Abstract
Persistent pain following breast cancer surgery is a significant problem. Both inherited and acquired mechanisms of inflammation appear to play a role in the development and maintenance of persistent pain. In this longitudinal study, growth mixture modeling was used to identify persistent breast pain phenotypes based on pain assessments obtained prior to and monthly for 6 months following breast cancer surgery. Associations between the “no pain” and “mild pain” phenotypes and single nucleotide polymorphisms (SNPs) spanning 15 cytokine genes were evaluated. The methylation status of the CpG sites found in the promoters of genes associated with pain group membership was determined using bisulfite sequencing. In the multivariate analysis, three SNPs (i.e., interleukin 6 (IL6) rs2069840, C-X-C motif chemokine ligand 8 (CXCL8) rs4073, tumor necrosis factor (TNF) rs1800610) and two TNF CpG sites (i.e., c.-350C, c.-344C) were associated with pain group membership. These findings suggest that variations in IL6, CXCL8, and TNF are associated with the development and maintenance of mild persistent breast pain. CpG methylation within the TNF promoter may provide an additional mechanism through which TNF alters the risk for mild persistent breast pain after breast cancer surgery. These genetic and epigenetic variations may help to identify individuals who are predisposed to the development of mild levels of persistent breast pain following breast cancer surgery.
Keywords: cytokine genes, breast cancer, DNA methylation, persistent pain, post-surgical pain, post-mastectomy pain, epigenetics, interleukin 6, C-X-C motif chemokine ligand 8, tumor necrosis factor
1. Introduction
Persistent pain in women following breast cancer surgery is common and associated with altered mood and sleep patterns, decreased quality of life, and disability [1, 2]. Persistent postsurgical pain may result from ongoing nociceptor activation and/or nerve injury [3]. During the early postoperative period, inflammatory mediators produce sensitization at the affected area. These reversible changes in sensitivity discourage stimulation which serves as a protective mechanism that facilitates healing. Sustained activation of nociceptors may lead to maintenance of central sensitization and maladaptive phenotypic changes that alter the normal stimulus-response relationship and produce persistent pain [4].
Persistent alterations within nociceptors include changes in gene expression [5]. Evidence suggests that ongoing activation of inflammatory cells plays a role in the establishment of persistent pain [6]. In addition, peripheral nerve injury elicits an inflammatory reaction that prompts the aggregation of immune cells and increases the local concentration of pro-inflammatory cytokines [7]. These mediators participate in the initiation and maintenance of persistent pain after nerve injury by generating ectopic activity [8], altering neuronal connectivity [9], and reducing the number of inhibitory neurons [10].
Despite a clear connection between immune mechanisms and persistent pain [7], few studies were identified that evaluated for associations between polymorphisms in cytokine pathways and cancerrelated pain [11–14]. Findings from these studies are difficult to interpret because pain was characterized using a dichotomized rating, the samples were small, and the number of polymorphisms evaluated was not comprehensive.
Emerging evidence suggests that acquired adaptations to genetic regulation (termed “epigenetics”) are pervasive in biology [15]. Deoxyribonucleic acid (DNA) methylation is an epigenetic mechanism that regulates gene expression [15]. Acquisition of methylation at CpG dinucleotides provides an adaptive capacity for the organism to adjust to sustained changes in its environment. The methylation status of CpG sites within gene promoters has emerged as a promising biomarker for risk stratification and detection of human disease [16].
Recent work from our group used growth mixture modeling (GMM) to identify subgroups of women with distinct persistent breast pain trajectories prior to and for six months following breast cancer surgery [17]. Three distinct classes were identified using patients’ ratings of worst pain in their breast. A fourth pain group was designated for those women who did not experience breast pain. The largest subgroup of women identified was the mild breast pain class (n= 173, 43.5%) which had a mean worst pain severity score of 2.5 (on a 0 to 10 numeric rating scale (NRS)). Of note, mild levels of persistent postsurgical pain are associated with diminished perceptions of overall health and reduced physical and social functioning [18]. Therefore, using data from women who were classified into the no breast pain and mild breast pain classes, the purposes of this study were to: 1) evaluate for associations between single nucleotide polymorphisms (SNPs) contained within cytokine genes and pain group membership and 2) determine the methylation status of CpG sites contained within the promoter of cytokine genes that harbored gene variations associated with pain group membership.
2. Methods
2.1. Patients and settings
This longitudinal study is part of a larger study of women who underwent breast cancer surgery [11, 17, 19]. Patients were recruited from a Comprehensive Cancer Center, two public hospitals, and four community practices. Patients were included if they: were female; ≥18 years of age; underwent unilateral breast cancer surgery; were able to read, write, and understand English; and gave written informed consent. Patients were excluded if they had bilateral breast cancer surgery and/or had distant metastasis. Of the 516 patients approached, 410 were enrolled in the study (response rate 79.4%). The major reasons for refusal were: being too busy, overwhelmed with the cancer diagnosis, or insufficient time available to do the baseline assessment prior to surgery.
2.2. Subjective measures
The Breast Symptoms Questionnaire (BSQ) and Post Surgical Pain Questionnaire, evaluated persistent and acute postoperative pain, respectively. Part 1 of the BSQ obtained information on the occurrence of pain and the occurrence of other symptoms in the breast scar area. Additional symptoms that were based on the work of Tasmuth and colleagues [20, 21]. Patients completed Part 2 of the BSQ if they had pain in the breast scar area. Patients rated the intensity of their average and worst pain using a 0 (no pain) to 10 (worst imaginable pain) NRS. The NRS is a valid and reliable measure of pain intensity [22].
Postsurgical Pain Questionnaire evaluated pain intensity, pain relief, and satisfaction with pain treatment in the first 24 to 48 hours after surgery. Average and worst pain intensity were rated on a 0 (no pain) to 10 (worst imaginable pain) NRS. Pain relief was rated on a 0% (no relief) to 100% (complete relief) scale. Satisfaction with pain treatment was rated on a 0 (not satisfied at all) to 10 (extremely satisfied) NRS. Patients completed this questionnaire during the month 1 study visit.
2.3. Study procedures
The Committee on Human Research at the University of California, San Francisco and the Institutional Review Boards at our other sites approved this study. During the patient’s preoperative visit, a clinician explained the study to the patient and determined her willingness to participate. For women who were willing to participate, clinicians introduced the patient to the research nurse. The research nurse determined the woman’s eligibility, obtained written informed consent prior to surgery, had the patients complete the enrollment questionnaires (Assessment 0).
Patients were contacted two weeks after surgery to schedule the first post-surgical appointment. Patients were seen either in their home or in the Clinical Research Center at 1, 2, 3, 4, 5, and 6 months after surgery. During each of the visits, the women completed the study questionnaires and provided information on new and ongoing treatments. The blood sample was collected at the time of enrollment or during one of the monthly study visits. Patients’ medical records were reviewed for disease and treatment information.
2.4. Characterization of the persistent breast pain phenotype
The characterization of the persistent breast pain phenotype used in the current study was described previously [17]. At each assessment, patients were asked, “Are you experiencing pain in your affected breast?” If the patient reported pain, she rated her “current pain at its worst” using a 0 (no pain) to 10 (worst pain) NRS. Prior to performing the GMM analysis, patients who reported no pain in their affected breast for all 6 assessments (i.e., enrollment and 2, 3, 4, 5, and 6 months) were identified (N=126; 31.7%). These patients were not included in the GMM analysis. The remaining 272 women’s ratings of worst breast pain were used in the GMM analysis. GMM was used to assign each individual into a latent class based on similarities in worst pain ratings at enrollment and at 2, 3, 4, 5, and 6 months after surgery. Pain ratings obtained at the 1-month follow-up assessment were excluded from the model because it reduced the variability in pain ratings among the patients. Attempts to determine the latent classes failed when the month 1 ratings were included in the GMM analysis.
The GMM methods are described in detail elsewhere [23]. In brief, a single growth curve that represented the “average” change trajectory was estimated for the total sample. Then the number of latent growth classes that best fit the data was identified using established guidelines [24–26]. Descriptive statistics and frequency distributions for the no breast pain and mild breast pain classes were generated for demographic and clinical characteristics using Stata version 12.1 (StataCorp, College Station, TX). Independent sample t-tests, Mann-Whitney U tests, and Chi square and Fisher’s Exact tests were used to evaluate for differences in demographic and clinical characteristics between the two breast pain classes. Adjustments were not made for missing data in comparisons between the GMM classes. Therefore, the cohort for each analysis was dependent on the largest set of available data across groups. A p-value of <0.05 was considered statistically significant.
