Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Jun 27.
Published in final edited form as: Biol Res Nurs. 2014 Oct 10;17(3):237–247. doi: 10.1177/1099800414550394

Associations Between Cytokine Genes and a Symptom Cluster of Pain, Fatigue, Sleep Disturbance, and Depression in Patients Prior to Breast Cancer Surgery

Sy-Huey Doong 1, Anand Dhruva 2, Laura B Dunn 2, Claudia West 1, Steven M Paul 1, Bruce A Cooper 1, Charles Elboim 3, Gary Abrams 2, John D Merriman 1, Dale J Langford 1, Heather Leutwyler 1, Christina Baggott 1, Kord Kober 1, Bradley E Aouizerat 1,4,*, Christine Miaskowski 1,*
PMCID: PMC5486211  NIHMSID: NIHMS864791  PMID: 25304131

Abstract

Pain, fatigue, sleep disturbance, and depression are common and frequently co-occurring symptoms in oncology patients. This symptom cluster is often attributed to the release of pro-inflammatory cytokines. The purposes of this study were to determine whether distinct latent classes of patients with breast cancer (n = 398) could be identified based on their experience with this symptom cluster, whether patients in these latent classes differed on demographic and clinical characteristics, and whether variations in cytokine genes were associated with latent class membership. Three distinct latent classes were identified: All Low (61.0%), Low Pain and High Fatigue (31.6%), All High (7.1%). Compared to patients in the All Low class, patients in the ALL High class were significantly younger, had less education, were more likely to be non-White, had a lower annual income, were more likely to live alone, had a lower functional status, had a higher comorbidity score, and had more advanced disease. Significant associations were found between interleukin (IL) 6 rs2069845, IL13 rs1295686, and tumor necrosis factor alpha rs18800610 and latent class membership. Findings suggest that variations in pro- and anti-inflammatory cytokine genes are associated with this symptom cluster in breast cancer patients.

Keywords: pain, fatigue, sleep disturbance, depression, symptom cluster, cytokines, candidate genes, interleukin, tumor necrosis factor alpha

Pain, fatigue, sleep disturbance, and depression are common and frequently co-occurring symptoms reported by oncology patients. Recent studies have revealed that these symptoms are related to each other and can form a symptom cluster (Barsevick, 2007; Kim, Barsevick, Beck, & Dudley, 2012). This symptom cluster is associated with significant decrements in oncology patients’ functional status and quality of life (Miaskowski et al., 2006; Pud et al., 2008). Evidence suggests that the model of cytokine-induced sickness behavior may be one mechanism to explain the occurrence of this symptom cluster in oncology patients (Gilbertson-White, Aouizerat, & Miaskowski, 2011).

Sickness behavior refers to the physiologic changes and associated behaviors that develop in individuals with an infection (Dantzer, 2001; Dantzer & Kelley, 2007) or through the administration of agents that trigger an inflammatory response (Eisenberger et al., 2010; Maier & Watkins, 2003). Sickness behavior occurs as a result of the release of pro-inflammatory cytokines from the central nervous system. A growing body of evidence suggests that some of the most common symptoms reported by oncology patients are associated with changes in the levels of pro- and anti-inflammatory cytokines (Miaskowski & Aouizerat, 2012; Seruga, Zhang, Bernstein, & Tannock, 2008).

For example, in recent studies of oncology patients, sleep disturbance and increased fatigue were associated with changes in a number of serum interleukins (Collado-Hidalgo, Bower, Ganz, Irwin, & Cole, 2008; Liu et al., 2012; Saligan & Kim, 2012; Schubert, Hong, Natarajan, Mills, & Dimsdale, 2007). Recent work from our research team identified associations between a polymorphism in tumor necrosis factor alpha (TNFA) and increased levels of fatigue and sleep disturbance in oncology patients and their family caregivers (Aouizerat et al., 2009). In addition, studies have shown that higher levels of interleukin 6 (IL6) and TNFα (Alesci et al., 2005; Dowlati et al., 2010; Sukoff Rizzo et al., 2012) and polymorphisms in interleukin 1 receptor 2 (IL1R2), interleukin 10 (IL10), and TNFA (Dunn et al., 2013) are associated with depressive symptoms. Finally, in another recent study, elevated plasma levels of pro-inflammatory cytokines, including IL1-β, IL6, and TNFα, were associated with the development and maintenance of neuropathic pain (Wang, Lehky, Brell, & Dorsey, 2012). Collectively, these findings suggest that variations in both pro- and anti-inflammatory cytokine genes may influence the symptom experience of oncology patients.

In a recent study of oncology patients and their family caregivers, we used latent class analysis (LCA) to identify subgroups of participants with distinct symptom profiles and evaluated for associations between polymorphisms in a number of cytokine genes and high levels of self-reported pain, fatigue, sleep disturbance, and depression. We found that individuals who were homozygous for the rare allele rs2243248 in IL4 were significantly more likely to be classified into the “all high” symptom group (Illi et al., 2012). However, our heterogeneous and relatively small sample limited our ability to identify additional genetic associations. In the present study, we sought to replicate this finding in a larger, different, and more homogeneous sample of patients with breast cancer.

While surgery can trigger the release of cytokines, several studies found that fatigue, pain, sleep disturbance, and depression can occur prior to and may trigger higher levels of cancer-related symptoms after breast cancer surgery (De Vries, Van der Steeg, & Roukema, 2009; Wright et al., 2009). For example, in one study, patients with chronic preoperative pain were three times more likely to report postoperative pain compared to patients without preoperative pain (Sheridan et al., 2012). In addition, patients who developed chronic pain after breast cancer surgery reported higher preoperative levels of depression and anxiety (Miaskowski, Cooper, Paul, et al., 2012).

To our knowledge, no studies have evaluated associations between the symptom cluster of pain, fatigue, sleep disturbance, and depression and variations in cytokine genes in patients prior to breast cancer surgery. Therefore, the purposes of this study were to determine whether distinct latent classes of patients with breast cancer could be identified based on their experience with the symptom cluster of pain, fatigue, sleep disturbance, and depression; determine whether patients in these latent classes differed on demographic and clinical characteristics; and determine whether genetic variations in pro- and anti- inflammatory cytokines were associated with latent class membership.

Methods

Patients and Settings

More detailed information on this study is published elsewhere (Miaskowski et al., 2013). In brief, patients were recruited from breast care centers located in a comprehensive cancer center, two public hospitals, and four community practices. A patient was eligible to participate if she was an adult woman (≥18 years) who would undergo breast cancer surgery on one breast; was able to read, write, and understand English; agreed to participate; and gave written informed consent. Patients were excluded if they were having breast cancer surgery on both breasts and/or had distant metastasis at the time of diagnosis. A total of 516 patients were approached to participate, 410 were enrolled in the study (response rate 79.5%), and 398 completed the study questionnaires.

Instruments

The demographic questionnaire obtained information on age, education, ethnicity, marital status, employment status, living situation, and financial status. The Karnofsky Performance Status (KPS) scale was used to evaluate patient’s functional status (Karnofsky, 1977). The Self-Administered Comorbidity Questionnaire (SCQ) was used to evaluate comorbidity with 13 common medical conditions. The SCQ has well-established validity and reliability and has been used in studies of patients with a variety of chronic conditions (Sangha, Stucki, Liang, Fossel, & Katz, 2003).

