There is increasing evidence illustrating the association of chronic pain and aberrant levels of neuroimmune mediators.1,2 However the immune and glial molecules responsible for different aspects of chronic pain still warrant identification and exploration. Using central neuroinflammation as a paramount factor in the creation and maintenance of chronic pain, this study aimed to investigate and describe the expression of peripheral and central nervous system (CNS) cytokines, chemokines and growth factors associated with the chronic pain continuum.
The objectives of the present study were to examine the central inflammatory cytokines, chemokines, and growth factors present in a chronic pain population and explore patterns of the same mediator molecules, over time, in response to intrathecal opioid therapy. Secondary objectives were to explore the relationship of central and peripheral neuroimmune mediators in a chronic pain population, while examining the psychobehavioral traits associated with chronic pain. Based on previous research, we hypothesized an increase in neuroimmunologic markers in the chronic pain population with a temporal rise in levels over treatment with an intrathecal analgesic pump.
Materials and Methods:
All study procedures were approved by the UCLA Institutional Review Board (IRB). Recruitment and participation in this study commenced between April 2013 and December 2014. The UCLA Cousins Center for Psychoneuroimmunology Inflammatory Biology Core Laboratory served as the specimen processing, storage and analysis site of this study.
Study participants: Inclusion criteria for this study required that participants be 18 years or older with a chronic pain diagnosis of > 6 months and were eligible for intrathecal analgesic pump. Exclusion criteria included meningitis, active upper respiratory infection/flu or febrile nature, acute medical or psychiatric disorders, acute psychological or physiologic danger/instability, altered mental status, pregnancy, the inability to communicate in English, a diagnosis of cancer, human immunodeficiency virus (HIV) infection, recent epidural injections for treatment of chronic pain (within six months), drug abuse, history of blood transfusion(s), and palliative pain treatment.
Study Design: The study consisted of two phases. The first was a cross-sectional, correlational study design. Collections of cerebrospinal fluid (CSF) from eight chronic pain study participants were analyzed for interleukin (IL)-1ß, IL-2, IL-4, IL-6, IL-8, IL-10, tumor necrosis factor-alpha (TNF-α), interferon gamma (IFN- γ), vascular endothelial growth factor (VEGF) and granulocyte-macrophage colony-stimulating factor (GM-CSF), and compared to CSF from 30 age- and gender-matched control subjects. Next, a longitudinal study design was utilized to explore the trends of the pro- and anti-inflammatory mediators during intrathecal analgesic pump therapy. Plasma and CSF samples were collected from chronic pain participants at the pump trial, pump implant and/or pump refill visits. Longitudinal evaluation of pain perception (SF-MPQ), anxiety (HAM-A), depression (BDI-II) and sleep quality (PSQI) in study participants allowed for correlational analysis of psychobehavioral and physiologic molecular patterns over the treatment course.
Statistical analysis: Descriptive statistics were calculated for all variables in the study and examined for significant responses. Data were tested for normal distribution and the level of significance was set at p < 0.05 for all statistical testing. All data are reported as mean ± standard deviation unless otherwise noted. The cross-sectional statistical analysis focused on comparing the distributions of detectable CSF neuroimmune markers between the chronic pain and control groups using the two sample t-test. The longitudinal statistical analysis used generalized estimating equation (GEE) models to assess the patterns of neuroimmune markers across time. Correlations were calculated to explore relationships between psychobehavioral measures, demographics, and levels of neuroimmune markers using Pearson’s r or Spearman’s Rho correlations. Calculations were accomplished with SPSS software version 22 (SPSS, IBM Inc., Armonk, NY).
Demographics: Of the eight chronic pain patients who participated in this study, five participants were female, and three participants were male. The age range for the study participants was 41-74 years, with a mean age of 53.50 ± 10.53 years. The mean body mass index (BMI) of study participants was 32.35 ± 4.21 kg/m2; 6/8 subjects were obese. The mean metabolic equivalents (METS) reported by participants was 3.38 ± 1.51 indicating participants could complete moderate intensity activities. The majority of the patients suffered from post-laminectomy syndrome chronic pain (n=6) with a mean length of diagnosis of 9.31 ± 7.75 years.