Logistic regression analysis was performed to evaluate for associations between phenotypic characteristics and pain group membership. Based on a review of the literature, all phenotypic characteristics that were identified in the bivariate analyses as being significantly different between the pain classes were evaluated for inclusion in the multivariate analysis. A backwards stepwise approach was used to create the most parsimonious model. Only predictors with a p-value of <.05 were retained in the final model. These same predictors were used in the models that evaluated the associations between genotype and pain group membership and CpG methylation level and pain group membership.
2.5. Genotype determination
Genomic DNA was extracted from peripheral blood mononuclear cells using the PUREGene DNA Isolation System (Invitrogen, Carlsbad, CA). DNA was available from 310 of the 398 patients. DNA samples were quantitated with a Nanodrop Spectrophotometer (ND-1000; Nanodrop Products, Wilmington, DE) and normalized to a concentration of 50 nanogram/microliter (ng/µL) (diluted in 10 mM Tris/1 mM EDTA). A combination of tagging SNPs and literature driven SNPs (i.e., reported as being associated with altered function and/or symptoms) were chosen for this study. Tagging SNPs needed to be common (i.e., had a minor allele frequency (MAF) of ≥.05 in public databases (e.g., HapMap)).
The genotyping was performed blinded to clinical status and incorporated positive and negative controls. Samples were genotyped using a custom built array on the Golden Gate platform (Illumina, San Diego, CA) and processed with a standard protocol using GenomeStudio (Illumina, San Diego, CA). Two blinded reviewers visually inspected signal intensity profiles and resulting genotype calls for each SNP. Disagreements were adjudicated by a third reviewer. Quality control filtering of SNPs was performed. SNPs with call rates <95%, Hardy-Weinberg P < .001, and/or a MAF of <5% were excluded.
As shown in Supplementary Table 1, a total of 82 SNPs from 15 cytokine genes (i.e., interferon gamma (IFNG): 5 SNPs; IFNG receptor 1 (IFNGR1): 1 SNP; IL1B: 12 SNPs; IL1 receptor 1 (IL1R1): 4 SNPs; IL1R2: 3 SNPs; IL2: 3 SNPs; IL4: 2 SNPs; IL6: 9 SNPs; C-X-C motif chemokine ligand 8 (CXCL8): 3 SNPs; IL10: 7 SNPs; IL13: 4 SNPs; IL17A: 5 SNPs; nuclear factor kappa beta-1 (NFKB1): 11 SNPs; NFKB2: 4 SNPs; tumor necrosis factor (TNF): 9 SNPs) passed all quality control filters and are included in subsequent analyses. These candidate genes were selected based on a review of the literature on the role of inflammatory mediators in the development of persistent pain [27–29]. PUPASuite 2.0 was used to evaluate potential functional roles for SNPs associated with persistent pain [30].
2.6. CpG methylation determination
The percentage of CpG methylation in promoter regions of genes harboring SNPs associated with the persistent pain phenotype in multivariate analyses (i.e., IL6, CXCL8, TNF) was determined using bisulfite sequencing. Genomic DNA from each patient was normalized to a concentration of 75 ng/µl to 125 ng/µl (diluted in 10 mM Tris/1 mM EDTA). Approximately 1.5 micrograms (µg) of genomic DNA was treated with sodium bisulfite and purified using the Methyl Detector Kit (Active Motif, Carlsbad, CA). One duplicate DNA sample was included within each bisulfite conversion group (n=10) to serve as a technical replicate.
Bisulfite treated DNA was used as the template for polymerase chain reaction (PCR) of the promoter regions using primers complementary to the bisulfite converted DNA sequences (Figure 1A, Figure 2A, Figure 3A). The promoter region of TNF was assessed using previously reported primers [31]. Two regions of the IL6 promoter and one region of the CXCL8 promoter were assessed with primers designed using MethPrimer (http://www.urogene.org/methprimer/). Primers were designed to assess methylation of CpG sites within the proximal promoter for CXCL8. Specific CpG sites within the IL6 promoter were selected for bisulfite sequencing based on prior evidence of differential methylation (i.e., c.-1162C [32], c.-727C (unpublished data)). Primer sequences and annealing temperatures for each amplicon are provided in Supplementary Table 2.
Figure 1.
Deoxyribonucleic acid (DNA) methylation at the promoter region of the interleukin 6 (IL6) gene. (A) Scaled schematic representation of the 5’ untranslated region of the IL6 gene showing the distribution of CpG sites (vertical lines) and regions (horizontal bars) amplified using polymerase chain reaction. Nucleotide positions in relation to predicted translation start site are provided from GRCh37.p9 Primary Assembly; NM_000600.3. Two regions of the IL6 gene promoter were assessed. Region I assessed the percent methylation of c.-1162C, c.-1159C, c.-1157C, c.-1132C, c.-1124C and c.-1120C. Region II assessed the percent methylation of c.-729C, c.-727C, c.-691C, c.-673C, and c.-637C. The c.-1064C locus was not assayed. The association of the methylation level at c.-1120C and pain group membership was not evaluated because this site was 100% methylated for all samples. Locations of the amplified regions of the bisulfite-modified genome are shown as horizontal bars above the promoter. (B) DNA methylation levels of the promoter regions for IL6.
Abbreviations: na, not assayed
Figure 2.
Deoxyribonucleic acid (DNA) methylation at the promoter region of the C-X-C motif chemokine ligand 8 (CXCL8) gene. (A) Scaled schematic representation of the 5’ untranslated region of the CXCL8 gene showing the distribution of CpG sites (vertical lines) and region (horizontal bars) amplified using polymerase chain reaction. Nucleotide positions in relation to predicted translation start sites are provided from GRCh37.p9 Primary Assembly; NM_000584.3. Primer binding sites are shown (arrows). (B) DNA methylation levels of the promoter region for CXCL8.
Figure 3.
Deoxyribonucleic acid (DNA) methylation at the promoter region of the tumor necrosis factor alpha (TNF) gene. (A) Scaled schematic representation of the 5’ untranslated region of the TNF gene showing the distribution of CpG sites (vertical lines) and region (horizontal bars) amplified using polymerase chain reaction. Nucleotide positions in relation to predicted translation start sites are provided from GRCh37.p9 Primary Assembly; NM_000594.3. Primer binding sites are shown (arrows). (B) DNA methylation levels of the promoter region for TNF. Asterisks indicates statistical significance at p<0.05. (C) Heat map showing the pairwise shared variance (r2) of methylation levels between CpG sites. The value of r2 multiplied by 100 is provided within each box. The degree of shared variance was color-coded to indicate high (red) to none (white) with intermediate values rendered in pink. 5’ 3’
Each 25 µl reaction contained 2 µl of bisulfite treated DNA and 2.5 units PfuTurbo Cx DNA Polymerase (Agilent, Santa Clara, CA). Optimal efficiency of PCR was achieved through touchdown thermocycling and titration of reaction components to minimize the formation of non-specific products. Excess primers and nucleotides were removed from the PCR products using ExoSAP-IT (USB Corp., Cleveland, OH). The PCR products were sequenced directly using the reverse primer with BigDye terminator sequencing chemistry (Applied Biosystems, Carlsbad, CA).
Quantitation of methylation at each CpG site was estimated from sequence trace files (i.e., .abi files) using Epigenetic Sequencing Methylation Analysis (ESME) software version 3.2.4 (Epigenomics AG, Berlin). ESME performs quality control, aligns the sequence trace file to the expected bisulfite converted genomic reference sequence, normalizes the signal intensities, corrects for incomplete bisulfite conversion, and calculates methylation levels for each CpG site by comparing the cytosine to thymine peaks [33].
2.7. Statistical analyses
2.7.1. Candidate gene analysis
Gene counting was used to determine allele and genotype frequencies. Hardy-Weinberg equilibrium was assessed by the Chi Square test. Measures of linkage disequilibrium (i.e., D’ and r2) were done using Haploview 4.2. Linkage disequilibrium (LD)-based haplotype block definition was based on D’ confidence interval method [34].
Haplotype analyses were conducted in order to localize the association signal within each gene and to determine if the identified haplotypes improved the strength of the association with the phenotype. Haplotypes were constructed using the program PHASE version 2.1 [35]. The haplotype construction procedure was done five times using different seed numbers each time. Only haplotypes that were inferred with probability estimates of ≥.85, across all five iterations, were retained for downstream analyses. Haplotypes were evaluated assuming a dosage model.
