Small bre neuropathy in patients with chronic pain and a previous diagnosis of Multiple Chemical Sensitivity syndrome

Enrico Fileccia (  enrico leccia@gmail.com ) IRCCS Istituto delle Scienze Neurologiche di Bologna, UOC Clinica Neurologica Alex Incensi IRCCS Istituto delle Scienze Neurologiche di Bologna, UOC Clinica Neurologica Francesco Ventruto Università di Bologna Giovanni Rizzo IRCCS Istituto delle Scienze Neurologiche di Bologna, UOC Clinica Neurologica Damiano Galimberti Università di Catania, Dipartimento di Biochimica e Patologia Clinica Giacomo Rao Sovrintendenza Sanitaria Centrale Settore Prevenzione Ricerca Direzione Generale INAIL Roma Fabrizio Salvi IRCCS Istituto delle Scienze Neurologiche di Bologna, UOC Clinica Neurologica Rocco Liguori Università di Bologna Vincenzo Donadio IRCCS Istituto delle Scienze Neurologiche di Bologna, UOC Clinica Neurologica


Introduction
Small ber neuropathy (SFN) is an often misdiagnosed condition characterised by either sensory and/or autonomic symptoms due to a selective involvement of Aδ and C-bres. Diagnosis is based mainly on skin biopsy, quantitative sensory testing and clinical questionnaires although other examination such as test of sudomotor function can also play a role. [1] Multiple chemical sensitivity (MCS) syndrome is a still not fully understood condition where patients reports the arising of different symptoms after exposure to several chemical substances. [2] Among the symptoms reported, pain is often present in patients who complain of MCS symptoms. [2] In this paper we report the results of a study conducted in a cohort of patients referred to our observation for the suspect of SFN because of chronic pain who had received a previous diagnosis of MCS syndrome.

Materials And Methods
We screening 25 patients referred to our observation because of chronic pain and a previous diagnosis of MCS from Italian tertiary centers particularly skilled in the diagnosis of MCS. However, 4 patients were excluded from the study because the presence of acquired causes predisposing a peripheral nerve involvement such as Sjogreen disease (2 patients) and vitamin B12 de ciency (2 different patients). For this study we selected a total of 21 patients (19 F, 2 M, mean age 50.8 years +-8.53) showing no predisposing causes for peripheral neuropathy after an extensive serum screening excluding diabetes, liver, kidney and thyroid dysfunctions, vitamin B12 and folate de ciencies, autoimmune disorders and infections. The MCS diagnosis was performed based on previous described criteria. [2] Diagnosis of the condition was based on the presence of at least 6 months of recurrent symptoms involving central nervous system (CNS) associated with self reported smell hypersensitivity and at least another organ system: the main feature of the condition is that symptoms are triggered by exposure to low levels of multiple chemicals substances and ameliorates or disappear after removal of the noxious stimuli. [2] Besides CNS involvement patients often complains of allergic diathesis, self-reported food or alcohol intolerance and symptoms regarding musculoskeletal, gastrointestinal, skin, respiratory, peripheral nerves and cardiovascular system. [2,3]. In addition, 16 out of 21 patients also received the diagnosis of bromyalgia. All patients underwent neurological clinical examination (including clinical scales for pain and autonomic disorders), nerve conduction studies and skin biopsy. Skin biopsy results were compared to age matched healthy controls (15 F, 4 M, mean age 57.22 years +-13.081). The procedures used were approved by the local Human Ethics Committee (Comitato Etico indipendente-AUSL Bologna, code number 12073) and followed the Helsinki Declaration related to clinical research of human beings. In addition, patients and healthy controls gave their written informed consent to participate in the study.

Skin biopsy
Three mm punch biopsies were taken from proximal thigh (15 cm above the patella) and distal leg (10 cm above the lateral malleolus) hairy skin sites. 50-microm-thick sections were obtained using a freezing sliding microtome (HM550; Thermo Scienti c, Waltham, MA, USA). Twelve free-oating sections were 3), were added for an overnight incubation. Skin sections were rstly viewed and analyzed under a Zeiss uorescent microscope. Furthermore, the sections were also analyzed using a confocal laser scanning microscope (Nikon confocal microscopy, Eclipse Ti A1, Japan) to study the innervation pattern. Digital images were collected in successive frames of 1-2 μm increments on a Z-stack plan, at the appropriate wavelengths for secondary antibodies, with a x20 or x40 plan apochromat objective and subsequently projected to obtain a double-stained digital image with a computerized system. Autonomic innervation density was quanti ed using the previously described automated method showing high interobserver and intraobserver reliability. [4,5] Epidermal nerve ber density (ENFD) was calculated by considering a single epidermal ber marked by PGP 9.5 crossings of the dermal-epidermal junction.