Patients were asked to rate the intensity of their average and worst pain, using a 0 (no pain) to 10 (worst imaginable pain) numeric rating scale (NRS). The NRS is a valid and reliable measure of pain intensity (Jensen, 2003).

The Center for Epidemiologic Studies-Depression (CES-D) scale was used to evaluate depressive symptoms. CES-D scores can range from 0 to 60, with scores of ≥16 indicating the need for individuals to seek clinical evaluation for major depression (Radloff, 1977). Cronbach’s alpha for the CES-D was .90.

The General Sleep Disturbance Scale (GSDS) was used to evaluate sleep disturbance. The GSDS total score can range from 0 (no disturbance) to 147 (extreme sleep disturbance). A GSDS total score of ≥ 43 indicates a significant level of sleep disturbance (Fletcher et al., 2008; Miaskowski, et al., 2006). Cronbach’s alpha for the GSDS total score was .86.

The Lee Fatigue Scale (LFS) was used to evaluate physical fatigue and energy (Lee, Hicks, & Nino-Murcia, 1991). Total fatigue and energy scores are calculated as the mean of the 13 fatigue items and the 5 energy items, respectively, with higher scores indicating greater fatigue severity and higher levels of energy. A cutoff score of ≥ 4.4 indicates high levels of fatigue. A cutoff score of ≥ 4.8 indicates low levels of energy (Dhruva et al., 2010). Cronbach’s alphas for fatigue and energy were .96 and .93, respectively.

Study Procedures

The Committee on Human Research at the University of California, San Francisco, and the Institutional Review Boards at each of the study sites approved the study. During the patient’s preoperative visit, a clinician explained the study to the patient, determined her willingness to participate, and introduced the patient to the research nurse. The research nurse determined eligibility and obtained written informed consent prior to surgery. After providing written informed consent, patients completed the questionnaires and a blood sample was obtained.

Methods of Analysis for Phenotypic Data

Descriptive statistics and frequency distributions were generated for the sample characteristics and symptom data. All calculations used actual values. No adjustments were made for missing data. Therefore, the cohort for each analysis was dependent on the largest set of available data across groups. A p-value of < .05 is considered statistically significant.

Latent class analysis (LCA) was used to identify subgroups of patients with similar experiences with the symptom cluster of pain, fatigue, sleep disturbance, and depression (i.e., latent classes; Vermunt & Magdison, 2002). The final number of latent classes was identified by evaluating the Bayesian information criterion (BIC), the parametric bootstrapped likelihood ratio test (BLRT), and entropy. With this analysis, the model that fits the data best has the lowest BIC and/or BLRT, which indicates that the estimated model is a better fit than the model with one fewer class (Nylund, Asparouhov, & Muthen, 2007). In addition, better-fitting models should produce higher entropy values (Celeux & Soromenho, 1996). Finally, well-fitting models “make sense” conceptually, and the estimated classes differ as might be expected on variables not used in the generation of the model (Nylund et al., 2007).

Latent class models often use categorical variables (Lanza, Flaherty, & Collins, 2003). As in the present study, when continuous variables are analyzed (i.e., pain, fatigue, sleep disturbance, and depression scores), LCA is called latent class profile analysis (LCPA). However, one of the continuous variables in the present study, namely “worst pain,” which was reported on a 0 to 10 NRS, had a large number of 0s because a number of the patients did not report pain prior to surgery. We accommodated this large number of 0s modeling worst pain as a “two-part” variable. In this type of model, the variable is examined with one part representing the difference between those who reported no pain compared to those who reported any pain and with the second part differentiating among those who reported any pain on the remaining portion of the NRS (i.e., the 1 to 10 part of the NRS; L. K. Muthen & Muthen, 1998–2010b).

The LCPA was performed using Mplus™ version 6 (L. K. Muthen & Muthen, 1998–2010a). Estimation was carried out with robust maximum-likelihood (MLR) and the expectation-maximization (EM) algorithm (B. Muthen & Shedden, 1999). Due to the inclusion of a categorical variable (i.e., the binary variable for the occurrence of pain versus no pain), Gauss-Hermite adaptive numeric integration with 20 integration points was employed. Subsequent analyses of differences among the identified classes were carried out with the Statistical Package for the Social Sciences (SPSS) version 19 for Windows™ (SPSS, 2012).

Methods of Analysis for Genomic Data

Gene selection

Cytokines and their receptors are classes of polypeptides that mediate inflammatory processes (Verri et al., 2006). Pro-inflammatory cytokines promote systemic inflammation and include interferon gamma (IFNG) 1, IFNG receptor 1 (IFNGR1), IL1R1, IL2, IL8, IL17A, nuclear factor kappa beta 1 (NFKB1), NFKB2, and TNFα. Anti-inflammatory cytokines suppress the activity of pro-inflammatory cytokines and include IL1R2, IL4, IL10, and IL13. Of note, IFNG1, IL1B, and IL6 perform pro- and anti-inflammatory functions (Seruga et al., 2008; Verri et al., 2006). The specific single nucleotide polymorphisms for each of the relevant genes are listed in Supplemental Table 1.

Blood collection and genotyping

Of the 398 patients who completed the baseline assessment, 302 provided a blood sample. Genomic deoxyribonucleic acid (DNA) was extracted from archived buffy coats using the PUREGene DNA Isolation System (Invitrogen, Carlsbad, CA). Study personnel performing genotyping were blinded to LCPA status, and positive and negative controls were included. Samples were genotyped using the GoldenGate genotyping platform (Illumina, San Diego, CA) and processed using GenomeStudio (Illumina, San Diego, CA).

Single-nucleotide polymorphism selection

A combination of tagging single-nucleotide polymorphisms (SNPs) and literature-driven SNPs were selected for analysis. We required that the tagging SNPs were common (i.e., estimated to have a minor allele frequency ≥ .05) in public databases. In order to ensure robust genetic association analyses, quality-control filtering of SNPs was performed. SNPs with call rates of < 95% or Hardy-Weinberg p-values of < .001 were excluded.

As shown in supplemental Table 1, a total of 82 SNPs among the 15 candidate genes passed all quality-control filters and were included in the genetic association analyses. Potential functional roles of SNPs associated with the symptoms of pain, fatigue, sleep disturbance, and depression were examined using PUPASuite 2.0 (Conde et al., 2006).

Statistical Analyses

Allele and genotype frequencies were determined by gene counting. Hardy-Weinberg equilibrium was assessed by the Chi-square or Fisher Exact tests. Measures of linkage disequilibrium ([LD] i.e., D’ and r2) were computed from the patients’ genotypes with Haploview 4.2. LD-based haplotype block definition was based on D’ confidence interval (Gabriel et al., 2002).

Haplotypes were constructed using the program PHASE version 2.1 (Stephens, Smith, & Donnelly, 2001). Only haplotypes that were inferred with probability estimates of ≥.85 across five iterations were retained for downstream analyses. Haplotypes were evaluated assuming a dosage model (i.e., analogous to the additive model).

Ancestry informative markers (AIMS) were used to minimize confounding due to population stratification (Halder, Shriver, Thomas, Fernandez, & Frudakis, 2008; Hoggart et al., 2003; Tian, Gregersen, & Seldin, 2008). Homogeneity in ancestry among patients was verified by principal component analysis (Price et al., 2006) using Helix Tree (Golden Helix, Bozeman, MT; data not shown). We included 106 AIMs in the analysis. The first three principal components (PCs) were selected to adjust for potential confounding due to population substructure (i.e., race/ethnicity) by including the three covariates in all regression models.