Cross Sectional Analysis: At baseline, CSF from the study population was compared to control samples of CSF to assess for significant differences in the previously mentioned neuroimmune mediators. Three cytokines yielded measurable quantities; IL-6, IL-8 and GM-CSF. Interleukin-1ß, IL-2, IL-4, IL-10, TNF-α, IFN- γ and VEGF were undetectable in the majority of CSF samples. Table 1 summarizes the cross-sectional neuroimmune mediator data.
Both the Wilcoxon test and two sample t-test demonstrated significance for between-group differences with respect to IL-8 (p < 0.001, Figure 1) but not IL-6 or GM-CSF (t-test p = 0.496 and p = 0.539 respectively). A matched pair analysis controlling for age and gender yielded similar results; IL-8 was significantly different (p = 0.001), IL-6 and GM-CSF were not (p = 0.717 and p = 0.862 respectively). Linear regression models determined similar findings; IL-8 was significantly different between chronic pain and control CSF samples (p < 0.001), IL-6 and GM-CSF were not (p = 0.78 and p = 0.70 respectively).
Longitudinal Analysis: Plasma and CSF patterns of IL-1ß, IL-2, IL-4, IL-6, IL-8, IL-10, TNF-α, IFN- γ, VEGF and GM-CSF were analyzed in the chronic pain study population over the trial, insertion and/or refill of an implantable intrathecal analgesic pump. Six of the eight chronic pain participants rendered at least two distinct time point specimen collections; two of the participants were ineligible for longitudinal inclusion secondary to pump trial failure or removal of implanted pump. Of the 10 neuroimmune markers assessed for in the CSF, only IL-6, IL-8 and GM-CSF yielded measurable quantities. Of the 10 cytokines assessed for in plasma, IL-1ß, IL-6, IL-8, IL-10, TNF-α, VEGF and GM-CSF were detectable. Using a linear mixed effect model with subject random effect and fixed gender and age effect, CSF levels of IL-8 and GM-CSF were found to significantly change over time (p = 0.008 and p = 0.012, respectively). Likewise, plasma levels of GM-CSF were also shown to change significantly over time (p = 0.024). Table 2 summarizes the longitudinal neuroimmune mediator data.
Psychobehavioral Measures: The eight study participants reported an initial mean pain rating index of 22.88 ± 7.10, mean present pain intensity of 3.57 ± 1.13, and a mean pain VNS score of 7.38 ± 1.32, indicative of severe pain. Moreover, at baseline, mean BDI-II and HAM-A scores indicated mild-moderate depression and anxiety symptomatology. Notably, all subjects reported poor sleep quality (PSQI > 5); the mean PSQI score for the whole group was strikingly high (15.50 ± 3.16). None of the subjective pain, psychopathology or sleep quality measures changed significantly over time. Pearson’s r and Spearman’s Rho correlations were analyzed to determine significance between psychobehavioral data and central or peripheral neuroimmunologic marker levels; no significant findings were made in the cross-sectional analysis.
In the present study, the most robust and significant finding was markedly higher CSF levels of IL-8 in chronic pain patients compared to age- and gender-matched controls (p < 0.001). To our knowledge, this is the first time that a longitudinal increase of this specific cytokine has been reported; a finding similar to the longitudinal increase of CSF IL-6 described by Zin et al.3 in a study with a similar pain cohort. It is not clear whether the significant increase in IL-8 over time may reflect the surgical implant procedure or opioid medication used in the intrathecal pump, as both could be confounding variables in the origination of higher IL-8 levels. It is interesting to note that this significant longitudinal increase was not appreciated in the peripheral plasma, and thus must reflect a CNS response. This increase in CSF IL-8 is likely attributable to increased CNS inflammation, as indicated by the temporal increase of CSF GM-CSF.
Importantly, this study provides further supportive evidence that there may be an etiological relationship between CNS inflammation and chronic pain states in humans; nevertheless, these results are preliminary and warrant replication in higher-powered studies. Notably, all of the study participants took at least one opioid medication for treatment of their pain and reported poor sleep quality, illustrating the need to investigate possible mechanisms of association. Future longitudinal clinical research, utilizing paradigms of proinflammatory cytokine and/or neurotrophic factor inhibition, may provide further detail to the interactions between neuroimmune mediators and chronic pain mechanisms. Such research may provide a greater understanding of chronic pain pathophysiology, and spur development of novel targeted treatments.