Nerve conduction studies
Sensory nerve conduction studies were recorded from sural and peroneal nerve bilaterally in the lower limbs, from median or ulnar nerve of the non-dominant hand in the upper limbs. Sensory conduction studies were performed antidromically. Motor bres conduction studies were performed recording from abductor hallucis brevis and extensor digitorum brevis (tibial and peroneal nerve respectively) bilaterally in the lower limbs and from abductor digiti minimi or abductor pollicis brevis in the upper limbs (ulnar or median nerve respectively).

Clinical assessment
All patients underwent complete neurological assessment and clinical questionnaires assessing neuropathic pain and autonomic symptoms (DN4 and Compass 31 scale, respectively). [6,7] Statistic Statistical analyses were performed using SPSS 25.0 for Windows (IBM Corp., Armonk, NY). Categorical variables were evaluated by chi-square test. For the analysis of continuous variables, we used the Kolmogorov-Smirnov test to verify the normal data distribution. Autonomic innervation density values were normally distributed, and an ANCOVA was used to evaluate differences between patients and healthy controls, using age and sex as covariates. Sensitivity and speci city for differentiating the groups were calculated using the optimal cut-off values determined by receiver operating characteristic (ROC) curve analysis. To search for correlation between skin biopsy data and clinical scores, we used Spearman rank test, as clinical scale values were not normally distributed. P-values < 0.05 were accepted as statistically signi cant.

Clinical assessment and nerve conduction studies
Patients and controls were matched for sex and age. Neurological examination did not nd any abnormal sign in patients although all of them complain of pain mainly at the extremity of the limbs while autonomic symptoms were complained of only by some of them ( Table 1). The mean duration of symptoms was 9.6 years for pain (range 1-40; SD 8.78) and 15,38 years (range 2-45; SD 13.73) for autonomic symptoms. DN4 score mean score was 6.5 (range 4-9; SD +-1.43). Mean COMPASS 31 score was 38.5 (range 0-65,7; SD 19.08). For what it concerns the latter, the more involved subdomains resulted the cardiovascular and gastrointestinal ones (adjusted mean score of 16.21 and 9.81 respectively).
Scoring of DN4, COMPASS 31 and clinical features of the patients are summarised in Table 2. Nerve conduction studies resulted normal in all the patients.

Skin innervation
ENFD was signi cantly decreased in both thigh and leg while the autonomic innervation did not differ from controls (Table 3; Figure 1). The appropriate cutoff ENFD score differentiating patients from controls was calculated by using the ROC curve analysis. An optimal cutoff score in the lower leg was 8.3 ENFD/mm (sensitivity 93% and speci city 100%) and 12.8 ENFD/mm (sensitivity 86% and speci city 100%) for the thigh. Adopting these scores a somatic small ber neuropathy was found in all patients: in 18 patients (86%) with a distal and proximal involvements, 2 patients (10%) showed only a proximal neuropathy while 1 patient (4%) presented only a distal involvement (Figure 2).
The correlation analysis disclosed no signi cant correlation correlation between ENFD (in both leg and thigh) and the degree of neuropathic pain evaluated by DN4 (p>0,05) or the pain duration (p>0,05) although a slight tendency toward an inverse correlation was found between thigh ENFD and DN4 score (r=-0,4; p=0,07). Moreover, no correlation between autonomic skin innervation and COMPASS 31 score, including the correlated subdomains, was found. Table 3. Skin biopsy somatic and autonomic innervations mean results of the patients and healthy subjects are shown. Epidermal ber nerve density (EFND) is calculated counting every single epidermal ber marked with PGP 9.5 that crosses the dermal-epidermal junction. EFND is calculated for every sample, the site (thigh or leg) EFND is then calculated as a mean of the three samples obtained from each site. Arrector pili muscle (APM) innervation was calculated counting the number of bres marked with Dopamine beta hydroxylase (noradrenergic marker) surrounding these structures of the hair follicles. Sweat glands (SG) innervation was calculated considering the number of bres marked with vasoactive intestinal peptide (cholinergic marker). We found no differences between patients and healthy control in autonomic innervation parameters.

