Up-regulation of Cathepsin G in the Development of Chronic Postsurgical Pain: An Experimental and Clinical Genetic Study

CHRONIC postsurgical pain is a well-recognized complication affecting at least 10% of patients undergoing common surgical operations, such as hernia repair, mastectomy, and joint replacement surgery.^1–6  Among those who experienced chronic postsurgical pain, a substantial proportion of patients reported severe pain that adversely affected their quality of life.²  Despite the magnitude of this problem, there is no consensus for managing chronic postsurgical pain.

Central sensitization, a phenomenon when nociceptive neurons in the spinal dorsal horns become sensitized by peripheral tissue inflammation, is important in the development of chronic pain. Further study on the molecular mechanisms of central sensitization is therefore highly warranted to identify the potential targets for the development of new biomarkers and analgesics that may be useful to predict and to treat chronic postsurgical pain, respectively.⁷  Maintenance of inflammatory milieu within the dorsal horns contributes to central sensitization. Emerging evidence suggests that inflammatory mediators act as modulators of pain signal transmission in the spinal dorsal horns.⁸  The proinflammatory cytokines, such as interleukin (IL) 1β and tumor necrosis factor-α, may sensitize the ascending nociceptive pathway with or without glial cell activation.⁸  For instance, the chemokine (C-X3-C motif) ligand 1 was up-regulated in spinal astrocytes after peripheral nerve injury,⁹  leading to the release of IL-1β from activated microglia and thus sensitization of nociceptive neurons.¹⁰  These processes are mediated through the signal transducer and activator of transcription 3 (Stat3)–dependent pathway. In this respect, selective Stat3 inhibition reduces the chemokine production and the associated mechanical allodynia.¹¹ 

Proteases modulate peripheral and central inflammation by regulating chemotaxis of immune cells and production of cytokines and chemokines.^12,13  Recently, several proteases have been identified as pain modulators in the spinal cord. Inhibition of spinal matrix metalloproteinase 2 and 9 reduced neuropathic pain after peripheral nerve injury.¹⁴  Similarly, reversal of neuropathic pain in rats could be achieved by selective inhibition of spinal microglial cathepsin S.¹⁵  These results suggest that proteases function as modulators of pain-related cytokines and chemokines. Targeting proteases may therefore represent a new approach to regulate nociceptive transmission in the spinal cord.

In this study, using a genome-wide screening approach, we identified cathepsin G (CTSG) as an up-regulated protease in the dorsal horns. We hypothesized that an increase in CTSG activity contributed to pain hypersensitivity via central sensitization in a rat model of chronic inflammatory pain. To establish the clinical relevance of these findings, we evaluated the novel association between CTSG gene polymorphisms and the risk of chronic postsurgical pain in a large cohort of surgical patients.

The experimental protocols and surgical procedures were approved by the Animal Experimentation Ethics Committee, the Chinese University of Hong Kong, Hong Kong Special Administrative Region, People’s Republic of China. Male Sprague–Dawley rats, weighting 200 to 250 g, were used in all experiments.

We used CTSG inhibitors—tosyl phenylalanyl chloromethyl ketone (TPCK; Sigma-Aldrich, USA)—and CTSG-specific inhibitor (CatGI; Merck KGaA, Germany) to block the activity of CTSG in our experiments.^16,17  Both reagents were dissolved in dimethyl sulfoxide (Sigma-Aldrich) at concentrations of 50 mM and were stored at −20°C before use. Purified human neutrophil CTSG protein was obtained from EMD4Biosciences (Merck KGaA). CTSG protein was dissolved in a solution (2 mU/μl) containing 100 mM sodium chloride, 50 mM sodium acetate (pH 5.5), and diluted with normal saline for administration in animal (8 mU per injection) and cell experiments (2.5 mU/ml).

