Correlation of a multi-cytokine panel with clinical disease activity in patients with rheumatoid arthritis

Clinical Biochemistry 43 (2010) 1309–1314

Contents lists available at ScienceDirect

Clinical Biochemistry j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / c l i n b i o c h e m

Correlation of a multi-cytokine panel with clinical disease activity in patients with rheumatoid arthritis Nataliya Milman a,b, Jacob Karsh a,b, Ronald A. Booth a,c,⁎ a b c

Faculty of Medicine, University of Ottawa, 451 Smyth Rd., Ottawa ON, Canada K1H 8M5 Department of Medicine, Division of Rheumatology, 1967 Riverside Dr. Ottawa ON, Canada K1H 7W9 Department of Pathology and Laboratory Medicine, Division of Biochemistry, The Ottawa Hospital, 501 Smyth Rd, Ottawa ON, Canada K1H 8L6

a r t i c l e

i n f o

Article history: Received 18 February 2010 Received in revised form 5 July 2010 Accepted 13 July 2010 Available online 23 July 2010 Keywords: Rheumatoid arthritis Cytokine Interleukin-6 (IL-6) Interferon-γ (IFN-γ) Health Assessment Questionnaire (HAQ) Standard 28-joint Disease Activity Score (DAS28) DMARD Infliximab Etanercept Adalimumab

a b s t r a c t Objective: Explore the potential use of a cytokine panel as biochemical markers of disease activity in rheumatoid arthritis (RA) patients. Design and methods: 57 adult RA patients were assessed using five validated clinical disease activity tools: Health Assessment Questionnaire (HAQ), standard 28-joint Disease Activity Score (DAS28), DAS28 using C-reactive protein (DAS28-CRP), Clinical Disease Activity Index (CDAI), and Simple Disease Activity Index (SDAI). Plasma cytokine levels (IL-2, IL-4, IL-6, IL-8, IL-10, VEGF, IFN-γ, TNF-α, IL1α, IL1β, MCP1, and EGF) were measured in 47 of the 57 patients and correlated with clinical indicators. Results: We found significant correlations between plasma levels of IL-6 and all clinical measures of disease activity; Spearman coefficients (p values) were: HAQ: 0.347(0.017); DAS28: 0.409(0.005); DAS-CRP: 0.378(0.011); CDAI: 0.312(0.033); SDAI: 0.310(0.039); ESR: 0.448(0.002); and CRP: 0.513(0.001). IFN-γ also correlated with DAS-CRP: 0.309(0.039) and SDAI: 0.301(0.044). Furthermore, the levels of IL-6 and IFN-γ increased significantly with worsening disease, as defined by the European League Against Rheumatism (EULAR) classification of disease activity. Conclusion: A significant correlation between plasma levels of IL-6 and clinical disease activity in patients with RA suggests a future role of IL6 as a disease activity marker. © 2010 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.

Introduction Rheumatoid arthritis (RA) is a systemic autoimmune inflammatory disorder primarily affecting peripheral joints. Left untreated, chronic inflammation leads to cartilage and bone erosions, destroying the joint architecture and causing disability. Disease modifying antirheumatic drugs (DMARDs) such as methotrexate, leflunomide, hydroxychloroquine, sulfasalazine, and prednisone have been shown to significantly slow joint destruction and reduce or sometimes prevent disability. More recently, various biologic agents, usually in combination with one or more DMARDs, were shown to be even more effective than the DMARDs used alone. The most common biologic agents used in RA are those that target tumour necrosis factor-α (TNFα), including infliximab (anti-TNF humanized mouse monoclonal antibody), etanercept (recombinant fusion protein containing 2 p75 TNF-R fused to the Fc portion of human IgG1), and adalimumab (recombinant human IgG1 anti-TNF monoclonal antibody). There is considerable variation in disease activity in treated patients with RA. Strategies employed to assess this variation include ⁎ Corresponding author. Division of Biochemistry, The Ottawa Hospital, 501 Smyth Rd, Ottawa, Ontario K1H 8L6, Canada. Fax: +1 613 737 8541. E-mail address: [email protected] (R.A. Booth).

identification of genetic factors [1], gene expression studies in peripheral blood cells [2] and synovial tissue [3], synovial cellularity and cytokine expression [4], and circulating levels and/or activity of cytokines [5]. Although all of these approaches have yielded some insights, they have limited use in everyday clinical practice because of the complex methodology involved in their measurements. In addition, there are many ways in which one could attempt to measure disease activity in RA. Most of the studies mentioned above used simple measures such as joint counts, radiographic damage, or inflammatory markers, which may not capture all aspects of disease activity in RA. For this reason, tools that incorporate several aspects of disease activity, including joint counts, visual analogue scales, questionnaires and inflammatory markers, have been validated for use both in everyday practice and in clinical studies of RA. The five most commonly used tools for measuring disease activity in RA are the Health Assessment Questionnaire (HAQ) [6], standard 28-joint Disease Activity Score (DAS28) [7], DAS28 using C-reactive protein (DAS28-CRP) [8], Clinical Disease Activity Index (CDAI) [9], and Simple Disease Activity Index (SDAI) [10]. Rheumatoid arthritis is driven by a complex interaction of multiple cytokines, with TNF-α and IL-6 as central mediators. In this pilot study we examined plasma levels of an array of 12 cytokines and chemokines in patients with RA on different treatment regimens,