Discussion
Multiple chemical sensitivity is a still not full known disorder which unfortunately to date lack of objective identi able alteration. The clinical picture is characterized by multitude of symptoms which are triggered by exposure to low concentrations of various chemical substances. The condition is also known as Idiopathic environmental intolerance. [2,8,9 ] However, the triggering substances and the concentration level required to provoke symptoms are not well de ned and there is an huge variability among patients. MCS is not rare and epidemiological studies suggest it can affect up to 13% of population. [10,11] The pathogenic mechanism is still not understood; a possibile role of in ammatory cytokines, immune cell abnormalities, metabolic vulnerability to oxidative stress and neural sensitization have been suggested. [12][13][14][15][16][17] Genetic polymorphisms of the superoxide dismutase (SOD) 2 gene has also been proposed to increase the risk of being affected by the condition. [18] Moreover, polymorphisms involving the gene encoding for CYP2D6 and NAT-2 (two enzymes involved in drugs metabolism) have also being reported as factors which increase the likelihood of being affected by MCS. [19] Among the most common symptoms reported by patients there are ortosthatic intolerance, subjective breathing di culties, rash and peripheral edema. [2,3,20] Psychiatric comorbodities are also very common. [21] In addition, patients with MCS very often complained of chronic pain which is a disabling symptom with a severe impact in quality of life. [2,3,20] Chronic pain is also very common in a bromyalgia disorder and chronic fatigue syndrome (CFS). [22] Fibromyalgia is a frequent disorder with an estimated prevalence of about 3-5% in normal population. [23,24] The condition is characterized by widespread pain and tenderness in speci c parts of the body (the so called "tender points"). The mechanism of pain in bromyalgia is mainly considered related to a central abnormal ampli cation of pain. However up to 60% of patients diagnosed with bromyalgia have a reduced intraepidermal nerve ber density. [25][26][27][28][29] Moreover microneurographic evaluations of patients affected by bromyalgia showed also an abnormal activities in C bres. [30] The main clinical feature of CFS is a disabling fatigue lasting at least 6 month which is neither provoked by exercise or relieved by rest, is present for the most part of the day and is associated with other symptoms including pain, concentration di culties and sleep disturbance. [31] Our data showed the presence of a somatic small ber neuropathy in all examined MCS patients complaining pain. This is the rst study in literature reporting the presence of SFN in patients affected by chronic pain and MCS. Although the lack of correlation between DN4 and ENFD could suggests that chronic pain in MCS patients is not related to the presence of an underlying small ber neuropathy, this result is not surprising as it has previously been described that neuropathic pain is not correlated to the epidermal innervation. [32] This nding could be due to the universally use of a pan-neural (i.e. PGP 9.5) not re ecting the complexity of nociceptor subgroups (i.e. damaged and regenerating bers) involved in SFN. [32] Furthermore, SFN has been described in conditions classically characterized by absence of pain as amyotrophic lateral sclerosis [33].
Notwithstanding the foregoing, the slight tendency towards an inverse correlation between thigh ENFD and DN4 score support a possible in uence of epidermal denervation in the pain symptoms complained by our MCS patients.
Moreover it should be mentioned that the majority of our patients had received also a previous diagnosis bromyalgia; as the presence of SFN can be found up to 60% of patients affected by bromyalgia but all of our patients showed a reduced EFND, an additional different mechanism behind the overlap underlying the pain in the two conditions is likely. Interestingly we found a sparing of the autonomic skin innervation component suggesting that autonomic symptoms complained by patients and assessed through COMPASS 31 scale are possibly due to a selective involvement of speci c autonomic subdivisions (such as the cardiovascular and gastrointestinal systems) sparing the skin or could have a different underlying pathogenic process.
We were unable to perform cardiovascular re exes to ascertain a cardiovascular autonomic involvement in our patients and this con gures a limitation of the study. The main reason that prevented us from performing cardiovascular re exes was the inability of the patients to stay in the hospital for a long time.
Notwithstanding the suggestive possibility of having found a pathological biomarker for pain complained by MCS patients, our data do not automatically imply that SFN is part of the MCS clinical picture; as we study a subset of patients who were referred to our observation for the presence of chronic pain an additional study involving a larger cohort of MCS patients, hopefully involving subjects without pain, is needed to establish the relevance of SFN in these patients.
The study of patients without pain could be interesting since SFN has been reported as an early pathological sign of other disease, such as diabetes. [34,35] In accordance with the foregoing, whether SFN is part of the clinical picture of MCS and if so whether it has a causal role in the pain complained by the patients still remained to be better clari ed. A functional study of peripheral nociceptors, by using microneurography [36], will be desirable to con rm the nociceptors involvement underlying ENFD abnormalities and pain complained by MCS patients.

Conclusions
We studied 21 patients with a previous diagnosis of MCS complaining of chronic pain disclosing somatic SFN on skin biopsy in all patients. Notwithstanding high mean score on COMPASS 31 scale we found a sparing of autonomic bres suggesting a lesser role of skin nerve for autonomic symptoms. This is the rst skin biopsy study conducted on MCS patients however to con rm whether somatic SFN is part of the condition future and larger studies should be conducted. Scatter plot leg and thigh ENFD Figure 2. Scatter plot showing epidermal nerve ber density (ENFD)/mm in both thigh and leg for patients and con-trols. Patients and controls showed different ENDF/mm values. The optimal cutoff score for differentiating the groups was 8,3 ENFD/mm (sensitivity 93% and speci city 100%) in the lower leg and 12,8 ENFD/mm (sensitivity 86% and speci city 100%) in the thigh. Adopting these scores a somatic small ber neuropathy was found in all patients: 17 patients (80%) disclosed a