2.7.2. DNA methylation analysis
The association between the percentage of methylation at each CpG site and the persistent breast pain phenotype was evaluated by t-test at a p-value of < 0.05 for blood samples collected within 6 months following surgery (n=120). Bland-Altman analysis was used to assess the level of agreement between technical replicates and to determine the quality control replicate range methylation estimates (i.e., measurement noise) [36]. The 95% limits of agreement are provided in Supplementary Table 3. All CpG methylation estimates are expressed as mean percentages ± standard deviations.
2.7.3. Multivariate logistic regression analyses
Ancestry informative markers (AIMs) were used to minimize confounding due to population stratification [37–39]. These AIMs were determined by principal component analysis [40] using Helix Tree (Golden Helix, Bozeman, MT). The number of principal components (PCs) was sought that distinguished the major racial/ethnic groups in the sample by visual inspection of scatter plots of orthogonal PCs. This procedure was repeated until no discernible clustering of patients by their self-reported race/ethnicity was possible (data not shown). The first three PCs were selected for use as covariates in all of the regression models to adjust for potential confounding due to population substructure (i.e., race/ethnicity). One hundred and six AIMs were used in the analysis.
Three genetic models were assessed for each SNP: additive, dominant, and recessive. The genetic model that best fit the data, by maximizing the significance of the p-value, was selected for each SNP. Logistic regression analysis, that controlled for significant covariates, as well as genomic estimates of and self-reported race/ethnicity, was used to evaluate for associations between genotype and pain group membership or CpG methylation level and pain group membership. The backwards stepwise approach was employed to create the most parsimonious model. Except for genomic estimates of and self-reported race/ethnicity, only predictors with a p-value of <.05 were retained in the final model. (Epi)genetic model fit and both unadjusted and adjusted odds ratios were estimated using STATA version 12.1.
As was done in all of our previous studies (e.g., [11, 41]), based on recommendations in the literature [42, 43], the implementation of rigorous quality controls for genomic data, the non-independence of SNPs/haplotypes in LD, as well as the exploratory nature of the analyses, adjustments were not made for multiple testing. The significant SNPs identified in the bivariate analyses were evaluated further using regression analyses. These analyses controlled for differences in phenotypic characteristics identified in our prior study (Table 1) [44], potential confounding due to population stratification, as well as variation in other SNPs/haplotypes within the same gene. Only the SNPs that remained significant are reported in this paper. Therefore, the significant independent associations reported in this paper are unlikely to be due solely to chance. Unadjusted (bivariate) associations are reported for all of the SNPs that passed quality control criteria in Supplementary Table 1 to allow for subsequent comparisons and meta-analyses.
Table 1.
Differences in demographic and clinical characteristics between the breast pain classes prior to surgery
| Demographic Characteristics | No Pain n=126 |
Mild Pain n=173 |
Statistics |
|---|---|---|---|
|
| |||
| Mean (SD) | Mean (SD) | ||
|
| |||
| Age (years) | 58.6 (11.4) | 53.4 (11.5) | t=3.84; p<0.001 |
|
| |||
| Education (years) | 15.8 (2.8) | 16.0 (2.6) | t=−0.79; p=0.430 |
|
| |||
| % (N) | % (N) | ||
|
| |||
| Ethnicity | |||
| White | 73.0 (92) | 63.7 (109) | Χ2=3.07; p=0.381 |
| Black | 7.1 (9) | 8.8 (15) | |
| Asian/Pacific Islander | 10.3 (13) | 12.9 (22) | |
| Hispanic/ mixed ethnic background/other | 9.5 (12) | 14.6 (25) | |
|
| |||
| Lives alone* | 20.8 (26) | 24.6 (42) | FE; p=0.486 |
|
| |||
| Marital status | |||
| Married/partnered | 41.3 (52) | 39.5 (68) | FE; p=0.811 |
| Single, separated, widowed, divorced | 58.7 (74) | 60.5 (104) | |
|
| |||
| Currently working for pay* | 52.0 (65) | 50.9 (87) | FE; p=0.906 |
|
| |||
| Total annual household income | |||
| < $10,000 to $19,999 | 8.5 (9) | 14.4 (21) | Z=−0.717; p=0.473 |
| $20,000 to $99,000 | 48.1 (51) | 43.8 (64) | |
| ≥ $100,000 | 43.4 (46) | 41.8 (61) | |
|
| |||
| Clinical Characteristics | Mean (SD) | Mean (SD) | |
|
| |||
| Body mass index (kg/m2) | 27.1 (7.0) | 25.9 (5.3) | t=1.62; p=0.107 |
|
| |||
| Karnofsky Performance Status score | 96.2 (8.7) | 93.6 (9.3) | t=2.42; p=0.016 |
|
| |||
| Self-Administered Comorbidity Scale score | 4.0 (2.3) | 4.0 (3.0) | t=−0.02; p=0.981 |
|
| |||
| Number of breast biopsies | 1.4 (0.7) | 1.6 (0.8) | Z=−2.45; p=0.014 |
|
| |||
| % (N) | % (N) | ||
|
| |||
| Occurrence of comorbid conditions (% and number of women who reported each comorbid condition from the Self-Administered Comorbidity Questionnaire)* | |||
| Heart disease | 4.0 (5) | 2.9 (5) | FE; p=0.747 |
| High blood pressure | 34.9 (44) | 24.3 (42) | FE; p=0.052 |
| Lung disease | 2.4 (3) | 3.5 (6) | FE; p=0.738 |
| Diabetes | 7.1 (9) | 5.8 (10) | FE; p=0.640 |
| Ulcer | 3.2 (4) | 4.0 (7) | FE; p=0.765 |
| Kidney disease | 1.6 (2) | 0.6 (1) | FE; p=0.575 |
| Liver disease | 3.2 (4) | 1.7 (3) | FE; p=0.460 |
| Anemia | 7.9 (10) | 5.2 (9) | FE; p=0.348 |
| Depression | 16.7 (21) | 22.5 (39) | FE; p=0.243 |
| Osteoarthritis | 17.5 (22) | 14.5 (25) | FE; p=0.522 |
| Back pain | 22.2 (28) | 27.2 (47) | FE; p=0.348 |
| Rheumatoid arthritis | 1.6 (2) | 2.9 (5) | FE; p=0.703 |
|
| |||
| Diagnosed with mastitis* | 11.2 (14) | 12.3 (21) | FE; p=0.856 |
|
| |||
| Diagnosed with fibrocystic disease* | 18.6 (22) | 22.4 (38) | FE; p=0.465 |
|
| |||
| Ever breast fed* | 48.0 (60) | 47.1 (81) | FE; p=0.907 |
|
| |||
| Surgery to affected breast unrelated to cancer* | 7.9 (10) | 14.5 (25) | FE; p=0.101 |
|
| |||
| Post-menopausal* | 71.0 (88) | 58.3 (98) | FE; p=0.027 |
|
| |||
| Received neoadjuvant chemotherapy* | 17.5 (22) | 18.6 (32) | FE; p=0.879 |
|
| |||
| On hormonal replacement therapy prior to surgery* | 19.0 (24) | 16.9 (29) | FE; p=0.648 |
|
| |||
| Stage of disease | |||
| Stage 0 | 17.5 (22) | 20.8 (36) | Z=−0.07; p=0.945 |
| Stage 1 | 41.3 (52) | 35.8 (62) | |
| Stage IIA and IIB | 35.7 (45) | 35.8 (62) | |
| Stage IIIA, IIIB, IIIC, and IV | 5.6 (7) | 7.5 (13) | |
|
| |||
| Pain in breast prior to surgery* | 2.4 (3) | 41.1 (69) | FE; p<0.001 |
|
| |||
| Swelling in affected breast* | 3.2 (4) | 6.4 (11) | FE; p=0.286 |
|
| |||
| Numbness in affected breast* | 2.4 (3) | 3.5 (6) | FE; p=0.738 |
|
| |||
| Strange sensations in affected breast* | 12.7 (16) | 34.7 (60) | FE; p<0.001 |
|
| |||
| Hardness in affected breast* | 7.9 (10) | 21.4 (37) | FE; p=0.002 |
|
| |||
| Surgical Characteristics | Mean (SD) | Mean (SD) | |
|
| |||
| Number of lymph nodes removed | 4.3 (4.7) | 6.0 (7.1) | t=−2.45; p=0.015 |
|
| |||
| Number of drains placed during surgery | 0.4 (0.7) | 0.5 (0.7) | t=−1.33; p=0.184 |
|
| |||
| % (N) | % (N) | ||
|
| |||
| Type of surgery | |||
| Breast conserving | 84.1 (106) | 77.5 (134) | FE; p=0.186 |
| Mastectomy | 15.9 (20) | 22.5 (39) | |
|
| |||
| Reconstruction at the time of surgery* | 15.9 (20) | 29.1 (50) | FE; p=0.009 |
|
| |||
| Surgical drain placed in breast at time of surgery* | 65.7 (23) | 64.2 (43) | FE; p=1.000 |
|
| |||
| Post-surgical Characteristics | Mean (SD) | Mean (SD) | |
|
| |||
| Number of postoperative complications | 0.2 (0.5) | 0.2 (0.5) | t=−0.49; p=0.625 |
|
| |||
| Severity of average postoperative pain | 2.8 (2.1) | 3.7 (2.2) | t=−3.46; p<0.001 |
|
| |||
| Severity of worst postoperative pain | 4.2 (2.6) | 5.0 (2.6) | t=−2.60; p=0.010 |
|
| |||
| % (N) | % (N) | ||
|
| |||
| Received radiation therapy during the 6 months* | 56.3 (71) | 54.3 (94) | FE; p=0.814 |
|
| |||
| Received adjuvant chemotherapy during the 6 months* | 31.7 (40) | 34.7 (60) | FE; p=0.621 |
|
| |||
| Received hormonal therapy during the 6 months* | 45.2 (57) | 41.6 (72) | FE; p=0.556 |
|
| |||
| Received biological therapy during the 6 months* | 8.7 (11) | 12.7 (22) | FE; p=0.351 |
|
| |||
| Received complementary therapy during the 6 months* | 23.8 (30) | 28.3 (49) | FE; p=0.427 |
|
| |||
| Had breast reconstruction during the 6 months* | 4.8 (6) | 10.4 (18) | FE; p=0.087 |
|
| |||
| Had re-excision or mastectomy during the 6 months* | 18.3 (23) | 31.2 (54) | FE; p=0.016 |
denotes the percentage of patients with the demographic or clinical characteristic
Abbreviations: FE = Fisher’s Exact; SD = standard deviation; kg = kilogram; m2 = meters squared
Adapted from Miaskowski et al. [1].
