Although peripheral neuropathy (PN) is a common complication in connective tissue diseases (CTD) and has been well studied, recent research has shown that PN is more diverse and frequent in different subtypes of CTD than was expected. The incidence of PN in Sjögren's syndrome and rheumatoid arthritis (RA) varies according to different disease subtypes, and the pathogenesis of neuropathic pain in different subtypes of eosinophilic granulomatosis with polyangiitis (EGPA) may also differ. Neurogenic inflammation, autoantibody-mediated changes, ischemia of the vascular wall and metabolic mechanisms have been shown to contribute to the pathogenesis of PN in CTD. Moreover, allergic inflammation has been recently identified as a possible new mechanism producing peripheral neuropathic pain associated with MPO-ANCA negative EGPA patients. Glucocorticoids are routinely used to relieve pain caused by PN. However, these steroids may cause hyperalgesia, exacerbate neuropathic pain, and activate the early phase of pain induction and produce hyperalgesia. Recently, neuroactive steroids, such as progesterone, tetrahydroprogesterone and testosterone, have been shown to exert protective effects for several PN symptoms, and in particular neuropathic pain. Neuroactive steroids will be an interesting topic for future research into PN in CTD.
Connective tissue diseases (CTD) are chronic inflammatory autoimmune diseases induced by antibodies or T-cell responses directed against self-antigens, which can affect all body systems, including the central nervous system (CNS) and peripheral nervous system (PNS) [1]. When the PNS is involved in CTD, peripheral neuropathy (PN) is the most common complication [2], which comprises a heterogeneous group of disorders, such as mononeuropathy, polyneuropathy and mononeuritis multiplex. PN may be a manifestation or a characteristic sign of immune system dysfunction, with variable prevalence and prognosis in CTD. Therefore, rapid recognition and treatment are essential. However, due to a varied complex spectrum of overlapping clinical manifestations, PN is an under-diagnosed complication in CTD and a particular challenge for rheumatologists and neurologists. Glucocorticoids and immunosuppressants are usually administered as basic and routine treatments of PN in CTD. However, as reported in experimental models of neuropathic pain, glucocorticoids may cause hyperalgesia, exacerbate neuropathic pain, and activate the early phase of pain induction and indeed produce hyperalgesia [3]. A possible strategy to find an effective treatment for PN is shifting the focus to new biological targets and relevant molecular events in the PNS; in particular, neuroactive steroids are a highly promising therapeutic option [4] as these steroids can modulate PNS functions.
Axonal sensory polyneuropathy and sensorimotor polyneuropathy can be characterized by paresthesia and defects (including mild touch, proprioception and vibration sensation) in the distal part of the symmetrical limb, mainly affecting the distal end of the lower limbs, and may be accompanied by burning pain in the feet. In addition to the above manifestations, motor weakness may be present in sensorimotor polyneuropathy, which is usually mild and limited to the extensor muscles of the toes or feet [5].
Small-fiber neuropathy (SFN) is an algetic esthesioneurosis that usually results in burning pain and arises in the early stage of several systemic diseases such as diabetes, amyloidosis and CTDs [6]. The main manifestations of small fiber neuropathy are numbness, burning sensation, electric pain, pricking, pruritus, involving the limbs, trunk or the proximal part of the face [5]. Motor neuron disease is characterised as paresis, atrophy and bundle fibrillation, mainly in the distal limb [5]. Besides, SFN consists of two different types, which may be underestimated. The first is called “length-dependent” SFN, which is a neuropathic pain arising in a distal “stocking-and-glove” distribution reported by the patients. The conventional model of this distal neuropathic pain is related to equivalent skin biopsy markers of the most distal axonal degeneration. These markers include reduced intra-epidermal nerve-fiber density (IENFD) of amyelinic nerves. Compared to the proximal leg, the fiber density is decreased at the distal leg. While, concerning the second type of the disease which is called “non-length-dependent” SFN, patients suffer from heterodox and atypical models of neuropathic pain which involves the face, truncus and proximal arms and legs. However, skin biopsy results reveal that this non-conventional model of proximal neuropathic pain is related to skin biopsy markers which show that neuronal degeneration affects the most proximal component in the PNS—the dorsal root ganglia (DRG). Under the circumstances, the IENFD in the distal leg is not decreased any more compared to that in the proximal leg [6].
As a heterogeneous group of neurological disorders, the reported prevalence and clinical manifestations of PN in CTD varies widely (Table 1). As shown in Table 1, the main studies during the last 5 years have focused on systemic lupus erythematosus (SLE), Sjögren's syndrome (SS), systemic sclerosis (SSc), antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis and rheumatoid arthritis (RA). It is still worth mentioning that up to one-third of PN cases have a non-CTD aetiology including infection, drug toxicity or metabolic diseases. The final attribution of PNS involvement in CTD is therefore a relevant and challenging clinical issue [7][8].
