Journal Logo

Research Paper

Systemic inflammatory markers in neuropathic pain, nerve injury, and recovery

Sandy-Hindmarch, Olivera; Bennett, David L.a; Wiberg, Akirab; Furniss, Dominicb; Baskozos, Georgiosa; Schmid, Annina B.a,*

Author Information
doi: 10.1097/j.pain.0000000000002386

1. Introduction

It is well established from preclinical models that neuroinflammation plays an important role in the initiation and maintenance of neuropathic pain.20,50 The presence of neuroinflammation has also been confirmed in patients with severe peripheral neuropathies, where nerve biopsies are warranted.22,63,73,74,82 However, the limited access to human neural tissue means that the clinical detection of inflammation mostly relies on indirect measures, such as the presence of systemic inflammatory markers in the blood. Most studies report changes in blood inflammatory markers in patients with neuropathic pain; however, the limited number of cytokines studied (eg, Tumor Necrosis Factor [TNF] TNF, IL-6, IL1β, IL-4, and IL-10) prevents a comprehensive overview and discovery of contributions of less studied cytokines in most studies.9,34,36,64,77

Along with the role of inflammation during initiation and maintenance of neuropathic pain, its contribution to recovery and resolution has gained increasing interest in recent years.17,38 Studying the role of the immune system in the resolution of neuropathic pain remains challenging in humans because many neuropathic pain conditions are chronic and treatments are often only modestly effective. There is some indication from analysis of serial blister fluid from patients with complex regional pain syndrome that inflammatory markers normalise over time; however, this does not seem to be related to treatment outcome or disease characteristics.40,78 There is also evidence of a role for inflammatory mediators in patients with sciatica, where changes in some serum inflammatory mediators over time show associations with pain and recovery.34 However, most studies restrict their analyses to a limited number of inflammatory markers (eg, IL-6, IL-8, IL-4, IL-1β, TNF, and C-reactive protein [CRP]).37,54,55,59,67,85 Comprehensive longitudinal analyses of human inflammatory changes at both gene and protein levels are needed to shed light on the role of inflammation in neuropathic pain maintenance and resolution.

Here, we use carpal tunnel syndrome (CTS) as a human model system to prospectively study inflammation in the context of neuropathic pain and its resolution. Carpal tunnel syndrome is the most common peripheral neuropathy and cause for neuropathic pain.1 Unlike many other neuropathic pain conditions, CTS can successfully be treated by a single and time-locked intervention: surgical decompression. CTS therefore represents an ideal model system that allows the prospective evaluation of inflammation from the active stage of nerve injury (presurgery) to recovery (postsurgery). There is growing evidence for a role of blood inflammatory mediators in the active stage of CTS,49,69 although conflicting results have also been reported.26,36,69 The most comprehensive cross-sectional study to date suggested that serum concentrations of C-C motif chemokine ligand 5 (CCL5), vascular endothelial growth factor (VEGF), CXCL8, and CXCL10 as well as the number of central and effector memory T-cell populations were significantly increased in patients compared with healthy controls, confirming the presence of systemic inflammation in CTS.49 Here, we analyse a comprehensive set of blood inflammatory mediators at both mRNA and protein levels (1) in the active stage of nerve injury (CTS presurgery vs healthy controls) and (2) in recovery (paired patient samples presurgery and postsurgery). We also explore associations of inflammatory markers with patients' symptoms and specifically neuropathic pain.

2. Methods

2.1. Participants

We used the data available from the prospective longitudinal Oxford CTS cohort.2 Patients with clinically and electrodiagnostically confirmed CTS were recruited from surgical waiting lists at Oxford University Hospitals NHS Foundation Trust. Patients were excluded if electrodiagnostic testing (EDT) revealed a nerve dysfunction other than CTS, if there was another medical condition affecting the upper limb or neck (eg, hand osteoarthritis or cervical radiculopathy), if there was a history of significant trauma to the upper limb or neck, or if they were pregnant. Patients with potentially confounding conditions such as autoimmune or inflammatory disease (eg, rheumatoid arthritis or multiple sclerosis), active infection (eg, hepatitis), other systemic illnesses (eg, diabetes or cancer), or those taking immunosuppressive medications were also excluded. Patients undergoing repeat CTS surgery were also excluded. Patients with CTS were assessed at baseline (before surgery) and 6 months after surgery, when functional and structural neural recovery as well as symptom improvement is apparent2 and systemic cytokine levels are unlikely to be influenced by a potential inflammatory reaction related to the surgical intervention.

Twenty-one, age-matched and gender-matched, healthy controls were also included in the study. They did not have any systemic illness, including potentially confounding conditions mentioned above, did not experience pain in the hand in the past 3 months, and EDT of the radial, ulnar, and median nerve was within normal limits. Healthy participants were recruited within the university department, through public notice boards and media advertisements. All healthy controls attended one assessment. Ethical approval was given for the project (Riverside London Ethics Committee Ref 10/H0706/35), and all participants provided informed written consent before participating.

2.2. Phenotypic data

A detailed description of the phenotypic data collected is available elsewhere.2 For this study, we included age, sex, height, weight, and body mass index as baseline variables. Symptom duration was recorded in months. Symptom severity was evaluated with the symptom subscale of the Boston Carpal Tunnel Questionnaire41 (0 = no symptoms to 5 = severe symptoms). Neuropathic pain severity was evaluated with the Neuropathic Pain Symptom Inventory (NPSI),8 which includes numerical rating scales (0 = no pain to 10 = worst pain imaginable) for burning pain, deep pressure pain, paraesthesia, paroxysmal pain, evoked pain, as well as a composite score (0-100). The severity of pain over the past 24 hours was recorded on a visual analogue scale (0 = no pain to 10 = worst pain imaginable). Surgical outcome was determined with the Global Rating of Change Scale (GROC), which ranges from −7 (a very great deal worse) to +7 (a very great deal better).32 A patient was considered to have a successful recovery after surgery if they reported a GROC score of ≥ +5 (a good deal better).35 Standard EDT of the median, ulnar, and radial nerve was performed with an ADVANCE system (NeuroMetrix, Waltham, MA). Electrodiagnostic test severity was graded on the scale derived by Bland6 as follows: normal (grade 0), very mild (grade 1), mild (grade 2), moderate (grade 3), severe (grade 4), very severe (grade 5), and extremely severe (grade 6). For a more detailed description of the EDT refer to the study by Schmid et al.60

2.3. Blood sampling and processing

Three millilitres of venous blood was sampled into RNA stabilising tubes (Tempus blood RNA tube, Fisher Scientific, Loughborough, United Kingdom) and stored at −20°C for batch processing. Blood serum was extracted from whole blood collected into a BD Vacutainer SST tube for serum collection (BD, Wokingham, United Kingdom). The blood was left to clot before being centrifuged at 3000 rpm for 10 minutes at 4°C. The serum fraction was aliquoted and stored at −80°C for batch processing.

2.4. Gene expression

RNA was extracted from blood following published protocols (Tempus Spin RNA Isolation Kit, Thermo Fisher, Paisley, United Kingdom). In brief, samples were defrosted and Phosphate Buffered Saline (PBS) added to each sample which was then vortexed and centrifuged. The RNA pellet was resuspended and purified using column filtration. RNA was converted into cDNA using the EvoScript Universal cDNA Master kit (Roche, Welwyn Garden City, United Kingdom). Custom made TaqMan array microfluidic cards (Thermo Fisher) were designed containing 44 markers implicated in inflammation or neuropathic pain as well as 3 housekeeping genes. TaqMan array cards were used because they are highly sensitive45 and use a preloaded assay format, with the remaining master mix and sample being added through specialised loading ports, which substantially reduces operator error. The gene list contained cytokines and chemokines, both anti-inflammatory and proinflammatory, and immune cell markers such as CD3D, CD16, and CD14 to detect the presence of T cells, neutrophils, and monocytes, respectively. The full list of genes can be found in Supplementary Table 1 (available at https://links.lww.com/PAIN/B422). The cards were run as per standard protocol. In brief, 60 µL of patient cDNA was mixed with 60 µL of TaqMan Fast Advanced Master Mix to achieve a final volume of 120 µL and cDNA concentration of 10 ng/µL. Paired patient samples from before and after surgery were processed on the same card with each assay being run in singlet. The cards were run on a QuantStudio 7 Flex Real-Time PCR System (Applied Biosystems, Waltham, MA). Cycle times (Ct) for each gene in each sample were recorded and used in future analyses. We included TRAP1 and DECR1 as housekeeping genes because they are stably expressed in human blood.52,65 On the microfluidic card, 18S was also included as a mandatory housekeeping gene. The average expression of TRAP1, DECR1, and 18S was subtracted from the expression of the genes of interest to provide normalised expression values.