3. Results
3.1. Demographic and clinical characteristics of the pain classes
As summarized in Table 1, patients in the mild pain class were younger and more likely to be premenopausal. In addition, these patients had a lower KPS score; had had a higher number of breast biopsies; and were more likely to report pain, strange sensations, and hardness in their breast prior to surgery. Patients in the mild pain class were more likely to have had a higher number of lymph nodes removed; to have had reconstruction at the time of surgery; to have had higher levels of postoperative pain; and to have had a re-excision or mastectomy in the 6 months following surgery [17].
3.2. Candidate gene analysis for pain group membership
As summarized in Supplementary Table 1, no associations with pain group membership were found for the SNPs in IFNGR1, IL1B, IL1R1, IL1R2, IL2, IL4, IL13, IL17A, NFKB1, and NFKB2. However, the genotype frequency was significantly different between the no pain and mild pain classes for 7 SNPs and 2 haplotypes among 5 genes (IFNG1: 1 SNP; IL6: 1 SNP, 1 haplotype; CXCL8: 3 SNPs, 1 haplotype; IL10: 1 SNP; TNF: 1 SNP). In order to better estimate the magnitude (i.e., odds ratio, OR) and precision (i.e., 95% confidence interval, CI) of genotype on pain group membership, multivariate logistic regression models were fit. As shown in Table 2, in addition to genotype and genomic estimates of and self-reported race/ethnicity, the phenotypic characteristics included in all of the regression models were: the presence of preoperative pain in the affected breast, hardness in the affected breast preoperatively, and occurrence of re-excision or mastectomy within the first 6 months following surgery. The genetic associations that remained significant were for IL6 rs2069840, CXCL8 rs4073, and TNF rs1800610 (Table 2).
Table 2.
Multiple logistic regression analyses for candidate gene polymorphisms
| GMM Class Comparison |
Predictor | Odds Ratio | Standard Error |
95% CI | Z | p-value |
|---|---|---|---|---|---|---|
| No pain versus mild pain (n=228) | IL6 rs2069840 | 0.21 | 0.12 | 0.07, 0.63 | −2.78 | 0.005 |
| Breast pain preoperatively | 28.75 | 18.85 | 7.96, 103.91 | 5.12 | <0.001 | |
| Hardness in the breast preoperatively | 3.42 | 1.97 | 1.10, 10.59 | 2.13 | 0.033 | |
| Re-excision/mastectomy | 3.17 | 1.27 | 1.45, 6.95 | 2.88 | 0.004 | |
| Overall model fit: χ2 = 81.67, p <0.0001; pseudo R2 = 0.2637 | ||||||
| No pain versus mild pain (n=228) | CXCL8 rs4073 | 0.40 | 0.16 | 0.18, 0.87 | −2.32 | 0.020 |
| Breast pain preoperatively | 29.95 | 19.48 | 8.37, 107.16 | 5.23 | <0.001 | |
| Hardness in the breast preoperatively | 3.02 | 1.71 | 1.00, 9.13 | 1.96 | 0.050 | |
| Re-excision/mastectomy | 2.52 | 0.97 | 1.19, 5.34 | 2.41 | 0.016 | |
| Overall model fit: χ2 = 78.55, p <0.0001; pseudo R2 = 0.2536 | ||||||
| No pain versus mild pain (n=228) | TNF rs1800610 | 0.37 | 0.17 | 0.15, 0.89 | −2.22 | 0.026 |
| Breast pain preoperatively | 26.52 | 17.02 | 7.54, 93.33 | 5.11 | <0.001 | |
| Hardness in the breast preoperatively | 4.14 | 2.39 | 1.34, 12.81 | 2.46 | 0.014 | |
| Re-excision/mastectomy | 2.72 | 0.86 | 1.08, 4.78 | 2.16 | 0.030 | |
| Overall model fit: χ2 = 78.15, p <0.0001; pseudo R2 = 0.2523 | ||||||
Notes: Multiple logistic regression analysis of the no pain and the mild pain GMM group on the worst pain rating. For each model, the first three principal components identified from the analysis of ancestry informative markers as well as self-report race/ethnicity were retained in all models to adjust for potential confounding due to race or ethnicity (data not shown). Predictors evaluated in each model included genotype (IL6 rs2069840: CC+CG versus GG; CXCL8 rs4073 TT+TA versus AA; TNF rs1800610: CC versus CT+TT), pain in the affected breast prior to surgery, hardness in the affected breast prior to surgery, and re-excision or mastectomy was performed within 6 months after the initial surgery for breast cancer.
Abbreviations: CI =confidence interval; CXCL8 = C-X-C motif chemokine ligand 8; GMM = growth mixture model; IL6 = interleukin 6; TNF = tumor necrosis factor alpha
In the regression analysis for IL6 rs2069840, patients homozygous the rare G allele (i.e., CC+CG versus GG) were 79% less likely to be in the mild breast pain class. In the regression analysis for CXCL8 rs4073, patients homozygous for the rare A allele (i.e., TT + TA versus AA) were 60% less likely to be in the mild breast pain class. In the regression analysis for TNF rs1800610, patients who were heterozygous or homozygous for the rare T allele (i.e., CC versus CT+TT) were 63% less likely to be in the mild breast pain class.
3.3. DNA methylation analysis
Eleven CpG sites were examined in the IL6 promoter. The distal CpG sites (i.e., c.-1162C, c.-1159C, c.-1157C, c.-1132C, c.-1124C, c.-1120C) were highly methylated (i.e, >90%) and the proximal CpG sites (i.e., c.-729C, c.-727C, c.-691C, c.-673C, c.-637C) were largely unmethylated (i.e., <15%). No statistically significant differences were found in the percentage of methylation at any of the IL6 CpG sites examined between the two breast pain groups (Figure 1B). Of note, the presence of the rare C allele in rs1800796 disrupted the CpG site at c.-637C (i.e., from CpG to CpC). However, when the common G allele was present, the c.-637C CpG site was unmethylated in both pain groups.
Six CpG sites were assayed in the CXCL8 promoter (i.e., c.-116C, c.-106C, c.-31, c.+46C, c.+125, c.+171C). All CpG sites were largely unmethylated. No statistically significant differences were found in the percentage of methylation at each of the CXCL8 sites examined between the two breast pain groups (Figure 2B).
Eleven CpG sites were assayed in the TNF promoter. The distal CpG sites (i.e., c.-484C, c.-425C, c.-419C) were highly methylated (i.e., >70%) and the proximal CpG sites (i.e., c.-350C, c.-344C, c.-342C, c.-327C, c.-300C, c.-253C, c.-230C, c.-219C) were largely unmethylated (i.e., <30%). Compared to the no breast pain class, a significantly higher percentage of the c.-350C, c.-344C, and c.-342C sites were methylated in the mild breast pain class (all p<0.05, Figure 3B). Of note, these three CpG sites are moderately correlated (Figure 3C).