0009-9120/$ – see front matter © 2010 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.clinbiochem.2010.07.012

1310

N. Milman et al. / Clinical Biochemistry 43 (2010) 1309–1314

including antil-TNF-α agents. We then attempted to correlate the levels of these cytokines with activity of RA as measured by the various standardized measures of disease activity mentioned above. Our aim was to identify patterns of cytokine expression in an attempt to identify potential biochemical markers of disease activity. We also compared expression of the studied cytokines between patients on conventional DMARDs and patients on the biologic agents. Methods Patient selection This study was approved by the Ottawa Hospital Research Ethics Board. Rheumatoid arthritis patients treated with DMARDs (e.g. azathioprine, sulfasalazine, and methotrexate) and/or one of biologic agents (abatacept, adalimumab, anakinra, etanercept or infliximab) attending the Arthritis Centre at the Ottawa Hospital, Riverside campus between December 18, 2007 and January 15, 2008, were recruited. A single member of the study group (NM) performed the clinical assessments on all patients. Consecutive patients with rheumatoid arthritis on biologic agents were selected from those registered at the rheumatology clinic. When only patients on DMARDs were registered at the time of the recruitment, consecutive patients on DMARDs were selected. Clinical data collection After written consent was obtained from the patient, demographic data were collected. Each patient completed a Health Assessment Questionnaire (HAQ), and marked patient global disease activity on a 10 cm visual analogue scale (VAS). The recruiter performed a 28-joint count and marked physician's global disease score on a 10 cm VAS. Scores were calculated for the five validated clinical disease activity tools (CDATs): HAQ [6], DAS28 [7], DAS28-CRP [8], CDAI [9], and SDAI [10]. Laboratory analysis Whole blood and plasma specimens were collected on the day of clinical assessment for erythrocyte sedimentation rate (ESR), Creactive protein (CRP), and cytokine analysis. CRP was analyzed by nephelometry (Beckman Coulter Image). The standard, non-highsensitive, assay was used for analysis which has a lower reportable limit of 1 mg/L and an interassay CV of b 4% at a level of 8 mg/L. Plasma for cytokine analysis was separated within 1 h of collection and stored frozen (−20 °C) until the time of analysis. Twelve cytokines (IL-2, IL-4, IL-6, IL-8, IL-10, VEGF, IFN-γ, TNF-α, IL1α, IL1β, MCP1, and EGF) were analyzed by immunoassay biochip microarray (Randox Evidence Investigator, UK). Each cytokine assay is performed on a 9 × 9 mm biochip with discrete test regions containing antibodies specific to each of the cytokines and is capable of simultaneously measuring all 12 cytokines from a single 100 μL sample. Measuring ranges for the analytes are: IL-2, 0–3000 pg/mL; IL-4, 0–900 pg/mL; IL-6, 0–900 pg/mL; IL-8, 0–3000 pg/mL; IL-10, 0–1000 pg/mL; IL-1α, 0–500 pg/mL; IL-1β, 0–250 pg/mL; IFNγ, 0–1500 pg/mL; EGF, 0– 900 pg/mL; MCP-1, 0–1500 pg/mL; TNFα, 0–1500 pg/mL; and VEGF, 0–3000 pg/mL, as reported by Randox. Interassay imprecision (CV) has been previously reported [11–13], and is less than 20% for all analytes measured. Statistical analysis Cytokine profiles were first compared in patients on different types of treatment (biologic agents versus DMARDs alone). The relationship between cytokine levels and the clinical disease status was then studied, as determined by the various CDATs in patients

