Background Neuropathic pain is common in peripheral neuropathy. Recent genetic studies have linked pathogenic voltage-gated sodium channel (VGSC) variants to human pain disorders. Our aims are to determine the frequency of SCN9A, SCN10A and SCN11A variants in patients with pure small fibre neuropathy (SFN), analyse their clinical features and provide a rationale for genetic screening.
Methods Between September 2009 and January 2017, 1139 patients diagnosed with pure SFN at our reference centre were screened for SCN9A, SCN10A and SCN11A variants. Pathogenicity of variants was classified according to established guidelines of the Association for Clinical Genetic Science and frequencies were determined. Patients with SFN were grouped according to the VGSC variants detected, and clinical features were compared.
Results Among 1139 patients with SFN, 132 (11.6%) patients harboured 73 different (potentially) pathogenic VGSC variants, of which 50 were novel and 22 were found in ≥ 1 patient. The frequency of (potentially) pathogenic variants was 5.1% (n=58/1139) for SCN9A, 3.7% (n=42/1139) for SCN10A and 2.9% (n=33/1139) for SCN11A. Only erythromelalgia-like symptoms and warmth-induced pain were significantly more common in patients harbouring VGSC variants.
Conclusion (Potentially) pathogenic VGSC variants are present in 11.6% of patients with pure SFN. Therefore, genetic screening of SCN9A, SCN10A and SCN11A should be considered in patients with pure SFN, independently of clinical features or underlying conditions.
- small fibre neuropathy
- painful neuropathy
- neuropathic pain
- voltage-gated sodium channels
- frequency of (potentially) pathogenic variants
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
- small fibre neuropathy
- painful neuropathy
- neuropathic pain
- voltage-gated sodium channels
- frequency of (potentially) pathogenic variants
According to the International Association for the Study of Pain (IASP), pain is defined as an unpleasant sensory and emotional experience associated with actual or potential tissue damage. As in neuropathic pain the somatosensory nervous system is affected,1 it is not surprising that pain is frequently reported in peripheral neuropathy, especially when the small diameter sensory nerve fibres are involved.2 Pure small fibre neuropathy (SFN) is a peripheral neuropathy in which the thinly myelinated Aδ-fibres and unmyelinated C-fibres are selectively affected, leading to sensory symptoms and autonomic dysfunction.2 In general, neuropathic pain is the main symptom, reflected by allodynia and hyperalgesia, but also thermal sensory loss, pinprick loss, restless legs syndrome, sicca syndrome, accommodation problems, hyperhidrosis or hypohydrosis, micturation disturbances, impotence and/or diminished ejaculation or lubrication, bowel disturbances (constipation, diarrhoea, irritability, gastroparesis, cramps), hot flushes, orthostatic dizziness and cardiac palpitations may be present.2 Pharmacotherapy for neuropathic pain in SFN is challenging since the efficacy of currently available medications is moderate and side-effects are often dose-limiting.3 4 A better understanding of genetic causes and pathophysiology of peripheral neuropathy may provide a basis for development of more effective, personalised treatment.
Voltage-gated sodium channels (VGSCs) are integral membrane polypeptides that are mainly present in excitable cells. The large α-subunit is constructed of four homologous domains (DI-DIV) with six transmembrane segments (S1-S6), forming an ion-selective pore. Associated smaller auxiliary β-subunits contribute to targeting and anchoring of the channel at specific sites in the plasma membrane, modulating gating properties of the α-subunit.5 VGSCs NaV1.7, NaV1.8 and NaV1.9, respectively encoded by SCN9A, SCN10A and SCN11A, are preferentially expressed in the small diameter dorsal root ganglion neurons (DRGs) and their peripheral axons. They play important roles in generation and conduction of action potentials in the physiological pain pathway.5 6 In addition, NaV1.7 is present in sympathetic ganglion neurons of the peripheral autonomic nervous system.5
Gain-of-function VGSC variants or dysregulated VGSC expression can cause pathological pain states characterised by spontaneous and prolonged pain.6 A causal link between pathogenic SCN9A variants and multiple human pain syndromes has been reported. SCN9A variants that increase the excitability of DRG were initially found in inherited erythromelalgia (IEM), an autosomal dominant disorder characterised by episodic painful red discoloured extremities due to warm temperatures and exercise.7 8 Shortly thereafter, pathogenic SCN9A variants were shown to be responsible for the episodic pain and autonomic features in the ocular, mandibular and sacral regions in paroxysmal extreme pain disorder (PEPD).9 10 Several years later, SCN9A gain-of-function variants were demonstrated in 28.6% (n=8/28) of a cohort of patients with skin biopsy-confirmed idiopathic pure SFN.11 Subsequently, pathogenic SCN10A12 and SCN11A13 variants were found in in patients with SFN.