In order to better estimate the magnitude (i.e., odds ratio, OR) and precision (i.e., 95% confidence interval, CI) of TNF promoter methylation on pain group membership, multivariate logistic regression models were fit. In addition to methylation percentage, genomic estimates of and self-reported race/ethnicity, the same phenotypic characteristics were included as was done for the candidate gene analyses. As shown in Table 3, two CpG sites (i.e., TNF c.-350C, TNF c.-344C) remained significant in the final models.
Table 3.
Multiple logistic regression analyses for TNF CpG promoter methylation
| GMM Class Comparison |
Predictor | Odds Ratio | Standard Error |
95% CI | Z | p-value |
|---|---|---|---|---|---|---|
| No pain versus mild pain (n=117) | TNF c.-180C methylation | 83.21 | 163.05 | 1.79, 3872.62 | 2.26 | 0.024 |
| Breast pain preoperatively | 31.09 | 32.84 | 3.92, 246.48 | 3.25 | 0.001 | |
| Hardness in the breast preoperatively | 2.87 | 2.17 | 0.65, 12.65 | 1.40 | 0.162 | |
| Re-excision/mastectomy | 2.30 | 1.26 | 0.79, 6.72 | 1.53 | 0.126 | |
| Overall model fit: χ2 = 44.06, p<0.0001; pseudo R2 = 0.2810 | ||||||
| No pain versus mild pain (n=116) | TNF c.-174C methylation | 215.90 | 543.46 | 1.55, 29987.25 | 2.14 | 0.033 |
| Breast pain preoperatively | 28.37 | 30.07 | 3.55, 226.47 | 3.16 | 0.002 | |
| Hardness in the breast preoperatively | 4.13 | 3.17 | 0.92, 18.57 | 1.85 | 0.065 | |
| Re-excision/mastectomy | 2.39 | 1.32 | 0.81, 7.05 | 1.58 | 0.113 | |
| Overall model fit: χ2 = 44.72, p<0.0001; pseudo R2 = 0.2870 | ||||||
Notes: Multiple logistic regression analysis of the no pain and the mild pain GMM group on the worst pain rating. For each model, the first three principal components identified from the analysis of ancestry informative markers as well as self-report race/ethnicity were retained in all models to adjust for potential confounding due to race or ethnicity (data not shown). Predictors evaluated in each model included CpG methylation level in 1% increments (i.e., TNFA c.-180C, TNFA c.-174C), pain in the affected breast prior to surgery, presence of hardness in the affected breast prior to surgery, and re-excision or mastectomy was performed within 6 months after the initial surgery for breast cancer.
Abbreviations: CI =confidence interval; GMM = growth mixture model; TNF = tumor necrosis factor alpha
4. Discussion
This study is the first to evaluate for associations between genetic and epigenetic variations in cytokine genes and the development of mild persistent breast pain in women following breast cancer surgery. While cytokines modulate nociceptive signaling during acute and chronic inflammation and following tissue injury [45], significant inter-individual variability exists in the development, intensity, and resolution of postoperative pain. In this study, three SNPs (i.e., IL6 rs2069840, CXCL8 rs4073, TNF rs1800610) were associated with pain group membership after adjusting for phenotypic characteristics.
IL6 is produced by a variety of cells at sites of tissue injury and inflammation. IL6 promotes the development of persistent pain through prolonged or permanent sensitization of nociceptors and spinal cord neurons; sympathetic sprouting and microglia activation; and modulation of nociceptive mediators that produce long-term potentiation [46]. In addition to its pro-inflammatory effects, IL6 exerts anti-inflammatory effects by promoting the release of TNFα and IL1 antagonists to attenuate pro-inflammatory signaling [47]. Our results suggest that the rare G allele of rs2069840 decreases the risk for mild persistent breast pain after breast cancer surgery. Recent work found that the rare G allele of rs2069840 is associated with lower plasma IL6 concentrations in survivors of myocardial infarction [48] and in patients with leprosy [49]. In addition, serum concentrations of IL6 in the immediate postoperative period were positively associated with the extent of tissue damage that occurred during surgery [50]. Preoperative administration of pentoxifylline, a phosphodiesterase inhibitor that inhibits the synthesis of IL6 [51], lowered serum IL6 concentrations and reduced analgesic consumption following cholecystectomy [52]. Our findings are consistent with these studies and suggest that the rare G allele of rs2069840, which was less frequent in the mild persistent breast pain group, is associated with decreased serum concentrations of IL6 which prevents the development of mild persistent breast pain following surgery.
CXCL8 is a chemokine and a prominent mediator of the innate immune response to inflammatory stimuli [53]. Our results suggest that the rare A allele of rs4073 decreases the risk for the development of mild persistent breast pain following breast cancer surgery. This SNP is located in the proximal promoter of CXCL8 and is known to be functional. The common T allele is associated with increased CXCL8 gene and protein expression in vitro [53, 54]. Increased levels of CXCL8 were associated with increased pain following disc herniation [55] and postherpetic neuralgia [56]. Our findings are consistent with these studies and suggest that the common T allele of rs4073, which was more frequent in the mild persistent breast pain group, is associated with increased serum concentrations of CXCL8 which promotes the development of mild persistent breast pain following surgery.
TNFα is a pleiotropic pro-inflammatory cytokine that promotes the production of other pro-inflammatory cytokines following nerve injury [57] and plays a prominent role in the development and maintenance of persistent pain [58]. Our results suggest that the rare T allele of rs1800610 decreases the risk for the development of mild persistent breast pain following breast cancer surgery. TNF rs1800610 is located in a non-coding region. The functional consequences of rs1800610 were investigated in vitro and during physiologic stimulation in patients with rheumatoid arthritis and healthy controls [59, 60]. In these studies, no differences were found in TNF precursor mRNA production between the rs1800610 alleles. Additional studies are needed to determine whether this SNP is in LD with other variations in the TNF locus that affect cytokine production or if tissue-specific influences of this SNP on TNF gene expression occur.
A gene containing polymorphisms associated with a specific phenotype may accrue epigenetic adaptation(s) that modulate gene expression. These epigenetic changes provide a dynamic mechanism to improve homeostasis in the setting of a sustained change in the environment. However, these attempts at homeostasis may be maladaptive when environmental contexts shift precipitously. Ongoing activation of inflammatory cells, and inadequate compensation by spinal inhibitory mechanisms, may promote the establishment of persistent pain following neuronal injury, at least in part, through changes in gene expression moderated by epigenetic processes. However, evidence of altered epigenetic processes associated with persistent pain is limited [61, 62]. In this study, the c.-350C and c.-344C sites of TNF had higher levels of methylation in the mild breast pain class compared to the no breast pain class. The effect sizes calculated for percentage methylation level at the TNF c.-350C and c.-344C sites were 0.63 and 0.66, respectively, by Cohen’s d which indicates a moderate to strong effect on pain group membership.
Regulation of TNF transcription occurs at transcription factor binding sites within the proximal TNF promoter [63–65]. The TNF proximal promoter contains putative binding sites for several transcription factors that alter TNFα production in response to lipopolysaccharide [65, 66]. The c.-350C site lies within the putative binding site of the specificity protein 1 (Sp1) transcription factor (i.e., 5′-CCGCCC-3′; c.-350C site underlined). The binding of Sp1 at this locus may be affected by methylation at c.-344C and c.-342C. Sp1 is a member of the specificity protein family of transcription factors that enhances or represses gene transcription in response to physiologic and pathologic stimuli [67]. For example, nitric oxide released in the course of an inflammatory response results in Sp1 binding to the TNF promoter and the initiation of TNF gene transcription in human leukocytes [68]. It is unclear how cytosine methylation at the Sp1 binding site within the TNF promoter may affect TNF gene expression. However, the observation that cytosine methylation within the putative Sp1 binding site impaired binding of nuclear proteins to the bromodomain containing 7 (BRD7) gene promoter in human nasopharyngeal carcinoma cells [69], suggests that methylation may be directly responsible for modulation of TNF gene expression at this site. Additional studies are needed to determine how methylation at these sites alters cytokine hemostasis and whether Sp1 binding occurs at these sites and is impacted by CpG methylation.
