Detection of systemic inflammation in severely impaired chronic pain patients and effects of a multimodal pain rehabilitation program

2.1 Patient settings and demographics

Sixty-eight consecutive patients referred to the Pain Clinic at the Uppsala University Hospital, Sweden were asked to participate in the study, of these 68 patients, 16 declined to participate. Thus, 52 gave written informed consent and were enrolled in the study and participated in the baseline sampling.

The group examined in this study were patients characterized in detail in a previously study [20]. They presented with severe mixed pain conditions: musculoskeletal pain, fibromyalgia, widespread pain, abdominal pain, pelvic pain, facial pain, migraine, complex regional pain syndrome (CRPS) and neuropathic pain of different origin. Of the 52 participants, 76% were diagnosed with a psychiatric disorder, the majority (69%) with depression or anxiety/depression disorder. All patients had in common a very complex expression of the pain disease and had severe functional limitations, with some entirely bedridden [20]. Another characterization of the group was their extreme symptom preoccupation, reporting in average 22 symptoms other than pain.

Forty-one of the 52 patients were accepted and participated in the PRP. In the study setting the patients were followed-up 1 year after treatment. At that time, data from 28 patients were collected (Table 1). Thirteen patients declined further participation in the study: four patients without declaring why and eight experienced the journey to the clinic too long and exhausting. One patient was excluded because of psychiatric illness. At follow-up, a median of 17 months (range 12–36) had elapsed from the first sampling. The large variation in the follow-up times can be explained by the waiting-list period before participation in the program.

The control group consisted of 49 age and sex-matched healthy blood donors, given that age and sex have been shown to influence the results on cytokines assays [22]. At follow-up the patients acted as their own controls: the baseline data for inflammatory activity were compared with 1-year follow up data.

2.2 The cognitive behavior therapy – pain rehabilitation program (PRP)

Forty-one of the originally 52 enrolled patients were accepted and treated in the PRP. The inclusion criterion for entering the program was that primary or secondary pain care, including multidisciplinary rehabilitation had failed to improve the chronic pain (CP) related disability. Exclusion criteria were active alcohol or illicit drug abuse, psychosis or suicidal behavior.

The intervention consisted of a multimodal, acceptance-based, individualized program with a psychiatrist, a physician specialized in pain and rehabilitation medicine, a physiotherapist, an occupational therapist, a phycologist and a nurse working with the patient as a team.

The program started with a 1-week team-based evaluation of physical and cognitive functioning, as well as a review of current drug treatment. If the patients were accepted for treatment, they participated in a 4-week in-hospital treatment. Many of the interventions included individualized exposure to avoided activities that the patients had stated as valuable for a meaningful life. Because our group of patients were labeled extreme “avoiders” this approach was a key component to the treatment program. The patients avoided emotions, thoughts and activities, all of which hinder the ability to live a productive, satisfying life despite the pain disease. Another focus in the treatment program centered on alleviating stress, tension and worrying by using mindfulness and methods for cognitive reorganization according to cognitive behavior therapy (CBT). The treatment was conducted 30 h per week. The program ended with a follow-up week 2 months after the treatment-period to evaluate the results of the intervention. The patients were supported during their time at home through phone and Internet contact. In all, the patients participated in rehabilitation for 6 months.

2.3 Blood sampling

Venous blood was collected from the patients and the plasma was analyzed. The blood samples were collected in ethylenediaminetetraacetic acid (EDTA) tubes and centrifuged at 2,400 g for 5 min. The resulting supernatant was collected and stored in aliquots at −70 °C until further analysis. The samples as well as the controls were handled according to the same protocol at baseline and at the 1-yearfollow-up.