A follow-up study of 393 patients diagnosed with SFN, based on typical clinical features (neuropathic pain and autonomic complaints) in combination with an abnormal intraepidermal nerve fibre density (IENFD) in skin biopsy and/or abnormal temperature threshold testing (TTT), showed that 9.1% (n=34/393) harboured an SCN9A variant, 4.2% (n=15/359) an SCN10A variant and 3.5% (n=12/345) an SCN11A variant.13 14
Over the past years our SFN-cohort has expanded to 1502 patients. Aims of the current study were to provide more precise data on SCN9A, SCN10A and SCN11A variant frequencies and a rationale for the genetic screening of patients with pure SFN and to compare the clinical features of patients with SFN with and without VGSC variants.
Study population and clinical characterisation
The retrospective study was conducted at the Departments of Neurology and Clinical Genetics of the Maastricht University Medical Centre+ (Maastricht UMC+), Maastricht, The Netherlands, a tertiary national referral centre for patients with clinical symptoms of SFN. Between September 2009 and January 2017, 1502 adult patients (age ≥18 years old) with SFN symptoms were examined in a structured day case setting. The following records were taken:
Demographic data; age of onset of complaints; duration of symptoms; altered pain sensation; presence of erythromelalgia symptoms, itch or cramps; influence of temperature, exercise or rest on pain; medical history and family history; neuropathic pain medication used at moment of presentation.
Neurological examination (muscle strength, pinprick sensation, vibration and position sense, tendon reflexes).
Nerve conduction studies (NCS; motor nerves: peroneal and tibial nerves to determine the compound muscle action potential amplitude, distal latency and conduction velocities; sensory nerves: ulnar and/or median and sural nerve to determine sensory nerve action potential amplitudes, distal latencies and conduction velocities).
Thermal threshold testing (TTT).15
Skin biopsy for determination of IENFD.16
Multiple questionnaires, including the Visual Analogue Scale to assess pain intensity17; Neuropathic Pain Scale (NPS) to evaluate 10 qualities of neuropathic pain18 and the SFN Symptom Inventory Questionnaire (SFN-SIQ) which includes 13 SFN specific symptoms.19
According to the international criteria, the diagnosis pure SFN was established when typical clinical symptoms were present in combination with a decreased IENFD in skin biopsy and/or abnormal TTT, without signs of large nerve fibre damage on neurological examination and/or NCS.2 20 Other underlying diseases or use of medication were excluded as possible causes. To search for associated conditions, patients diagnosed with pure SFN underwent extensive blood analyses as described previously.21
Genomic DNA was extracted from whole blood using NucleoSpin8 Blood Isolation kit (Macherey-Nagel, Düren, Germany) according to manufacturer’s instructions. Coding exons and exon-flanking intronic regions of SCN9A, SCN10A and SCN11A were amplified by PCR and sequenced by Sanger sequencing. Sequences were compared with reference sequence GRCh37. Variants detected were annotated according to guidelines of the Human Genome Variation Society (http://www.hgvs.org/mutnomen/). Variants which were located in functional domain of the protein and/or at a highly conserved amino acid in mammalian paralogues/human VGSC orthologues were classified and reported according the Practice Guidelines of the Association for Clinical Genetic Science (ACGS) and recommendations of Waxman et al.22 23 Cosegregation of (potentially) pathogenic variants with the disease was tested, if possible, in cases with a positive family history.