In this sample, our group previously identified associations between variations in two cytokine genes (i.e., IL1R1, IL13) and preoperative pain in the affected breast [11]. Presurgical pain in cancer patients may be associated with sensitization of peripheral nerves by inflammatory mediators released from cells within the tumor and/or nerve injury due to compression by the tumor [70, 71]. Subsequently, mechanical trauma during surgery produces significant tissue injury that releases inflammatory mediators and intense nociceptive activity in peripheral nociceptors. Three SNPs (i.e., IL1R2 rs11674595; IL4 rs2243248; IL13 rs1800925) and one haplotype (i.e., IL10 haplotype A8) among four genes predicted membership in the severe pain class [72]. In the current study, we evaluated associations between the mild breast pain class and the same cytokine genes. Three SNPs (i.e, IL6 rs2069840, CXCL8 rs4073, TNF rs1800610) in three different genes predicted membership in the mild pain class.
The different subsets of genes associated with these distinct pain phenotypes suggest different underlying mechanisms for these pain conditions. Strong evidence supports the involvement of immune cells in the development and maintenance of persistent pain [27–29]. Sequential production and release of pro- and anti-inflammatory cytokines orchestrate the inflammatory response following tissue injury. Pro-inflammatory cytokines induce their production through positive feedback and act synergistically to amplify inflammatory signals. The local release of pro-inflammatory cytokines must be balanced by an adequate anti-inflammatory response. Aberrant release of cytokines from immune cells may affect the cascade of events that are initiated by tissue injury, lead to alterations in gene expression and processing of afferent signals. Therefore, variations within cytokine genes may alter the balance between pro- and anti-inflammatory cytokine production and response which may ultimately predispose an individual to the development of persistent pain.
Study limitations must be acknowledged. First, no direct measurements of serum cytokines were done to provide additional data on the mechanisms that underlie the development of mild persistent pain. Second, future studies are needed to validate the IL6 and TNF SNP associations. Third, methylation was evaluated using peripheral leukocytes which may not reflect changes in the tissues at the site of injury. Finally, DNA methylation was the only epigenetic mechanism evaluated in this study. The expression of IL6, CXCL8, and TNF may be modulated by other epigenetic mechanisms and/or DNA methylation outside of the regions evaluated in this study.
In conclusion, findings from this study suggest a role for polymorphisms within the IL6, CXCL8, and TNF genes and changes in methylation in the TNF gene promoter in the development of mild persistent breast pain in women following breast cancer surgery. The genes, SNPs, and methylation sites identified in this study may help to identify individuals who are predisposed to the development of mild persistent postsurgical breast pain. Future studies are needed to confirm our findings and determine if these associations are seen in other persistent postsurgical pain syndromes.
Supplementary Material
Highlights.
Variations in IL6, CXCL8, and TNF are associated with the development and maintenance of mild persistent breast pain.
CpG methylation within the TNF promoter may be an additional mechanism through which TNF alters the risk for mild persistent breast pain after breast cancer surgery.
These genetic and epigenetic variations may help to identify patients who are at greater risk for mild levels of persistent breast pain following breast cancer surgery.
Acknowledgments
This study was funded by grants from the National Cancer Institute (NCI, CA107091 and CA118658). C. Miaskowski is an American Cancer Society Clinical Research Professor and is supported by a K05 award from the NCI (CA168960). K. Stephens was supported by an institutional training grant (T32 NR07088), a Sigma Theta Tau Research Award, and a F32 National Research Service Award (NR015728). This project was supported by NIH/NCRR UCSF-CTSI grant number UL RR024131. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health (NIH).
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of interest statement: The authors have no conflicts of interest to declare.
References
- 1.Andersen KG, Kehlet H. Persistent pain after breast cancer treatment: a critical review of risk factors and strategies for prevention. J Pain. 2011;12(7):725–46. doi: 10.1016/j.jpain.2010.12.005. [DOI] [PubMed] [Google Scholar]
- 2.Schreiber KL, Kehlet H, Belfer I, Edwards RR. Predicting, preventing and managing persistent pain after breast cancer surgery: the importance of psychosocial factors. Pain Manag. 2014;4(6):445–59. doi: 10.2217/pmt.14.33. [DOI] [PubMed] [Google Scholar]
- 3.Macrae WA. Chronic post-surgical pain: 10 years on. Br J Anaesth. 2008;101(1):77–86. doi: 10.1093/bja/aen099. [DOI] [PubMed] [Google Scholar]
- 4.von Hehn CA, Baron R, Woolf CJ. Deconstructing the neuropathic pain phenotype to reveal neural mechanisms. Neuron. 2012;73(4):638–52. doi: 10.1016/j.neuron.2012.02.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Scholz J, Woolf CJ. Can we conquer pain? Nat Neurosci. 2002;5(Suppl):1062–7. doi: 10.1038/nn942. [DOI] [PubMed] [Google Scholar]
- 6.McMahon SB, Cafferty WB, Marchand F. Immune and glial cell factors as pain mediators and modulators. Exp Neurol. 2005;192(2):444–62. doi: 10.1016/j.expneurol.2004.11.001. [DOI] [PubMed] [Google Scholar]
- 7.Marchand F, Perretti M, McMahon SB. Role of the immune system in chronic pain. Nat Rev Neurosci. 2005;6(7):521–32. doi: 10.1038/nrn1700. [DOI] [PubMed] [Google Scholar]
- 8.Gold MS. Spinal nerve ligation: what to blame for the pain and why. Pain. 2000;84(2–3):117–20. doi: 10.1016/s0304-3959(99)00309-7. [DOI] [PubMed] [Google Scholar]
- 9.Janig W, Levine JD, Michaelis M. Interactions of sympathetic and primary afferent neurons following nerve injury and tissue trauma. Prog Brain Res. 1996;113:161–84. doi: 10.1016/s0079-6123(08)61087-0. [DOI] [PubMed] [Google Scholar]
- 10.Harvey RJ, Depner UB, Wassle H, Ahmadi S, Heindl C, Reinold H, Smart TG, Harvey K, Schutz B, Abo-Salem OM, Zimmer A, Poisbeau P, Welzl H, Wolfer DP, Betz H, Zeilhofer HU, Muller U. GlyR alpha3: an essential target for spinal PGE2-mediated inflammatory pain sensitization. Science. 2004;304(5672):884–7. doi: 10.1126/science.1094925. [DOI] [PubMed] [Google Scholar]
- 11.McCann B, Miaskowski C, Koetters T, Baggott C, West C, Levine JD, Elboim C, Abrams G, Hamolsky D, Dunn L, Rugo H, Dodd M, Paul SM, Neuhaus J, Cooper B, Schmidt B, Langford D, Cataldo J, Aouizerat BE. Associations between pro- and anti-inflammatory cytokine genes and breast pain in women prior to breast cancer surgery. J Pain. 2012;13(5):425–37. doi: 10.1016/j.jpain.2011.02.358. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Reyes-Gibby CC, Shete S, Yennurajalingam S, Frazier M, Bruera E, Kurzrock R, Crane CH, Abbruzzese J, Evans D, Spitz MR. Genetic and nongenetic covariates of pain severity in patients with adenocarcinoma of the pancreas: assessing the influence of cytokine genes. J Pain Symptom Manage. 2009;38(6):894–902. doi: 10.1016/j.jpainsymman.2009.04.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Reyes-Gibby CC, Spitz M, Wu X, Merriman K, Etzel C, Bruera E, Kurzrock R, Shete S. Cytokine genes and pain severity in lung cancer: exploring the influence of TNF-alpha-308 G/A IL6-174G/C and IL8-251T/A. Cancer Epidemiol Biomarkers Prev. 2007;16(12):2745–51. doi: 10.1158/1055-9965.EPI-07-0651. [DOI] [PubMed] [Google Scholar]