2.4 Analysis of inflammatory biomarkers in blood samples

Using the multiplex proximal extension assay (PEA), a panel of 92 proteins was simultaneously analyzed in a multiplex assay (Proseek Multiplex Inflammation, Olink Bioscience, Uppsala, Sweden) [23]. One microliter of plasma was mixed with 3 μL of incubation mix containing 94 probe pairs. Each of the probe pairs contained two target-specific antibodies with unique barcoded DNA and oligonucleotides attached. The mixture was incubated at 8 °C overnight and then added to a 96 μL extension mix containing PEA enzyme and polymerase chain reaction reagents and incubated for 5min at room temperature before the plate was transferred to a thermal cycler for 17 cycles of DNA amplification. A 96.96 Dynamic Array IFC (Fluidigm, CA, USA) was prepared and primed according to the manufacturer’s instructions. 2.8 μL of a sample mixture was mixed with 7.2 μL detection mix in a new 96-well plate and 5 μL were loaded into the right side of the primed 96.96 Dynamic Array IFC. The unique primer pairs for each cytokine were loaded into the left side of the 96.96 Dynamic Array IFC, and then the protein expression program was run in a Fluidigm Biomark reader following the instructions for Proseek [23].

The CRP level was measured at baseline in 45 patients, using the standard procedure at the clinic.

2.5 Verification of results using another quantitative method for analysing of inflammatory biomarkers

To verify the key-results these samples were investigated once again using another technique, a multi-array technology that allows quantitative estimation in which electrochemiluminescence detection is combined with patterned array MSD Cytokine Assays method (Meso Scale Diagnostics, Rockville, MD, USA) [24]. The proteins examined were CRP, CXCL5, CX3CLI, CXXLI, CXCL10, CCL23 and CCL19. Both baseline and follow-up data were analyzed.

4.1 Plasma analysis of inflammatory biomarkers in pain patients vs. healthy controls using multiplex PEA-technology

There were 73 biomarkers with enough measurements above the detection limit. Using 5% as our FDR, there were significant differences between the two groups in 43 of the 73 (58.9%) biomarkers, with higher values generally found in the pain patients (Fig. 1).

Fig. 1:

To visualize group differences, we plotted the first principal component against the scores from the linear discriminant analysis (essentially a weighted average of the combined biomarker level in each patient). There were 73/92 biomarkers with enough measurements above the detection limit. Using 5% as our FDR, there were significant differences between the two groups in 43/73 biomarkers.

In this study there were eight biomarkers (CXCL5, AXIN1, SIRT2, CXCL1, ST1A1, CXCL6, STAMBP, CASP8) in which the expression levels of the pain patients were more than four-fold higher than those of the controls. Three of those epithelial cell-derived neutrophil-activating peptide (CXCL5), silent information regulator protein 2 (SIRT2) and Axin-related protein 1 (AXIN1) had expression levels more than eight times higher (Fig. 1, Table 2).

There were lower expression levels in four biomarkers among the pain patients compared with the control: Transforming Growth Factor-alfa (TGF-alfa), Interleukin 18 (IL18), Cluster of differentiation 6 (CD6) and Oncostatin M (OSM).

C-reactive protein (CRP) were identified in 45 patients at baseline. The CRPs had a range from 0.2 to 9.5 with an average of 2.6, indicating normal values.

4.2 Inflammatory biomarkers at follow-up vs. baseline, with the patients serving as their own controls

At follow-up data were collected from 28 of 41 patients who had participated in the multimodal PRP. Mean follow-up was 17 months (range 12–36 months). The large variation in the follow-up times can be explained by the waiting-list period before participation in the program. At follow-up, 28 of the biomarkers differed from their baseline values (Fig. 2, Table 3). For all these 28 biomarkers, the expression levels were higher at baseline than at follow-up. Some illustrative examples of the difference between the baseline and follow-up values are shown in Fig. 3.

Fig. 2:

Differences between baseline and the 1-year follow-up. Data were collected for 28/52 patients at the 1-year follow-up. The values of 28 biomarkers at follow-up were lower than those at baseline. The first principal component is plotted against the scores from the linear discriminant analysis.

Fig. 3:

Boxplot of differences between the baseline values and those at the 1-year follow-up. For each patient and each biomarker, the difference between baseline and follow-up values was computed. p-Values for testing differences between the two time points were computed using the Wilcoxon one-sample test (shown in Table 3). Tests were adjusted for multiple testing using the Benjamini-Hochberg procedure.

4.3 Verification of results using a quantitative method

The change in the level of biomarkers were confirmed for the key findings by another established method for analyzing inflammatory proteins to allow quantitative estimate: MSD Cytokine Assays method [24] (data not shown). Seven proteins (CRP, CXCL5, CX3CLI, CXXLI, CXCL10, CCL23, CCL19) included in the PEA panel were tested and confirmed the same pattern over time as in the PEA-test.