Peripheral inflammation was induced by intraplantar injection of complete Freund’s adjuvant (CFA; Sigma-Aldrich) into the left hind paw during isoflurane anesthesia. Rats in the control group were injected with 100 μl of normal saline under the same experimental conditions.

Intrathecal catheterization was performed as previously described.¹¹  In brief, during isoflurane anesthesia, an 18-gauge guiding needle was inserted into the intrathecal space between lumbar vertebrae 5 and 6. A PE10-polyethylene catheter was then implanted through the guiding needle until it reached the lumbar enlargement of the spinal cord. The rats were housed individually after surgery and allowed to recover for 7 days before further treatment. Drugs (CatGI, CTSG protein, or saline, 10 μl per injection) were injected through the indwelling catheter over 5 min.

Thermal hyperalgesia was determined by using the plantar analgesia meter (Model 390G; IITC Life Science, USA). Each rat was placed in a Plexiglas chamber on a glass plate located above a light box. A beam of nonnoxious light was aimed onto the mid-plantar surface of left hind paw. Innocuous heat, generated by increasing light power was then applied to the hind paw. The light beam was immediately switched off when hind paw withdrawal was observed. The duration between the start of the noxious heating and paw withdrawal was defined as paw withdrawal latency (PWL). We set a maximum time limit of 20 s to prevent inadvertent tissue damage. PWL for each animal at any time point was averaged from at least three tests separated by an interval of 5 min. Investigators who determined pain behavioral response were blinded to group assignment.

Total RNA was extracted with RNeasy lipid tissue mini kit (QIAGEN, Germany). Ten microgram of RNA extracted from each tissue was used as template for the synthesis of complementary DNA (cDNA). In vitro transcription (cRNA synthesis and labeling) was performed using low RNA input fluorescent linear amplification kit (Agilent Technologies, USA) in the presence of cyanine 5-cytosine triphosphate. Synthesized fluorescence-labeled cRNA was then used for oligomicroarray hybridization. The hybridization solution was prepared according to in situ hybridization kit plus (Agilent Technologies). Hybridization was performed using Agilent 4 × 44 k whole rat genome microarray (Agilent Technologies) in a hybridization oven at 60°C for 18 h. The microarray scanner system (Agilent Technologies) was used for data analysis. The expression level of specific genes was determined by signal intensities. Fold change of specific gene was made between the average expression levels of two rats from the CFA group and two rats in the control group. A priori cutoff was set to filter out the differential genes according to the average expression levels of each data set. Accordingly, the differential gene list was selected based on the fold change (≥ 1.5) and P value (< 0.05) of unadjusted Student t test. Hierarchical cluster analyses for the whole data sets and the differential gene list were performed with GeneSpring software (Agilent Technologies). Cluster on both entities and condition and Pearson centered distance metric was considered for all the data sets during clustering.

Primary astrocyte culture was performed as previously described.¹¹  In brief, the cortex of neonatal (1 to 2 days old) rats was collected and used for single-cell isolation. Cells were seeded onto 75 cm² flasks and grew to confluency. To remove the contamination of other glial cells, flasks were shaken (speed at 225 rpm) at 37°C for 8 h. Enriched astrocyte cultures were grown in Dulbecco’s modified Eagle’s medium containing 10% heat-inactivated fetal bovine serum (Life Technologies, USA) and 1% antimycotic at 37°C.