For association tests, three genetic models were assessed for each SNP: additive, dominant, and recessive. Barring trivial improvements (i.e., delta < 10%), the genetic model that best fit the data, by maximizing the significance of the p-value, was selected for each SNP. Logistic regression analysis, which controlled for significant covariates, as well as genomic estimates of and self-reported race/ethnicity were used to evaluate the relationship between genotype and LCPA group membership. A backwards stepwise approach was used to create the most parsimonious model. Genetic model fit and both unadjusted and covariate-adjusted odds ratios were estimated using STATA version 9 (StataCorp, 2005).

As was done in our previous studies (Illi et al., 2012; McCann et al., 2012; Miaskowski, Cooper, Dhruva, et al., 2012), based on recommendations in the literature (Hattersley & McCarthy, 2005; Rothman, 1990), the implementation of rigorous quality controls for genomic data, the non-independence of SNPs/haplotypes in LD, and the exploratory nature of the analyses, adjustments were not made for multiple testing. In addition, significant SNPs identified in the bivariate analyses were evaluated further using regression analyses that controlled for differences in phenotypic characteristics, potential confounding due to population stratification, and variation in other SNPs/haplotypes within the same gene. Only those SNPs that remained significant were included in the final presentation of the results. Therefore, the significant independent associations reported are unlikely to be due solely to chance. Unadjusted (bivariate) associations are reported for all SNPs passing quality-control criteria in Supplementary Table 1 to allow for subsequent comparisons and meta-analyses.

Results

Latent Class Analysis

Using LCPA, we identified three distinct classes of patients based on their experiences with the symptoms of pain, fatigue, sleep disturbance, and depression. The fit indices for the candidate models are shown in Table 1. The three-class solution was selected because its BIC was lower than the BIC for both the two- and four-class solutions. As summarized in Table 2, the largest percentage of patients (61.0%) was classified in the “All Low” class and had mean scores for all symptoms that were below the clinically meaningful cutoff scores. A second group, which comprised 7.1% of the patients, was classified as the “All High” class. All four of the symptom scores were above the clinically meaningful cutoff scores for members of this class. The third class, comprised of 31.6% of the sample, was classified as the “Low Pain and High Fatigue” class.

Table 1.

Fit indices for two- through four- latent class solutions

Model LL AIC BIC BLRT Entropy
2 Class −2487.97 5017.95 5101.40 69.83† .84
3 Class −2469.03 4992.06 5099.35 37.89† .75
4 Class −2455.45 4976.89 5108.03 27.17* .81
Open in a new tab

Note. AIC = Akaike information criterion; BIC = Bayesian information criterion; BLRT = parametric bootstrapped likelihood ratio test for the K vs. K-1 model; LL = log-likelihood.

*

p < .05;

†

p < .0001.

Table 2.

Symptom severity scores at enrollment by latent classa

Symptom score All Low (1)
n = 241 (61.3%)		 Low Pain and High Fatigue (2)
n = 124 (31.6%)		 All High (3)
n = 28 (7.1%)		 Omnibus test p-value and Sidak post hoc contrasts
Global Sleep Disturbance Scale 40.1 (18.4) 59.8 (19.9) 67.6 (19.7) p < .00005
1 < 2 and 3
Center for Epidemiological Studies–Depression Scale 10.7 (7.7) 16.5 (9.0) 28.0 (12.6) p < .00005
1 < 2 < 3
Worst Pain Severity (0–10)b 2.4 (1.1) 2.9 (1.2) 7.6 (1.6) p < .00005
1 and 2 < 3
Lee Fatigue Scale 1.5 (1.2) 5.8 (1.3) 4.6 (1.7) p < .00005
2 > 3 > 1
Open in a new tab

Note. Values reported for symptom scores are mean (SD).

a

Predicted classes based on the most likely latent class membership.

b

Means and tests are only for patients who reported pain: Class 1 (n = 20), Class 2 (n = 31), Class 3 (n = 48).

Differences in Demographic and Clinical Characteristics Among the Three Latent Classes

As shown in Table 3, we found significant differences among the three latent classes in age, years of education, ethnicity, living arrangements, annual income, KPS score, estrogen and progesterone receptor status, and receipt of neoadjuvant chemotherapy (CTX).

Table 3.

Demographic and clinical characteristics by latent class.

Characteristic All Low (1) n = 241 (61%) Low Pain and High Fatigue (2) n = 124 (31.6%) All High (3) n = 28 (7.1%) Comparison Among 1,2, and 3 Comparison Between 1 & 3
Mean (SD) Statistic and p-value
Age (years) 56.9 (12.1) 52.0 (9.4) 50.7 (9.4) F(2,390) = 9.72, p < .0001
1 > 2 and 3		 t = 2.60; p = .01
Education (years) 15.7 (2.7) 16.0 (2.6) 14.0 (2.6) F(2,385) = 6.5, p = .002
1 and 2 > 3		 t = 3.18; p = .002
Number biopsies in past year 1.5 (0.7) 1.5 (0.9) 1.7 (1.1) NS NS
Karnofsky Performance Status score 95.6 (8.2) 90.4 (11.8) 84.4 (12.5) F(2,284) = 23.0, p < .0001
1 > 2 >3		 t = 4.53; p < .0001
Self-administered Comorbidity Questionnaire score 4.1 (2.7) 4.5 (3.1) 5.1 (2.7) NS t = - 2.00 ; p = .047
Body mass index (kg/m2) 26.5 (6.1) 27.5 (6.2) 26.9 (6.7) NS NS
% (N)
Lives alone 22.9 (55) 23.0 (28) 46.2 (12) Χ2 = 7.08, p = .029 FE; p = .016
Married/partnered 42.3 (102) 38.2 (47) 60.0 (15) NS NS
Ethnicity Χ2 = 24.3, p = .<.0001 X2 = 17.55; p = .001
 White 65.4 (157) 69.9 (86) 32.1 (9)
 Black 8.8 (21) 7.3 (9) 28.6 (8)
 Asian/Pacific Islander 13.3 (32) 8.1 (10) 28.6 (8)
 Hispanic/Mixed 12.5 (30) 14.6 (18) 10.7 (3)
Income Χ2 = 19.5, p = .001 X2 = 17.89; p = < .0001
 < $30,000 17.7 (35) 21.9 (23) 57.1 (12)
 $30,000–$99,000 43.9 (87) 35.2 (37) 28.6 (6)
 ≥ $100,000 38.4 (76) 42.9 (45) 14.3 (3)
Stage of disease NS MW = .018
 Stage 0 19.9 (48) 16.9 (21) 14.3 (4)
 Stage I 38.2 (92) 40.3 (50) 21.4 (6)
 Stage IIA and IIB 35.3 (85) 33.9 (42) 42.9 (12)
 Stage IIIA, IIIB, IIIC, and IV 6.6 (16) 8.9 (11) 21.4 (6)
Estrogen receptor positive 82.1 (197) 69.4 (86) 67.9 (19) Χ2 = 8.9, p = .012 NS
Progesterone receptor positive 75.4 (181) 61.3 (76) 64.3 (18) Χ2 = 8.3, p = .016 NS
HER2/neu receptor positive 13.1 (28) 21.9 (25) 18.5 (5) NS NS
Gone through menopause 65.8 (158) 60.2 (118) 64.0 (16) NS NS
Received neoadjuvant therapy 16.3 (39) 27.4 (34) 21.4 (6) Χ2 = 6.4, p = .041 NS
Received hormone replacement therapy prior to cancer diagnosis 17.5 (42) 17.1 (21) 10.7 (3) NS NS
Open in a new tab

Note. FE = Fisher’s exact test; HER2 = human epidermal growth factor receptor 2; MW = Mann Whitney U; NS = not significant; SD = standard deviation.