5.1 The main finding

The main finding was the high expression of inflammatory-related proteins in plasma in a cohort of severely impaired patients with mixed pain conditions compared with a healthy reference group (Fig. 1, Table 2). The finding indicates an ongoing systemic inflammatory process detectable in plasma that previously has not been adequately described.

Another important finding was the decrease, in comparison with baseline, in 28 inflammatory-related biomarkers at the 1-year follow-up after participation in the CBT-based individualized multimodal PRP (Figs. 2 and 3, Table 3).

5.2 Analysis of inflammatory biomarkers in blood samples

The new multiplex PEA technology used in this study allowed a simultaneously analysis of 92 inflammatory-related proteins. The panel consisting mainly of cytokines and chemokines [23] has the advantage of permitting a pattern of inflammatory biomarkers to be observed, yielding an inflammatory plasma profile rather than merely analyzing a few isolated cytokines, chemokines or other inflammatory-related proteins.

We confirmed the fundamental results through a test using multi-array technology that allows quantitative estimations (MSD Cytokine Assays method) [24], which produces absolute values. This second confirmative test strengthens the results found in the multiplex PEA-assay.

5.3 The role of inflammatory activity related to pain pathophysiology in chronic pain

Already Hippocrates [28] observed that infection as well as tissue damage presents with dolor, calor, rubor, tumor and function laesa, all of which are nowadays known as signs of activity in the inflammatory response system. With the help of new analytic tools [23], [24], the role of the immune system and the inflammatory reactions in the development and maintenance of CP can be better understood [9], [10], [11], [12], [13], [14], [15], [19].

Upregulation of the inflammatory system as an answer to infection, tissue damage or autoimmune disease is well described [29]. A model that can explain similar upregulation of components in the immune system involved in inflammation as a response to stress is suggested by Slavich and Irwin in the publication “From stress to Inflammation and Major Depressive Disorder: A social Signal Transduction Theory of Depression” [30]. Experiences of social threat and adversity, which are perceived as stress, upregulate components of the immune system. Stress activates the dorsal anterior cingulate cortex and anterior insula [31], regions that project to lower level brain areas (e.g. hypothalamus, autonomic control nuclei in the brainstem) that initiate and modulate inflammatory activity via the hypothalamic-pituitary-adrenal (HPA) axis, and the sympathetic nervous system [32]. As an answer to the stimulation, immune cells will start the production of pro-inflammatory interleukins (e.g. Interleukin 1, IL1; tumor necrosis factor alfa; TNF-alfa; and Interleukin 6, IL6), resulting in an inflammatory reaction. Cytokines mediate information to the central nervous system (CNS) via humoral and neuronal transport and initiate activation in central immunocompetent cells (such as microglia and astrocytes), resulting in a production of pro-inflammatory cytokines in the CNS. These centrally produced cytokines promote a change in behavior [9], [10], i.e. the individual develops an adaptive response characterized by decreased mood, social and physical isolation, anhedonia, increased anxiety and heightened pain sensitivity [9], [10] the “sickness behavior syndrome” [9], [10], [29]. The “sickness behavior syndrome” is represented at an extreme level in the most impaired pain patients examined in this study, making this theory applicable to the understanding of CP as a stressor.

5.4 Changes in the level of inflammatory biomarkers in the present study

In all, 43 proteins were expressed on a different level in the patients compared with the controls (39 of the proteins increased and four decreased). Moen et al. studying patients with persistent pain after disc herniation using the same multiplex PEA technology as used in our study, also reported increased levels of biological markers in the pain patients: CXCL5 was increased by 217% and chemokine (C-X-C motif) ligand 1 (CXCL1) by 53% [19]. In the present study CXCL5 was increased by 861% and CXCL1 by 529% (Table 2).

CXCL5 is a chemokine that recruits and activates neutrophils mainly during acute inflammatory responses [33], [34].

AXIN1, one of the three cytokines heightened more than eight times in the CP patients, is an enzyme involved in the regulation of the Wnt signaling pathway, a process whereby a chemical or physical signal is transmitted through a cell [35]. Wnt signaling regulates different functions in the immune system. Activated Wnt signaling induces production of pro-inflammatory cytokines (IL-18 and TNF-alfa).