grouped by the treatment type, as well as in all patients taken together. Finally, we compared cytokine profiles of patients with different severities of RA according to the European League Against Rheumatism (EULAR) classification of disease activity (low disease activity: DAS28 b 3.2, moderate disease activity: DAS28 between 3.2 and 5.1, high disease activity: DAS28 N 5.1 [14]). Microsoft Excel® and Sigma Stat® software applications were used for statistical analysis of data. Spearman's rank correlation coefficient was calculated to determine correlations between clinical measures and cytokine levels. NCSS/PASS statistical software was used for graphical representation of the correlations. The Mann– Whitney Rank Sum test was used to compare non-parametric values (ESR, CRP, RF, anti-CCP, and all cytokines except MCP1) between patients on biologic agents and those on DMARDs alone. Student's ttest was used to perform the same comparison for values with normal distribution (HAQ, DAS, DAS-CRP, CDAI, SDAI, and MCP1). Cytokine and clinical measures with normal distribution (see above) between 3 different EULAR groups were compared using 1-way ANOVA. A similar comparison for non-parametric values was performed using 1-way ANOVA on ranks. Results Demographics 57 patients were recruited. Cytokine analysis data was only available for 47 out of 57 patients, 35 of which were on biologic agents (with or without a DMARD), and 12 on DMARDs alone. Among these 47 patients, 1 patient did not have CRP measured and thus was excluded from the statistical tests involving CRP and DAS28-CRP, and 1 patient did not have ESR and CRP values measured and thus was excluded from analyses involving ESR, CRP, DAS28, and DAS28-CRP, as well as comparisons by EULAR classes (determined from DAS28). Demographic data as well as average values for clinical parameters for the 47 patients used for cytokine analysis are depicted in Table 1. Table 1 Demographics and baseline clinical characteristics of patients enrolled in the study. Mean (range)a Age (years) Female (%) Disease duration (years) Swollen joint count (range 0–28) Tender joint count (range 0–28) Patient VASb (range 0–100) Physician VAS (range 0–100) HAQ DAS DAS-CRP CDAI SDA1 ESR (mm/h) CRP(mg/L) Treatment, N Biologics (±DMARDs), N (%) Etanercept Adalimumab Infliximab Abatacept Anakinra DMARDs alonec, N (%) Methotrexate Hydroxych1oroquine Leflunomide Sulfasalazine Prednisone a b c

54.3 (24–81) 78.7 13.7 (1–40) 6.0 (0–22) 12.0 (0–28) 46.3 (0–91) 41.6 (0–88) 1.3 (0–2.5) 4.3 (0–8.3) 4.4 (1,0–7.9) 27.1 (0–63.0) 27.9 (0.76.5) 14.5 (0–125) 12.5 (0–231)

35 (74.5) 17 13 3 1 1 12 (25.5) 10 10 3 1 1

Means are shown with ranges in brackets unless specified otherwise. VAS, visual analogue scale. Some patients are on multiple DMARDs.

N. Milman et al. / Clinical Biochemistry 43 (2010) 1309–1314

1311

Correlation between clinical disease and cytokine levels We found a significant correlation between plasma level of IL-6 and all of the clinical measures of disease activity in all patients taken together, irrespective of the treatment type (Table 2, Fig. 1). The correlations (r) for CDATs with the corresponding significance levels (p) are: HAQ, r = 0.347, p = 0.017; DAS28, r = 0.409, p = 0.005; DASCRP, r = 0.378, p = 0.011; CDAI, r = 0.312, p = 0.033; and SDAI, r = 0.310, p = 0.039. IL-6 also correlated with the swollen joint count (r = 0.367, p = 0.012) and physician's global disease assessment (r = 0.341, p = 0.0192) (data not shown). The correlations were strongest for IL-6 versus the laboratory-based markers of inflammation: ESR r = 0.448 (p = 0.002) and CRP, r = 0.513 (p b 0.001). IFN-γ also correlated with several CDATs, specifically DAS-CRP (r = 0.309, p = 0.039) and SDAI (r = 0.301, p = 0.044). Other cytokines did not show a significant correlation with clinical disease activity, with the exception of the acute phase reactant CRP, which correlated with IL-8 (r = 0.309, p = 0.039) and IL-10 (r = 0.327, p = 0.028). The same correlations were similar when only patients on biologic agents were studied (n = 35); the correlations between IL-6 and the corresponding measures of clinical disease activity are as follows: for HAQ: r = 0.331, p = 0.052; DAS28: r = 0.449, p = 0.008; DAS28-CRP: r = 0.400, p = 0.021; CDAI: r = 0.375, p = 0.027; SDAI: r = 0.376, p = 0.031; ESR: r = 0.489, p = 0.004; and CRP: r = 0.475, p = 0.005. The correlations for IFN-γ were: for DAS28-CRP, r = 0.355, p = 0.043 and for SDAI, r = 0.327, p = 0.063. The correlations were not significant in the small groups of patients (n = 12) on DMARDs alone. Cytokine levels in different EULAR groups When patients were grouped according to EULAR classification of disease activity, we found significant increases in the levels of IL-6 and IFN-γ with worsening disease (Table 3). No statistically significant differences were found between EULAR groups in the other studied cytokines. Fig. 1. Graphical representation of correlations between plasma levels of IL6 (pg/mL) and the DAS (Disease Activity Score) or ESR. A, Log10IL-6 versus DAS28; B, Log10IL-6 versus ESR. r, Spearman's rank correlation coefficient; p, significance value.