The primary analysis was the comparison of clinical variables between patients with pure SFN with and without VGSC variant. For categorical variables, the χ² test was used or the Fisher’s exact test when necessary. For continuous variables, the independent student’s t-test was chosen. Equal variances between two groups were tested with the Levene’s test. For these analyses, a significance level of 0.05 was used.
Posthoc analyses were performed to investigate whether differences between patients with specific VGSC variants and without variants were present. In total, six posthoc analyses per variable were executed. The analyses were performed in the same way as the primary analyses; however, the significance level was adjusted for multiple testing with the Bonferroni correction (0.05/6). Therefore, posthoc analyses were compared with a significance level of 0.0083. Missing values were not imputed or estimated.
The diagnosis of pure SFN was established in 1139 of 1502 patients (75.8%) referred to our centre (figure 1). More females (59.2%) were present than males. The mean age at presentation in our referral centre was 52.1 years (SD 13.2 years), with an age of onset of symptoms at 46 years (SD 14.6 years). In 25.9% of patients the diagnosis of pure SFN was based on the clinical picture in combination with both an abnormal IENFD and TTT, in 6.9% of the patients only the IENFD was abnormal, while in 67.2%, solely an abnormal TTT was found.
In 61.1% (n=696) of patients with pure SFN, additional workup revealed no associated conditions. Autoimmune diseases were found in 21.7% (n=247), glucose intolerance in 10.6% (n=121), vitamin B12 deficiency in 6.6% (n=75), diabetes mellitus in 5.2% (n=59), alcohol abuse in 2.7% (n=31), chemotherapy in 2.2% (n=25), monoclonal gammopathy of undetermined significance in 1.1% (n=13) and haemochromatosis in 0.6% (n=7) of patients, which is in line with our previous reports.21 Multiple associated conditions can be found within one patient.
Genetic screening of SCN9A
Among 1139 patients with pure, 28 different (potentially) pathogenic heterozygous SCN9A variants were detected in 58 patients (5.1%, figure 1 and table 1). Six variants have already been published as pathogenic,11 24–27 one variant as probably pathogenic,11 and two variants as risk factor.23 Eighteen variants were novel and classified as possibly pathogenic (n=3) or of uncertain clinical significance (VUS, n=15; table 1). Eleven SCN9A variants were found in >1 patient. Nine patients harboured more than one variant in SCN9A (table 1). Finally, one patient was heterozygous for two SCN9A VUSs, c.1555G>A and c.2271G>A and the pathogenic SCN11A c.3473T>C variant.
For nine variants cell electrophysiology showed a gain-of-function of the NaV1.7 channel.11 24–29 Cosegregation with pain in the family was demonstrated for six variants, was inconclusive for three variants and not supportive for two variants (table 1).
Nine (potentially) pathogenic SCN9A variants detected in our cohort of patients with pure SFN have been reported in patients with IEM, PEPD, paroxysmal itch, painful diabetic neuropathy (PDN) and Dravet syndrome.30–34 Two patients with SFN, positive for one of these SCN9A variants, had a history consistent with erythromelalgia (variant c.554G>A), two patients with SFN suffered from diabetes mellitus (one patient with variant c.1552G>T and one patient with variant c.2215A>G) and three patients with SFN complained of itch (variant c.2215A>G). None of the other patients carrying ≥1 of these SCN9A variants were positive for erythromelalgia, PEPD, paroxysmal itch, PDN or Dravet syndrome.
The clinical features of each individual patient harbouring (potentially) pathogenic SCN9A variants are shown in the online supplementary table 1.
Genetic screening of SCN10A
In SCN10A, 25 different (potentially) pathogenic heterozygous variants were detected in 42 patients with pure SFN (3.7%, n=45/1,139, figure 1 and table 2). One variant was already published as pathogenic35 and three variants as probably pathogenic12 36 Twenty-one variants were novel for SFN and classified as probably pathogenic (n=2), possibly pathogenic (n=2) or VUS (n=17; table 2). Five SCN10A variants were present in >1 patient. Two patients harboured two SCN10A variants. For three variants, cell electrophysiology showed a gain-of-function of the NaV1.8 channel and for one variant, DRG neuron hyperexcitability was seen (table 2).12 35 36 Cosegregation tested for three patients was only positive for one variant (table 2).