Differences in Demographic and Clinical Characteristics Between the All Low and the All High Latent Classes

Because the subsequent genomic analyses were done using an extreme phenotype approach (Li, Lewinger, Gauderman, Murcray, & Conti, 2011), we have summarized differences in demographic and clinical characteristics between the All Low and the All High latent classes in Table 3. Compared to the All Low class, patients in the All High class were significantly younger, had fewer years of education, and more likely to be non-White, have a lower annual household income, and live alone. In addition, compared to the All Low class, patients in the All High class had a lower functional status score, a higher comorbidity score, and a more advanced stage of disease at the time of diagnosis.

Candidate Gene Analyses for the Two Latent Classes

As summarized in Supplementary Table 1, the genotype frequency was significantly different between the two latent classes for nine SNPs and two haplotypes spanning seven genes (i.e., IL1R1, IL6, IL13, IL17A, NFKB1, NFKB2, TNFA). One haplotype (HapA2, p= .010) was identified in IL1R1. For the two SNPs identified in IL6 (rs1554606, rs2069845), a recessive model (both p = .044) fit the data best. Two SNPs (rs1295686, rs20541) and one haplotype (HapA1, p < .0001) were identified in IL13. For these two SNPs, a dominant model fit the data best (p < .0001, p = .041, respectively). For the SNP in IL17A (rs2275913), a dominant model fit the data best (p = .048). For the two SNPs identified in NFKB1 (rs4648110, rs4648141), an additive model fit the data best (p = .009, p = .031, respectively). For the one SNP identified in NFKB2 (rs1056890), a dominant model fit the data best (p = .026). For the one SNP identified in TNFA (rs1800610), a dominant model fit the data best (p = .039).

Regression Analyses of IL1R1, IL6, IL13, IL17A, NFKB1, NFKB2, and TNFA Genotypes and Haplotypes and Latent Class Membership

In order to better estimate the magnitude (i.e., odds ratio [OR]) and precision (95% confidence interval [CI]) of the association between genotype and latent class membership, multivariate logistic regression models were fit. In addition to genotype, the phenotypic variables included in the regression models were genomic estimates of and self-reported race/ethnicity (i.e., White, Black, Asian, Hispanic/mixed ethnic background/other), age, KPS score, and living arrangements.

The only genetic associations that remained significant in the multivariate regression analyses were for IL6 rs2069845, IL13 rs1295686, and TNFA rs1800610 (Table 4). In the regression analysis for IL6 rs2069845, carrying two doses of the rare G allele (i.e., AA+AG versus GG) was associated with a 14-fold increase in the odds of being in the All High latent class. Of note, IL6 rs1554606 and IL6 rs2069845 were completely collinear (i.e., mutual surrogates). Therefore, only one of the two SNPs (i.e., rs2069845) was selected to represent these two SNPs.

Table 4.

Multiple Logistic Regression Analyses for IL6, IL13, and TNFA Genotypes Comparing the Latent Classes All Low vs All High (n = 196).

Predictor OR SE 95% CI Z p-value
IL6 genotype 14.22 15.125 1.768, 114.378 2.50 0.013
Age 0.72 0.131 0.505, 1.030 −1.80 0.072
KPS score 0.36 0.120 0.186, 0.689 −3.07 0.002
Lives alone 9.12 7.567 1.793, 46.375 2.66 0.008
Overall model fit: χ2 = 36.05, p =.0001, R2 = 0.3768
IL13 genotype 28.53 38.358 2.045, 397.932 2.49 0.013
Age 0.68 0.148 0.444, 1.040 −1.78 0.075
KPS score 0.44 0.135 0.242, 0.805 −2.67 0.008
Lives alone 10.89 9.960 1.815, 65.376 2.61 0.009
Overall model fit: χ2 = 40.32, p <.0001, R2 = 0.4215
TNFA genotype 4.97 3.871 1.079, 22.879 2.06 0.040
Age 0.71 0.130 0.500, 1.026 −1.84 0.065
KPS score 0.45 0.133 0.253, 0.806 −2.69 0.007
Lives alone 10.77 8.864 2.145, 54.050 2.89 0.004
Overall model fit: χ2 = 33.50, p =.0002, R2 = 0.3502
Open in a new tab

Note. For each model, the first three principal components identified from the analysis of ancestry informative markers as well as self-report race/ethnicity (i.e., White, Black, Asian/Pacific Islander, Hispanic/mixed ethnic background/other) 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 rs2069845: AA + AG versus GG; IL13 rs1295686: GG versus GA + AA; TNFA rs1800610: CC versus CT + TT), age (in 5-year increments), KPS score (in 10-point increments), and whether the patient lives alone (No/Yes). CI =confidence interval; KPS = Karnofsky Performance Status; IL = interleukin; TNFA = tumor necrosis factor alpha.

In the regression analysis for IL13 rs1295686, carrying one or two doses of the rare A allele (i.e., GG versus GA+AA) was associated with a 29-fold increase in the odds of being in the All High latent class. In the regression analysis for TNFA rs1800610, carrying one or two doses of the rare T allele (i.e., CC versus CT+TT) was associated with a 5-fold increase in the odds of being in the All High latent class.

Discussion

To the best of our knowledge, this study is the first to identify distinct subgroups of women prior to breast cancer surgery based on their experience with the symptom cluster of pain, fatigue, sleep disturbance, and depression. The LCPA identified three relatively distinct latent classes: patients who reported low levels of all four symptoms (i.e., All Low, 61%), those who reported high levels of all four symptoms (i.e., All High, 7.1%), and those who reported low levels of pain and high levels of fatigue (31.6%). While our previous study of oncology patients and their family caregivers identified three distinct latent classes (Illi et al., 2012), only the All High classes were similar to each other in symptom severity scores and constituted a similar percentage of the study participants across studies (i.e., 12.3% in previous study versus 7.1% in the current study). While some of the differences in the symptom-severity scores associated with latent classes may be attributable to differences in sample characteristics, two characteristics were associated with membership in the All High class in both studies: compared to participants in the other classes, participants in the All High class were younger and had lower KPS scores. An emerging and consistent finding in our research (Dodd, Cho, Cooper, & Miaskowski, 2010; Illi et al., 2012; Miaskowski, Cooper, Dhruva et al., 2012) and that of others (Chen & Tseng, 2006; Cheung, Le, Gagliese, & Zimmermann, 2011) is that younger age and poorer functional status are associated with a higher symptom burden.

Three SNPs, among three cytokine genes (i.e., IL6, IL13, TNFA), were associated with latent class membership. Taken together, these findings support our hypothesis that cytokine genes are involved in the symptom cluster of pain, fatigue, sleep disturbance, and depression that mimics sickness behavior in animals and humans (Dantzer & Kelley, 2007; Tizard, 2008).