- 14.Reyes-Gibby CC, Spitz MR, Yennurajalingam S, Swartz M, Gu J, Wu X, Bruera E, Shete S. Role of inflammation gene polymorphisms on pain severity in lung cancer patients. Cancer Epidemiol Biomarkers Prev. 2009;18(10):2636–42. doi: 10.1158/1055-9965.EPI-09-0426. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Robertson KD. DNA methylation and human disease. Nat Rev Genet. 2005;6(8):597–610. doi: 10.1038/nrg1655. [DOI] [PubMed] [Google Scholar]
- 16.Laird PW. The power and the promise of DNA methylation markers. Nat Rev Cancer. 2003;3(4):253–66. doi: 10.1038/nrc1045. [DOI] [PubMed] [Google Scholar]
- 17.Miaskowski C, Cooper B, Paul SM, West C, Langford D, Levine JD, Abrams G, Hamolsky D, Dunn L, Dodd M, Neuhaus J, Baggott C, Dhruva A, Schmidt B, Cataldo J, Merriman J, Aouizerat BE. Identification of patient subgroups and risk factors for persistent breast pain following breast cancer surgery. J Pain. 2012;13(12):1172–87. doi: 10.1016/j.jpain.2012.09.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Haythornthwaite JA, Raja SN, Fisher B, Frank SM, Brendler CB, Shir Y. Pain and quality of life following radical retropubic prostatectomy. J Urol. 1998;160(5):1761–4. [PubMed] [Google Scholar]
- 19.Miaskowski C, Dodd M, Paul SM, West C, Hamolsky D, Abrams G, Cooper BA, Elboim C, Neuhaus J, Schmidt BL, Smoot B, Aouizerat BE. Lymphatic and angiogenic candidate genes predict the development of secondary lymphedema following breast cancer surgery. PLoS One. 2013;8(4):e60164. doi: 10.1371/journal.pone.0060164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Tasmuth T, von Smitten K, Hietanen P, Kataja M, Kalso E. Pain and other symptoms after different treatment modalities of breast cancer. Ann Oncol. 1995;6(5):453–9. doi: 10.1093/oxfordjournals.annonc.a059215. [DOI] [PubMed] [Google Scholar]
- 21.Tasmuth T, von Smitten K, Kalso E. Pain and other symptoms during the first year after radical and conservative surgery for breast cancer. Br J Cancer. 1996;74(12):2024–31. doi: 10.1038/bjc.1996.671. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Jensen MP. The validity and reliability of pain measures in adults with cancer. J Pain. 2003;4(1):2–21. doi: 10.1054/jpai.2003.1. [DOI] [PubMed] [Google Scholar]
- 23.Dunn LB, Aouizerat BE, Langford DJ, Cooper BA, Dhruva A, Cataldo JK, Baggott CR, Merriman JD, Dodd M, West C, Paul SM, Miaskowski C. Cytokine gene variation is associated with depressive symptom trajectories in oncology patients and family caregivers. Eur J Oncol Nurs. 2013;17(3):346–53. doi: 10.1016/j.ejon.2012.10.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Jung T, Wickrama KAS. An introduction to latent class growth analysis and growth mixture modeling. Social and Personality Psychology Compass. 2008;2(1):302–317. [Google Scholar]
- 25.Nylund KL, Asparoutiov T, Muthen BO. Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Struct Equ Modeling. 2007;14(4):535–569. [Google Scholar]
- 26.Tofighi D, Enders CK. Identifying the correct number of classes in growth mixture models. Information Age Publishing; Charlotte, NC: 2008. [Google Scholar]
- 27.Austin PJ, Moalem-Taylor G. The neuro-immune balance in neuropathic pain: involvement of inflammatory immune cells, immune-like glial cells and cytokines. J Neuroimmunol. 2010;229(1–2):26–50. doi: 10.1016/j.jneuroim.2010.08.013. [DOI] [PubMed] [Google Scholar]
- 28.Calvo M, Dawes JM, Bennett DL. The role of the immune system in the generation of neuropathic pain. Lancet Neurol. 2012;11(7):629–42. doi: 10.1016/S1474-4422(12)70134-5. [DOI] [PubMed] [Google Scholar]
- 29.Scholz J, Woolf CJ. The neuropathic pain triad: neurons, immune cells and glia. Nat Neurosci. 2007;10(11):1361–8. doi: 10.1038/nn1992. [DOI] [PubMed] [Google Scholar]
- 30.Conde L, Vaquerizas JM, Dopazo H, Arbiza L, Reumers J, Rousseau F, Schymkowitz J, Dopazo J. PupaSuite: finding functional single nucleotide polymorphisms for large-scale genotyping purposes. Nucleic Acids Res. 2006;34(Web Server issue):W621–5. doi: 10.1093/nar/gkl071. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Zhang Q, Wang HY, Bhutani G, Liu X, Paessler M, Tobias JW, Baldwin D, Swaminathan K, Milone MC, Wasik MA. Lack of TNFalpha expression protects anaplastic lymphoma kinase-positive T-cell lymphoma (ALK+ TCL) cells from apoptosis. Proc Natl Acad Sci U S A. 2009;106(37):15843–8. doi: 10.1073/pnas.0907070106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Nile CJ, Read RC, Akil M, Duff GW, Wilson AG. Methylation status of a single CpG site in the IL6 promoter is related to IL6 messenger RNA levels and rheumatoid arthritis. Arthritis Rheum. 2008;58(9):2686–93. doi: 10.1002/art.23758. [DOI] [PubMed] [Google Scholar]
- 33.Lewin J, Schmitt AO, Adorjan P, Hildmann T, Piepenbrock C. Quantitative DNA methylation analysis based on four-dye trace data from direct sequencing of PCR amplificates. Bioinformatics. 2004;20(17):3005–12. doi: 10.1093/bioinformatics/bth346. [DOI] [PubMed] [Google Scholar]
- 34.Gabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, Higgins J, DeFelice M, Lochner A, Faggart M, Liu-Cordero SN, Rotimi C, Adeyemo A, Cooper R, Ward R, Lander ES, Daly MJ, Altshuler D. The structure of haplotype blocks in the human genome. Science. 2002;296(5576):2225–9. doi: 10.1126/science.1069424. [DOI] [PubMed] [Google Scholar]
- 35.Stephens M, Smith NJ, Donnelly P. A new statistical method for haplotype reconstruction from population data. Am J Hum Genet. 2001;68(4):978–89. doi: 10.1086/319501. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Bland JM, Altman DG. Applying the right statistics: analyses of measurement studies. Ultrasound Obstet Gynecol. 2003;22(1):85–93. doi: 10.1002/uog.122. [DOI] [PubMed] [Google Scholar]
- 37.Halder I, Shriver M, Thomas M, Fernandez JR, Frudakis T. A panel of ancestry informative markers for estimating individual biogeographical ancestry and admixture from four continents: utility and applications. Hum Mutat. 2008;29(5):648–58. doi: 10.1002/humu.20695. [DOI] [PubMed] [Google Scholar]
- 38.Hoggart CJ, Parra EJ, Shriver MD, Bonilla C, Kittles RA, Clayton DG, McKeigue PM. Control of confounding of genetic associations in stratified populations. Am J Hum Genet. 2003;72(6):1492–1504. doi: 10.1086/375613. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Tian C, Gregersen PK, Seldin MF. Accounting for ancestry: population substructure and genome-wide association studies. Hum Mol Genet. 2008;17(R2):R143–50. doi: 10.1093/hmg/ddn268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet. 2006;38(8):904–9. doi: 10.1038/ng1847. [DOI] [PubMed] [Google Scholar]
- 41.Illi J, Miaskowski C, Cooper B, Levine JD, Dunn L, West C, Dodd M, Dhruva A, Paul SM, Baggott C, Cataldo J, Langford D, Schmidt B, Aouizerat BE. Association between pro- and anti-inflammatory cytokine genes and a symptom cluster of pain, fatigue, sleep disturbance, and depression. Cytokine. 2012;58(3):437–47. doi: 10.1016/j.cyto.2012.02.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Hattersley AT, McCarthy MI. What makes a good genetic association study? Lancet. 2005;366(9493):1315–23. doi: 10.1016/S0140-6736(05)67531-9. [DOI] [PubMed] [Google Scholar]
- 43.Rothman KJ. No adjustments are needed for multiple comparisons. Epidemiology. 1990;1(1):43–6. [PubMed] [Google Scholar]
- 44.Langford DJ, Paul SM, West CM, Dunn LB, Levine JD, Kober KM, Dodd MJ, Miaskowski C, Aouizerat BE. Variations in potassium channel genes are associated with distinct trajectories of persistent breast pain after breast cancer surgery. Pain. 2015;156(3):371–80. doi: 10.1097/01.j.pain.0000460319.87643.11. [DOI] [PubMed] [Google Scholar]
- 45.Watkins LR, Milligan ED, Maier SF. Glial activation: a driving force for pathological pain. Trends Neurosci. 2001;24(8):450–5. doi: 10.1016/s0166-2236(00)01854-3. [DOI] [PubMed] [Google Scholar]
- 46.De Jongh RF, Vissers KC, Meert TF, Booij LH, De Deyne CS, Heylen RJ. The role of interleukin-6 in nociception and pain. Anesth Analg. 2003;96(4):1096–103. doi: 10.1213/01.ANE.0000055362.56604.78. [DOI] [PubMed] [Google Scholar]
- 47.Tilg H, Trehu E, Atkins MB, Dinarello CA, Mier JW. Interleukin-6 (IL-6) as an anti-inflammatory cytokine: induction of circulating IL-1 receptor antagonist and soluble tumor necrosis factor receptor p55. Blood. 1994;83(1):113–8. [PubMed] [Google Scholar]
- 48.Picciotto S, Forastiere F, Pistelli R, Koenig W, Lanki T, Ljungman P, Pitsavos C, Ruckerl R, Sunyer J, Peters A. Determinants of plasma interleukin-6 levels among survivors of myocardial infarction. Eur J Cardiovasc Prev Rehabil. 2008;15(6):631–8. doi: 10.1097/HJR.0b013e3283069d9a. [DOI] [PubMed] [Google Scholar]
- 49.Sousa AL, Fava VM, Sampaio LH, Martelli CM, Costa MB, Mira MT, Stefani MM. Genetic and immunological evidence implicates interleukin 6 as a susceptibility gene for leprosy type 2 reaction. J Infect Dis. 2012;205(9):1417–24. doi: 10.1093/infdis/jis208. [DOI] [PubMed] [Google Scholar]
- 50.Cruickshank AM, Fraser WD, Burns HJ, Van Damme J, Shenkin A. Response of serum interleukin-6 in patients undergoing elective surgery of varying severity. Clin Sci (London) 1990;79(2):161–5. doi: 10.1042/cs0790161. [DOI] [PubMed] [Google Scholar]
- 51.Liu J, Feng X, Yu M, Xie W, Zhao X, Li W, Guan R, Xu J. Pentoxifylline attenuates the development of hyperalgesia in a rat model of neuropathic pain. Neurosci Lett. 2007;412(3):268–72. doi: 10.1016/j.neulet.2006.11.022. [DOI] [PubMed] [Google Scholar]
- 52.Wordliczek J, Szczepanik AM, Banach M, Turchan J, Zembala M, Siedlar M, Przewlocki R, Serednicki W, Przewlocka B. The effect of pentoxifiline on post-injury hyperalgesia in rats and postoperative pain in patients. Life Sci. 2000;66(12):1155–64. doi: 10.1016/s0024-3205(00)00419-7. [DOI] [PubMed] [Google Scholar]
- 53.Ahn MH, Park BL, Lee SH, Park SW, Park JS, Kim DJ, Jang AS, Park JS, Shin HK, Uh ST, Kim YK, Kim YW, Han SK, Jung KS, Lee KY, Jeong SH, Park JW, Choi BW, Park IW, Chung MP, Shin HD, Song JW, Kim DS, Park CS, Shim YS. A promoter SNP rs4073T>A in the common allele of the interleukin 8 gene is associated with the development of idiopathic pulmonary fibrosis via the IL-8 protein enhancing mode. Resp Res. 2011;12:73. doi: 10.1186/1465-9921-12-73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Lee WP, Tai DI, Lan KH, Li AF, Hsu HC, Lin EJ, Lin YP, Sheu ML, Li CP, Chang FY, Chao Y, Yen SH, Lee SD. The-251T allele of the interleukin-8 promoter is associated with increased risk of gastric carcinoma featuring diffuse-type histopathology in Chinese population. Clin Cancer Res. 2005;11(18):6431–41. doi: 10.1158/1078-0432.CCR-05-0942. [DOI] [PubMed] [Google Scholar]
- 55.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(1):62–6. doi: 10.1007/s005860100306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Kotani N, Kudo R, Sakurai Y, Sawamura D, Sessler DI, Okada H, Nakayama H, Yamagata T, Yasujima M, Matsuki A. Cerebrospinal fluid interleukin 8 concentrations and the subsequent development of postherpetic neuralgia. Am J Med. 2004;116(5):318–24. doi: 10.1016/j.amjmed.2003.10.027. [DOI] [PubMed] [Google Scholar]
- 57.Shamash S, Reichert F, Rotshenker S. The cytokine network of Wallerian degeneration: tumor necrosis factor-alpha, interleukin-1alpha, and interleukin-1beta. J Neurosci. 2002;22(8):3052–60. doi: 10.1523/JNEUROSCI.22-08-03052.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58.Leung L, Cahill CM. TNF-alpha and neuropathic pain--a review. J Neuroinflammation. 2010;7:27. doi: 10.1186/1742-2094-7-27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59.Kaijzel EL, Bayley JP, van Krugten MV, Smith L, van de Linde P, Bakker AM, Breedveld FC, Huizinga TW, Verweij CL. Allele-specific quantification of tumor necrosis factor alpha (TNF) transcription and the role of promoter polymorphisms in rheumatoid arthritis patients and healthy individuals. Genes Immun. 2001;2(3):135–44. doi: 10.1038/sj.gene.6363747. [DOI] [PubMed] [Google Scholar]
- 60.Verweij CL. Tumour necrosis factor gene polymorphisms as severity markers in rheumatoid arthritis. Ann Rheum Dis. 1999;58(Suppl 1):I20–6. doi: 10.1136/ard.58.2008.i20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61.Denk F, McMahon SB. Chronic pain: emerging evidence for the involvement of epigenetics. Neuron. 2012;73(3):435–44. doi: 10.1016/j.neuron.2012.01.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62.Geranton SM. Targeting epigenetic mechanisms for pain relief. Curr Opin Pharmacol. 2012;12(1):35–41. doi: 10.1016/j.coph.2011.10.012. [DOI] [PubMed] [Google Scholar]
- 63.Falvo JV, Uglialoro AM, Brinkman BM, Merika M, Parekh BS, Tsai EY, King HC, Morielli AD, Peralta EG, Maniatis T, Thanos D, Goldfeld AE. Stimulus-specific assembly of enhancer complexes on the tumor necrosis factor alpha gene promoter. Mol Cell Biol. 2000;20(6):2239–47. doi: 10.1128/mcb.20.6.2239-2247.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 64.Rhoades KL, Golub SH, Economou JS. The regulation of the human tumor necrosis factor alpha promoter region in macrophage, T cell, and B cell lines. J Biol Chem. 1992;267(31):22102–7. [PubMed] [Google Scholar]
- 65.Yao J, Mackman N, Edgington TS, Fan ST. Lipopolysaccharide induction of the tumor necrosis factor-alpha promoter in human monocytic cells. Regulation by Egr-1, c-Jun, and NF-kappaB transcription factors. J Biol Chem. 1997;272(28):17795–801. doi: 10.1074/jbc.272.28.17795. [DOI] [PubMed] [Google Scholar]
- 66.Tsai EY, Falvo JV, Tsytsykova AV, Barczak AK, Reimold AM, Glimcher LH, Fenton MJ, Gordon DC, Dunn IF, Goldfeld AE. A lipopolysaccharide-specific enhancer complex involving Ets, Elk-1, Sp1, and CREB binding protein and p300 is recruited to the tumor necrosis factor alpha promoter in vivo. Mol Cell Biol. 2000;20(16):6084–94. doi: 10.1128/mcb.20.16.6084-6094.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 67.Suske G. The Sp-family of transcription factors. Gene. 1999;238(2):291–300. doi: 10.1016/s0378-1119(99)00357-1. [DOI] [PubMed] [Google Scholar]
- 68.Wang S, Wang W, Wesley RA, Danner RL. A Sp1 binding site of the tumor necrosis factor alpha promoter functions as a nitric oxide response element. J Biol Chem. 1999;274(47):33190–3. doi: 10.1074/jbc.274.47.33190. [DOI] [PubMed] [Google Scholar]
- 69.Liu H, Zhang L, Niu Z, Zhou M, Peng C, Li X, Deng T, Shi L, Tan Y, Li G. Promoter methylation inhibits BRD7 expression in human nasopharyngeal carcinoma cells. BMC Cancer. 2008;8:253. doi: 10.1186/1471-2407-8-253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 70.Cleeland CS, Bennett GJ, Dantzer R, Dougherty PM, Dunn AJ, Meyers CA, Miller AH, Payne R, Reuben JM, Wang XS, Lee BN. Are the symptoms of cancer and cancer treatment due to a shared biologic mechanism? A cytokine-immunologic model of cancer symptoms. Cancer. 2003;97(11):2919–25. doi: 10.1002/cncr.11382. [DOI] [PubMed] [Google Scholar]
- 71.Mantyh PW, Clohisy DR, Koltzenburg M, Hunt SP. Molecular mechanisms of cancer pain. Nat Rev Cancer. 2002;2(3):201–9. doi: 10.1038/nrc747. [DOI] [PubMed] [Google Scholar]
- 72.Stephens K, Cooper BA, West C, Paul SM, Baggott CR, Merriman JD, Dhruva A, Kober KM, Langford DJ, Leutwyler H, Luce JA, Schmidt BL, Abrams GM, Elboim C, Hamolsky D, Levine JD, Miaskowski C, Aouizerat BE. Associations between cytokine gene variations and severe persistent breast pain in women following breast cancer surgery. J Pain. 2014;15(2):169–80. doi: 10.1016/j.jpain.2013.09.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.