2.5. Protein levels

U-PLEX plate custom biomarker multiplex assay kits (Meso Scale Diagnostics LLC, Rockville, MD) were custom designed to detect 18 selected cytokines or chemokines related to the gene expression data with high sensitivity (Supplementary Table 1, available at https://links.lww.com/PAIN/B422).12,16 Processing followed the standard manufacturer protocols. Each capture antibody was combined with a specific linker molecule and incubated at room temperature for 30 minutes. The linking reaction was inhibited with the addition of stop solution. Linked capture antibodies were pooled and 50 µL was added to each well and incubated at 4°C overnight on a shaker. The next day, the capture antibody solution was removed and the plate was washed 3 times with PBS +0.05% Tween 20, followed by addition of 25 µL of assay buffer along with either 25 µL of serum or assay standards. The plates were incubated on a shaker at room temperature for 1 hour. Patient samples and standards were removed, and the wells were washed 3 times with wash buffer. Fifty microliters of detection solution was added to each well before incubation for 1 hour at room temperature on a shaker. The plates were washed 3 times with wash buffer, and 150 µL of read buffer was added before the plates were read according to the manufacturer's instructions on a MESO QuickPlex SQ 120 plate reader (Meso Scale Diagnostics LLC).

For detection of transforming growth factor (TGF)-β, the same procedure was followed, but acidification was required as per standard protocol: samples were treated with 20 µL of 1M HCl per 100 mL and incubated at room temperature for 10 minutes before neutralising by the addition of 14 µL of 1.2M NaOH in 0.5M HEPES buffer per 100 µL of sample.

As CCL5 (RANTES) was not available in the U-PLEX panel, the detection of CCL5 was performed with R-PLEX plates (Meso Scale Diagnostics LLC). The standard manufacturer protocol was used. The procedure was the same as described above with the exception that the serum samples were diluted 1:50. No specific linker molecules were used; instead, streptavidin-coated plates were used to bind biotinylated anti-CCL5 capture antibodies.

To detect human CRP in serum, we used a CRP Quantikine ELISA kit (R&D Systems, Minneapolis, MN) and followed the standard protocol. In brief, 100 µL of assay diluent was added to each well, followed by 50 µL of either standard or sample (serum samples were diluted 100-fold in calibrator diluent). We ran the samples along with an 8-point standard curve in duplicate. The plates were sealed and incubated at room temperature for 2 hours. Samples or standards were then aspirated off, and wells were washed 4 times with 400 µL wash buffer. Two hundred microliters of human CRP conjugate (secondary antibody with horseradish peroxidase activity) was added to each well and incubated at room temperature for 2 hours. The plates were then washed 4 times with 400 µL buffer and 200 µL of substrate solution was added. The plates were incubated at room temperature for 30 minutes in the dark. Fifty microliters of stop solution was added, and the plates were read on a BMG FLUOstar Omega (BMG Labtech Ltd, Aylesbury, United Kingdom) with the wavelength set to 450 nm.

All patient samples were run in duplicate, and paired patient samples (before and after surgery) as well as standards were processed on the same plate. Standards were composed of known concentrations of each cytokine or chemokine being analysed and were used to calculate the protein concentrations in each patient sample. Quantification of the target inflammatory mediators was based on duplicates of an 8-point calibration curve that was calculated automatically using discovery workbench software for MSD plates or interpolated from the standard curve in graphpad prism 9 for CRP.

2.6. Statistical analysis

This analysis is a secondary analysis of a published cohort2 of exploratory character and did therefore not include an a priori sample size calculation.29 Genomic data were analysed with the statistics package R,70 using the software package limma57 to determine differential gene expression. Batch correction was conducted on normalised data sets. The batch corrected data were input into an empirical Bayesian statistical model to shrink variance within the data. Data were then fitted into a mixed linear model to determine differential gene expression using patient as blocking factor. Differential expression was investigated 1) between patients before surgery and healthy controls (active stage of nerve injury) and 2) between patients before and after surgery (recovery). Results were corrected for false discovery rate (FDR) with Benjamini–Hochberg correction where an adjusted P value of <0.05 was considered significant.

Mann–Whitney U tests were used to compare serum protein levels between healthy participants and patients with CTS before surgery (active stage of nerve injury). Wilcoxon signed-rank tests were conducted to compare serum protein levels between patients with CTS before and after surgery (recovery). Results were corrected for FDR using Benjamini–Hochberg correction where an adjusted P value of <0.05 was considered significant. Any protein remaining significant after FDR correction was included in further analyses.

To determine associations between gene or protein expression and clinical phenotype, Spearman rank correlation analyses were conducted. Non-normal and zero-inflated data dictated the usage of the nonparametric and not sensitive to outliers Spearman rank-order correlation to assess the strength of monotonic relationships. Both before and after surgery data were used for most analyses. The only exception was for duration of symptoms (where only presurgery data were used) and GROC scores (where only postsurgery data were used). To limit the number of potential correlations and therefore type 1 error, only the 2 genes with the largest log2 fold change (Log2FC) and that were significant after FDR correction in each analysis were used in further analyses to determine associations with phenotype. In some instances, either the mRNA expression value or the protein level could not be determined for a particular inflammatory mediator, in any given patient. As such, some data points in the correlation analysis were not available. Scatter plots between mRNA expression or protein levels and clinical phenotype scores were first inspected to determine a monotonic relationship. The Spearman correlation was only conducted for monotonic data. Several of the clinical scores included zero values. These values represent genuine sampling points for instance reflecting complete symptom resolution and were thus retained. The Spearman correlation can cope well with zero-inflated data of up to 30%.30 However, to fully determine the effect of zero values, we used a hurdle model consisting of a truncated Poisson model fitted to nonzero scores and a binomial model fitted to zero scores. We used the same approach to determine potential associations between preoperative and postoperative changes in inflammatory markers and clinical phenotypes comparing Log2FCs of gene or protein levels and clinical phenotype scores. As these investigations were exploratory in nature, used a small sample size, did not involve repetitive hypothesis testing, and because exploratory correlations were performed for interdependent variables, FDR correction was not suitable in this instance.

The effects of age, sex, and body mass index (BMI) on the mRNA and protein expression of inflammatory mediators were found to be very limited (Supplementary Tables 2 and 3, available at https://links.lww.com/PAIN/B422) after fitting a mixed linear model or preforming a two-way analysis of variance including covariates for age, sex, and BMI on mRNA and protein expression data, respectively. As these covariates were uninformative, they were not included in the differential expression analysis or in correlations between inflammatory mediator expression and clinical phenotype data. Their inclusion would have added unnecessary complexity to the models, reducing power and diluting effects. In the case of postsurgery vs presurgery comparison, the usage of a mixed linear model controls for individual differences in baseline expression and consequently for variance due to differences in individuals' age, sex, or BMI.

3. Results

Baseline and clinical data of the 55 patients with CTS and 21 healthy controls can be found in Table 1. Patients and controls were comparable for age, sex, height, weight, and BMI (Table 1). After surgery, 47 patients (85%) were classified as successfully recovered as they reported a GROC score of ≥ +5.