Eight (potentially) pathogenic SCN10A variants detected in our cohort of patients with pure SFN have been reported in patients with Brugada syndrome (BrS), atrial fibrillation (AF), sudden infant death syndrome (SIDS), Lennox-Gastaut syndrome (LGS), febrile infection-related epilepsy syndrome (FIRES) and autism.37–40 Only the patients with SFN with variant c.41G>T had arrhythmia. The other previously reported conditions were not seen in our cohort of patients with ≥1 SCN10A variants.
The individually clinical data of the 42 patients with pure SFN with (potentially) pathogenic SCN10A variants are shown in the online supplementary table 2.
Genetic screening of SCN11A
We found 20 different (potentially) pathogenic heterozygous variants in 33 patients with pure SFN (2.9%, n=33/1,139, figure 1 and table 3). Three variants were already published as probably pathogenic,13 41 and one variant as possibly pathogenic and five variants as VUS.13 Eleven variants were novel and classified as possibly pathogenic (n=5) or VUS (n=6; table 3). Six SCN11A variants were detected in >1 patient. Only one patient was heterozygous for two SCN9A VUSs (c.1555G>A and c.2271G>A) and the pathogenic SCN11A c.3473T>C variant.
Cell electrophysiology showed a gain-of-function of three SCN11A variants,13 41 while a loss-of-function of the NaV1.9 channel was seen for one variant.41 Cosegregation was tested for two variants and only supportive for one (table 3). All 25 (potentially) pathogenic SCN11A variants detected in our cohort were specific for SFN.
The clinical features per patient harbouring (potentially) pathogenic SCN11A variant are shown in the online supplementary table 3.
Patients with pure SFN with or without VGSC variant
The number of patients with SFN with a VGSC variant and a decreased IENFD was not significantly higher than that of patients without a VGSC variant (37.9% vs 32.6%; p=0.328). Furthermore, in both groups TTT was almost equally abnormal (92.4% with VGSC variant vs 93.2% without VGSC variant; p=0.741). Patients that harbour an SCN9A variant reported significantly more often erythromelalgia-like symptoms compared with patients with an SCN10A variant or without VGSC variant (43.9% vs 16.7%; p=0.004% and 43.9% vs 26.4%; p=0.004). The proportion of patients with SFN that experienced an aggravation of the pain by warm temperature was significantly higher in those with a VGSC variant compared with those without VGSC variant (45.2% vs 30.9%; p=0.014). No differences were seen in other symptoms, obtained by history taking and various questionnaires (figure 2 and figure 3). In case patients were harbouring multiple VGSC variants, in the same channel or in different channels, the type of complaints and severity was similar.
In patients with a VGSC variant, family history for SFN-related symptoms was more frequently positive than in the other patients with pure SFN (33.9% vs 24.6%; p=0.027; online supplementary table 4). In 15.1% (n=66/436) of patients with pure SFN with an underlying condition a (potentially) pathogenic VGSC variant was found. In the 132 patients with SFN with a VGSC variant the following underlying conditions were demonstrated: in 22.0% (n=29) an immunological disease, in 6.8% (n=9) glucose intolerance, in 5.3% (n=7) vitamin B12 deficiency, in 3.0% (n=4) diabetes mellitus, in 2.3% (n=3) alcohol abuse, in 0.8% (n=1) a history of chemotherapy and in 0.8% (n=1) a monoclonal gammopathy of undetermined significance. The number of patients per specific underlying condition was too small to study potential relationships between these conditions and particular VGSC variants.