In the present study, carriers who were homozygous for the rare allele in IL6 rs2069845 had increased odds of belonging to the All High class. While this SNP is located in the intronic region of the IL6 gene, it resides in region that is a likely binding site for two ribosomal binding proteins (RBPs; UCSC Genome Bioinformatics Website at genome.ucsc.edu). In addition, recent findings suggest that it plays a role in the modulation of inflammatory responses. In one study (Sousa et al., 2012), carriers of the rare allele were at higher risk for the occurrence of leprosy Type 2 reactions. These Type 2 reactions presented as disseminated painful erythematous tender nodules, with or without neuritis, usually followed by systemic symptoms (e.g., fever, malaise) and were associated with elevated levels of TNFα. In another study (Tabassum et al., 2012), the minor allele was associated with an increased risk for childhood obesity. Interestingly, a systemic inflammatory response appears to occur in children with obesity that is related to elevated levels of inflammatory cytokines being produced by excessive amounts of adipose tissue (Schwarzenberg & Sinaiko, 2006). In a study of patients with myocardial infarction (Ljungman et al., 2009), the rare allele was associated with increased circulating levels of IL-6. Finally, findings from our research group and others suggest that additional SNPs in IL6 are associated with the symptoms of fatigue (Miaskowski et al., 2010), sleep disturbance (Miaskowski, Cooper, Dhruva, et al., 2012; Miaskowski et al., 2010), and depression (Dowlati et al., 2010; Sukoff Rizzo et al., 2012). Taken together, these findings suggest that variations in the IL6 gene contribute to inflammatory responses and symptoms associated with sickness behavior.

In terms of TNFA, patients who were homozygous for the rare allele in rs1800610 were more likely to be in All High class. This SNP is located in an intronic region of the gene. Multiple lines of evidence suggest that TNFA rs1800610 is involved in gene regulation. It resides in an RBP binding site, a transcription-factor binding site, and a regulation-sensitive region of the TNFA gene (UCSC Genome Bioinformatics). We were able to identify no studies that evaluated the association between variations in this SNP and common symptoms experienced by oncology patients. However, TNFα is an inducer of a diverse number of processes that initiate and perpetuate inflammatory responses. A number of lines of evidence suggest that changes in TNFα levels are associated with pain, fatigue, sleep disturbance, and depression. For example, the expression of this pro-inflammatory cytokine increases in neurons following a painful stimulus (Andrade et al., 2011). In addition, a recent meta-analysis reported significantly higher serum concentrations of TNFα in patients with major depression (Dowlati et al., 2010).

In a previous study of oncology patients and their family caregivers (Aouizerat et al., 2009), we found an association between a different SNP in TNFA (rs1800629) and the severity of sleep disturbance and morning fatigue. Compared to participants who were homozygous for the common allele, participants who were heterozygous or homozygous for the rare allele reported lower levels of sleep disturbance and morning fatigue. In addition, in the same sample, participants with one or two doses of the rare allele were less likely to be categorized into a subsyndromal depression class (i.e., CES-D scores of ~15) compared to patients in the Resilient class (i.e., CES-D scores of ~5; Dunn et al., 2013). While in vitro evidence suggests that TNFA rs1800629 is a functional SNP, studies of the magnitude and direction of gene expression associated with this SNP have yielded conflicting results. In the current study, it is not clear why this SNP was not associated with membership in the All High class. Research is needed to identify additional SNPs in TNFA that are associated with higher and lower levels of pain, fatigue, sleep disturbance, and depression in oncology patients.

In the present study, an SNP in one anti-inflammatory cytokine gene was associated with membership in the All High class. Patients who were heterozygous or homozygous for the rare allele in IL13 rs1295686 were 29 times more likely to be classified in the All High class compared to those without the rare allele. It is not readily apparent why this SNP was not associated with membership in the All High class in our previous study (Illi et al., 2012). However, in the present sample of women with breast cancer, we found that each dose of the rare A allele in IL13 rs1295686 was associated with a 1.6-fold increase in the odds of reporting pain in the breast prior to surgery (McCann et al., 2012).

While rs1295686 is located in intron 3 of the IL13 gene, it resides in a region of the gene that is a binding site for the NFKB transcription factor (Hirota et al., 2012; Moffatt et al., 2010) and occurs within a region of IL13 that undergoes DNA methylation. Preclinical studies suggest that IL13 has antinociceptive effects in mice. This effect may be mediated through inhibition of the release of TNFα (Vale et al., 2003) and IL1β (Karam, Al-Kouba, Bazzi, Smith, & Leung, 2011). In addition, in a preclinical study by Karam et al. (2011), mice that were infected with a low dose of Leishmania (which induces a sustained hyperalgesia) and received concomitant administration of IL-13 (which reduced the hyperalgesic response) produced increased levels of IL6. Karam et al. concluded that the anti-hyperalgesia effect of IL13 was independent of its effect on IL6 release. Given the associations between latent class membership and variations in IL6 and IL13 found in the current study, additional research with larger samples is warranted to examine the relationships between these two molecular markers and the severity of pain, fatigue, sleep disturbance, and depression, not only as a symptom cluster, but also as individual symptoms.

In terms of clinical studies, associations between IL13 rs1295686 and immunoglobulin E (IgE) dysfunction were evaluated in the Framingham Heart Study (Granada et al., 2012) and in a study that evaluated total IgE concentrations in cord-blood samples (Hong et al., 2010). In both of these studies, the rare allele rs1295686 in IL13 was associated with decreases in IgE concentrations. While we found no additional clinical studies on associations between this SNP and other symptoms, the preclinical and clinical findings suggest that additional research is warranted on the role of the IL13 gene, in conjunction with other pro- and anti-inflammatory cytokine genes, in the symptom experience of oncology patients.

In our previous study of oncology patients and their family caregivers (Illi et al., 2012), we found an association between another anti-inflammatory cytokine gene (i.e., IL4 rs2243248) and membership in the All High symptom class. Participants who were heterozygous or homozygous for the rare allele were 6 times more likely to be classified in the All High class than those without the rare allele. In the current study, we found no association between this SNP and the All High class. The reasons why this SNP was not associated with latent class membership in the current study may be related to differences in demographic (e.g., gender) and clinical (e.g., cancer diagnoses) characteristics between the two samples, as well as differences in the symptom characteristics of the latent classes in the two studies. Additional research is needed with larger samples to explore the role of variations in IL4 and common symptoms in oncology patients.

Several study limitations need to be acknowledged. First, most of the patients were Caucasian, middle-aged, well-educated, mildly overweight, married, and postmenopausal. The relative homogeneity of this sample of women limits the generalizability of the study findings. Second, the exact etiologies for the symptom cluster of pain, fatigue, sleep disturbance and depression warrants additional investigation. In the present study, we hypothesized that sickness behavior associated with inflammation was the mechanism for this symptom cluster. However, sickness behavior is not the only potential mechanism for these four symptoms. The specific interactions that occur among IL-6, IL-13, and TNFα and how they trigger these four symptoms remain to be determined. In addition, each cytokine may contribute in a differential manner to the development of these four symptoms. Therefore, additional research is warranted to evaluate each single gene’s role in the development of sickness behavior and to evaluate for interactions among multiple genes in order to understand the causes of symptom clusters. Third, patients in the current study reported a comorbidity score of between 4 and 5. Therefore, we cannot rule out the fact that some of the associations identified in this study might be attributable to other chronic conditions. Fourth, we did not measure serum cytokine levels in these patients. In addition, future studies need to evaluate if the polymorphisms identified in this study affect protein production and/or binding of these cytokines to their receptors. Finally, it is possible that we did not identify all of the SNPs in the cytokine genes that are associated with the symptom cluster due to the small sample size in the All High class. A larger sample is needed to increase the power to detect differences in other cytokine genes in future studies.