Authors | Prevalence/ Constituent Ratio (%) |
Patients (N) | Type of Study | Main Electrodiagnostic Tests Pattern | Main Form of PN Manifestation | |
---|---|---|---|---|---|---|
Systemic lupus erythematosus | ||||||
Xianbin et al. | [8] | 1.5% | 4924 | Cross-sectional | Sensory (67.5%), motor (49.3%) | Polyneuropathy +++ Mononeuropathy ++ Cranial neuropathy ++ Myasthenia gravis ++ |
Toledano et al. | [10] | 17.7% | 524 | Cross-sectional | Sensory-motor (56%), axonal 80.3% | Polyneuropathy +++ Mononeuropathy ++ Cranial neuropathy + |
Saigal et al. | [11] | 36% | 50 | Cross-sectional | Sensory-motor, axonal | - |
Bortoluzzi et al. | [12] | 6.9% | 1224 | Cross-sectional | Sensory-motor (25%) | Polyneuropathy +++ Cranial neuropathy +++ Mononeuropathy ++ Mononeuritis multiplex + |
Hanly et al. | [13] | 7.6% | 1827 | Cohort | Sensory-motor (71%), sensory (16.1%) axonal (41.7%), demyelination (21.7%) |
Polyneuropathy +++ Mononeuropathy ++ Cranial neuropathy ++ Mononeuritis multiplex ++ |
Fargetti et al. | [14] | 1.8% | 2074 | Cohort | Sensory-motor (68.4%), axonal (49.3%) | Polyneuropathy +++ Mononeuropathy ++ Polyradiculoneuropathy + Cranial neuropathy + |
Sjögren’s syndrome | ||||||
Ye W et al. | [17] | 19% pSS 31.1% sSS |
415 pSS 151 sSS |
Cross-sectional | - | - |
Seeliger et al. | [18] | 44 SS + PNP | 108 PNP | Cross-sectional +case-control | Motor (100%), sensory (89%) axonal (36%), demyelinating (23%), both (41%) |
- |
Carvajal Alegria et al. | [19] | 16% | 392 | Cohort | Sensory (57%), sensory-motor (33%) | Mononeuritis multiplex Polyneuropathy Cranial neuropathy |
Przyńska-Mazan et al. | [20] | 63.9% | 61 pSS | Cross-sectional | Sensory-motor axonal (47.5%), demyelination, both (5.1%) | Polyneuropathy +++ Mononeuropathy +++ Entrapment neuropathy ++ Mononeuritis multiplex ++ |
Sireesha et al. | [21] | - | 20 pSS 1 sSS |
Cross-sectional | - | Mononeuritis multiplex +++ Ganglionopathy ++ Trigeminal neuropathy ++ |
Jaskólska et al. | [22] | 72% | 50 pSS | Cross-sectional | Sensory-motor axonal (22%) | Entrapment neuropathy +++ Mononeuropathy ++ Cranial neuropathy + |
Jaskólska et al. | [23] | 46% | 50 pSS | Cross-sectional | Sensory-motor (47%) | Mononeuropathy ++ Cranial neuropathy ++ |
Systemic sclerosis (scleroderma) | ||||||
Raja et al. | [27] | 36.6% | 60 | Cross-sectional | Sensory (65%), motor (53%) | Polyneuropathy +++ Mononeuropathy ++ Entrapment neuropathy ++ |
Paik et al. | [28] | 28% | 60 | Cross-sectional | Sensory-motor axonal, no demyelinating |
- |
* Yagci et al. | [29] | 29.2% | 24 | Cross-sectional | - | Entrapment neuropathy Polyneuropathy |
* Sriwong et al. | [30] | 38% | 50 | Cohort | - | Median neuropathy at the wrist |
Polyarteritis nodosa | ||||||
Sharma et al. | [34] | 88.9% | 27 | Cross-sectional | Axonal sensory-motor (81.8%) | Mononeuritis multiplex |
Eosinophilic granulomatosis with polyangiitis | ||||||
Bischof et al. | [38] | 19% 23% 65% |
572 GPA 218 MPA 165 EPGA |
Cross-sectional | Sensory-motor (32%), sensory (16%), motor (5%) | Mononeuritis multiplex +++ |
Zhang et al. | [39] | 46.4% | 110 EPGA | Retrospective cohort | - | Polyneuropathy +++ Mononeuritis multiplex ++ |
Cho et al. | [40] | 75% | 61 EPGA | Retrospective cohort | Sensory (44/46), motor (24/46) | Mononeuritis multiplex +++ Mononeuropathy ++ Polyneuropathy ++ |
Nishi et al. | [41] | - | 82 EPGA | Retrospective | Axonal | - |
Rheumatoid Arthritis | ||||||
Kaeley et al. | [49] | 75.28% | 89 | Cross-sectional | Asymmetrical sensorimotor axonal neuropathy, pure motor | Mononeuritis multiplex Entrapment neuropathy |
Kumar et al. | [50] | 34.4% (seropositive) 15.38% (seronegative) |
60 | Cross-sectional | - | - |