Table 1 - Baseline and clinical data.
Healthy CTS pre CTS post
Number of participants 21 55 55
Age (y) 63 [21] 64 [16]
Female gender, n (%) 14 (66.7%) 37 (67.3%)
Mean height (SD) [cm] 169.12 (9.34) 168.19 (8.80)
Weight (kg) 70.4 [15.75] 68.7 [16.8]
BMI (kg/m2) 26 [5] 24 [5]
Duration of symptoms (mo) 36 [42]
EDT grade 3 [2] 2 [2]
 Normal, n (%) 0 (0) 7 (13)
 Very mild, n (%) 4 (7) 18 (33)
 Mild, n (%) 7 (13) 9 (17)
 Moderate, n (%) 18 (33) 11 (20)
 Severe, n (%) 11 (20) 7 (13)
 Very severe, n (%) 13 (24) 2 (4)
 Extremely severe, n (%) 2 (4) 1 (2)
Boston symptom score 2.55 [1.12] 1.27 [0.45]
Boston function score 2.13 [1.19] 1.25 [0.63]
VAS pain 1.7 [4.05] 0 [0]
NPSI total score 21 [25] 1 [7.5]
 Burning pain 0 [4] 0 [0]
 Deep pain 1.5 [2.75] 0 [0]
 Evoked pain 0.33 [2.67] 0 [1]
 Paraesthesia 6 [4.75] 0 [0]
 Paroxysmal pain 0 [3] 0 [0]
GROC score 7 [1]
Data are presented as median with interquartile range [square brackets] unless indicated otherwise.
CTS, carpal tunnel syndrome; IQR, interquartile range; EDT, electrodiagnostic testing; VAS, visual analogue scale; NPSI, Neuropathic Pain Symptom Inventory; GROC, Global Rating of Change; SD, Standard Deviation.

3.1. Systemic inflammatory changes in the active stage of nerve injury

Table 2 contains the results of the gene expression analyses in the active stage of CTS compared with healthy controls. Differential expression analysis between healthy controls and patients with CTS presurgery revealed expression of a single gene, PTGES2 encoding a membrane bound enzyme which catalyses the conversion of prostaglandin H2 to prostaglandin E2, to be significantly decreased in patients compared with healthy controls (adjusted P = 0.013).

Table 2 - Gene expression changes in the active stage of carpal tunnel syndrome.
Gene LogFC AveExpr P Adj.p.Val
PTGES2 0.72 7.35 0.000 0.013
FCGR3B −0.99 6.71 0.002 0.052
IL-4 −0.64 11.08 0.004 0.055
CXCL5 −0.88 6.21 0.005 0.055
IL23A 0.36 9.26 0.011 0.076
CCL5 −0.42 1.06 0.013 0.076
CXCL8 −0.76 6.59 0.012 0.076
IL12B −0.94 16.60 0.008 0.076
IL1B −0.43 5.96 0.022 0.095
TLR4 −0.43 5.15 0.019 0.095
TGFB1 −0.22 2.59 0.058 0.224
PDGFA −0.39 9.65 0.061 0.224
CXCL10 0.50 10.69 0.071 0.245
IL-9 −0.61 16.86 0.079 0.254
IFNG −0.33 9.73 0.114 0.303
MMP9 −0.36 4.82 0.103 0.303
IL-6 −0.97 11.83 0.116 0.303
CX3CL1 −2.56 15.64 0.123 0.303
IL22 −4.55 14.22 0.126 0.303
CRP 2.04 15.74 0.156 0.357
CCL2 0.40 13.04 0.181 0.377
CCL21 5.59 15.30 0.212 0.424
CHI3L1 −0.41 6.04 0.240 0.462
NOS2 −0.69 15.83 0.253 0.467
IL10 0.28 13.87 0.278 0.480
CD80 −0.22 11.57 0.280 0.480
IL7 −0.16 9.58 0.423 0.700
TNF −0.20 6.43 0.443 0.709
CD14 0.11 8.27 0.471 0.729
VEGFA 0.09 9.72 0.512 0.744
CCL4 −0.15 6.95 0.500 0.744
CXCL11 0.13 11.72 0.528 0.746
CD3D 0.08 4.39 0.546 0.746
IL13 −0.18 16.15 0.606 0.746
IL17A −1.15 15.82 0.605 0.746
CCL11 −8.98 7.21 0.594 0.746
TAC1 4.44 −1.08 0.640 0.749
NGF 3.83 8.19 0.674 0.770
IL18 −0.15 6.83 0.716 0.799
IL1RN 1.05 11.79 0.808 0.881
IL2 −0.04 13.30 0.849 0.891
CXCL9 −0.05 11.30 0.854 0.891
CSF3 0.39 14.03 0.913 0.932
IL5 0.09 13.66 0.974 0.974
Changes in gene expression in patients with CTS (presurgery) compared with healthy controls. Genes are ranked in the descending order based on adjusted P values. Significant dysregulation (adjusted P < 0.05) is shown in bold.
AveExpr, average expression (average normalised Ct value across all samples); Adj.p.Val, adjusted P value; CRP, C-reactive protein; CTS, carpal tunnel syndrome; LogFC = log2 fold change.

The concentrations of inflammatory proteins in patients with CTS before surgery compared with healthy participants revealed 2 mediators that were significantly different. Both TGF-β (adjusted P = 0.016) and CCL5 (adjusted P = 0.047) were increased in patients with CTS presurgery compared with healthy controls (Fig. 1 and Table 3).

F1
Figure 1.:
Significant changes in serum inflammatory protein levels. (A) Graphs show significantly increased serum protein levels for TGF-β and CCL5 in patients with CTS presurgery (pink) compared with healthy controls (green). (B) Graph shows significantly downregulated IL-4 in patients with CTS after (red) compared with before surgery (pink). Data are shown as violin plots with median, quartiles, and single data points. Significant dysregulation (FDR-corrected Mann–Whitney U tests and Wilcoxon tests, respectively) is indicated with *P < 0.05. CTS, carpal tunnel syndrome; FDR, false discovery rate.
Table 3 - Serum level changes in the active stage of carpal tunnel syndrome.
Healthy CTS (presurgery)
Inflammatory mediator Median IQR Median IQR P Adj.p.Value
TGF-β 42,447.69 15,597.64 55,410.50 20,339.69 0.0008 0.016
CCL5 93,212.10 72,600.04 140,804.24 100,719.45 0.005 0.047
IL-4 0.01 0.05 0.05 0.06 0.040 0.267
CXCL10 190.02 114.23 233.83 158.86 0.092 0.460
CCL2 263.16 82.47 288.87 122.57 0.127 0.488
VEGF 74.75 50.40 94.88 81.73 0.147 0.488
IL-10 0.20 0.16 0.17 0.13 0.198 0.523
CXCL8 7.90 4.14 9.03 5.13 0.209 0.523
IL-6 0.80 0.31 0.66 0.62 0.370 0.812
IL-9 0.31 0.23 0.23 0.27 0.406 0.812
IL-2 0.00 0.00 0.00 0.00 0.491 0.857
IFN-γ 6.69 7.49 6.65 5.67 0.546 0.857
IL-17 0.00 0.00 0.00 0.00 0.693 0.857
IL-1β 0.08 0.07 0.09 0.07 0.727 0.857
CRP 1.27 2.70 1.49 1.54 0.728 0.857
GM-CSF 0.00 0.01 0.00 0.04 0.743 0.857
TNF-α 0.56 0.29 0.50 0.35 0.751 0.857
CXCL5 1565.00 1467.27 1533.24 1068.58 0.771 0.857
IL-12 0.11 0.30 0.00 0.30 0.872 0.899
Fractalkine 6061.66 1333.26 6149.32 2099.72 0.899 0.899
Differences in serum inflammatory marker levels in patients with CTS (presurgery) compared with healthy controls. Inflammatory mediators are ranked in descending order based on adjusted P values. Significant dysregulation (P < 0.05) is shown in bold. Values are given in pg/mL apart from CRP which is provided in mg/L
Adj.p.Value, adjusted P value; CRP, C-reactive protein; CTS, carpal tunnel syndrome; IQR, interquartile range.

3.2. Systemic inflammatory changes associated with resolution of the disease

Table 4 contains the paired gene expression analysis related to the resolution of CTS. A total of 12 genes were differentially expressed: IL-9, CCL5, PDGFA, IL-1β, CXCL5, TGFB1, VEGFA, IL-4, TLR4, FCGR3B, IL-6, and CD3D. Of these, the 2 genes with the largest Log2FC were IL-9 and IL-6, where IL-9 mRNA was increased after surgery (Log2FC = −1.099) and IL-6 mRNA was decreased after surgery (Log2FC = 0.92). These 2 genes were included in further analyses.