TNF-α levels are higher in patients on biologic agents We did not find significant differences in cytokine profiles between patients on biologic agents and those on DMARDs alone, except for the level of TNF-α. TNF-α was significantly higher in patients on biologic agents (12.780 [6.455, 23.230] for median [quartile 1, quartile 3]) compared to those on DMARDs alone (5.465 [3.705, 8.585]), p = 0.022. When the specific biologic agent was assessed, we found significantly elevated level of TNF-α in patients on etanercept (18.520[13.673, 26.883]) compared to those on adalimumab (6.170[4.520, 17.358]) or DMARDs alone (5.465 [3.705, 8.585]), p b 0.05 for both comparisons.

Discussion We have demonstrated a significant and consistent correlation between plasma level of IL-6 and all studied measures of clinical disease. Importantly, such correlation was found not only for basic measures of disease activity, such as joint counts, global disease activity scores, and acute phase reactants (ESR and CRP), but also for 5 well-validated and widely used composite measures of disease activity (HAQ, DAS, DAS-CRP, CDAI, and SDAI). To our knowledge,

Table 2 Correlations between clinical measures of disease activity and serum cytokine levels*.

IL-2, pg/mL IL-4, pg/mL IL-6, pg/mL IL-8, pg/mL IL-10, pg/mL VEGF,pg/mL INF-γ,pg/mL INF-α,pg/mL IL-1α,pg/mL IL-1β,pg/mL MCPI,pg/mL EGF,pg/mL

HAQ

DAS28

DAS28-CRP

CDAI

SDAI

ESR, mm/hr

CRP, mg/L

0.018 0.151 0.347 (0.017) 0.189 0.150 0.122 0.160 −0.034 0.048 −0.017 0.026 0.099

−0.005 0.191 0.409 (0.005) 0.234 0.226 0.142 0.252 −0.009 0.084 0.012 0.145 0.112

0.021 0.256 0.378 (0.011) 0.276 0.256 0.241 0.309(0.039) 0.018 0.107 0.034 0.218 0.169

0.003 0.175 0.312 (0.033) 0.250 0.212 0.228 0.269 0.034 0.075 0.003 0.123 0.148

0.024 0.222 0.310 (0.039) 0.257 0.219 0.275 0.301 (0.045) 0.053 0.054 0.023 0.179 0.169

0.036 0.092 0.448 (0.002) 0.227 0.254 0.006 0.161 −0.003 0.147 0.051 0.145 0.041

0.037 0.173 0.513 (<0.001) 0.309 (0.039) 0.327 (0.028) 0.140 0.146 0.000 0.218 0.055 0.272 0.139

*Spearman rank order coefficients are shown in the corresponding cells with p values in brackets for statistically significant correlations (underlined).

1312

N. Milman et al. / Clinical Biochemistry 43 (2010) 1309–1314

Table 3 Cytokine levels based on EULAR classification of disease activity*.

IL-2, pg/mL IL-4, pg/mL IL-6, pg/mL IL-8, pg/mL IL-10, pg/mL VEGF, pg/mL INF-γ, pg/mL INF-α, pg/mL IL-α, pg/mL IL-β, pg/mL MCP, pg/mL EFG, pg/mL

EULAR low activity median (Q1,Q3), n = 13**

EULAR moderate activity median (Q1,Q3), n = 15

EULAR high activity median (Q1,Q3), n = 18

P value

10.400 (5.055, 19.375) 4.990 (4.150, 5.768) 2.180 (1.203, 5.355) 8.080(4.633, 14.730) 1.920 (0.920, 3.330) 15.760 (12.185, 18.723) 2.130 (1.072, 2.790) 13.290 (4.902, 24.670) 1.430 (0.610, 1.903) 1.890 (0.645, 5.630) 92.010 (78.362, 110.868) 1.470 (1.067, 3.793)

8.430 5.330 3.870 8.080 3.080 13.810 3.450 8.710 1.050 1.860 120.830 1.980

9.665 (4.210, 23.330) 6.120 (4.700, 8.262) 5.560 (2.860, 21.300) 11.185 (5.300, 16.100) 1.950 (1.50, 5.910) 18.290 (13.060, 33.960) 3.725 (2.130, 7.670) 8.585 (5.140, 18.650) 0.895 (0.570, 3.30) 1.765 (0.960, 6.650) 111.135(83.590, 151.890) 2.015 (1.230, 4.840)

NS NS 0.042 NS NS NS 0.043 NS NS NS NS NS

(4.492, 11195) (4.700, 8.262) (1.660, 5.475) (5.090, 11.145) (1.595, 4.672) (11.458, 26.762) (2.458, 4.520) (5.242, 19.960) (0.495, 1.620) (0.705, 3.765) (84.715, 130.818) (1.080, 2.870)