An increase in proinflammatory cytokine concentrations has been found in patients with vascular neuropathy [7]. Nociceptors located in nerve endings can sense IL-1β and TNF-α directly and induce activation of MAP kinases, resulting in increased membrane excitability. In addition, MAP activation leads to the release of different neuropeptides such as calcitonin gene-related protein, substance P, nitric oxide and chemokines, which subsequently cause vasodilatation, increases in vascular permeability and cell trafficking [7]. On the other hand, these mediators released from sensory neurons in the periphery directly attract and activate immune innate cells and adaptative immune cells such as T lymphocytes [7]. Nerve growth factor and prostaglandin E2 are major inflammatory mediators released from immune cells that act on sensory neurons inducing peripheral sensitization and hyperalgesic phenomena [7].
The sural nerve biopsy specimens of PN-EGPA patients with negative MPO-ANCA were mainly characterized by eosinophil infiltration [41], suggesting that allergic inflammation was an underlying mechanism in MPO-ANCA negative EGPA patients. Through animal experiments, Fujii et al. [73] found that peripheral nerve damage caused by allergic inflammation can induce Nep. As for allergic individuals, increased humoral immunity may lead to anti-plexin D1 antibody production via molecular mimicry with environmental allergens. Anti-plexin D1 antibodies can damage primary pain-conducting neurons, thus inducing neuropathic pain [73]. In addition, the overproduction of ET-1 in inflamed skin tissues and sera may induce blood-brain barrier (BBB) hyperpermeability and activate microglia and astroglia through the ET-1/EDNRB pathway in allergic inflammation, thus causing NeP [73].
The most possible pathogenesis of vasculitis-related PN is inflammation of precapillary arteries in the nerves [74]. Deposition of immune complexes or T cell-mediated immunity plays a major role in inducing the immunological inflammation and necrosis of vessel wall [74]. The end result of both processes is the induction of immunological inflammation and necrosis of blood vessel walls, which eventually leads in addition to focal or multifocal, axonal, ischemic neuropathy [74].
A process of molecular mimicry may act as the starting motif to target different specific antigens within the structure of a nerve [7]. Nodes of Ranvier may be a vulnerable target for autoimmunity due to the intrinsically elevated number of potential antigens and the crucial permeability of the blood-nerve barrier in nodal and juxtaparanodal structures [7].
IgG and IgM anticardiolipin antibodies were detected in the serum of CTD patients [75]. The existence of anti-ganglioside antibodies in PN-SLE patients has been found frequently [7], while the chronic inflammatory demyelinating polyneuropathy (CIDP) associated with IgG4 antibodies to neurofascin-155 (NF155) was recently described [76]. These immune antibody markers have been not only proven to be useful in clinical practice but also uncovered novel pathophysiological mechanisms, clinical phenotypes, therapeutic responses and prognosis indicators.
PNS involvement in CTD can also be caused by metabolic disorders secondary to aggressive therapy, multiorgan pathology and endocrine abnormalities. Metabolic disorders may induce a reaction of demyelinating neuropathy and axon dystrophy in severe cases [75].
NCS can reveal the asymmetric or multifocal nature of neuropathy. Electromyogram (EMG)/NCS can establish whether patients have sensory-motor neuropathy, sensory neuronopathy or motor neuronopathy. Moreover, EMG/NCS can be applied to distinct polyradiculopathy with mononeuritis multiplex, as both patterns have very similar diffuse, non-length dependent neuropathies. Importantly, EMG/NCS can also be used to identify whether the affected structure of nerves is the axon itself or demyelination. In addition, needle electromyography is very useful in estimating the disease course, the extent of the injury and the existence of a superimposed myopathy [77].
The sural nerve is most frequently used for biopsy, which contributes to the determination of the nature and extent of PN [5]. Bischof et al. [38] performed nerve biopsies in 31 PN-AVV patients and showed that 55% of patients had definite vasculitis. Similarly, about 80% of PN-EGPA patients had extravascular eosinophils and 77% of the patients had vasculitis, while no extravascular granuloma was observed in the total of 44 patients [40].
