CCL25 in the cerebrospinal fluid is negatively correlated with fatigue in chronic pain patients

Background and aim

Chronic pain is often accompanied by other symptoms such as fatigue and sleep disturbance, and these symptoms all correlate with neuroimmune activation. However, their relation to one another on a neuroimmune axis remains elusive. Based on a recent review, cytokines in the cerebrospinal fluid (CSF) seem to be generally upregulated in patients with chronic pain compared to controls, disregarding pain type. Some of these have the possilibity of altering blood–brain barrier (BBB) permeability. Therefore, cytokine levels in serum and CSF, as well as BBB permeability, were measured in a cohort of patients suffering from either degenerative disc disease (DDD), lumbar disc herniation (LDH) or osteoarthritis (OA). In this exploratory study, we were interested in whether cytokines in the serum or CSF are associated with sleep disturbance or fatigue, with special consideration of the effect of BBB permeability, and whether functional clusters can be found among these cytokines.

Method

One-hundred-twenty patients with DDD/LDH/OA, all awaiting surgery, were included. Blood and CSF were collected on the day of surgery. Pain was measured with a visual analog scale 0–100 mm, sleep disturbance was assessed using Pittsburgh Sleep Quality Index (PSQI), and fatigue was measured using the Multidimensional Fatigue Inventory (MFI). A 92-protein multiplex panel (OLINK, Sweden) was used to analyze cytokine expression in serum and CSF, respectively. CSF-serum albumin quotient was measured using ELISA. Non-parametric statistics were used for univariate analyses, and a false discovery rate (FDR) < 0.10 was considered statistically significant. Bonferroni correction was applied to all multivariable protein analyses to obtain conservative effect estimates.

Main results

There was an association between BBB permeability and serum-CSF dynamics: thirty-one cytokines showed significant CSF-serum correlation, and BBB permeability was significantly correlated to the quotients of 35 cytokines and to the CSF levels of 11 cytokines. Several cytokines were negatively correlated to both pain at rest and general fatigue. No correlations were found between sleep disturbance and cytokines. Network analyses of serum and CSF cytokines that were correlated with fatigue revealed functional clusters in both compartments. Anxiety, depression, and pain during rest were important regressors for sleep disturbance with an R² = 0.41. In addition to depression and pain during rest, CSF levels of CCL25 was a significant regressor regarding general fatigue, with an R² = 0.47.

Discussion and conclusion

In this exploratory study of immune profiles in chronic pain cohorts awaiting surgery, the importance of BBB dynamics on serum-CSF cytokine dynamics, and to a lesser extent on central levels of cytokines, is highlighted. Surprisingly, there were no associations between any cytokines in serum or CSF and sleep disturbance, despite a high prevalence of clinically significantly disturbed sleep. In contrast, several associations between general fatigue and cytokine levels in both serum and CSF were found. Cytokines of note are CXCL11 and CCL25 in the CSF, especially because of their direct functional association. CXCL11 has been found to exert neuroprotective effects in animal models, while CCL25 is known as a proinflammatory cytokine and is the only cytokine to fall out as a significant negative contributor to degree of fatigue. While causality cannot be addressed, the negative correlation between CCL25 and fatigue in both univariate and multivariate analyses implies that neuroimmune activity might have an ameliorating effect on the degree of fatigue.

This study further adds to existing evidence that centrally acting cytokines are associated with severity of symptomatology, and highlights that pain and fatigue seem to have slightly different cytokine profiles. Sleep disturbance needs to be further addressed, ideally using both subjective and objective assessment methods. CCL25 and CXCL11 are interesting biomarkers for future research on pain and associated symptoms.