate the therapeutic application of VR in various conditions. These include recovery from stroke [104], improving motor function in cerebral palsy [105], managing post-traumatic stress disorder [106], alleviating perioperative pain and anxiety [107], treatment of phobias [108] and management of acute and chronic pain [109] especially phantom-limb pain [110]. The effectiveness of VR in pain management, commonly termed VR analgesia, can be generally attributed to distraction, or the shifting of attention away from pain, with potential affective aspects. Early evidence demonstrated that increased pain vigilance and awareness in patients with chronic pain is associated with higher feelings of distress and disability [111] while distraction through cognitively demanding tasks reduces perceived pain intensity and neuronal activity in brain structures associated with pain processing [112] and produces even greater analgesia in high catastrophizing patients [113]. Indeed, the actual process of “pain perception” is not solely a somatic reflection but rather dependent on emotion, cognition and attention as well [114,115,116,117,118]. Accordingly, the use of VR demonstrates significant analgesic efficacy, during or immediately following the “VR experience”, in different types of acute and chronic pain [119,120,121,122]. However, targeting central sensitization and brain remodeling of chronic pain through VR would essentially require evidence of long-lasting improvement of perceived pain intensity. Indeed, VR produced lasting analgesia in patients with fibromyalgia at 6 months follow up [123] and chronic headache pediatrics at 3 months post-treatment [124]. Furthermore, Mehesz and colleagues (2021) showed that an immersive VR experience in healthy participants is able to produce efficient conditioned pain modulation and, in a surrogate central sensitization model, alleviate mechanical pain sensitivity [125]. Additionally, a recent case report by Orakpo and colleagues (2021) showed that VR, fused with neurofeedback therapy, achieved adequate analgesia that was sustained for 1 year in a patient with chronic spondylolisthesis pain, indicating further neuromodulation promise of VR in centralized pain syndromes [126]. Moreover, immersive VR is shown to not only reduce perception of capsaicin-induced ongoing pain, but also to elevate pain thresholds of corresponding secondary hyperalgesia [127]. These observations provide direct evidence supporting the effectiveness of VR in the management of central sensitization and modulation of pain processing. It should be noted; however, that effective patient distraction would entail being comfortable with and willing to use VR, which might vary across different demographics, available VR hardware and simulated VR environments. Accordingly, the production of a “VR pharmacy” to provide individualized or patient-tailored experiences was previously proposed [128]. On the other hand, the use of VR in phantom limb pain and pain associated with certain musculoskeletal disorders relies on additional mechanisms other than distraction. Phantom limb pain is a form of neuropathic pain that is highly prevalent among amputees, which results from representational mismatching and subsequent central pain mechanisms [129]. Indeed, phantom limb pain is associated with reduced thermal pain thresholds in various body parts, indicating central alterations [130], correlated with mechanical wind-up pain and thermal allodynia [131], and the altered pain processing and wind-up of phantom limb pain are positively correlated with catastrophizing indicating roles for cognitive and emotional sensitization [132]. Furthermore, phantom limb pain involves reorganizations or regional, amputated limb, boundary re-mapping; however, maladaptive plasticity of preserved representation and activity despite the lack of sensory input results in multiple painful and non-painful, illusory, amputated limb perceptions [133]. In order to correct, or account for, the representational mismatching in phantom limp pain; various techniques, mainly based on enhanced visual input, have been developed including mirror therapy, motor imagery, and virtual visual feedback, all of which are able to reduce phantom limb pain [134]. Through immersive VR systems, embodiment of a virtual limb or body part allows the modulation of perceptual disturbances and control of phantom limb pain and other types of chronic pain [135]. Accordingly, somatic VR experiences represent a novel form of rehabilitation. Indeed, the use of immersive VR in phantom limb pain patients is shown in various studies and case reports to elevate pain thresholds [136], decrease pain and improve anxiety [137] and provide sustained pain reductions [138,139,140]. These findings support a significant promise for VR in the modulation of central processing in chronic pain and management of phantom limb pain; however, larger studiers are still required.