Fresh lumbar spinal cords were harvested and fixed with 4% paraformaldehyde overnight at 4°C. Spinal cords were dehydrated in 30% sucrose solutions for 2 days (or until the tissues sink). The spinal cords were then embedded in optimal cutting temperature compound (Sakura Finetek, USA), and 20-μm sections were cut on a cryostat (Leica, Germany). A free-floating section staining method was adopted. All the sections were blocked for unspecific antibody binding sites using 5% normal calf serum in 0.01 M phosphate-buffered saline with 0.3% Triton x-100 (Sigma-Aldrich). After blocking for 1 h at room temperature, sections were incubated with the primary antibody solution overnight at 4°C, followed by incubation with the secondary antibody at room temperature for 1 h. Then, sections were transferred to the glass slides and mounted in ProLong Gold Antifade Reagent (Life Technologies). All the antibodies including antiglial fibrillary acidic protein (anti-GFAP, 1 μg/ml; Santa Cruz, USA), antimyeloperoxidase (1:300 dilution; Santa Cruz), anti-S100A9 (1:300; Santa Cruz), anti-IL-1β (1:300; R&D System, USA), and the fluorochrome-conjugated secondary antibody (1:1,000; Molecular Probes, USA) were prepared in 1% bovine serum albumin—phosphate-buffered saline with Triton x-100. All images were collected on a Zeiss laser scanning microscope (Carl Zeiss, Germany) using 20X or 40X objective with the same microscope acquisition parameters.

Total RNA was extracted using trizol/phenol/chloroform following a standard protocol. cDNA was prepared from 2 μg of total RNA using the high-capacity cDNA reverse transcription kit (Life Technologies). Gene levels were determined by SYBR green master mixture on an HT7900 system (Applied Biosystems, USA), with β-actin as the internal control.

Tissues or cells were lysed in lysis buffer (50 mM Tris, 150 nM NaCl, 1% NP-40, 1 mM EDTA, pH 7.6) containing 10% protease inhibitor cocktail. After blocking with 5% skim milk for 1 h, membranes were incubated with primary antibodies against phospho-inhibitor of κB (IκB) (1:1,000; Santa Cruz), phospho-P38 mitogen–activated protein kinases (p38MAPK) (1:1,000; Cell Signaling Technology, USA), p38MAPK (1:1,000; Cell Signaling Technology), phospho-Stat3 (1:1,000; Cell Signaling Technology), Stat3 (1:2,000; Santa Cruz), IL-1β (1:300; R&D System), or β-actin (1:3,000; Santa Cruz) overnight at 4°C. The membranes were incubated with horseradish peroxidase– conjugated anti-mouse or rabbit secondary antibody (1:3,000; Cell signaling Technology) for 1 h at room temperature. The protein bands were visualized using the enhanced chemiluminescence detection system (GE Healthcare Biosciences, USA).

Paw withdrawal latency from behavioral assessment were analyzed among groups using ANOVA with repeated measures. Intergroup difference was tested using unpaired t test. Data from the reporter assay and quantitative reverse transcription polymerase chain reaction were compared between groups using unpaired t test. Multiple comparisons were adjusted with Bonferroni correction. All reported P values were two sided, and a P value less than 0.05 was considered statistically significant.

We collected blood samples before surgery from patients who were enrolled in an observational cohort study of chronic postsurgical pain. The details of the study objectives, design, and methods are reported at the Chinese clinical trial registry (ChiCTR-ONC-10001099). We recruited adult patients, aged 18 yr old or older who were scheduled for a wide variety of surgical procedures that include a skin incision. Patients were not eligible if they only required a pure endoscopic or radiologic procedure. Patients were also excluded if they were previously enrolled in the study or who were not expected to be available for, or to cooperate with, postoperative interviews. Patients were contacted at 12 months after surgery and were asked to rate their experience of pain over the surgical site using the brief pain inventory.¹⁸  In addition, patients also reported their current health status based on the EQ-5D scale.¹⁹  The primary outcome was pain over the surgical site that has persisted for 12 months after the index surgery.²⁰  We genotyped single-nucleotide polymorphisms (SNPs) with predicted minor allele frequency greater than 5% in Chinese Han population. Genotype frequencies were analyzed using the Hardy–Weinberg equilibrium analysis. We performed a multivariable logistic regression analysis to determine the effect of CTSG gene polymorphism on chronic postsurgical pain. The prespecified independent variables included patient’s age, sex, smoking habits, history of education and employment, and pain syndrome before the index surgery. In addition, we tested the effects of anesthetic techniques (regional block, combined general and regional anesthesia vs. general anesthesia alone), severity of acute postoperative pain, and opioid consumption during early postoperative period on the development of chronic postsurgical pain. Given that the reported incidence of chronic postsurgical pain was at least 10%,^1–6  we determined that a cohort of more than 1,100 patients would ensure a stable regression model for evaluation of 10 variables.^21,22  Similarly, the same sample size would provide 81% power to identify SNPs with minor allele frequency 20% or greater for an odds ratio 1.4 or greater. The collection of genomic DNA and genetic analysis were approved by the Clinical Research Ethics Committee. All patients gave written informed consent.