*“low” disease activity: DAS28 b 3.2, “moderate” activity: DAS28 between 3.2 and 5.1, “high” activity: DAS28 N 5.1. **EULAR class could not be determined for 1 patient. Comparison between three groups was performed using ANOVA on Ranks test, and significant p values are shown (values highlighted). NS, non-significant.

the relationship between IL-6 and composite measures of disease activity has not been studied before. Also of note is that this crosssectional study demonstrated the correlation across a heterogeneous group of RA patients on different treatments (including biologic agents) and at different time points in their disease course (Table 1). In addition, we have found a statistically significant increase in IL-6 levels with worsening disease based on EULAR classification of disease activity, further supporting the correlation between IL-6 and clinical disease. IL-6 plays a central role in the pathogenesis of RA. IL-6-deficient mice are either completely resistant to the development of antigeninduced arthritis (AIA) and collagen-induced arthritis (CIA), or develop delayed and less severe CIA [15]. Furthermore, the level of plasma IL-6 is elevated in RA patients compared to healthy controls [16–18], and therapy directed at blocking IL-6 activity with tocilizumab, a humanized antihuman IL-6 receptor monoclonal antibody, has been shown to be effective in treating patients with RA in several phase III clinical trials [19,20]. A correlation between both synovial and serum IL-6 levels and clinical disease activity has been suggested by several previous sources. Miltenburg et al. demonstrated a correlation (r = 0.717; p b 0.001) between synovial fluid (SF) IL-6 level and the local joint activity score (scored using joint temperature, swelling, and pain) in patients with different arthritides [21]. Madhok et al. [22] found an association between serum IL-6 and the Ritchie Articular Index (r = 0.3, p = 0.01), duration of morning stiffness (r = 0.2, p = 0.03), and CRP (r = 0.2, p = 0.05) in 93 patients with RA referred for initiation of DMARDs. Multiple additional sources [18,23] have demonstrated a correlation between serum IL-6 and the acute phase reactants, consistent with the known function of IL-6 as a stimulator of hepatic synthesis of acute phase reactants. Van Leeuwen studied IL-6 in relation to acute phase reactants and clinical disease in 51 patients with early RA prospectively over 3 years, as well as in cross-section [24]. Over 3 years, within-individual plasma IL-6 levels correlated with acute phase reactants (mean correlation for CRP was 0.595, for ESR 0.575) as well as with clinical markers (mean correlation for swollen joint count 0.470, tender joint count 0.423, and for Ritchie articular index 0.484). Cross sectional between-patient correlations of IL-6 and acute phase reactants were even stronger (Spearman coefficient for CRP was 0.605 and for ESR was 0.744), while the cross-sectional correlation between time integrated level of IL-6 and radiological progression of disease was found to be nonsignificant (r = 0.226, p b 0.40). Several sources have demonstrated decreases in IL-6 levels in response to various therapies. Braun-Moscovici et al. studied levels of IL-6 following administration of infliximab [25]. They found a sustained decrease in IL-6 levels in patients who responded to therapy, whereas a transient decrease was followed by increase back to baseline in non-responders. In a different study, Knudsen et al.

found a significant and sustained decrease in circulating IL-6 level in 13 patients with RA who responded (according to American College of Rheumatology (ACR) 20 response) to infliximab and methotrexate therapy, while no significant change in IL-6 level or DAS28 in 4 nonresponders [26]. Interestingly, the same authors found a correlation between baseline IL-6 level and total Sharp score at week 52 of infliximab therapy, identifying IL-6 as a potential marker of subsequent disease progression. Similar decreases in circulating IL-6 paralleled by clinical improvement were demonstrated with leflunomide [27], methotrexate [28], and gold [29]. The present study was a cross-sectional study and therefore was not designed to study IL-6 level in response to therapy. However, the decrease in IL-6 levels with clinical improvement following various RA therapies demonstrated by the above mentioned studies supports the correlation between serum IL-6 and clinical disease demonstrated in this study. In the future, IL-6 may become a clinically useful early marker of response to treatment and disease progression, however further research is required to clarify its role. We have also identified a correlation between plasma IFN-γ and 2 of the measured CDATs (specifically, DAS-CRP and SDAI). Furthermore, a significant increase in IFN-γ was seen with worsening EULAR disease activity status. Studies attempting to identify the role of IFN-γ in rheumatoid arthritis have yielded mixed results. Initially, recombinant IFN-γ was attempted as a therapy for RA. While a trial by Cannon et al. failed to show effectiveness of IFN-γ [30], several later trials showed benefit of IFN-γ compared to placebo [31,32]. Recently, the opposite strategy, namely neutralization of IFN-γ, was tested by a group of Russian investigators who suggested it as a possible alternative to anti-TNFα agents in patients with resistant RA [33]. Synovial expression of IFN-γ (as well as its receptor) is higher in patients with RA compared to controls with non-inflammatory arthritis [34]. In addition, an association between dinucleotide repeats in noncoding region of IFN-γ gene and susceptibility to, as well as severity of rheumatoid arthritis, was suggested by a case-control study by KhaniHajani et al. [35], although a subsequent prospective study by Constantin et al. failed to confirm such an association [36]. An association between serum IFN-γ and clinical disease was tested by Verhoef et al. by measuring serum IFN-γ in two groups of patients with different disease severities [37]. In contrast to the present study, this group did not find a correlation between serum IFN-γ and the extent of radiographic joint damage or any of the other clinical variables. We found increased plasma levels of TNF-α in patients on biologic agents compared to those on DMARDs alone. The increased levels of TNF-α seen in patients on biologic agents may be due to the worse disease in those patients compared to patients on DMARDs alone, however previous reports have suggested that administration of antiTNF-α agents may increased measured TNF-α levels. Charles et al. studied serum TNF-α levels after administration of increasing doses of infliximab and found a dose-depended increase in serum TNF-α