Table 4 - Gene expression changes associated with resolution of carpal tunnel syndrome.
Gene LogFC AveExpr P Adj.p.Val
IL-9 −1.10 16.26 0.0003 0.014
IL-6 0.92 12.02 0.008 0.034
CXCL5 −0.55 5.69 0.002 0.027
FCGR3B −0.45 6.22 0.006 0.031
IL-1β −0.33 5.68 0.002 0.027
TLR4 −0.31 4.88 0.006 0.031
IL-4 −0.31 10.75 0.006 0.031
PDGFA −0.27 9.41 0.003 0.027
CCL5 −0.24 0.83 0.003 0.027
VEGFA −0.20 9.64 0.007 0.031
CD3D 0.17 4.50 0.010 0.039
TGFβ1 −0.16 2.46 0.005 0.031
IL10 −0.29 13.80 0.026 0.095
IL13 −0.38 15.89 0.029 0.100
MMP9 −0.25 4.59 0.031 0.101
CHI3L1 −0.28 5.79 0.058 0.175
CXCL10 −0.24 10.71 0.064 0.181
NOS2 −0.86 15.23 0.109 0.275
CXCL8 −0.18 6.29 0.149 0.357
CX3CL1 1.48 15.62 0.181 0.414
IL18 −0.32 6.63 0.227 0.495
CSF3 2.02 15.55 0.260 0.543
IL22 2.21 13.88 0.306 0.612
CCL11 15.43 9.36 0.394 0.727
TNF 0.10 6.42 0.449 0.799
TAC1 6.55 0.49 0.545 0.837
IL2 −0.07 13.26 0.540 0.837
CCL2 −0.08 13.11 0.558 0.837
IL12B −0.18 16.10 0.587 0.854
NGF 1.05 10.73 0.773 0.863
IL1RN 0.97 12.98 0.694 0.863
CRP 0.13 16.46 0.716 0.863
CXCL11 0.03 11.77 0.769 0.863
CD14 0.03 8.31 0.729 0.863
IL7 −0.03 9.52 0.739 0.863
CD80 −0.04 11.49 0.717 0.863
CXCL9 −0.05 11.26 0.708 0.863
PTGES2 −0.05 7.53 0.638 0.863
IL17A −0.72 14.82 0.642 0.863
IL5 0.19 13.78 0.903 0.975
CCL21 0.09 16.52 0.955 0.975
CCL4 0.01 6.91 0.947 0.975
IFNG −0.01 9.63 0.921 0.975
IL23A 0.00 9.36 0.978 0.978
Changes in gene expression in patients with CTS before compared with after surgery. Genes are ranked in descending order based on adjusted P values. Significant change (adjusted P < 0.05) is shown in bold where differentially expressed genes are ranked by log2FC.
Adj.p.Val; adjusted P value; AveExpr, average expression (average normalised Ct value across all samples); CRP, C-reactive protein; CTS, carpal tunnel syndrome; Log2FC; log2 fold change.

Paired analysis of serum inflammatory protein levels in patients with CTS from before and after surgery identified IL-4 as being increased presurgery (adjusted P = 0.002, Fig. 1 and Table 5) with no other marker being significantly different. The cytokine IL-9, which showed significantly different expression at the mRNA level, displayed a similar trend at the protein level, although this was not significant after stringent FDR correction (adjusted P = 0.09). IL-6 was not found to be significantly different at the protein level (adjusted P = 0.24).

Table 5 - Serum level changes associated with resolution of carpal tunnel syndrome.
Presurgery Postsurgery
Inflammatory mediator Median IQR Median IQR P Adj.p.Value
IL-4 0.05 0.06 0.00 0.04 0.0001 0.002
Fractalkine 6149.32 2099.72 6246.77 1809.57 0.008 0.078
IL-9 0.23 0.27 0.39 0.20 0.014 0.091
IL-12 0.00 0.30 0.00 0.14 0.034 0.171
IL-6 0.66 0.62 0.70 0.86 0.065 0.242
IL-1β 0.09 0.07 0.06 0.08 0.073 0.242
GM-CSF 0.00 0.04 0.00 0.09 0.129 0.368
IL-10 0.17 0.13 0.17 0.12 0.250 0.626
CCL2 288.87 122.57 299.20 154.29 0.482 0.817
TNF-α 0.50 0.35 0.53 0.33 0.493 0.817
IL-2 0.00 0.00 0.00 0.00 0.500 0.817
IL-17 0.00 0.00 0.00 0.12 0.566 0.817
CRP 1.49 1.54 1.77 1.96 0.581 0.817
VEGF 94.88 81.73 103.40 78.38 0.592 0.817
CXCL5 1533.24 1068.58 1438.87 991.59 0.644 0.817
CCL5 140,804.24 100719.45 134049.52 107239.00 0.656 0.817
CXCL8 9.03 5.13 10.68 4.08 0.694 0.817
TGF-β 55,410.50 20,339.69 54,827.34 20,182.99 0.762 0.846
CXCL10 233.83 158.86 231.69 192.15 0.918 0.952
IFN-γ 6.65 5.67 6.21 4.03 0.952 0.952
Differences in serum levels in patients with CTS before compared with after surgery. Inflammatory mediators are ranked in descending order based on adjusted P values. Significant dysregulation (P < 0.05) is shown in bold. Values are given in pg/mL apart from CRP which is provided in mg/L
Adj.p.Value, adjusted P value; CRP, C-reactive protein; CTS, carpal tunnel syndrome; IQR, interquartile range.

Supplementary Table 4 (available at https://links.lww.com/PAIN/B422) provides a summary of mRNA expression and protein levels in healthy controls and patients with CTS before and after surgery.

3.3. Systemic inflammatory mediators correlate with clinical pain phenotypes

The results of the mRNA clinical phenotype correlation analyses can be found in Figure 2. Correlation analyses revealed a significant correlation of IL-9 mRNA expression with several pain scores. There was a negative correlation of IL-9 mRNA with the Boston symptom score, pain visual analogue scale, and the NPSI composite as well as several subscores (burning pain and paraesthesia). IL-9 mRNA expression also negatively correlated with EDT grade. For IL-6 and PGTES2 mRNA expression, no monotonic relationships were present with clinical phenotypes and so no correlation analyses were conducted for these mediators.

F2
Figure 2.:
Significant correlations of inflammatory gene expression and clinical phenotypes. IL-9 negatively correlated with a range of symptom scores (Boston symptom questionnaire, pain VAS, NPSI total score, and NPSI subscores for burning pain and paraesthesia) as well as electrodiagnostic test severity (EDT grade). Presurgery and postsurgery data were included in the analyses. The Spearman rank correlation was used with a P < 0.05 being considered significant. A smoothed spline has been added to highlight the trend of the data. EDT, electrodiagnostic testing; NPSI, Neuropathic Pain Symptom Inventory, VAS, visual analogue scale.

We next conducted correlation analyses between protein concentrations and patient phenotype scores. No monotonic relationships were found between protein levels and patient phenotype scores for CCL5 or TGF-β and so analyses were not conducted for these mediators. IL-4 correlated positively with EDT grade and several pain scores including the Boston symptom questionnaire and the NPSI total score (Fig. 3). Interestingly, when correlations were performed using IL-9 protein concentrations, it was again found to negatively correlate with the total NPSI score as well as 2 NPSI subscores (paraesthesia and paroxysmal pain) (Supplementary Fig. 1, available at https://links.lww.com/PAIN/B422).

F3
Figure 3.:
Significant correlations of protein levels and clinical phenotypes. IL-4 protein levels positively correlated with the Boston symptom score as well as the NPSI total score and its subdomain paraesthesia. In addition, there was a positive correlation between IL-4 protein levels and electrodiagnostic test severity (EDT grade). Presurgery and postsurgery data were included in the analyses. The Spearman rank correlation was used with a P < 0.05 being considered significant. A smoothed spline has been added to highlight the trend of the data. EDT, electrodiagnostic testing; NPSI, Neuropathic Pain Symptom Inventory, VAS, visual analogue scale.

To determine the effect of zero inflation in these correlations, a hurdle modelling approach consisting of a truncated Poisson model for nonzero scores and a binomial model for zero scores was fit to the data (Supplementary Table 5, available at https://links.lww.com/PAIN/B422). In general, this modelling was in good agreement with the Spearman correlation analysis, especially for IL-9 mRNA expression, where the odds ratio showed that higher IL-9 mRNA expression was associated with patients scoring zero for symptom severity. However, in most of these cases the correlation for the zero-inflated model did not reach significance.

To determine whether changes in mRNA expression or protein level were associated with changes in clinical symptoms, we analysed the Log2FCs of IL-6, IL-9, and PTGES2 mRNA expression and CCL5, TGF-β, IL-4, and IL-9 protein levels against the Log2FCs of clinical symptoms. However, no monotonic relationships were present for any of these mediators with clinical scores, and thus, no correlation analyses were conducted with Log2FC data.