As all of the patients in our cohort suffered from painful SFN and no patients with painless SFN were included, it was not possible to investigate if the presence of VGSC variants in patients with an associated condition is related to the development of pain. Since this study had a retrospective design, it was not possible to collect data about the use of pain medication in a standardised way to provide reliable information on the response to treatment. However, the data on pain features and intensity, indirectly suggest the poor efficacy of the drugs (online supplementary tables 1–3).
In our retrospective cohort, 132 of 1139 (11.6%) patients with pure SFN harbour potentially pathogenic heterozygous variants in SCN9A, SCN10A and/or SCN11A. SCN9A variants were found more frequently (5.1%, n=58/1139 patients) than SCN10A (3.7%, n=41/1139 patients) and SCN11A (2.9%, n=38/1139 patients) variants. Fifty variants were novel for SFN and classified as probably pathogenic (n=2), possibly pathogenic (n=10) or VUS (n=38). In this cohort, erythromelalgia showed a significant relationship with the presence of SCN9A variants. Furthermore, warmth-induced pain was significantly increased with the presence of potentially pathogenic VGSC variants. Other clinical features of pure SFN, such as abnormal TTT, abnormal IENFD, abnormal pain sensation, itch, cramp and cold-induced, exercise-induced and rest-induced pain, were not significant different for patients with and without VGSC variants (figure 2). Also, the NPS and SFN-SIQ revealed comparable results for patients with and without VGSC variants (figure 3).
The frequencies of variants in the SCN10A and SCN11A genes in our patients with pure SFN are slightly lower than those reported before, respectively, 4.2%–4.8% and 3.5%–3.8%.13 14 21 For SCN9A, however, the frequency has decreased over the years. At first, eight gain-of-function variants were identified in 28 subjects (28.6%) with biopsy-confirmed idiopathic pure SFN.11 Subsequently, in the cohort of 393 consecutive patients diagnosed with SFN, 17 potentially pathogenic variants were found in 34 patients (9.1%).13 14 Then, we reported in an extended cohort of 921 patients diagnosed with idiopathic pure SFN, 78 patients have ≥1 potentially pathogenic SCN9A variants (8.5%),21 and here we report 58 subjects with ≥1 potentially pathogenic SCN9A variants in 1139 patients with pure SFN (5.1%). The decline in frequency for SCN9A in the expanded cohort can be explained by less stringent inclusion criteria for the current cohort, more stringent variant classification criteria, extension of functional data obtained by cell electrophysiology and data of cosegregation analysis. For instance, variants with no change in channel function and/or cosegregation with the disease in affected family members, like c.3734A>G (p.Asn1245Ser) and c.3799C>G (p.Leu1267Val), were classified as unlikely to be pathogenic in the current study, while they were classified as VUS with frequencies, respectively, of 1.5% and 1.3% in the cohort of 393 patients with SFN,13 14 and 1.1% and 0.9% in the cohort of 921 patients with idiopathic pure SFN.
To date, only one other cohort of patients with painful neuropathy (n=217) has been tested for SCN9A, SCN10A and SCN11A gene variations.42 In this cohort, the number of patients with ≥1 low-frequency (minor allele frequency (MAF)<5% in the NHLBI Exome Sequencing Project Exome Variant Server, European American (EVS-EA) population) missense variant in SCN9A, SCN10A and SCN11A was respectively 25%, 21% and 13%. From these low-frequency missense variants, 8.7% SCN9A (n=19/217), 0.9% SCN10A (n=2/217) and 0.9% SCN11A (n=2/17) variants have been previously reported in patients with IEM, SFN or PDN, including SCN9A variant c.3734A>G (n=7/217, 3.2%) and c.3799C>G (n=1/217, 0.5%).42 Compared with our cohort, the incidence of potentially pathogenic variants in the 217 patients with painful neuropathy was ~4–5-fold higher. The discrepancies in frequencies of reported variants between both cohorts are mainly caused by different variant filtering strategies (ie, MAF<5% vs<1%) and variant classification approaches (ie, all missense mutations vs highly conserved missense mutation). Different patient inclusion criteria and used sequencing platforms may also have an effect the frequencies of variants detected.