N. Milman et al. / Clinical Biochemistry 43 (2010) 1309–1314

followed by a gradual decrease towards baseline levels after a single dose of infliximab [23]. The authors found that this increased TNF-α was not biologically active, and postulated that the measured levels represented a high molecular weight complex between TNF-α overproduced during active disease and the monoclonal antibody (infliximab) administered during treatment. Of interest, we found higher TNF-α levels in patients on etanercept compared to patients on adalimumab. The clinical significance of this observation is not clear, and may simply be the result of variable detection of the high molecular weight drug–TNF-α complexes by the assay used in our study.

[2]

[3]

[4]

[5]

Study limitations This study was performed in search of potential biochemical markers and/or predictors of disease activity. We realize that ideally to look for such predictors one would also want to look at withinpatient variability in cytokine expression in relation to disease activity through a longitudinal study, preferably with multiple assessments before as well as after initiation of therapy. This, in contrast, was a cross-sectional study involving a single clinical assessment and biochemical analysis for each patient. Therefore, we looked for differences in cytokine profiles between patients with different levels of disease control. Due to the small sample size, our statistical analysis was limited to basic correlations and comparisons; multivariate and cluster analyses (which could be useful in the analysis of subsets of cytokines), would require a larger number of patients to identify meaningful results. The correlations that we found, although statistically significant, are not strong. The correlations are strongest between IL-6 and the acute phase reactants; however the difference in IL-6 accounts for only roughly 26% of the variability of CRP. The correlations between the cytokines and the CDATs are somewhat weaker, with IL-6 accounting for 17% of the variation in DAS28. We have performed a large number of statistical tests on the data taken from a relatively small number of patients, which raises the possibility that some of the significant results occurred due to chance alone. However, consistency of the correlation between IL-6 and clinical disease across all clinical measures suggests that this association is, indeed, significant. Conclusion and future directions We have found a significant and consistent correlation between plasma level of IL-6 and all studied measures of clinical disease activity in a cross-sectional study of a heterogeneous group of RA patients on different treatments and with different disease severities. We have also demonstrated a significant elevation in the level of plasma IL-6 with worsening disease based on EULAR classification of disease activity. This observation not only supports the increasing awareness of the key importance of IL-6 in the pathogenesis of RA, but also suggests an additional role of IL-6 as a potential marker of disease activity. A larger scale prospective study looking at plasma levels of IL6 before and after the initiation of various RA therapies would be necessary to confirm this hypothesis.

[6] [7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15] [16]

[17]

[18]

[19]

[20]

[21]

[22]

Acknowledgment

[23]

This study was supported by a grant from the Ottawa Hospital Pathology and Laboratory Medicine Academic Enrichment Fund.

[24]

References [25] [1] Chatzikyriakidou A, Georgiou I, Voulgari PV, Venetsanopoulou AI, Drosos AA. Combined tumour necrosis factor-alpha and tumour necrosis factor receptor genotypes could predict rheumatoid arthritis patients' response to anti-TNF-alpha