In conclusion, findings from the present study partially confirm our previous findings of subgroups of patients with distinct experiences with the symptom cluster of pain, fatigue, sleep disturbance, and depression (Miaskowski et al., 2006; Pud et al., 2008). In addition, our findings provide new evidence that genetic variations in IL6, IL13, and TNFA are associated with the symptom cluster of pain, fatigue, sleep disturbance, and depression. The association between pro- and anti-inflammatory cytokine genes and this symptom cluster provides additional support for the role of inflammatory mechanisms in the development of these symptoms.

Acknowledgments

This study was funded by grants from the National Cancer Institute (NCI, CA107091 and CA118658). Dr. Dunn received funding from the Mount Zion Health Fund. Dr. Miaskowski is an American Cancer Society Clinical Research Professor and is funded by a K05 award from the NCI (CA168960). Dr. Dhruva is funded through an NIH Mentored Patient-Oriented Research Career Development Award (K23 AT005340). Dr. Langford is supported by a Department of Defense Breast Cancer Research Program Postdoctoral Fellowship. Dr. Merriman was supported by an NINR fellowship (F31 NR012604), an ACS Doctoral Degree Scholarship (DSCN-10-087), an Oncology Nursing Society Doctoral Scholarship, and a UCSF Nursing Alumni Association Scholarship. Dr. Baggott is funded by an American Cancer Society Mentored Research Scholar Award (MRSG 12-01-PCSM). Dr. Leutwyler is funded by the KL2 Program (RR624130). This project is supported by NIH/NCRR UCSF-CTSI grant number UL1 RR024131. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.

Footnotes

Declaration of Conflicting Interests: The authors declared no conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