4. Discussion

Using CTS as a model system, we have explored changes in systemic blood inflammatory markers associated with the presence of nerve injury, neuropathic pain, and recovery. In the active stage of CTS, PTGES2 mRNA expression was significantly lower and TGF-β and CCL5 protein levels higher in patients compared with healthy controls; however, a large variation was observed in protein levels within groups. During recovery, 12 genes were significantly differentially expressed, among which IL-9 (increased postsurgery) and IL-6 (decreased postsurgery) showed the most pronounced changes. At the protein level, IL-4 was significantly increased presurgery. Intriguingly, correlation analyses identified IL-9 to be negatively correlated with several pain scores at both the mRNA and protein level, while protein concentrations of IL-4 positively correlated with patients' pain scores. Our findings highlight the potential role of systemic immune dysregulation in focal nerve injury and neuropathic pain.

4.1. Systemic inflammatory changes associated with recovery

Carpal tunnel syndrome provided a unique opportunity to study changes in inflammatory markers associated with resolution of nerve injury and neuropathic pain. The most striking finding of the presurgery to postsurgery comparison was that of IL-9, which showed a significant increase in mRNA expression postsurgery, a similar trend at the protein level and consistent negative associations with a range of pain severity scores. The postoperative phase of CTS is reflective of a state of recovery, where symptoms have largely resolved and the affected nerve is in the process of repair.2 The postoperative increase in IL-9 may therefore highlight a role for this cytokine in symptom resolution and nerve repair or regeneration. IL-9 is a relatively understudied cytokine with limited literature describing a pleiotropic function with both proinflammatory and anti-inflammatory capacity.27,33,46,66 Studies in patients with inflammatory bowel disease report that increased IL-9 serum protein concentration correlated with a less favourable prognosis and increased disease severity.19,71 By contrast, but in line with our findings, other studies point towards a proresolution effect of IL-9. Compelling preclinical data implicate IL-9 in the activation and activity of Treg cells.21,46,56 A proresolution role of IL-9 has also been confirmed in patients with rheumatoid arthritis in clinical remission56 and in patients with lumbar radicular pain, where IL-9 was higher in patients with mild compared with severe disk herniations.31 These recent discoveries, together with our findings, suggest that IL-9 may be an interesting candidate with a proresolution role in the context of neuropathic pain. Further work may seek to determine whether IL-9 has therapeutic utility in neuropathic pain.

In contrast to IL-9, IL-6 gene but not protein expression was higher before compared with after CTS surgery. IL-6 is a well-studied proinflammatory cytokine with a clear link to pain, including neuropathic pain.84 Although the literature on IL-6 levels in serum of patients with CTS is conflicting,26,36,49,69 its contribution to other entrapment neuropathies such as lumbar radicular pain and its severity is well established.37,54,59,61,77 The therapeutic utility of IL-6 in the treatment of neuropathic pain has already started to be explored. For instance, tocilizumab, a humanised antibody which binds the IL-6 receptor, has shown promising short-term effects in preliminary clinical trials in patients with sciatica.51,58

IL-4 protein levels were also increased before compared with after surgery. Preclinical literature suggests that IL-4 is a key mediator in reducing neuropathic pain behaviours in models of peripheral nerve injury.7,75 This contrasts with our findings of IL-4 downregulation in the resolution phase and its positive correlation with several pain scores. Of note, IL-4 was changed at both mRNA and protein levels albeit in opposing directions. This discrepancy in mRNA expression and protein levels is commonly observed18,48 and is thought to be driven in large part by protein half-life.3,81 Cytokines have typically very short half-lives,10,79 which could explain the observed discrepancy. However, the serum levels were low (<0.5 pg/mL) and the clinical relevance of IL-4 remains therefore unclear.

4.2. Systemic inflammatory changes in the active stage of nerve injury

A limited number of systemic inflammatory mediators were altered in patients with CTS compared with healthy controls. This is most likely attributed to the vast variation in cytokine expression in humans,42 which requires either large effects or large samples to detect subtle changes. Indeed, CTS represents a focal and more subtle nerve injury compared with more severe neuropathic conditions such as phantom limb pain or systemic inflammatory neuropathies where proinflammatory cytokines are often dysregulated.13,53 Nevertheless, we observed higher CCL5 and TGF-β at protein levels and lower PTGES2 mRNA levels in patients with CTS. In line with our findings, CCL5 has previously been identified as being increased in the serum of patients with CTS compared with healthy controls.49 We did not detect any significant correlations for CCL5 with any pain phenotypes. CCL5 has however previously been implicated in the generation of neuropathic pain in experiments treating mice with the CCL5 antagonist met-RANTES43 or by using CCL5 knockout mice.44 Interestingly, however, a previous study has found that serum CCL5 levels in CTS negatively correlated with neuropathic pain severity.49 Given our replication of changes to CCL5 in CTS as a model system of neuropathic pain and the growing preclinical literature, the role of CCL5 in the context of nerve injury and neuropathic pain deserves more attention.

TGF-β protein levels were also higher in patients with CTS compared with healthy controls. TGF-β has previously been implicated in the pathogenesis of CTS, but mainly with regards to fibrotic changes.15,23,28,62 The here identified higher TGF-β levels may therefore be acting more as a driver for fibrotic change in CTS, rather than protecting against neuropathic pain. Of note, we have identified TGFB3 as a causal gene in a genome-wide association study of CTS,80 further corroborating the importance of this pathway.

Prostaglandin E2 has a well-established association with acute and chronic neuropathic pain in preclinical models.47 In contrast to our findings of downregulated PTGES2 mRNA expression in patients with CTS compared with controls, serum protein levels have been reported to be unchanged.26 Nevertheless, analysis of tenosynovial tissue indicated that prostaglandin E2 is increased in patients with CTS compared with healthy controls.5,26,72 The apparent discrepancy between our findings and that of the literature may be due to differences in the biological samples used and the method of analysis. We analysed mRNA expression instead of protein levels, which often do not correlate.18,48

4.3. Many differentially expressed cytokines are involved in naive CD4+ T-cell differentiation

An interesting feature of several of the dysregulated inflammatory mediators identified here (TGF-β, IL-4, and IL-6) is that they are known to be involved in naive CD4+ helper T-cell differentiation.4,11,14,24,25,39,68,76,83 They may therefore be working in combination to orchestrate specific subpopulations of CD4+ T cells. Indeed, flow cytometric analysis of peripheral blood in patients with CTS identified increased CD4+ T-cell effector memory and central memory populations in patients with CTS compared with healthy controls.49 These findings further fuel the increasing interest in T cells in the context of neuropathic pain.38

4.4. Limitations

A limitation of this study is that the cellular sources of the cytokines and the target cells on which the cytokines act remain unknown. This knowledge would prove useful in fully defining the role of inflammation in neuropathic pain and its recovery. Another issue with measuring cytokine expression systemically in the blood is that it may not accurately reflect the local environment at the lesion site and may miss markers that do not circulate at high levels in the blood. Nevertheless, our findings demonstrate that peripheral blood analysis can be informative in the context of neuropathic pain if lesioned tissues are not available. The protein analyses in particular showed high variation within groups, which may reflect differences in disease presentation as well as inherent differences in baseline expression of these mediators. The lower number of healthy participants may have contributed in part to this variation. Caution should thus be taken when interpreting the role of these mediators. However, as this study did not intend to identify biomarkers of CTS, the identification of specific biological pathways and mediators associated with the active stage of the disease and its recovery may still prove insightful. Finally, although the list of inflammatory mediators investigated here is more comprehensive than in previous studies, it is not exhaustive and may have missed relevant markers. Our initial data will be an important resource to guide future validation studies.

5. Conclusions

Investigating the systemic expression of inflammatory mediators in patients with CTS revealed a role both during the active stage as well as resolution of nerve injury and neuropathic pain. Intriguingly, IL-9 was upregulated during recovery and consistently negatively correlated with symptom scores, suggesting a proresolution role in the context of nerve injury and neuropathic pain. PTGES2 mRNA as well as TGF-β and CCL5 protein levels were associated with the active stage of nerve injury, and IL-6 and IL-4 mRNA and protein levels were upregulated presurgery compared with postsurgery, respectively. Our findings implicate specific cytokines to play a role in neuropathic pain associated with focal nerve injury and its recovery.