We evaluated the gene expression in a chronic inflammatory pain model by intraplantar injection of CFA (100 μl per rat). In this model, PWL was significantly reduced after CFA injection (P = 0.0013, ANOVA), confirming the development of chronic pain (fig. 1A). Ipsilateral lumbar dorsal horn at L4 to L6 was then harvested for microarray analysis. To preserve the migrated blood-borne cells, perfusion was avoided before tissue harvest.²³  Total tissue RNA was then extracted, and the expression level of specific genes was determined according to the average signal intensities (fig. 1B). A total of 169 annotated genes exhibited 1.5 or greater fold change between groups. Hierarchical cluster analysis demonstrated distinct gene expression patterns in dorsal horn tissues harvested from CFA- and saline-injected rats (fig. 1C). Among these, 87 genes were up-regulated and 82 genes were down-regulated after CFA injection. Ontology analysis revealed that 9 of the 169 genes (5.3%) were proteases or protease inhibitors (Supplemental Digital Content 1, http://links.lww.com/ALN/B188). Although none of the candidate genes had reached statistical significance in a Bonferroni-adjusted analysis, CTSG expression exhibited the highest fold change (6.17 fold), and further experiments were conducted to determine its role in spinal cord pain modulation.

As an independent platform to validate the findings from our microarray analysis, the mRNA expression of four selected genes, including CTSG, was measured by reverse transcription polymerase chain reaction (fig. 1D). CTSG gene was up-regulated (4.7 fold, P < 0.001, Bonferroni-adjusted t test) in the dorsal horns of CFA-injected rats (n = 4) compared with controls (n = 4). Furthermore, experiment was performed to explore the changes of CTSG gene expression over time (n = 4 in each time point). Figure 1E shows that CTSG expression was only increased 4 days after CFA injection and persisted on day 7. This result suggested that up-regulation of CTSG in the spinal cord contributed to the maintenance of chronic inflammatory pain. Interestingly, CTSG expression was not altered in a rat neuropathic pain model with chronic constriction injury to the sciatic nerve (Supplemental Digital Content 2, http://links.lww.com/ALN/B189). Finally, up-regulation of genes that are known to be associated with chronic pain, namely prodynorphin (4.0 fold, P = 0.027), corticotropin-releasing hormone (2.4 fold, P = 0.009), and allograft inflammatory factor 1-like (1.8 fold, P = 0.021), was confirmed in our CFA model.

We then evaluated the effect of CTSG inhibition on the development of chronic inflammatory pain. Rats were randomly assigned to receive subcutaneous injection of either vehicle or CTSG inhibitor TPCK^16,17  1 h before CFA injection (n = 12 in each group). Heat hyperalgesia was detected within 30 min after intraplantar CFA injection and persisted for the subsequent 5 days. Pretreatment with a single subcutaneous injection of TPCK reduced hyperalgesia compared with vehicle (fig. 2A). The data suggested that CTSG contributed to pain hypersensitivity in CFA-induced chronic inflammation.