Nine (potentially) pathogenic SCN9A variants detected in our cohort of patients with pure SFN have been reported as disease-causing variants in other pain phenotypes and Dravet syndrome.30–34 Multigenerational segregation with the disease in SFN families and/or functional testing in DRG neurons by cell electrophysiology support that these variants are causative for SFN (table 1). As patients with SFN share clinical features with other pain-phenotypes, it is not surprising that these variants are reported for SFN and IEM, paroxysmal itch and painful diabetic neuropathy. Features of Dravet syndrome have not been seen in our cohort, and no abnormal pain sensation have been reported for patients with Dravet syndrome and a (potentially) pathogenic SCN9A variant.43 For SCN10A variants, eight (potentially) pathogenic variants have been report as disease-causing in BrS, AF, SIDS, LGS, FIRES and autism.37–40 Multigenerational segregation with the disease in SFN families and/or functional testing in DRG neurons by cell electrophysiology indicate that the majority of the variants are causative for SFN (table 2). Variant c.41G>T, p.(Arg14Leu), which has been described in patients with BrS and AF,37 38 was identified four times in our cohort of patients with SFN. Only one patient had SFN and arrhythmia. SIDS, LGS, FIRES and autism have not been seen in our patients with pure SFN. For SCN11A, none of potentially pathogenic variants identified in the current study have been associated with other inherited disorders. Taken together, these findings suggest that one variant can produce multiple different disease outcomes, depending on cell-type-specific expression11 24 25 presence of additional disease-causing variants or modifiers23 and/or additional underlying conditions (eg, diabetes mellitus or vitamin B12 deficiency).21
For variants where segregation with disease in multigenerations was confirmed, this should be considered supporting, though not definite evidence for pathogenicity. A definite conclusion would require at least 10 meioses, which is generally not the case. In this study, cosegregation of the variant with disease was only tested for 15 potentially pathogenic variants in 20 families, because most families were too small to test segregation with disease properly or the proband was not in contact with their relatives.
About two-third of the potentially pathogenic SCN9A variants identified in our cohort of patients with pure SFN have been localised to domain I (DI) and II (DII) and the intracellular linker between DI-DII and DII-domain III (DIII). This is in contrast with our findings for SCN10A, where approximately 70% of the SCN10A variants were localised to DIII and domain IV (DIV), the intracellular linker between DIII-DIV and C-terminus of the protein. For SCN11A, the distribution of potentially pathogenic variants was all across the gene. No possible hotspots were identified in this gene.
Patch clamp studies have shown that SCN9A variants associated with SFN produce gain-of-function channel changes, ranging from impaired slow-inactivation to depolarised slow-inactivation and fast-inactivation, and induce DRG neurons hyperexcitability.11 24–29 Compared with PEPD, where most SCN9A variants were located in the intracellular linker between DIII-DIV and intracellular loop linking segments S4-S5 of DIII and DIV, effects on fast-inactivation of SFN SCN9A variants were relatively mild. IEM SCN9A variants exhibit hyperpolarised activation or enhanced ramp currents and were mainly localised to transmembrane segment S4, S5 and S6 and the intracellular loop linking segments S4-S5.44 Functional IEM characteristics were not demonstrated for SFN SCN9A variants. Although most potentially pathogenic SCN9A variants in this cohort were localised to DI and DII and the intracellular linkers DI-DII and DII-III, it is still unclear how this type of NaV1.7 channel dysfunction causes SFN. However, certain SFN SCN9A variants have been shown to impair regeneration and/or degeneration of sensory axons, suggesting that enhanced sodium channel activity and reverse Na-Ca exchange may contribute to a decrease in length of peripheral sensory axons.45
Besides SCN9A, variants in SCN10A and SCN11A have been shown to participate in the pathophysiology of SFN. DRG neurons expressing SCN10A or SCN11A mutant channels exhibit increased excitability and abnormal spontaneous firing activity.12 13 35 36 41 One SCN10A variant and nine SCN11A variants have been associated with other inherited pain disorders.46–50 Although different cell electrophysiology properties were seen for several of these varians,12 13 35–37 41 46 functional data for SCN10A and SCN11A variants are too limited to correlate channel phenotype with clinical phenotype.