Conflict of interest statement

The authors have no conflicts of interest to declare.

Appendix A. Supplemental digital content

Supplemental digital content associated with this article can be found online at https://links.lww.com/PAIN/B422.

Acknowledgements

The authors thank all participants for taking part in this study. The help with recruitment by the hand surgeons at Oxford University Hospital NHS Trust and the Clinical Research Network Thames Valley is gratefully acknowledged. Thanks also go to Dr Dan Blat for specialist knowledge and guidance.

The Oxford carpal tunnel cohort was supported by an advanced postdoc mobility fellowship from the Swiss National Science Foundation (P00P3-158835 to A.B.S.) and an early career research grant from the International Association for the Study of Pain (to A.B.S.). A.B. Schmid is supported by a Wellcome Trust Clinical Career Development Fellowship (222101/Z/20/Z) and the National Institute for Health Research (NIHR) Oxford Biomedical Research Centre (BRC). The views expressed are those of the authors and not necessarily those of the NHS, the NIHR, or the Department of Health. G. Baskozos is supported by Diabetes United Kingdom (19/0005984). Immunocore Ltd supported a studentship (OSH). D.L. Bennett is a senior Wellcome clinical scientist (202747/Z/16/Z). A. Wiberg is supported by an NIHR Clinical Lectureship. D. Furniss is supported by the National Institute for Health Research (NIHR) Oxford Biomedical Research Centre (BRC). This research was funded in whole, or in part, by the Wellcome Trust [222101/Z/20/Z, 202747/Z/16/Z].

Immunocore Ltd supported a studentship (O.S.-H.) but did not have any input into the design, data collection, analysis, interpretation, and write-up of this project. D.L. Bennett has acted as a consultant on behalf of Oxford Innovation in the last 2 years for Amgen, CODA therapeutics, Bristows, Lilly, Munidpharma, and Theranexus.

Author contributions: A.B. Schmid, O. Sandy-Hindmarch, and D.L. Bennett conceptualized the study. A.B. Schmid collected the data. A. Wiberg and D. Furniss helped with recruitment. O. Sandy-Hindmarch performed laboratory experiments. O. Sandy-Hindmarch and G. Baskozos analysed the data. O. Sandy-Hindmarch and A.B. Schmid wrote the first draft of the manuscript, and all authors provided input to the final manuscript.

References

[1]. Attal N, Lanteri-Minet M, Laurent B, Fermanian J, Bouhassira D. The specific disease burden of neuropathic pain: results of a French. Nationwide Surv Pain 2011;152:2836–43.
[2]. Baskozos G, Sandy-Hindmarch O, Clark AJ, Windsor K, Karlsson P, Weir GA, McDermott LA, Burchall J, Wiberg A, Furniss D, Bennett DLH, Schmid AB. Molecular and cellular correlates of human nerve regeneration: ADCYAP1/PACAP enhance nerve outgrowth. Brain 2020;143:2009–26.
[3]. Belle A, Tanay A, Bitincka L, Shamir R, O'Shea EK. Quantification of protein half-lives in the budding yeast proteome. Proc Natl Acad Sci U S A 2006;103:13004–9.
[4]. Bettelli E, Carrier Y, Gao W, Korn T, Strom TB, Oukka M, Weiner HL, Kuchroo VK. Reciprocal developmental pathways for the generation of pathogenic effector TH17 and regulatory T cells. Nature 2006;441:235–8.
[5]. Bianchi E, Tauronei S, Leopizzi M, Bardella L, Di Liddo R, Nucci F, Pasture FS, Vitale M, Mazzotti A, Artico M. Immunohistochemical profile of VEGF, PGE2 and TGF-β in inflammatory tenosynovitis of carpal tunnel syndrome. Eur J Inflamm 2012;10:491–9.
[6]. Bland JD. A neurophysiological grading scale for carpal tunnel syndrome. Muscle Nerve 2000;23:1280–3.
[7]. Bobinski F, Teixeira JM, Sluka KA, Santos ARS. IL-4 mediates the analgesia produced by low-intensity exercise in mice with neuropathic pain. PAIN 2018;159:437–50.
[8]. Bouhassira D, Attal N, Fermanian J, Alchaar H, Gautron M, Masquelier E, Rostaing S, Lanteri-Minet M, Collin E, Grisart J, Boureau F. Development and validation of the neuropathic pain symptom inventory. PAIN 2004;108:248–57.
[9]. Brisby H, Olmarker K, Larsson K, Nutu M, Rydevik B. Proinflammatory cytokines in cerebrospinal fluid and serum in patients with disc herniation and sciatica. Eur Spine J 2002;11:62–6.
[10]. Castell JV, Geiger T, Gross V, Andus T, Walter E, Hirano T, Kishimoto T, Heinrich PC. Plasma clearance, organ distribution and target cells of interleukin-6/hepatocyte-stimulating factor in the rat. Eur J Biochem 1988;177:357–61.
[11]. Chabaud M, Durand JM, Buchs N, Fossiez F, Page G, Frappart L, Miossec P. Human interleukin-17: a T cell-derived proinflammatory cytokine produced by the rheumatoid synovium. Arthritis Rheum 1999;42:963–70.
[12]. Chakraborty S, Brubaker J, Harro C, Weirzba T, Sack D. Development of a novel multiplex electrochemiluminescent-based immunoassay to aid enterotoxigenic Escherichia coli vaccine development and evaluations. J Immunol Methods 2019;470:6–14.
[13]. Chamessian A, Van de Ven T, Buchheit T, Hsia HL, McDuffie M, Gamazon ER, Walsh C, Bruehl S, Buckenmaier C', Shaw A. Differential expression of systemic inflammatory mediators in amputees with chronic residual limb pain. PAIN 2017;158:68–74.
[14]. Chen W, Jin W, Hardegen N, Lei KJ, Li L, Marinos N, McGrady G, Wahl SM. Conversion of peripheral CD4+CD25− naive T cells to CD4+CD25+ regulatory T cells by TGF-β induction of transcription factor Foxp3. J Exp Med 2003;198:1875–86.
[15]. Chikenji T, Gingery A, Zhao C, Passe SM, Ozasa Y, Larson D, An KN, Amadio PC. Transforming growth factor-β (TGF-β) expression is increased in the subsynovial connective tissues of patients with idiopathic carpal tunnel syndrome. J Orthop Res 2014;32:116–22.
[16]. Chowdhury F, Williams A, Johnson P. Validation and comparison of two multiplex technologies, Luminex® and Mesoscale Discovery, for human cytokine profiling. J Immunol Methods 2009;340:55–64.
[17]. Davies AJ, Rinaldi S, Costigan M, Oh SB. Cytotoxic immunity in peripheral nerve injury and pain. Front Neurosci 2020;14:142.
[18]. de Sousa Abreu R, Penalva LO, Marcotte EM, Vogel C. Global signatures of protein and mRNA expression levels. Mol Biosyst 2009;5:1512–26.
[19]. Defendenti C, Sarzi-Puttini P, Saibeni S, Bollani S, Bruno S, Almasio PL, Declich P, Atzeni F. Significance of serum Il-9 levels in inflammatory bowel disease. Int J Immunopathol Pharmacol 2015;28:569–75.
[20]. Ellis A, Bennett DL. Neuroinflammation and the generation of neuropathic pain. Br J Anaesth 2013;111:26–37.
[21]. Elyaman WB, Bradshaw EM, Uyttenhove C, Dardalhon V, Awasthi A, Imitola J, Bettelli E, Oukka M, van Snick J, Renauld JC, Kuchroo VK, Khoury SJ. IL-9 induces differentiation of TH17 cells and enhances function of FoxP3+ natural regulatory T cells. PNAS 2009;106:12885–90.
[22]. Empl M, Renaud S, Erne B, Fuhr P, Straube A, Schaeren-Wiemers N, Steck AJ. TNF-alpha expression in painful and nonpainful neuropathies. Neurology 2001;56:1371–7.
[23]. Ettema AM, Amadio PC, Zhao C, Wold LE, An KN. A histological and immunohistochemical study of the subsynovial connective tissue in idiopathic carpal tunnel syndrome. J Bone Joint Surg Am 86 2004;7(1458–66).
[24]. Fantini M, Becker C, Monteleone G, Pallone F, Galle PR, Neurath MF. Cutting edge: TGF-β induces a regulatory phenotype in CD4+CD25− T cells through Foxp3 induction and down-regulation of Smad7. J Immunol 2004;172:5149–53.
[25]. Ferraro A, Socci C, Stabilini A, Valle A, Monti P, Piemonti L, Nano R, Olek S, Maffi P, Scavini M, Secchi A, Staudacher C, Bonifacio E, Battaglia M. Expansion of Th17 cells and functional defects in T regulatory cells are key features of the pancreatic lymph nodes in patients with type 1 diabetes. Diabetes 2011;60:2903–13.
[26]. Freeland AE, Tucci MA, Barbieri RA, Angel MF, Nick TG. Biochemical evaluation of serum and flexor tenosynovium in carpal tunnel syndrome. Microsurgery 2002;22:378–85.
[27]. Gerlach K, Hwang Y, Nikolaev A, Atreya R, Dornhoff H, Steiner S, Lehr HA, Wirtz S, Vieth M, Waisman A, Rosenbauer F, McKenzie AN, Weigmann B, Neurath MF. TH9 cells that express the transcription factor PU.1 drive T cell-mediated colitis via IL-9 receptor signaling in intestinal epithelial cells. Nat Immunol 2014;15:676–86.
[28]. Gingery A, Yang TH, Passe SM, An KN, Zhao C, Amadio PC. TGF-β signaling regulates fibrotic expression and activity in carpal tunnel syndrome. J Orthop Res 2014;32:1444–50.
[29]. Hoenig J, Heisey DM. The abuse of power: the pervasive fallacy of power calculations for data analysis. Am Stat 2001;55:1–6.
[30]. Huson L. Performance of some correlation coefficients when applied to zero-clustered data. J Mod Appl Stat Methods 2007;6:530–6.
[31]. Jacobsen H, Khan AN, Levine ME, Filippi CG, Chahine NO. Severity of intervertebral disc herniation regulates cytokine and chemokine levels in patients with chronic radicular back pain. Osteoarthritis Cartilage 2020;28:1341–50.
[32]. Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the minimal clinically important difference. Control Clin Trials 1989;10:407–15.
[33]. Jones CP, Gregory LG, Causton B, Campbell GA, Lloyd CM. Activin A and TGF-β promote T(H)9 cell-mediated pulmonary allergic pathology. J Allergy Clin Immunol 2012;129:1000–e3.
[34]. Jungen MJ, Ter Meulen BC, van Osch T, Weinstein HC, Ostelo RWJG. Inflammatory biomarkers in patients with sciatica: a systematic review. BMC Musculoskelet Disord 2019;20:156.
[35]. Kamper S. Global rating of change scales. Aust J Physiother 2009;55:289.
[36]. Karimi N, AbedianKenari S, Darvari F. Serum levels of inflammatory cytokines in patients with idiopathic carpal tunnel syndrome. Int J Neurosci 2020;131:228–32.
[37]. Kraychete DC, Sakata RK, Issy AM, Bacellar O, Jesus RS, Carvalho EM. Proinflammatory cytokines in patients with neuropathic pain treated with Tramadol. Rev Bras Anestesiol 2009;59:297–303.
[38]. Laumet G, Ma J, Robison AJ, Kumari S, Heijnen CJ, Kavelaars A. Cells as an emerging target for chronic pain therapy. Front Mol Neurosci 2019;12:216.
[39]. Le Gros G, et al. Generation of interleukin 4 (IL-4)-producing cells in vivo and in vitro: IL-2 and IL-4 are required for in vitro generation of IL-4-producing cells. J Exp Med 1990;172:921–9.
[40]. Lenz M, Uçeyler N, Frettlöh J, Höffken O, Krumova EK, Lissek S, Reinersmann A, Sommer C, Stude P, Waaga-Gasser AM, Tegenthoff M, Maier C. Local cytokine changes in complex regional pain syndrome type I (CRPS I) resolve after 6 months. PAIN 2013;154:2142–9.
[41]. Levine DW, Simmons BP, Koris MJ, Daltroy LH, Hohl GG, Fossel AH, Katz JN. A self-administered questionnaire for the assessment of severity of symptoms and functional status in carpal tunnel syndrome. J Bone Joint Surg Am 1993;75:1585–92.
[42]. Li Y, Oosting M, Smeekens SP, Jaeger M, Aguirre-Gamboa R, Le KTT, Deelen P, Ricaño-Ponce I, Schoffelen T, Jansen AFM, Swertz MA, Withoff S, van de Vosse E, van Deuren M, van de Veerdonk F, Zhernakova A, van der Meer JWM, Xavier RJ, Franke L, Joosten LAB, Wijmenga C, Kumar V, Netea MG. Functional genomics approach to understand variation in cytokine production in humans. Cell 2016;167:1099–110.e1014.
[43]. Liou JT, Mao CC, Ching-Wah Sum D, Liu FC, Lai YS, Li JC, Day YJ. Peritoneal administration of met-RANTES attenuates inflammatory and nociceptive responses in a murine neuropathic pain model. J Pain 2013;14:24–35.
[44]. Liou J, Yuan HB, Mao CC, Lai YS, Day YJ. Absence of C–C motif chemokine ligand 5 in mice leads to decreased local macrophage recruitment and behavioral hypersensitivity in a murine neuropathic pain model. PAIN 2012;153:1283–91.