Similarly, we tested whether CTSG inhibition might reverse chronic inflammatory pain in a posttreatment model. Rats (n = 14 in each group) were randomly assigned to receive vehicle or TPCK 4 days after intraplantar CFA injection. Treatment was given twice daily with a subcutaneous injection 1 h before behavioral testing and 12 h after the first injection. Figure 2B shows that PWL was similar between groups before treatment on day 4. After TPCK injections, heat hyperalgesia was significantly reduced on subsequent days. The data confirmed the therapeutic effect of CTSG inhibition on chronic inflammatory pain.

Because CTSG has been implicated in neutrophil chemotaxis,²⁴  we hypothesized that CTSG contributes to the development of pain hypersensitivity by promoting neutrophil-mediated inflammation in the spinal cord. In this experiment, we evaluated the effects of CTSG inhibition on heat hyperalgesia (fig. 3A) and neutrophil trafficking (fig. 3, B and C) in the spinal dorsal horns. Using an implanted intrathecal microcatheter, CatGI (a specific CTSG inhibitor) or saline (vehicle) was injected twice daily to rats with CFA-induced chronic inflammatory pain (n = 10 in each group). In addition, rats with intraplantar injection of saline was used as negative controls (n = 10). CatGI instead of TPCK was used for intrathecal injection because of its higher water solubility.

Compared with negative controls, the number of neutrophils in the dorsal horns, as revealed by myeloperoxidase immunofluorescence staining, was significantly increased after intraplantar CFA injection (fig. 3, B and C). The changes corresponded to the induction of heat hyperalgesia (fig. 3A). Under fluorescence microscopy, neutrophils were distributed over both the superficial and deep layers of the spinal dorsal horn (fig. 3B). In addition, we found a substantial proportion of neutrophils expressed calgranulin B (S100A9; fig. 4A), indicating that these neutrophils were involved in the production of proinflammatory mediators.

In contrast, heat hyperalgesia was significantly decreased with CTSG inhibition (fig. 3A). Similarly, the number of neutrophils was reduced in the spinal dorsal horns after intrathecal injection of CatGI (fig. 3, B and C; n = 10). Using Western blot analysis, we demonstrated that total IL-1β level in the spinal dorsal horns was reduced (fig. 4, B–D; n = 10). These data suggested that CTSG is important in the modulation of neutrophilic infiltration to the spinal dorsal horns after peripheral inflammation. These neutrophils act as a source of inflammatory molecules such as IL-1β in the spinal cord that is essential to the maintenance of chronic inflammatory pain (fig. 4B and Supplemental Digital Content 3, http://links.lww.com/ALN/B190).

Previous studies have demonstrated the role of astrocytes in maintaining an inflammatory milieu during the development of chronic pain.^25,26  In this experiment, we tested whether CTSG could activate spinal astrocytes to release proinflammatory mediators. Using cultured spinal astrocytes, treatment with CTSG protein increased phosphorylation of IκB, p38MAPK, and Stat3 in a time-dependent manner (fig. 5A). All of these are important inflammatory mediators necessary for the induction and maintenance of chronic pain.⁸ 

We next tested whether CTSG treatment could activate astrocytes in vivo. After implantation of an intrathecal microcatheter, rats were randomly allocated to receive purified CTSG protein or an equivalent volume of saline (n = 10 per group). Four days later, rats were sacrificed and spinal sections were harvested for GFAP immunofluorescence staining. Concordant with our in vitro data, intrathecal injection of CTSG protein aggravated the increases in the number and cell size of spinal astrocytes in rats previously injected with CFA (fig. 5, B and C). However, it is noteworthy that intrathecal administration of CTSG protein per se had no effect on GFAP density in the spinal dorsal horn in the absence of CFA injection (fig. 5, B and C). In CFA-injected rats, intrathecal administration of CTSG protein was also associated with an increase in IL-1β mRNA expression (data not shown). The protein expression of IL-1β in the spinal dorsal horns was also increased and exhibited colocalization with GFAP (fig. 5, D–G). Overall, the data suggested that CTSG aggravated the activation of spinal astrocytes by peripheral inflammation.