In conclusion, in this cohort of 1139 patients with pure SFN, the overall frequency of potentially pathogenic SCN9A, SCN10A and SCN11A variants is 11.6%. Erythromelalgia and warmth-induced pain were the only SFN-related clinical features that showed a significant relationship with the presence of VGCS variants. As genetic screening for all patients with pure SFN is debatable, we believe that in the future, as the number of well-characterised variants of Nav channels increases, and an interexpert concordance on variant classification is reached, the utility of genetic screening for clinical care will rise and tailored treatments with specific sodium channel blockers can be developed. Furthermore, we have seen that certainty about the origin of symptoms as well as genetic counselling by which the patient and relatives are informed about the possibility of developing and transmitting the condition, is of great importance for patients with pure SFN. Therefore, genetic screening of SCN9A, SCN10A and SCN11A should be considered for patients with pure SFN, independently of clinical features or underlying conditions.
This study was approved by the Medical Ethics Committee and Board of Directors of the Maastricht UMC+.
IE and MS contributed equally.
Correction notice Since this paper was first published online a funding statement has been added to the proof.
Contributors Study concept and design: IE, MS, JGJH, BTADG, HJMS, ISJM, CGF, MMG. Acquisition, analysis or interpretation of the data: all authors. Statistical analysis: IE, MS, BTADG, PL, ISJM. Drafting of the manuscript: IE, MS, JH, BTADG, MMG. Review of the manuscript: RA, MM, JV, HJMS, SGW, GL, ISJM, CGF. Study supervision: JGJH, HJMS, ISJM, CGF, MMG.
Funding This work was supported in part by a grant from European Union’s Horizon 2020 research and innovation programme Marie Sklodowska-Curie grant for PAIN-Net, Molecule-to-man pain network (grant no. 721841).
Competing interests JGJH reports personal fees from Pfizer Inc. (travel funding and speakers’ honorarium) and grants from Prinses Beatrix Spierfonds (W.OK17-09), outside the submitted work. SGW reports grants from European Union’s Horizon 2020 research and innovation programme Marie Sklodowska-Curie grant for PAIN-Net, Molecule-to-man pain network (grant no. 721841) and European Union 7th Framework Programme (grant no. 602273) for the PROPANE study and was supported in part by the Rehabilitation Research and Development Service and Biomedical Laboratory Research Service, Department of Veterans Affairs, outside the submitted work. GL reports grants from European Union’s Horizon 2020 research and innovation programme Marie Sklodowska-Curie grant for PAIN-Net, Molecule-to-man pain network (grant no. 721841) and European Union 7th Framework Programme (grant n°602273) for the PROPANE study and participates in Steering committees/advisory boards for studies in small fibre neuropathy of Biogen/Convergence, Vertex and Chromocell, outside the submitted work. ISJM reports grants from Talecris Talents program, GSB CIDP Foundation International, Prinses Beatrix Spierfonds (W.OR12-01, W.OR15-25) and European Union 7th Framework Programme (grant no. 602273), participates Steering committees of the Talecris ICE Study, LFB, CSL Behring, Novartis, Grifols and Octapharma, serves on the editorial board of the Journal of Peripheral Nervous system and is a member of the Inflammatory Neuropathy Consortium (INC) and Peripheral Nerve Society, outside the submitted work. CGF reports grants from European Union’s Horizon 2020 research and innovation programme Marie Sklodowska-Curie grant for PAIN-Net, Molecule-to-man pain network (grant no. 721841), European Union 7th Framework Programme (grant n°602273) for the PROPANE study, Prinses Beatrix Spierfonds (W.OR12-01, W.OR15-25), Grifols and Lamepro for a trial on IVIg in small fibre neuropathy and participates in Steering committees/advisory boards for studies in small fibre neuropathy of Biogen/Convergence, Vertex and Chromocell, outside the submitted work. Other authors have no conflicts of interests to declare.
Patient consent Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.