SIRT2, the third cytokine with an eight-time heightened level in the CP patients, is an enzyme expressed in many cells and tissues and interacts with a member of the nuclear factor-kappa B (NF-kappaB) family. NF-kappa B proteins are transcription factors regulating the expression of various genes involved in the immune and inflammatory processes and are activated by toll-like receptors (TLRs) in the acute-phase response to injury and infection leading to the production of pro-inflammatory cytokines that are supposed to play an important role in aging and inflammatory diseases [36], [37].

There were four cytokines with a lower level in the CP patients: TGF-alfa, CD6, OSM and IL6. TGF-alfa, which induces epithelial development, is a member in a family of epidermal growth factors produced in macrophages, brain cells and keratinocytes [38]. CD6, a protein on the outer membrane of T-lymphocytes, is important to the cell’s action [39]. IL18 induces cell-mediated immunity [40]. OSM is a member of the IL6 group of cytokines initially known by its unique activity to inhibit the proliferation of tumor cells. Recent research has demonstrated its role in inflammation, hematopoiesis and development [41].

The present findings support the concept that a significant systemic inflammation is present in CP patients with mixed pain conditions. We suggest that inflammation may actively contribute to CP in conditions not traditionally looked upon as inflammatory diseases, which is thus in contrast to “real” inflammatory diseases such as rheumatoid arthritis and ankylosing spondylitis. Similar findings has recently been reported in studies on patients with wide-spread pain, including fibromyalgia [11], [13], neuropathic pain [14], [17], [18] and in patients suffering from CP after disc herniation [19].

5.5 Chronic systemic inflammation

Chronic systemic inflammation is thought to mediate in part the development of depression [10], [42], fatigue [43], sleep disorder [44], rheumatoid arthritis [45], cardiovascular diseases [46] and certain cancers [47], [48].

For neuropathic pain, several studies confirm that pro-inflammatory cytokines are elevated in blood samples and CSF [14], [17], [18], [19]. Results from two studies using PEA technology in patients with chronic widespread pain [13] and fibromyalgia [11] also found changes in biomarkers indicating systemic or neuroinflammation.

Data from the Dunedin Longitudinal Study, New Zeeland, a cohort study in which the participants are followed from birth to adulthood, showed that persons with childhood maltreatment confers increased risk for the expression of inflammatory activity [49], [50]. Several studies have shown that maltreatment in childhood is a risk factor for developing CP [51], [52], [53], [54]. A clinical observation is that maltreatment in childhood and bullying are frequent in patients seen at our clinic.

All this implicates a risk not only for comorbidities between CP patients and social-psychiatric disorders but also for somatic diseases [48]. The inflammatory reaction may be a common denominator to understand these processes.

5.6 Decreased levels of inflammatory biomarkers 1 year after participation in an CBT-based pain rehabilitation program

To our knowledge, this is the first study demonstrating that a set of inflammatory-associated biomarkers show a decrease 1 year after the patients took part in a PRP. Participation in a PRP for about 6 months is thought to have a major impact on, for example, the ability to cope with pain, tiredness, depressed mood and physical and social functioning [55]. These factors are associated with changes in the expression of inflammatory activities [42], [43], [44] and can be interpreted as a possible reason for the consistent reduction in inflammatory biomarkers seen at follow-up in our study. Several studies on unimodal treatment interventions (e.g. mindfulness and physical activity) have reported at least a trend towards a reduction in inflammatory biomarkers post-treatment [56], [57], [58], [59]. Studies by Moen et al. [19] and Pedersen et al. [17] reported changes in inflammatory-related biomarkers in patients followed over time. The studies described a reduction in inflammatory biomarkers in a group of patients that recovered from painful radiculopathy but not in a group that continued to suffer from pain at a 1-year follow-up.

We are certainly aware of the possibility that other changes in lifestyle factors than the PRP may have influenced our results (such as change of medication, BMI, socioeconomic status or smoking habits). Still, the clinical impression is that there was not a significant change in these lifestyle factors in our patients before or after the PRP.