1313

therapy and explain controversies of studies based on a single polymorphism. Rheumatology 2007;46:1034–5. Lequerre T, Gauthier-Jauneau AC, Bansard C, Derambure C, Hiron M, Vittecoq O, et al. Gene profiling in white blood cells predicts infliximab responsiveness in rheumatoid arthritis. Arthritis Res Ther 2006;8:R105, doi:10.1186/ar1990. van der Pouw Kraan TC, Wijbrandts CA, van Baarsen LG, Rustenburg F, Baggen JM, Verweij CL, et al. Responsiveness to anti-TNFalpha therapy is related to pretreatment tissue inflammation levels in rheumatoid arthritis patients. Ann Rheum Dis 2008;67(4):563–6. Wijbrandts CA, Dijkgraaf MGW, Kraan MC, Vinkenoog M, Smeets TJ, Dinant H, et al. The clinical response to infliximab in rheumatoid arthritis is in part dependent on pre-treatment TNF-alpha expression in the synovium. Ann Rheum Dis 2008;67 (8):1139–44. Marotte H, Maskinski W, Miossec P. Circulating tumour necrosis factor-alpha bioactivity in rheumatoid arthritis patients treated with infliximab: link to clinical response. Arthritis Res Ther 2005;7(1):R149–55. Fries JF, Spitz PW, Kraines RG, Holman HR. Measurement of patient outcomes in arthritis. Arthritis Rheum 1980;23:137–45. van der Heijde DM, Van't Hof MA, van Riel PL, Theunisse LA, Lubberts EW, van Leeuwen MA, et al. Judging disease activity in clinical practice in rheumatoid arthritis: first step in the development of a disease activity score. Ann Rheum Dis 1990;49:916–20. Prevoo MLL, van't Hof MA, Kuper HH, Van Leeuwen MA, Van De Putte LBA, Van Riel PLCM. Modified disease activity scores that include twenty-eight-joint counts. Arthritis Rheum 1995;38:44–8. Aletaha D, Nell VPK, Stamm T, Uffmann M, Pflugbeil S, Machold K, et al. Acute phase reactants add little to composite disease activity indices for rheumatoid arthritis: validation of a clinical activity score. Arthritis Res Ther 2005;7: R796–806. Smolen JS, Breedveld FC, Schiff MH, Kalden JR, Emery P, Eberl G, et al. A simplified disease activity index for rheumatoid arthritis for use in clinical practice. Rheumatology 2003;42:244–57. Kavsak PA, Lee A, Hirte H, Young E, Gauldie J. Cytokine elevations in acute coronary syndrome and ovarian cancer: a mechanism for the up-regulation of the acute phase proteins in these different disease etiologies. Clin Biochem 2008;41:607–10. Kavsak PA, Ko DT, Newman AM, Lustig V, Palomaki GE, et al. Vascular versus myocardial dysfunction in acute coronary syndrome: are the adhesion molecules as powerful as NT-proBNP for long-term risk stratification? Clin Biochem 2008;41 (6):436–9. Kavsak PA, Ko DT, Newman AM, Palomaki GE, Lustig V, Macrae AR, et al. Upstream markers provide for early identification of patients at high risk for myocardial necrosis and adverse outcomes. Clin Chim Acta 2008;387:133–8. Van Gestel AM, Prevoo MLL, van't Hof MA, et al. Development and validation of the European League Against Rheumatism response criteria for rheumatoid arthritis. Artritis Rheum 1996;39:34–40. Park JY, Pillinger MH. Interleukin-6 in the pathogenesis of rheumatoid arthritis. Bull NYU Hosp Jt Dis 2007;65(Suppl1):S4–S10. Knudsen LS, Christensen IJ, Lottenburger T, Svendsen MN, Nielsen HJ, Nielsen L, et al. Pre-analytical and biological variability in circulating interleukin 6 in healthy subjects and patients with rheumatoid arthritis. Biomarkers 2008;13(1):59–78. Knudsen LS, Ostergaard M, Balund B, Narvestad E, Petersen J, Nielsen HJ, et al. Plasma IL-6, plasma VEGF, and serum YKL-40: relationship with disease activity and radiographic progression in rheumatoid arthritis patients treated with infliximab and methotrexate. Scand J Rheumatol 2006;35:489–91. Lacki JK, Klama K, Mackiewicz SH, Mackiewicz U, Muller W. Circulating interleukin 10 and interleukin-6 serum levels in rheumatoid arthritis patients treated with methotrexate or gold salts: preliminary report. Inflamm Res 1995;44(1):24-24. Nishimoto N, Miyasaka N, Yamamoto K, Kawai S, Takeuchi T, Azuma J. Long-term safety and efficacy of tocilizumab, an anti-interleukin-6 receptor monoclonal antibody, in monotherapy, in patients with rheumatoid arthritis (the STREAM study): evidence of safety and efficacy in a 5-year extension study. Ann Rheum Dis 2009;68:1580–4. Nishimoto N, Miyasaka N, Yamamoto K, Kawai S, Takeuchi T, Azuma J, Kishimoto T. Study of active controlled tocilizumab monotherapy for rheumatoid arthritis patients with an inadequate response to methotrexate (SATORI): significant reduction in disease activity and serum vascular endothelial growth factor by IL-6 receptor inhibition therapy. Mod Rheumatol 2009;19(1):12–9. Miltenburg AMM, van Laar JM, de Kuiper R, Daha MR, Breedveld FC. Interleukin-6 activity in paired samples of synovial fluid. Correlation of synovial fluid interleukin-6 levels with clinical and laboratory parameters of inflammation. British J Rheumatol 1991;30:186–9. Madhok R, Crilly A, Watson J, Capell HA. Serum interleukin 6 levels in rheumatoid arthritis: correlations with clinical and laboratory indices of disease activity. Ann Rheum Dis 1993;52:232–4. Charles P, Elliott MJ, Davis D, Potter A, Kalden JR, Antoni C, Breedveld FC, Smolen JS, Eberl G, de Woody K, Feldmann M, Maini RN. Regulation of cytokines, cytokine inhibitors, and acute-phase proteins following anti-TNF-alpha therapy in rheumatoid arthritis. J Immunol 1999;163:1521–8. Van Leeuwen MA, Westra J, Limburg PC, van Riel PLCM, van Rijswijk MH. Clinical significance of interleukin-6 measurement in early rheumatoid arthritis: relation with laboratory and clinical variables and radiological progression in a three year prospective study. Ann Rheum Dis 1995;54:674–7. Braun-Moscovici Y, Markovits D, Zinder O, Schapira D, Rozin A, Ehrenburg M, Dain L, Hoffer E, Menahem Nahir A, Balbir-Gurman A. Anti-cyclic citrullinated protein antibodies as a predictor of response to anti-tumor necrosis factor-alpha therapy in patients with rheumatoid arthritis. J Rheumatol 2006;33:497–500.