[45]. Liu J, Kabir F, Manneh J, Lertsethtakarn P, Begum S, Gratz J, Becker SM, Operario DJ, Taniuchi M, Janaki L, Platts-Mills JA, Haverstick DM, Kabir M, Sobuz SU, Nakjarung K, Sakpaisal P, Silapong S, Bodhidatta L, Qureshi S, Kalam A, Saidi Q, Swai N, Mujaga B, Maro A, Kwambana B, Dione M, Antonio M, Kibiki G, Mason CJ, Haque R, Iqbal N, Zaidi AKM, Houpt ER. Development and assessment of molecular diagnostic tests for 15 enteropathogens causing childhood diarrhoea: a multicentre study. Lancet Infect Dis 2014;14:716–24.
[46]. Lu LF, Lind EF, Gondek DC, Bennett KA, Gleeson MW, Pino-Lagos K, Scott ZA, Coyle AJ, Reed JL, Van Snick J, Strom TB, Zheng XX, Noelle RJ. Mast cells are essential intermediaries in regulatory T-cell tolerance. Nature 2006;442:997–1002.
[47]. Ma W, Chabot JG, Vercauteren F, Quirion R. Injured nerve-derived COX2/PGE2 contributes to the maintenance of neuropathic pain in aged rats. Neurobiol Aging 2010;31:1227–37.
[48]. Maier T, Güell M, Serrano L. Correlation of mRNA and protein in complex biological samples. FEBS Lett 2009;583:3966–73.
[49]. Moalem-Taylor G, Baharuddin B, Bennett B, Krishnan AV, Huynh W, Kiernan MC, Shin-Yi Lin C, Shulruf B, Keoshkerian E, Cameron B, Lloyd A. Immune dysregulation in patients with carpal tunnel syndrome. Nat Scientific Rep 2017;7:8218.
[50]. Moalem G, Tracey DJ. Immune and inflammatory mechanisms in neuropathic pain. Brain Res Rev 2006;51:240–64.
[51]. Ohtori S, Miyagi M, Eguchi Y, Inoue G, Orita S, Ochiai N, Kishida S, Kuniyoshi K, Nakamura J, Aoki Y, Ishikawa T, Arai G, Kamoda H, Suzuki M, Takaso M, Furuya T, Kubota G, Sakuma Y, Oikawa Y, Toyone T, Takahashi K. Efficacy of epidural administration of anti-interleukin-6 receptor antibody onto spinal nerve for treatment of sciatica. Eur Spine J 2012;21:2079–84.
[52]. Pachot A, Blond JL, Mougin B, Miossec P. Peptidylpropyl isomerase B (PPIB): a suitable reference gene for mRNA quantification in peripheral whole blood. J Biotechnol 2004;114:121–4.
[53]. Parkitny L, McAuley JH, Di Pietro F, Stanton TR, O'Connell NE, Marinus J, van Hilten JJ, Moseley GL. Inflammation in complex regional pain syndrome: a systematic review and meta-analysis. Neurology 2013;80:106–17.
[54]. Pedersen LM, Schistad E, Jacobsen LM, Røe C, Gjerstad J. Serum levels of the pro-inflammatory interleukins 6 (IL-6) and -8 (IL-8) in patients with lumbar radicular pain due to disc herniation: a 12-month prospective study. Brain Behav Immun 2015;46:132–6.
[55]. Rathod TN, Chandanwale A, Ladkat KM, Chavan S, Chavan A, Bhosale PB. High sensitive C-reactive protein-Effective tool in determining postoperative recovery in lumbar disc disease. Indian J Orthop 2014;48:354–9.
[56]. Rauber S, Luber M, Weber S, Maul L, Soare A, Wohlfahrt T, Lin NY, Dietel K, Bozec A, Herrmann M, Kaplan MH, Weigmann B, Zaiss MM, Fearon U, Veale DJ, Cañete JD, Distler O, Rivellese F, Pitzalis C, Neurath MF, McKenzie ANJ, Wirtz S, Schett G, Distler JHW, Ramming A. Resolution of inflammation by interleukin-9-producing type 2 innate lymphoid cells. Nat Med 2017;23:938–44.
[57]. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, Smyth GK. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 2015;43:e47.
[58]. Sainoh T, Orita S, Miyagi M, Inoue G, Yamauchi K, Suzuki M, Sakuma Y, Kubota G, Oikawa Y, Inage K, Sato J, Nakata Y, Aoki Y, Takahashi K, Ohtori S. Single intradiscal injection of the interleukin-6 receptor antibody tocilizumab provides short-term relief of discogenic low back pain; prospective comparative cohort study. J Orthopaedic Sci 2016;21:2–6.
[59]. Schistad EI, Espeland A, Pedersen LM, Sandvik L, Gjerstad J, Røe C. Association between baseline IL-6 and 1-year recovery in lumbar radicular pain. Eur J Pain 2014;18:1394–401.
[60]. Schmid AB, Bland JD, Bhat MA, Bennett DL. The relationship of nerve fibre pathology to sensory function in entrapment neuropathy. Brain 2014;137:3186–99.
[61]. Shamji M, Guha D, Paul D, Shcharinsky A. Systemic inflammatory and Th17 immune activation among patients treated for lumbar radiculopathy exceeds that of patients treated for persistent postoperative. Neuropathic Pain Neurosurg 2017;81:537–44.
[62]. Sharma D, Jaggi AS, Bali A. Clinical evidence and mechanisms of growth factors in idiopathic and diabetes-induced carpal tunnel syndrome. Eur J Pharmacol 2018;837:156–63.
[63]. Sommer C, Koch S, Lammens M, Gabreels-Festen A, Stoll G, Toyka KV. Macrophage clustering as a diagnostic marker in sural nerve biopsies of patients with CIDP. Neurology 2005;65:1924–9.
[64]. Sommer C, Leinders M, Üçeyler N. Inflammation in the pathophysiology of neuropathic pain. PAIN 2018;159:595–602.
[65]. Stamova BS, Apperson M, Walker WL, Tian Y, Xu H, Adamczy P, Zhan X, Liu DZ, Ander BP, Liao IH, Gregg JP, Turner RJ, Jickling G, Lit L, Sharp FR. Identification and validation of suitable endogenous reference genes for gene expression studies in human peripheral blood. BMC Med Genomics 2009;2:49.
[66]. Stanko K, Iwert C, Appelt C, Vogt K, Schumann J, Strunk FJ, Ahrlich S, Schlickeiser S, Romagnani C, Jürchott K, Meisel C, Willimsky G, Kühl AA, Sawitzki B. CD96 expression determines the inflammatory potential of IL-9-producing Th9 cells. Proc Natl Acad Sci U S A 2018;115:E2940–9.
[67]. Stürmer T, Raum E, Buchner M, Gebhardt K, Schiltenwolf M, Richter W, Brenner H. Pain and high sensitivity C reactive protein in patients with chronic low back pain and acute sciatic pain. Ann Rheum Dis 2005;64:921–5.
[68]. Swain SL, Weinberg AD, English M, Huston G. IL-4 directs the development of th2-like helper effectors. J Immunol 1990;145:3796–806.
[69]. Takasu S, Takatsu S, Kunitomo K, Kokumai Y. Serum hyaluronic acid and interleukin-6 as possible markers of carpal tunnel syndrome in chronic hemodialysis patients. Artif Organs 1994;18:420–4.
[70]. Team RC. R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing, 2014.
[71]. Tian L, Li Y, Zhang J, Chang R, Li J, Huo L. IL-9 promotes the pathogenesis of ulcerative colitis through STAT3/SOCS3 signaling. Biosci Rep 2018;38.
[72]. Tucci MA, Barbieri RA, Freeland AE. Biochemical and histological analysis of the flexor tenosynovium in patients with carpal tunnel syndrome. Biomed Sci Instrum 1997;33:246–51.
[73]. Üçeyler N, Braunsdorf S, Kunze E, Riediger N, Scheytt S, Divisova Š, Bekircan-Kurt CE, Toyka KV, Sommer C. Cellular infiltrates in skin and sural nerve of patients with polyneuropathies. Muscle Nerve 2016;55:884–93.
[74]. Üçeyler N, Riediger N, Kafke W, Sommer C. Differential gene expression of cytokines and neurotrophic factors in nerve and skin of patients with peripheral neuropathies. J Neurol 2015;262:203–12.
[75]. Üçeyler N, Topuzoğlu T, Schiesser P, Hahnenkamp S, Sommer C. IL-4 deficiency is associated with mechanical hypersensitivity in mice. PLoS One 2011;6:e28205.
[76]. Veldhoen M, Hocking RJ, Atkins CJ, Locksley RM, Stockinger B. TGFbeta in the context of an inflammatory cytokine milieu supports de novo differentiation of IL-17-producing T cells. Immunity 2006;24:179–89.
[77]. Wang K, Bao JP, Yang S, Hong X, Liu L, Xie XH, Wu XT. A cohort study comparing the serum levels of pro- or anti-inflammatory cytokines in patients with lumbar radicular pain and healthy subjects. Eur Spine J 2016;25:1428–34.
[78]. Wesseldijk F, Huygen FJ, Heijmans-Antonissen C, Niehof SP, Zijlstra FJ. Six years follow-up of the levels of TNF-alpha and IL-6 in patients with complex regional pain syndrome type 1. Mediators Inflamm 2008;2008:469439.
[79]. Whiteside TL. Cytokines and cytokine measurements in a clinical laboratory. Clin Diagn Lab Immunol 1994;1:257–60.
[80]. Wiberg A, Ng M, Schmid AB, Smillie RW, Baskozos G, Holmes MV, Künnapuu K, Mägi R, Bennett DL, Furniss D. A genome-wide association analysis identifies 16 novel susceptibility loci for carpal tunnel syndrome. Nat Commun 2019;10:1030.
[81]. Wu G, Nie L, Zhang W. Integrative analyses of posttranscriptional regulation in the yeast Saccharomyces cerevisiae using transcriptomic and proteomic data. Curr Microbiol 2008;57:18–22.
[82]. Yamamoto M, Ito Y, Mitsuma N, Li M, Hattori N, Sobue G. Parallel expression of neurotrophic factors and their receptors in chronic inflammatory demyelinating polyneuropathy. Muscle Nerve 2002;25:601–4.
[83]. Zheng S, Gray JD, Ohtsuka K, Yamagiwa S, Horwitz DA. Generation ex vivo of TGF-β-producing regulatory T cells from CD4+CD25− precursors. J Immunol 2002;169:4183–9.
[84]. Zhou YQ, Liu Z, Liu ZH, Chen SP, Li M, Shahveranov A, Ye DW, Tian YK. Interleukin-6: an emerging regulator of pathological pain. J Neuroinflammation 2016;13:141.
[85]. Zu B, Pan H, Zhang XJ, Yin ZS. Serum levels of the inflammatory cytokines in patients with lumbar radicular pain due to disc herniation. Asian Spine J 2016;10:843–9.
Keywords:

Pain; Neuropathic pain; Carpal tunnel syndrome; Entrapment neuropathy; Neuroinflammation; Inflammation; Human; Cytokine; IL-9; IL-6; CCL5; TGF-beta; PTGES2

Supplemental Digital Content

© 2021 International Association for the Study of Pain