Effects of Spinal Manipulative Treatment on Inflammatory Markers in Low Back Pain Patients

Background: Spinal manipulative treatment(SMT) may work by reducing mechanical irritation to joint tissues (1) and thereby diminishing local inflammation. Evidence for an anti-inflammatory effect of instrument-assisted SMT has been observed in an animal model (2). These findings are consistent with the recently proposed hypothesis that the origin of all pain may be associated with inflammation and augmented production of inflammatory mediators (cytokines), principally TNFα (3). TNFα plays a major role in the pathophysiology of neuropathic pain-associated inflammation (4) and has been implicated in spinal pain syndromes, including intervertebral disc-related low back pain (5, 6) and sciatica (7,8). Our recent studies demonstrated that production of inflammatory mediators is elevated in patients with chronic cervical spine pain, both in vitro and in vivo, and accompanied by up-regulation of chemokine (CC) synthesis (9).

Recent reports on clinical (10), animal (2) and human in vitro models (11) suggest that SMT may exert an anti-inflammatory effect. Thus, Song et al. (2) found that SMT reduced inflammatory neuropathic pain. Teodorczyk-Injeyan et al. (11) demonstrated a significant attenuation of pro-inflammatory cytokine production in vitro, and no changes in serum substance P (SP) levels following SMT. Other studies from our laboratory showed that SMT may enhance both the production of and the response to the immunomodulatory cytokine, IL-2, and IL-2-dependent antibody synthesis (12, 13). Reduction in serum TNFα levels has been reported for cervicogenic headache patients (n=2, case studies) (14, 15). These observations suggest that SMT effects may be transduced into cellular components of the immune system.

Thus far, the clinical relevance of the effect of SMT as a modulator of inflammatory mediator production is unknown. Despite evidence of inflammatory pathophysiology of spinal pain (16), including a subtype of non-specific spinal pain (17), only one clinical study (18) evaluated the correlation between serum TNFα levels, pain intensity and back function.

In this proposed study we intend to use a clinical model to investigate the baseline levels of proinflammatory cytokines in acute and chronic low back pain patients and explore the potential anti-inflammatory effect of SMT following a course of manipulative treatments. Thus, Aim 1 is to determine baseline pro-inflammatory cytokine levels in individuals experiencing acute or chronic lower spinal pain of mechanical etiology, and compare them with asymptomatic controls'. Aim 2 is to explore the relationships between SMT, pain level and functional impairment, and the production of inflammatory mediators relative to baseline.

Anti-inflammatory cytokines (IL-10 and IL-1ra) have been found to be produced alongside, and in parallel, with their respective pro-inflammatory counterparts (TNFα and IL-1β), and act in concert to sustain/restore homeostasis (19). The proposed study will include determinations, in addition to pro-inflammatory cytokine (IL-1, TNFα, IL-6) levels, of the levels of IL-1 receptor antagonist (IL-1ra) and IL-10. Assessment of IL-10 synthesis is particularly relevant as this cytokine up-regulates the production of IL-1ra, which competes with active IL-1 for binding to IL-1 receptors, and which acts as a potent natural anti-inflammatory protein (19, 20).

Study design; Subject Recruitment: Subjects (volunteers) of both sexes, between the ages 20 and 60 years, experiencing acute (less than 4 weeks in duration) or chronic (12 weeks or longer in duration) mechanical low back pain (experienced between spinal levels L1- L5, with or without sacroiliac [SI] joint involvement) will be recruited by Canadian Memorial Chiropractic College (CMCC) personnel and posters from CMCC's outpatient clinics, and from the general public through newspaper advertisements . CMCC is in the Greater Toronto Area (GTA), where the population is a diverse ethnic mix (21). Potential participants who have presented to one of the CMCC clinics for the purpose of treatment will be encouraged/recruited by interns/clinicians prior to commencement of treatments. Others who present to participate in the research study will first be assigned to one of the clinic pods for initial assessment prior to entering the study. In no case, however, will participants have received SMT 4 weeks prior to commencement of the study.

Candidates will be interviewed in order to determine eligibility (see exclusion criteria, Appendix 1). They will complete the research intake form (Appendix 2) and be given detailed explanations about the research protocol (Appendix 3). A cohort of matching healthy asymptomatic subjects, recruited from the general population, will serve as controls for the determination of baseline cytokine levels.

Sample size determination: Data published for TNFα levels in chronic neck pain patients versus asymptomatic controls (9) were used to calculate a sample size estimate for this study. The study is powered for the primary outcome measure related to Question 1, looking at the difference between symptomatic and asymptomatic low back pain subjects at baseline (see Statistical Analyses). From Cohen's table (22), based on a power of 0.8, a two-tailed test with a p value < 0.05, the sample size was estimated to be 17 per group. In order to account for drop-outs and errors that may arise in the blood cultures, the sample size will be increased to 20 per group. As a result, there will be 40 symptomatic subjects and 20 asymptomatic controls, which should provide a more than adequate sample to test the primary outcome.

Subject assessment and group assignment: Qualified subjects will be scheduled. They will be greeted by one of two investigators (depending on recruitment site) , who will brief them and review the intake/eligibility form (Appendix 3) to confirm their eligibility before asking them to sign the informed consent form (Appendix 4). All subjects will then be asked to indicate their pain intensity level on the 10-point visual analogue scale (VAS) and complete the Oswestry functional disability questionnaire. They will then be assessed with standard chiropractic, orthopaedic and neurological tests by their respective chiropractic interns and supervising clinicians, who will assign them to the acute or chronic LBP groups, and who will formulate and explain a treatment plan.. For the duration of the study, treatments will consist of manual SMT and, where needed, manual (not instrument-assisted) soft tissue work provided by the clinician. Other treatment modalities will not be used.

Patients will receive 6 treatments over the course of 2 weeks. (comprised of manipulation of one lumbar or sacroiliac articulation, with or without soft tissue therapy). SMT will consist of a single high velocity, low amplitude (HVLA) thrust intended to cavitate and restore mobility to the joint. If at initial presentation a manipulative segment could not be identified, the patient will be excluded from the study. If on the other hand a manipulable segment is not found at subsequent visits, the clinician will limit the treatment to palpation and some soft tissue work as indicated . In all cases a blood sample will be drawn (see below) after initial assessment and prior to commencement of treatments of patients, in order to establish a baseline level of cytokines for each participant (primary outcome). At their 7th visit (at least 2 days after last treatment), a blood sample will be taken prior to commencement of any further treatment, and they will be asked to complete the exit questionnaire including a VAS (Appendix 5). Should a patient recover following a few treatments, as assessed by subjective feedback from the patient and a VAS result of less than 3/10, then a blood sample will be taken earlier and such will be duly noted. Should a patient require continued treatments beyond the 6 stipulated by the study, s/he will be free to do so under the direction of the clinician in charge of their case.

Interventions: Manipulation: As noted above, each spinal manipulation will consist of a HVLA thrust to an affected segment (23). A clinician will deliver treatments according to his/her assessment findings on a given day.

Venipuncture: On the day of admission into the study and at the completion of SMT therapy, an experienced phlebotomist will perform venipuncture using standard procedures (antecubital fossa, 21gauge needle) in the seated position. Heparinized blood samples (10 ml each) will be collected and transferred (at room temperature) to the laboratory within an hour of collection for the preparation of whole blood cultures as described below .

Laboratory studies

1. Induction of inflammatory cytokines To assess cytokine production in vitro, a whole blood (WB) culture system will be utilized (19). Briefly, multiple sets of WB cultures representing different treatment/culture conditions for each subject will be prepared. Cultures will be stimulated at initiation with 10 μg/ml lipopolysaccharide (LPS) for the induction of TNFα and IL-1β production. Phytohemagglutinin (PHA, 10 µg/ml) alone and in combination with LPS will be used to induce production of the anti-inflammatory cytokine, IL-10 and 2 chemokines, CCL2 and CCL3 . Cultures will be maintained at 37C (Celsius) in a humidified 5% C02 incubator. As clinical conditions involving inflammatory responses may cause a time shift in the capacity for cytokine production (19, 27), the levels of the studied mediators (TNFα, IL-1β, IL-1ra, IL-10, CCL2 and CCL3) will be examined at 24 h intervals in culture supernatants harvested between 24-72h post-initiation. Aliquots of the supernatants will be stored at -76C (Celsius) until tested. This model will allow investigation of the relation between the releases of pro- and anti-inflammatory mediators.
2. Determinations of cytokine levels by enzyme-linked immunosorbent assay The levels of TNFα, IL-1β , and IL-10 and IL-1ra in supernatants from whole blood cultures will be determined by specific enzyme-linked immunosorbent assays (ELISA) using DuoSet ELISA development system for natural and recombinant human cytokines (R&D Systems, Minneapolis, MN) as described previously (9, 11), and Quantikine Immunoassay Kits will be used for CCL2 and CCL3 determinations. All immunoassay procedures, including reagent and sample diluent preparations, will be carried out as per manufacturer's recommendations. Each of the studied samples will be tested at least twice at 2-4 different dilutions.

Statistical analyses All of the studied mediators will be measured for the LBP patients and asymptomatic subjects at baseline and at the completion of the treatment period. For Question 1 (the primary outcome), baseline comparison for each TNFα and IL-1β will be compared between symptomatic subjects (acute and chronic) versus asymptomatic control subjects. Two unpaired t-tests will be used to test the associated hypothesis. For Question 2, the mean pre-post difference scores will be compared for each of TNFα and IL-1β between the three groups (acute, chronic and asymptomatic) using an ANOVA. If baseline values are significantly different between groups (as hypothesized from Question 1), then ANCOVA will be used to account for this difference. It is anticipated that the sample size is sufficiently large to accommodate this analysis. However, after determining the results for Question 2, a power analysis will be completed. If the study is insufficiently powered for this question, then another sample size estimate will be calculated to inform future work.

For other pro-inflammatory and anti-inflammatory cytokines measured in this study, regression modeling will be used in an attempt to predict cytokine responses. It is understood that any model created during this process will need to be confirmed in a future investigation. Both models created and descriptive data derived from the investigation will be used to inform future work.

Time frame Based on experience, up to 5 subjects per week can be tested. However, based on condition prevalence, we project that patient recruitment and sample collection will take 8-12 months. Cytokine level determinations in cell cultures derived from the three groups of participants will lag behind by 3-4 months, and data analysis will be completed within another 2 months. Preparation of manuscripts will require an additional 4-6 months. Thus we anticipate completing this project within 18 -24 months of initiation of the study.