  1. Alesci S, Martinez PE, Kelkar S, Ilias I, Ronsaville DS, Listwak SJ, Gold PW. Major depression is associated with significant diurnal elevations in plasma interleukin-6 levels, a shift of its circadian rhythm, and loss of physiological complexity in its secretion: Clinical implications. Journal of Clinical Endocrinology and Metabolism. 2005;90:2522–2530. doi: 10.1210/jc.2004-1667. [DOI] [PubMed] [Google Scholar]
  2. Andrade P, Visser-Vandewalle V, Hoffmann C, Steinbusch HW, Daemen MA, Hoogland G. Role of TNF-alpha during central sensitization in preclinical studies. Neurological Sciences. 2011;32:757–771. doi: 10.1007/s10072-011-0599-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Aouizerat BE, Dodd M, Lee K, West C, Paul SM, Cooper BA, Miaskowski C. Preliminary evidence of a genetic association between tumor necrosis factor alpha and the severity of sleep disturbance and morning fatigue. Biological Research for Nursing. 2009;11:27–41. doi: 10.1177/1099800409333871. [DOI] [PubMed] [Google Scholar]
  4. Barsevick AM. The elusive concept of the symptom cluster. Oncology Nursing Forum. 2007;34:971–980. doi: 10.1188/07.ONF.971-980. [DOI] [PubMed] [Google Scholar]
  5. Celeux G, Soromenho G. An entropy criterion for assessing the number of clusters in a mixture model. Journal of Classification. 1996;13:195–212. [Google Scholar]
  6. Chen ML, Tseng HC. Symptom clusters in cancer patients. Supportive Care in Cancer. 2006;14:825–830. doi: 10.1007/s00520-006-0019-8. [DOI] [PubMed] [Google Scholar]
  7. Cheung WY, Le LW, Gagliese L, Zimmermann C. Age and gender differences in symptom intensity and symptom clusters among patients with metastatic cancer. Supportive Care in Cancer. 2011;19:417–423. doi: 10.1007/s00520-010-0865-2. [DOI] [PubMed] [Google Scholar]
  8. Collado-Hidalgo A, Bower JE, Ganz PA, Irwin MR, Cole SW. Cytokine gene polymorphisms and fatigue in breast cancer survivors: Early findings. Brain, Behavior and Immunity. 2008;22:1197–1200. doi: 10.1016/j.bbi.2008.05.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Conde L, Vaquerizas JM, Dopazo H, Arbiza L, Reumers J, Rousseau F, Dopazo J. PupaSuite: Finding functional single nucleotide polymorphisms for large-scale genotyping purposes. Nucleic Acids Research. 2006;34(Web server issue):W621–625. doi: 10.1093/nar/gkl071. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dantzer R. Cytokine-induced sickness behavior: Mechanisms and implications. Annals of the New York Academy of Sciences. 2001;933:222–234. doi: 10.1111/j.1749-6632.2001.tb05827.x. [DOI] [PubMed] [Google Scholar]
  11. Dantzer R, Kelley KW. Twenty years of research on cytokine-induced sickness behavior. Brain, Behavior and Immunity. 2007;21:153–160. doi: 10.1016/j.bbi.2006.09.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. De Vries J, Van der Steeg AF, Roukema JA. Determinants of fatigue 6 and 12 months after surgery in women with early-stage breast cancer: A comparison with women with benign breast problems. Journal of Psychosomatic Research. 2009;66:495–502. doi: 10.1016/j.jpsychores.2008.11.008. [DOI] [PubMed] [Google Scholar]
  13. Dhruva A, Dodd M, Paul SM, Cooper BA, Lee K, West C, Miaskowski C. Trajectories of fatigue in patients with breast cancer before, during, and after radiation therapy. Cancer Nursing. 2010;33:201–212. doi: 10.1097/NCC.0b013e3181c75f2a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dodd MJ, Cho MH, Cooper BA, Miaskowski C. The effect of symptom clusters on functional status and quality of life in women with breast cancer. European Journal of Oncology Nursing. 2010;14:101–110. doi: 10.1016/j.ejon.2009.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Dowlati Y, Herrmann N, Swardfager W, Liu H, Sham L, Reim EK, Lanctot K. A meta-analysis of cytokines in major depression. Biological Psychiatry. 2010;67:446–457. doi: 10.1016/j.biopsych.2009.09.033. [DOI] [PubMed] [Google Scholar]
  16. Dunn LB, Aouizerat BE, Langford DJ, Cooper BA, Dhruva A, Cataldo JK, Miaskowski C. Cytokine gene variation is associated with depressive symptom trajectories in oncology patients and family caregivers. European Journal of Oncology Nursing. 2013;17:346–353. doi: 10.1016/j.ejon.2012.10.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Eisenberger NI, Berkman ET, Inagaki TK, Rameson LT, Mashal NM, Irwin MR. Inflammation-induced anhedonia: Endotoxin reduces ventral striatum responses to reward. Biological Psychiatry. 2010;68:748–754. doi: 10.1016/j.biopsych.2010.06.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Fletcher BS, Paul SM, Dodd MJ, Schumacher K, West C, Cooper B, Miaskowski C. Prevalence, severity, and impact of symptoms on female family caregivers of patients at the initiation of radiation therapy for prostate cancer. Journal of Clinical Oncology. 2008;26(4):599–605. doi: 10.1200/JCO.2007.12.2838. [DOI] [PubMed] [Google Scholar]
  19. Gabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, Altshuler D. The structure of haplotype blocks in the human genome. Science. 2002;296:2225–2229. doi: 10.1126/science.1069424. [DOI] [PubMed] [Google Scholar]
  20. Gilbertson-White S, Aouizerat BE, Miaskowski C. Methodologic issues in the measurement of cytokines to elucidate the biological basis for cancer symptoms. Biological Research for Nursing. 2011;13:15–24. doi: 10.1177/1099800410379497. [DOI] [PubMed] [Google Scholar]
  21. Granada M, Wilk JB, Tuzova M, Strachan DP, Weidinger S, Albrecht E, O’Connoor GT. A genome-wide association study of plasma total IgE concentrations in the Framingham Heart Study. Journal of Allergy and Clinical Immunology. 2012;129(3):840–845, e821. doi: 10.1016/j.jaci.2011.09.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. 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. Human Mutation. 2008;29:648–658. doi: 10.1002/humu.20695. [DOI] [PubMed] [Google Scholar]
  23. Hattersley AT, McCarthy MI. What makes a good genetic association study? Lancet. 2005;366:1315–1323. doi: 10.1016/S0140-6736(05)67531-9. doi: S0140-6736(05)67531-9. [DOI] [PubMed] [Google Scholar]
  24. Hirota T, Takahashi A, Kubo M, Tsunoda T, Tomita K, Sakashita M, Tamari M. Genome-wide association study identifies eight new susceptibility loci for atopic dermatitis in the Japanese population. Nature Genetics. 2012;44:1222–1226. doi: 10.1038/ng.2438. [DOI] [PubMed] [Google Scholar]
  25. Hoggart CJ, Parra EJ, Shriver MD, Bonilla C, Kittles RA, Clayton DG, McKeigue PM. Control of confounding of genetic associations in stratified populations. American Journal of Human Genetics. 2003;72(6):1492–1504. doi: 10.1086/375613. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Hong X, Tsai HJ, Liu X, Arguelles L, Kumar R, Wang G, Wang X. Does genetic regulation of IgE begin in utero? Evidence from T(H)1/T(H)2 gene polymorphisms and cord blood total IgE. Journal of Allergy and Clinical Immunology. 2010;126:1059–1067, e1051. doi: 10.1016/j.jaci.2010.08.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Illi J, Miaskowski C, Cooper B, Levine JD, Dunn L, West C, Aouizerat BE. Association between pro- and anti-inflammatory cytokine genes and a symptom cluster of pain, fatigue, sleep disturbance, and depression. Cytokine. 2012;58:437–447. doi: 10.1016/j.cyto.2012.02.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Jensen MP. The validity and reliability of pain measures in adults with cancer. Journal of Pain. 2003;4:2–21. doi: 10.1054/jpai.2003.1. [DOI] [PubMed] [Google Scholar]
  29. Karam MC, Al-Kouba JE, Bazzi SI, Smith CB, Leung L. Interleukin-13 reduces hyperalgesia and the level of interleukin-1beta in BALB/c mice infected with Leishmania major with an up-regulation of interleukin-6. Journal of Neuroimmunology. 2011;234:49–54. doi: 10.1016/j.jneuroim.2011.02.003. [DOI] [PubMed] [Google Scholar]
  30. Karnofsky D. Performance scale. New York: Plenum Press; 1977. [Google Scholar]
  31. Kim HJ, Barsevick AM, Beck SL, Dudley W. Clinical subgroups of a psychoneurologic symptom cluster in women receiving treatment for breast cancer: A secondary analysis. Oncology Nursing Forum. 2012;39(1):E20–30. doi: 10.1188/12.ONF.E20-E30. [DOI] [PubMed] [Google Scholar]
  32. Lanza ST, Flaherty BP, Collins LM, editors. Latent class and latent transition analysis. Vol. 2. Hoboken, NJ: John Wiley & Sons, Inc; 2003. [Google Scholar]
  33. Lee KA, Hicks G, Nino-Murcia G. Validity and reliability of a scale to assess fatigue. Psychiatry Research. 1991;36:291–298. doi: 10.1016/0165-1781(91)90027-m. [DOI] [PubMed] [Google Scholar]