Given that CTSG acted as a pronociceptive mediator in our animal experiments, we conducted a candidate gene association study to determine whether variants in the CTSG gene were associated with an altered risk of chronic pain after surgery. Between January 2011 and November 2012, we recruited a total of 1,313 patients to an observational cohort study on the characteristics and epidemiology of chronic postsurgical pain. Among these patients, 1,152 were also enrolled to a genetic substudy (fig. 6). Patient characteristics, surgical details, severity of acute pain, and opioid consumption during the 3 days after surgery are summarized in table 1. A total of 246 patients (21.4%) reported persistent wound pain since the index surgery. At 12-month follow-up, the average (± SD) pain score was 3.2 ± 2.0 (0 = pain free, 10 = worst imaginable pain). Chronic postsurgical pain produced significant interference to all attributes of daily activities (table 2), and this was associated with poorer quality of life (table 3). All chronic pain patients were taking gabanoids, tricyclic antidepressants, nonsteroidal antiinflammatory drugs, and tramadol regularly. Eighty-five patients (34.6%) were undergoing cognitive behavioral therapy training.

Genotyping was performed for all SNPs in the CTSG gene with reported minor allele frequency greater than 5% in the Chinese Han population. Only two SNPs (rs2070697 and rs2236742) fulfilled the inclusion criteria. Both SNPs were significantly associated with chronic postsurgical pain. Patients who were homozygous carriers of the A allele for either of the SNPs had a lower rate of chronic postsurgical pain compared with wild-type (table 4). In a multivariable regression model, the protective effect of both SNPs remained significant when adjusted for age, sex, smoking habit, employment history, education level, prior pain syndrome, and severity of acute pain after index surgery. The area under the receiver operating characteristic curve for our multivariable regression model was 0.71 (95% CI, 0.67 to 0.85).

In this study, we have identified CTSG as a pronociceptive mediator in the development of pain hypersensitivity in an animal model of chronic inflammatory pain. Upon peripheral inflammation, CTSG expression in the spinal dorsal horn was significantly up-regulated. Inhibition of CTSG attenuated chronic inflammation-associated hyperalgesia, and this was accompanied with a decrease in neutrophilic infiltration and a lower level of IL-1β in the spinal dorsal horn. We further demonstrated the relevance of CTSG gene polymorphisms in the development of chronic postsurgical pain. The risk of chronic postsurgical pain was significantly reduced in patients carrying homozygous Aalleles of SNP rs2236742 or rs2070697 in the CTSG gene. Taken together, our data suggest that CTSG modulates pain hypersensitivity and is a potential target for pain management.

We found an increase in spinal cord neutrophils after induction of peripheral inflammation. This is consistent with the large body of evidence showing infiltration of mast cells, lymphocytes, and neutrophils in the dorsal root ganglion and spinal dorsal horn during the development of chronic pain.^8,23,27,28  In addition, we observed a subpopulation of neutrophils expressing high levels of S100A9 and IL-1β, indicating that neutrophils constitute an important source of inflammatory mediators in the initiation and maintenance of central sensitization.²³  Our study suggests that inhibition of CTSG could reduce the pain signaling, manifested as a decrease in heat hyperalgesia, by lowering neutrophil trafficking and IL-1β levels in the spinal cord. In the current study, we did not evaluate the effects of CTSG on mechanical stimulation. Nevertheless, CTSG inhibition would be expected to decrease allodynia because of its suppressive effect on IL-1β production in the spinal cord, which in turn, has been reported as an important modulator of mechanical allodynia.¹⁴  Previous studies have demonstrated the chemotactic properties of CTSG.^29–31  Interestingly, CTSG released from the spinal cord neutrophils has been shown to modulate the secretion of other chemokine, such as chemokine (C-X3-C motif) ligand 2, that are important for further neutrophil migration.³²  In this regard, it is highly likely that CTSG serves as a proinflammatory mediator for signal amplification during the development of central sensitization in the spinal cord. It should be noted that CTSG expression was not changed in dorsal root ganglion after intraplantar CFA injection. Furthermore, although local subcutaneous administration of CatGI reduced paw edema after CFA injection, intrathecal injection of CatGI did not produce the same effect (Supplemental Digital Content 4, http://links.lww.com/ALN/B191), despite significant attenuation of heat hyperalgesia (fig. 3A). Taken together, these findings suggest that CTSG act as a mediator of central sensitization in chronic inflammatory pain.

Reactivation of spinal astrocytes is one of the hallmarks in the development of chronic pain.^8,25,26  Our in vitro study in astrocytes showed that treatment with CTSG increased IκB phosphorylation—a marker of astrocyte reactivation. In these astrocytes, CTSG also increased the phosphorylation levels of Stat3 and p38MAPK in a time-dependent manner. It is commonly observed that activation of transmembrane receptor signaling results in Stat3 and p38MAPK phosphorylation.^11,33  Because previous studies have shown that CTSG is capable of transactivating transmembrane receptors including formyl peptide receptor and epidermal growth factor receptor,^34,35  we believe that CTSG may activate spinal astrocytes by a similar mechanism to initiate the Stat3 and p38MAPK signaling pathway. In this regard, administration of purified CTSG aggravated astrocyte activation induced by peripheral inflammation. Because CTSG administration in the absence of CFA injection had negligible effect on astrocyte activation, it is highly possible that other cofactors are involved in this process.

Chronic postsurgical pain occurred in 21.4% of patients in our cohort. This finding is comparable with previous reports showing that 10 to 30% of patients experienced persistent pain after surgery.^1,2,5  We have shown that patients carrying homozygous A allele of SNPs rs2070697 and rs2236742 in the CTSG gene were associated with a lower risk for chronic postsurgical pain. These results are consistent with our animal data. Our finding, however, cannot precisely locate the SNPs that regulate the transcription of CTSG because both SNPs are found in the introns of the gene (Supplemental Digital Content 5, http://links.lww.com/ALN/B192). Therefore, they may merely represent positional markers of the underlying functional SNP. Resequencing of the exon regions is currently underway to reveal the locus that may interfere with functional characteristics of the CTSG protein. In addition to CTSG polymorphisms, our multivariate regression model showed that male patients and patients younger than 65 yr were at a higher risk for chronic postsurgical pain. This is consistent with some studies but not with others.^4,36  Patients reporting severe pain early after surgery also tended to develop chronic postsurgical pain, but this did not reach statistical significance in our multivariate model. Both unemployment and low level of education were thought to be risk factors in chronic pain syndrome.³⁷  However, we were unable to demonstrate the impact of these factors on chronic postsurgical pain, and this may be related to a cultural issue where the local Chinese community is less dependent on unemployment benefits.

In summary, we have shown that CTSG was up-regulated in spinal dorsal horns after peripheral inflammation. CTSG contributed to the development of pain hypersensitivity owing to its chemotactic properties, leading to an increase in neutrophils and release of inflammatory mediators in the spinal cord. CTSG also promoted astrocyte activation upon peripheral inflammation. Variations in the CTSG gene were associated with reduced risks of chronic postsurgical pain in two independent groups of patients undergoing surgical procedures. CTSG represents a potential target for chronic pain intervention. Furthermore, preoperative determination of CTSG gene polymorphisms may facilitate perioperative physicians to formulate an appropriate plan to prevent chronic postsurgical pain.

This research was supported by General Research Fund (grant no. 461513) from Research Grant Council, Hong Kong (Hong Kong Special Administrative Region, People’s Republic of China) and Academic enhancement grant (grant no. AEG13/001) and Project grant (grant no. 11/012) from the Australian and New Zealand College of Anaesthetists (Melbourne, Australia).

The authors declare no competing interests.