Cognitive therapy

As discussed previously, cognition and emotion are important factors influencing the process of pain perception. In addition, catastrophizing and maladjusted pain cognitions are associated with higher pain scores, anxiety, central sensitization and maladaptive processing of pain [141,142,143,144]. This is also observed in neuropathic pain conditions; for instance, catastrophizing is commonly observed in patients with orofacial neuropathic pain, for which only select pharmacological options are available, and is associated with higher pain intensity [145,146,147]. Accordingly, various studies investigated the potentials of cognitive-based therapies in the management of chronic pain conditions. These mainly include cognitive behavioral therapy (CBT), mindfulness-based therapies (MBT) and acceptance and commitment therapy (ACT). Current evidence indicates that these thee approaches lead to incremental but statistically significant reductions in chronic pain scores [148,149,150,151]. Despite these improvements, the aim of cognitive therapy should be to affect pain processing and modulate central mechanisms of sensitization to improve responses to pharmacological therapy. Indeed, CBT is found to decrease induced-pain unpleasantness but not intensity; however, it significantly reduced secondary hyperalgesia; thus, central sensitization [152]. Accordingly, extensive research has been recently conducted to evaluate the neural mechanisms of cognitive therapies. It was shown that catastrophizing is associated with higher functional connectivity between the insula and primary somatosensory (S1) cortex in fibromyalgia patients. However, CBT intervention led to significant and long-term improvements in pain intensity and catastrophizing, which were associated with restorations of lower resting-state functional connectivity levels between the insula and S1 cortex [153]. In addition, chronic pain is associated with reduced grey matter volume of the prefrontal cortex [154] while CBT intervention causes increased grey matter volume in various cortical regions, and the volume increase in the prefrontal and somatosensory cortices is associated with reduced catastrophizing [155]. These findings are functionally reflected as well in fibromyalgia patients undergoing CBT therapy in whom CBT led to significant elevation in pain-evoked neuronal activity in the prefrontal cortex with suggested alterations in pain processing loops relating to pain reappraisal [156]. Further neuroimaging evidence shows that 11 weeks of CBT in chronic pain patients caused significant elevations in connectivity between the somatosensory cortex and basal ganglia while causing reductions in connectivity of default mode network with limbic regions such as the amygdala, which were accompanied with clinical improvements and improved pain-coping [157]. The connectivity alterations of CBT in chronic pain patients also involve resting-state brain networks, especially the orbitofrontal cortex, which has important roles in the cognitive processing of pain [158]. On the other hand, MBCT is another form of psychotherapy that relates to CBT but focuses on mindfulness through certain interventions such as meditation and other practices. It was found that cognitive therapies including CBT and MBCT, in patients with various chronic pain conditions, alter neuronal function throughout brain networks and reduce affective aspects of the pain experience [159]. In addition, mindfulness meditation in chronic pain, when compared to sham controls and placebo analgesia, is found to cause significantly higher reductions in pain intensity and unpleasantness and cause different brain activity alterations. These include enhanced activity of cognition-dependent pain-modulating cortical regions including the anterior insular, orbitofrontal and subgenual anterior cingulate cortices [160]. Therefore, MBCT-induced modulation of pain is different from and relies on different mechanisms compared to placebo analgesia. Positive findings are also observed with ACT interventions in relation to pain, behavior and connectivity alterations across emotion, cognition and pain processing networks [161, 162]. These novel findings provide key insights into the neural plasticity mechanisms by which cognitive therapies modulate central pain processing. Lastly, some reports indicate that perioperative CBT can decrease postsurgical pain and catastrophizing [163, 164], which in principle, and based on preliminary findings [165], aid in the prevention of post-surgical acute-to-chronic pain transition; however, further investigations are needed.

Exercise rehabilitation

Rehabilitation encompasses a multitude of interventions; however, in relation to pain management it mainly includes physical or exercise therapy, dietary control, stress management and other lifestyle modifications. Within the scope of this review, the focus on rehabilitation will be directed towards physical or exercise therapy in chronic pain. It is well established that exercise, within appropriate limits, has beneficial impact on pain and associated symptoms [166]. In addition, exercise-induced analgesia is a known phenomenon; however, the underlying mechanisms are complex and multiple hypotheses have been proposed [167]. On the other hand, various reports suggest that patients with chronic pain may not benefit from post-exercise analgesia as healthy individuals [168]. The pattern and not necessarily type of exercise; however, is a major outcome determinant such that sudden bouts of heavy exercise result in pain exacerbation, while regular moderate physical activity improves pain, decreases central neuronal excitability and promotes central inhibition [169]. In relation to central sensitization, various studies investigated the effects of exercise on pain sensitivity in patients with chronic pain and accumulating evidence demonstrates beneficial effects for exercise-induced hypoalgesia. In osteoarthritis, education and exercise lead to pain reduction and lower analgesic use post-exercise, while additional strength exercise reduces hyperalgesia but attenuates pain reductions [170, 171]. In chronic back pain, aerobic exercise results in significant reductions of chronic pain intensity, induced-pain sensitivity and interference, potentially due to activation of endogenous opioid analgesia [172]. Other specific types of exercise are also effective for chronic low-back pain; for instance, McKenzie exercise program was found more effective than conventional physiotherapy and led to significant reduction of central sensitization markers, pain intensity and disability; however, trunk muscular endurance did not improve [173]. However, effective exercise-induced recruitment of endogenous analgesia is not observed in all chronic pain conditions; for instance, exercise is effective in rheumatoid arthritis but not in chronic fatigue syndrome and fibromyalgia [174]. Therefore, a moderate physical activity, unless contraindicated, can be generally recommended; however, specific rehabilitation and exercise programs should be selected in an individualized manner.

Discussion and clinical considerations

Accumulating evidence indicates a significant role for maladaptive plasticity in the pathophysiology of various forms of chronic pain through functional and structural connectivity alterations. In this regard, non-pharmacological interventions including the discussed neuromodulation techniques, cognitive therapies and rehabilitation carry significant potentials to counteract maladaptive plasticity to help alleviate chronic pain or prevent acute-to-chronic pain transition. However, the functional and structural plasticity alterations associated with chronic pain show significant discrepancies across a wide array of chronic pain conditions. In addition, the molecular mechanisms by which different neuromodulation techniques impact neuronal plasticity vary widely as well; thus, each intervention would have differential efficacy across different pain conditions. Furthermore, inter-individual variability as well as associated psychocognitive factors must be taken into account as not all patients develop central sensitization, exhibit connectivity alterations or equally respond, or develop tolerance, to the various therapeutic interventions. Therefore, the importance of individualized treatment and patient-tailored selection of appropriate treatment options must be stressed. Clinical tools such as the central sensitization inventory [175] have been developed, which can help identify patients with central components of sensitization and corresponding severity [176], and shown to be valid even in the outpatient setting [177]. The choice of treatment intervention should be based on guideline recommendations derived from clinical evidence supporting the application of each treatment modality. The use of brain stimulation techniques such as rTMS and tDCS largely remains investigational with weak or inconclusive recommendations in neuropathic pain, fibromyalgia and spinal cord injury pain [60, 61]. This is due to inconsistent clinical evidence mainly attributable to randomized controlled trials (RCTs) with low study sample sizes [178]. On the other hand, an expert consensus panel in 2020 recommended the use of rTMS, applied to the M1 cortex, for neuropathic pain, post-traumatic brain injury-related headache, postoperative pain and prevention of migraine [179]. Other neurostimulation techniques have been more widely applied in the clinical setting such as high-frequency SCS, which is approved by the U.S. Food and Drug Administration as aid for the management of chronic back and limb pain as well as diabetic neuropathy [180]. Multiple clinical trials on the use of high-frequency SCS have been done with robust evidence to support its use for persistent back and radicular pain especially following failed back surgery [181], as also recommended by the National Institute for Health and Care Excellence for chronic neuropathic pain [182]. In relation to psychological therapy, particularly CBT, and exercise therapy, alone or as part of multi-disciplinary rehabilitation programs, clinical evidence supports slight improvements of function and pain scores over short (< 6 months), intermediate and long-term (> 12 months) follow-up in various chronic pain conditions including fibromyalgia [183]. Despite that rehabilitation and cognitive therapies provide modest improvements of pain scores, their psychological impact on pain cognition and brain connectivity could prove to be essential for patients with centralized pain syndromes. Therefore, CBT and exercise should be considered for all adult patients with primary chronic pain as recommended by the National Institute for Health and Care Excellence [184]. Lastly, the use of VR is yet to be approved for pain management as more robust clinical evidence is required.

Conclusions

Over the last two decades, the impact of maladaptive plasticity of central sensitization and brain remodeling has been highlighted and identified as a major component of various chronic pain conditions. Accordingly, neuromodulation research targeting maladaptive plasticity has been gaining momentum and shown tremendous usefulness in managing various forms of chronic pain that would otherwise be considered intractable and unresponsive. While pharmacological agents are still considered the cornerstone in the treatment of acute and chronic pain, novel neuromodulation techniques and protocols are continuously advancing with significant future potentials. Further large clinical trials are required to establish the long-term clinical safety and efficacy of these techniques, the results of which could reshape the scope of pain management in various chronic pain conditions.