Somatic and autonomic functional nervous evaluation of SFN involves the sympathetic and parasympathetic autonomic functions, which is realized by determining the psychophysical sensory thresholds (e.g. cold and heat) through quantitative sensory testing (QST), pain-related tests and recording of laser-evoked potentials (LEP), single axon recording utilizing microneurography and tests [78]. In another hand, in 2010, the European Federation of Neurological Societies and the Peripheral Nerve Society joint amended the Guidelines on the Application of Skin Biopsy in the Diagnosis of PN. And a conclusion was made that distal leg skin biopsy with quantification of the linear density of IENFD, adopting universally accepted counting rules, is a reliable and efficient technology to evaluate the diagnosis of SFN [79]. In fact, the process toward the determination of the diagnosis of SFN in individual patients, beginning from the chief complaints of sensory symptoms, is on the basis of the clues from skin biopsy and/or QST results. The combination of clinical signs and abnormal QST and/or IENFD findings can reliably be used to diagnose SFN compared with the combination of abnormal QST and IENFD findings without clinical signs [80].
What’s more, current diagnostic technologies for SFN are also composed of quantitative sensory testing with determination of warm and cold detection thresholds (WDT, CDT), recording of LEP and sympathetic skin responses (SSRs), and measurement of electrochemical skin conductance (ESC) utilizing Sudoscan(®) device [81].
Modern imaging methods allow the precise localization of peripheral nerve damage. Sonoelastography [29] and ultrasonography investigations of the median nerve [82] are emerging techniques to image the median nerve. Researchers have used the ultrasonography investigation of the median nerve to visualize the median nerve in the carpal tunnel, revealing an increased median nerve cross-sectional area (CSA) and decreased echogenicity due to neural edema within the carpal tunnel [51]. Sonoelastography, previously applied to document decreased skin elasticity in patients with SSc, now is also used for median nerve imaging [51].
Diffusion-weighted magnetic resonance neurography (DW-MRN) is another emerging technique that can exploit the greater water diffusion anisotropy in peripheral nerves for improved visualization [83]. Based on the concept of background body and vascular signal suppression for improved visualization of stationary fluid or cellular structures, DW-MRN is applicable to improved visualization of extremity nerves and their lesions in the wrist and palm with adequate image quality, thus providing a supplementary method for conventional magnetic resonance imaging [83].
In 2010, the guidelines of the Society for Neurology of USA recommended (with evidence level B) vasculitic peripheral neuropathy shock therapy with glucocorticoids, possibly in combination with immunosuppression as basic therapy [84]. Major treatment recommendations were: (1) corticosteroid (CS) monotherapy for at least 6 months as first-line treatment; (2) combination therapy applied to rapidly progressive and non-systemic vasculitic neuropathy and patients who progress on CS monotherapy; (3) immunosuppressive options including cyclophosphamide, azathioprine, and methotrexate; (4) cyclophosphamideapplies to severe neuropathies, generally administered in IV pulses to reduce cumulative doses and the associated side effects; (5) combination therapy for patients achieving clinical remission, followed by continued maintenance therapy for 18-24 months with azathioprine or methotrexate [84].
Glucocorticoids are routinely used to relieve pain. However, as reported in experimental models of neuropathic pain, glucocorticoids may contribute to hyperalgesia, exacerbate neuropathic pain, activate the early phase of pain induction and induce hyperalgesia [3]. Therefore, concerns have been raised about the use of glucocorticoids for pain treatment [3] and a new strategy is needed. Neuroactive steroids (i.e., steroid hormones synthesized by peripheral glands and those directly synthesized in the nervous system) represent critical physiological regulators of PNS function and can have protective impacts on several symptoms of PN including neuropathic pain [4]. Falvo et al. [4] summarized the role of neuroactive steroids for the treatment of neuropathic pain thus: progesterone prevents pain-related behaviors, like allodynia and hyperalgesia, in different models of neuropathic pain; tetrahydroprogesterone reduces thermal and mechanical hyperalgesia by enhancing the activity of the γ-amino butyric acid-A receptor and by blocking T-type Ca2+ channels; testosterone exerted anti-nociceptive effects in neuropathic rats; 17β-estradiol caused pain attenuation and a decrease of neuropathy-induced gliosis after sciatic nerve ligature; moreover, it exerted protective effects on neuropathic pain via estrogen receptors by inhibiting microglia activation and the production of inflammatory mediators. These observations support the view that neuroactive steroids may provide an interesting topic for neuropathic pain research in the near future.