1314

N. Milman et al. / Clinical Biochemistry 43 (2010) 1309–1314

[26] Knudsen LS, Ostergaard M, Baslund B, Narvestad E, Petersen J, Nielsen HJ, Ejbjerg BJ, et al. Plasma IL-6, plasma VEGF, and serum YKL-40: relationship with disease activity and radiographic progression in rheumatoid arthritis patients treated with infliximab and methotrexate. Scand J Rheumatol 2006;35:489–95. [27] Litinsky I, Paran D, Levartovsky D, Wigler I, Kaufman I, Yaron I, et al. The effects of leflunomide on clinical parameters and serum levels of IL-6, IL-10, MMP-1 and MMP-3 in patients with resistant rheumatoid arthritis. Cytokine 2006;33:106–10. [28] Crilly A, McInness IB, McDonald AG, Watson J, Capell HA, Madhok R. Interleukin 6 (IL-6) and soluble IL-2 receptor levels in patients with rheumatoid arthritis treated with low dose oral methotrexate. J Rheumatol 1995;22(2):224–6. [29] Madhok R, Crilly A, Murphy E, Smith J, Watson J, Capell HA. Gold therapy lowers serum interleukin 6 levels in rheumatoid arthritis. J Rheumatol 1993;20(4): 630–3. [30] Cannon GW, Pincus SH, Emkey RD, Denes A, Cohen SA, Wolfe F, et al. Double-blind trial of recombinant γ-interferon versus placebo in the treatment of rheumatoid arthritis. Arthritis Rheum 1989;32(8):S79–88. [31] German Lymphokine Study Group. Double blind controlled phase III multicenter clinical trial with interferon gamma in rheumatoid arthritis. Rheumatol Int 1992;12:175–85.

[32] Machold KP, Neumann K, Smolen JS. Recombinant human interferon γ in the treatment of rheumatoid arthritis: double blind placebo controlled study. Ann Rheum Dis 1992;51:1039–43. [33] Nasonova VA, Kukina GV, Sigidin IA. Neutralisation of interferon gamma — a new trend in therapy of rheumatoid arthritis. Ter Arkh 2008;80(5):30–7. [34] Dolhain RJEM, Ter Haar NT, Hoefakker S, Tak PP, De Ley M, Claassen E, et al. Increased expression of interferon (IFN)-gamma receptor in the rheumatoid synovial membrane compared with synovium of patients with osteoarthritis. Rheumatology 1996;35(1):24–32. [35] Khani-Hanjani A, Lacaille D, Hoar D, Chalmers A, Horsman D, Anderson M, et al. Association between dinucleotide repeat in non-coding region of interferongamma gene and susceptibility to, and severity of, rheumatoid arthritis. Lancet 2000;356(9232):820–5. [36] Constantin A, Navaux F, Lauwers-Cances V, Abbal M, van Meerwijk JP, Mazieres B, et al. Interferon gamma gene polymorphism and susceptibility to, and severity of, rheumatoid arthritis. Lancet 2001;358(9298):2051–2. [37] Verhoef CM, van Roon JA, Vianen ME, Bijlsma ME, Lafeber FP. Interleukin 10(IL10), not IL-4 or interferon-gamma production, correlates with progression of joint destruction in rheumatoid arthritis. J Rheum 2001;28(9):1960–6.