  34. Li D, Lewinger JP, Gauderman WJ, Murcray CE, Conti D. Using extreme phenotype sampling to identify the rare causal variants of quantitative traits in association studies. Genetic Epidemiology. 2011;35:790–799. doi: 10.1002/gepi.20628. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Liu L, Mills PJ, Rissling M, Fiorentino L, Natarajan L, Dimsdale JE, Ancoli-Israel S. Fatigue and sleep quality are associated with changes in inflammatory markers in breast cancer patients undergoing chemotherapy. Brain, Behavior and Immunity. 2012;26:706–713. doi: 10.1016/j.bbi.2012.02.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Ljungman P, Bellander T, Nyberg F, Lampa E, Jacquemin B, Kolz M, Airgene Study Group DNA variants, plasma levels and variability of interleukin-6 in myocardial infarction survivors: Results from the AIRGENE study. Thrombosis Research. 2009;124(1):57–64. doi: 10.1016/j.thromres.2008.10.009. [DOI] [PubMed] [Google Scholar]
  37. Maier SF, Watkins LR. Immune-to-central nervous system communication and its role in modulating pain and cognition: Implications for cancer and cancer treatment. Brain, Behavior and Immunity. 2003;17(Suppl 1):S125–131. doi: 10.1016/s0889-1591(02)00079-x. [DOI] [PubMed] [Google Scholar]
  38. McCann B, Miaskowski C, Koetters T, Baggott C, West C, Levine JD, Aouizerat BE. Associations between pro- and anti-inflammatory cytokine genes and breast pain in women prior to breast cancer surgery. Journal of Pain. 2012;13:425–437. doi: 10.1016/j.jpain.2011.02.358. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Miaskowski C, Aouizerat BE. Biomarkers: Symptoms, survivorship, and quality of life. Seminars in Oncology Nursing. 2012;28:129–138. doi: 10.1016/j.soncn.2012.03.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Miaskowski C, Cooper B, Paul SM, West C, Langford D, Levine JD, Aouizerat BE. Identification of patient subgroups and risk factors for persistent breast pain following breast cancer surgery. Journal of Pain. 2012;13(12):1172–1187. doi: 10.1016/j.jpain.2012.09.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Miaskowski C, Cooper BA, Dhruva A, Dunn LB, Langford DJ, Cataldo JK, Aouizerat BE. Evidence of associations between cytokine genes and subjective reports of sleep disturbance in oncology patients and their family caregivers. PLoS One. 2012;7(7):e40560. doi: 10.1371/journal.pone.0040560. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Miaskowski C, Cooper BA, Paul SM, Dodd M, Lee K, Aouizerat BE, Bank A. Subgroups of patients with cancer with different symptom experiences and quality-of-life outcomes: A cluster analysis. Oncology Nursing Forum. 2006;33:E79–89. doi: 10.1188/06.ONF.E79-E89. [DOI] [PubMed] [Google Scholar]
  43. Miaskowski C, Dodd M, Lee K, West C, Paul SM, Cooper BA, Aouizerat BE. Preliminary evidence of an association between a functional interleukin-6 polymorphism and fatigue and sleep disturbance in oncology patients and their family caregivers. Journal of Pain and Symptom Management. 2010;40:531–544. doi: 10.1016/j.jpainsymman.2009.12.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Miaskowski C, Dodd M, Paul SM, West C, Hamolsky D, Abrams G, 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]
  45. Moffatt MF, Gut IG, Demenais F, Strachan DP, Bouzigon E, Heath S, Gabriel Consortium A large-scale, consortium-based genomewide association study of asthma. New England Journal of Medicine. 2010;363:1211–1221. doi: 10.1056/NEJMoa0906312. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Muthen B, Shedden K. Finite mixture modeling with mixture outcomes using the EM algorithm. Biometrics. 1999;55(2):463–469. doi: 10.1111/j.0006-341x.1999.00463.x. [DOI] [PubMed] [Google Scholar]
  47. Muthen LK, Muthen BO. Mplus (Version 6) Los Angeles, CA: Author; 1998–2010a. [Google Scholar]
  48. Muthen LK, Muthen BO. Mplus user’s guide. 6th. Los Angeles, CA: Author; 1998–2010b. [Google Scholar]
  49. Nylund KL, Asparouhov T, Muthen BO. Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Structural Equation Modeling. 2007;14:535–569. [Google Scholar]
  50. Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal components analysis corrects for stratification in genome-wide association studies. Nature Genetics. 2006;38:904–909. doi: 10.1038/ng1847. [DOI] [PubMed] [Google Scholar]
  51. Pud D, Ben Ami S, Cooper BA, Aouizerat BE, Cohen D, Radiano R, Miaskowski C. The symptom experience of oncology outpatients has a different impact on quality-of-life outcomes. Journal of Pain and Symptom Management. 2008;35:162–170. doi: 10.1016/j.jpainsymman.2007.03.010. [DOI] [PubMed] [Google Scholar]
  52. Radloff LS. The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement. 1977;1:385–401. [Google Scholar]
  53. Rothman KJ. No adjustments are needed for multiple comparisons. Epidemiology. 1990;1:43–46. [PubMed] [Google Scholar]
  54. Saligan LN, Kim HS. A systematic review of the association between immunogenomic markers and cancer-related fatigue. Brain, Behavior and Immunity. 2012;26(6):830–848. doi: 10.1016/j.bbi.2012.05.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Sangha O, Stucki G, Liang MH, Fossel AH, Katz JN. The Self-Administered Comorbidity Questionnaire: A new method to assess comorbidity for clinical and health services research. Arthritis and Rheumatism. 2003;49:156–163. doi: 10.1002/art.10993. [DOI] [PubMed] [Google Scholar]
  56. Schubert C, Hong S, Natarajan L, Mills PJ, Dimsdale JE. The association between fatigue and inflammatory marker levels in cancer patients: A quantitative review. Brain, Behavior and Immunity. 2007;21:413–427. doi: 10.1016/j.bbi.2006.11.004. [DOI] [PubMed] [Google Scholar]
  57. Schwarzenberg SJ, Sinaiko AR. Obesity and inflammation in children. Paediatric Respiratory Reviews. 2006;7(4):239–246. doi: 10.1016/j.prrv.2006.08.002. [DOI] [PubMed] [Google Scholar]
  58. Seruga B, Zhang H, Bernstein LJ, Tannock IF. Cytokines and their relationship to the symptoms and outcome of cancer. Nature Reviews Cancer. 2008;8:887–899. doi: 10.1038/nrc2507. [DOI] [PubMed] [Google Scholar]
  59. Sheridan D, Foo I, O’Shea H, Gillanders D, Williams L, Fallon M, Colvin L. Long-term follow-up of pain and emotional characteristics of women after surgery for breast cancer. Journal of Pain and Symptom Management. 2012;44:608–614. doi: 10.1016/j.jpainsymman.2011.10.021. [DOI] [PubMed] [Google Scholar]
  60. 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. Journal of Infectious Diseases. 2012;205:1417–1424. doi: 10.1093/infdis/jis208. [DOI] [PubMed] [Google Scholar]
  61. SPSS. IBM SPSS for Windows (Version 21) Chicago: SPSS, Inc; 2012. [Google Scholar]
  62. StataCorp. Stata statistical software: Release 9. College Station, TX: Author; 2005. [Google Scholar]
  63. Stephens M, Smith NJ, Donnelly P. A new statistical method for haplotype reconstruction from population data. American Journal of Human Genetics. 2001;68:978–989. doi: 10.1086/319501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Sukoff Rizzo SJ, Neal SJ, Hughes ZA, Beyna M, Rosenzweig-Lipson S, Moss SJ, Brandon NJ. Evidence for sustained elevation of IL-6 in the CNS as a key contributor of depressive-like phenotypes. Translational Psychiatry. 2012;2:e199. doi: 10.1038/tp.2012.120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Tabassum R, Mahendran Y, Dwivedi OP, Chauhan G, Ghosh S, Marwaha RK, Bharadway D. Common variants of IL6, LEPR, and PBEF1 are associated with obesity in Indian children. Diabetes. 2012;61:626–631. doi: 10.2337/db11-1501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Tian C, Gregersen PK, Seldin MF. Accounting for ancestry: Population substructure and genome-wide association studies. Human Molecular Genetics. 2008;17(R2):R143–150. doi: 10.1093/hmg/ddn268. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Tizard I. Sickness behavior, its mechanisms and significance. Animal Health Research Review. 2008;9(1):87–99. doi: 10.1017/S1466252308001448. [DOI] [PubMed] [Google Scholar]
  68. Vale ML, Marques JB, Moreira CA, Rocha FA, Ferreira SH, Poole S, Ribeiro A. Antinociceptive effects of interleukin-4, -10, and -13 on the writhing response in mice and zymosan-induced knee joint incapacitation in rats. Journal of Pharmacology and Experimental Therapeutics. 2003;304:102–108. doi: 10.1124/jpet.102.038703. [DOI] [PubMed] [Google Scholar]
  69. Vermunt JK, Magdison J. Latent class cluster analyses. New York: Cambridge University Press; 2002. [Google Scholar]
  70. Verri WA, Jr, Cunha TM, Parada CA, Poole S, Cunha FQ, Ferreira SH. Hypernociceptive role of cytokines and chemokines: Targets for analgesic drug development? Pharmacological Therapeutics. 2006;112:116–138. doi: 10.1016/j.pharmthera.2006.04.001. [DOI] [PubMed] [Google Scholar]
  71. Wang XM, Lehky TJ, Brell JM, Dorsey SG. Discovering cytokines as targets for chemotherapy-induced painful peripheral neuropathy. Cytokine. 2012;59:3–9. doi: 10.1016/j.cyto.2012.03.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Wright CE, Bovbjerg DH, Montgomery GH, Weltz C, Goldfarb A, Pace B, Silverstein JH. Disrupted sleep the night before breast surgery is associated with increased postoperative pain. Journal of Pain and Symptom Management. 2009;37:352–362. doi: 10.1016/j.jpainsymman.2008.03.010. doi: S0885-3924(08)00386-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES