Article Text

Download PDFPDF

Systemic disease and the heart
Lowering interleukin-1 activity with anakinra improves myocardial deformation in rheumatoid arthritis
  1. I Ikonomidis1,
  2. S Tzortzis1,
  3. J Lekakis1,
  4. I Paraskevaidis1,
  5. I Andreadou2,
  6. M Nikolaou1,
  7. T Kaplanoglou3,
  8. P Katsimbri3,
  9. G Skarantavos3,
  10. P Soucacos3,
  11. D Th Kremastinos1
  1. 1
    2nd Cardiology Department, Attikon Hospital, University of Athens, Greece
  2. 2
    Department of Pharmaceutical Chemistry, University of Athens School of Pharmacy, Athens, Greece
  3. 3
    1st Department of Orthopaedics, Attikon Hospital, University of Athens, Greece
  1. Correspondence to Dr Ignatios Ikonomidis, University of Athens, Perikleous 19, N Chalkidona, Athens, 14343, Greece; ignoik{at}


Objective: Inhibition of interleukin-1 activity improves nitro-oxidative stress, endothelial and coronary function. The authors investigated (a) the association of nitro-oxidative stress and endothelial function with myocardial deformation, (b) the effects of anakinra, an interleukin-1a receptor antagonist on myocardial deformation in patients with rheumatoid arthritis (RA).

Methods: The authors compared 46 RA patients to 23 normal controls. 23 patients received anakinra (150 mg subcutaneously once daily) and 23 patients a 5-mg increase of prednisolone dose for 30 days. At baseline and post-treatment this study assessed (a) the left ventricular (LV) longitudinal, circumferential and radial strain and strain rate, using speckle tracking echocardiography, (b) the coronary flow reserve (CFR), (c) the flow-mediated endothelial-dependent dilation of the brachial artery (FMD) and (d) nitrotyrosine (NT) and malondialdehyde blood levels.

Results: Patients had impaired baseline myocardial deformation indices compared to controls (p<0.05). CFR and NT levels were related to longitudinal strain, systolic and diastolic strain rate, circumferential strain and systolic strain rate (p<0.05). FMD was related to longitudinal and circumferential diastolic strain rate (p<0.01). Compared to baseline, anakinra-treated patients increased the longitudinal strain (−17.8% (3.7%) vs −22.1% (3.5%)), systolic (−1.02 (0.23) l/s vs −1.25 (0.23) l/s) and diastolic (0.96 (0.37) l/s vs 1.20 (0.39) l/s) longitudinal strain rate, circumferential strain and strain rate (p<0.05 for all comparisons). No significant changes were observed among prednisolone-treated patients

Conclusions: Myocardial deformation is impaired in RA patients and is related to nitro-oxidative stress and endothelial dysfunction. Chronic inhibition of IL-1 improves LV deformation in parallel with endothelial function and nitro-oxidative stress.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

The inflammatory processes in patients with rheumatoid arthritis (RA) are strongly linked to an excess risk of cardiovascular disease and mortality.1 In this process interleukin-1 (IL-1) activity may cause myocardial cell damage and endothelial dysfunction leading to an impaired coronary flow reserve (CFR).2 3 The adverse effects of IL-1 on myocardial and endothelial cells are mediated by an enhanced nitro-oxidative stress4 5 and production of other inflammatory mediators.6 7 8

Anakinra, a recombinant form of human IL-1 receptor antagonist (IL-1ra), is commonly used for the treatment of RA.9 We have previously shown that treatment with anakinra reduces IL-1-mediated nitro-oxidative stress, leading to an improvement in endothelial function (as expressed by CFR and flow-mediated dilation of the brachial artery (FMD)) as well as in tissue Doppler-derived parameters of left ventricular (LV) function.10 However, tissue Doppler imaging (TDI) velocities of the mitral annulus may offer information only for the longitudinal component of LV function and are angle and load dependent. Conversely, speckle tracking echocardiography is considered a reliable method for the assessment of longitudinal, radial and circumferential LV myocardial deformation during systole and diastole11 and is angle and load independent. Studies have shown that LV myocardial systolic strain and long axis function are impaired even in the presence of preserved ejection fraction.12 13 14

In the present study, we hypothesised that nitro-oxidative stress in RA patients affects endothelial function, coronary microcirculation and myocardial function leading to impaired LV myocardial deformation. Thus, we investigated (a) whether LV myocardial deformation as assessed by two-dimensional (2D) speckle tracking echocardiography11 is impaired in RA patients compared to normal controls; (b) the association of endothelial function (as assessed by FMD15 and CFR), markers of nitro-oxidative stress (malondialdehyde and nitrotyrosine) on LV myocardial deformation parameters; and (c) the effects of chronic treatment with anakinra in comparison to the respective effects of treatment with corticosteroids on LV myocardial deformation.


Study population and protocol

We examined 46 patients (mean age 56 (SD 16) years, 31 females) with RA (American Rheumatism Association criteria16) who had an inadequate response to disease modifying antirheumatic drugs (DMARDs) and corticosteroids. All patients were on methotrexate 7.5 mg once per week, leflunamide 20 mg once daily and prednisolone 5 mg once daily. None of our patients was being treated with non-steroidal anti-inflammatory drugs (NSAIDS) within the past year. Out of 46 RA patients four (9%) were had been on stable treatment with statins and 11 (24%) on stable treatment with cardioactive medications for the past 6 months (table 1). We used the following equation to calculate the composite inflammatory disease activity score (DAS), which utilises C-reactive protein (CRP), the visual analogue score (VAS) of wellbeing and the number of tender and swollen joints (from a total of 28 joints assessed): DAS = √(0.56×number of tender joints) + √(0.28×number of swollen joints) + [0.70×ln(CRP)] + (0.014×VAS).17 To assess changes in physical activity at baseline and after 30 days of treatment, we used the Duke Activity Status Index (DASI). DASI is a brief, self-administered questionnaire designed to estimate patients’ ability to perform daily activities and provides an estimation of exercise capacity measured in metabolic equivalents (METs).18

Table 1

Clinical characteristics, conventional echocardiographic, vascular and biological markers of the study population

Of the 46 patients, a group of 23 patients (mean age 57 (SD 17) years, 17 females) received anakinra treatment (150 mg subcutaneously once daily) for 30 days. A second group of 23 RA patients were selected on 1:1 basis to have similar baseline age, sex and inflammatory disease activity as assessed by DAS with the anakinra-treated group. The prednisolone-treated RA patients also had an inadequate response to treatment with DMARDs and prednisolone but were treated with an increase of their initial dose of prednisolone by 5 mg for 30 days according to standard clinical practice.10 Patients were examined in the outpatient clinic every 15 days to assess clinical status and CRP. To examine the patients’ compliance with therapy, we asked these patients to provide the used ampoules of anakinra and the used cartridges of prednisolone tablets at each visit.

None of the patients had cardiovascular or renal disease or ischaemia during thallium scintigraphy or dobutamine stress. Patients with known or suspected coronary artery disease were excluded to ensure that measurement of CFR would reflect the status of coronary microcirculation and not the effects of epicardial coronary artery stenosis. None of our patients had a history of steroid induced diabetes.

Twenty-three asymptomatic subjects of similar age and sex as the RA patients and with a normal ECG, echocardiogram and treadmill test were selected as healthy control subjects among subjects attending the cardiology outpatient clinic.

Both subgroups of RA patients had measurements of biochemical, vascular and LV function markers at baseline and after 30 days of treatment. The healthy control subjects had a single baseline measurement of the examined markers.

The study protocol was approved by the institute’s ethics committee, and written informed consents were obtained from all patients.


Studies were performed using a Vivid 7 (GE Medical Systems, Horten, Norway) ultrasound system. All studies were digitally stored in a computerised station (Echopac GE, Horten, Norway) and were analysed by two observers blinded to clinical and laboratory data. All patients had adequate images for analysis.

2D echocardiography

We measured the following parameters from cross-sectional echocardiographic images of the left ventricle (LV): (1) end-diastolic (LVD), end-systolic diameter (LVS) (mm) and fractional shortening (FS) (%), and (2) ejection fraction (EF) (%) using the Simpson’s method of discs.

2D strain measurements

Using a dedicated software package (Echopac, GE Medical systems, Horten, Norway), two-dimensional strain was measured.11 19 Apical four-chamber and parasternal midventricular short-axis views of the LV were obtained at end-expiratory apnoea and three cardiac cycles were stored from each view in cineloop format for subsequent offline analysis. Sectors were adjusted to achieve frame rates ⩾50 frames/s (50–82 frames/s). By tracing the endocardiac contour from an end-diastolic frame the software automatically tracked the contour on subsequent frames. Adequate tracking can be verified in real time and corrected by adjusting the region of interest or manually correcting the contour to ensure optimal tracking.11 19 Mitral and aortic valve opening and closure were defined by anatomical M-mode. The average longitudinal (L), circumferential (C) and radial (R) peak systolic strain (S), as well as longitudinal, circumferential and radial peak systolic, and early diastolic strain rate (SR) were measured.11 17

The inter-observer and intra-observer variability for all measurements were ⩽8% and ⩽10%, respectively.

Coronary flow

Coronary flow velocities in the left anterior descending coronary artery were obtained with colour-guided pulse-wave Doppler from long-axis apical projections with a 7-MHz transducer.10 The velocity time integral of the diastolic component of coronary flow wave was measured at baseline and after adenosine infusion (140 μg/kg/min) for 3 minutes and the ratio of hyperaemic to resting measurement was used to calculate the coronary flow reserve (CFR).10 Measurements from three cardiac cycles were averaged. All subjects abstained from alcohol, caffeine and food for 8 hours and stopped taking any vasoactive medications for 24 hours before the study. Inter-observer and intra-observer variability of these measurements were 5% and 2%, respectively.

Endothelial function

FMD and nitrate-induced vasodilatation of the brachial artery were determined according to a previously published methodology.10 15 Before the study, all subjects abstained from alcohol, caffeine and food for 8 hours and stopped taking any vasoactive medications for 24 hours. Inter-observer and intra-observer variability of the brachial artery diameter were 0.08 (0.19) mm and 0.1 (0.12) mm, and the day-to-day variability of FMD was 1.1 (1)%.15

Laboratory assays

C-reactive protein (CRP) was measured by a high-sensitivity particle-enhanced immunonephelometry (Dade Behring, Marburg, Germany, measurement range: 0.175–1100 mg/l). Malondialdehyde (MDA) was determined spectrophotometrically using a commercial kit (Oxford Biomedical Research Colorimetric Assay for lipid peroxidation, measurement range: 1–20 nM), as previously described.10 ELISA was used to determine nitrotyrosine (NT) (Hycult Biotechnology BV, Uden, Netherlands; measurement range 2–1500 nmol/l).

Statistical analysis

To examine whether patients included in two treatment groups (anakinra-treated and prednisone-treated) were adequately balanced for the presence of atherosclerosis, we calculated the logit propensity score in each patient. For this calculation, we used a logistic regression model which included as dependent variables age, sex, hypertension, hyperlipidaemia, smoking, cardioactive medication, statins and baseline CFR and FMD as markers related to atherosclerosis. Propensity scores between anakinra-treated and prednisolone-treated patients were compared using a two-tailed t test.20

Categorical data were compared between patients treated with anakinra, prednisolone and normal controls by contingency tables (p value in table 1) and between each treatment group and normal controls by the χ2-test or Fisher exact test when five patients or fewer were included in each cell (table 1). Continuous variables were tested for normality using the Kolmogorov-Smirnov test. Normally distributed variables are given as mean (SD). Spearman correlation analysis was used to determine bivariate correlations. Because biomarkers had a non-normal distribution, data are expressed as median (interquartile range) and were analysed after transformation into ranks.21

Analysis of variance (ANOVA) (general linear model, SPSS 13) for repeated measurements was applied to compare the effects of anakinra versus prednisolone, with measurements at baseline and 30 days post-treatment used as a within-subject factor and type of treatment as between-subject factor. The F and p values of the interaction between time of measurement of the examined markers and type of treatment were calculated. The Greenhouse-Geisser correction was used when the sphericity assumption, as assessed by Mauchly’s test, was not met. Post-hoc comparisons were performed with Bonferroni’s correction.

Comparisons between normal controls and each treatment group at baseline or at 30 days were performed using the unpaired t test (vascular and LV deformation indices) and Mann-Whitney test (biomarkers), as well as the percentage changes of the examined indices between baseline and 30 days in the anakinra group versus changes in the prednisolone group. Statistical significance was considered as p<0.05.


Patients and controls had similar age sex, body mass index (BMI), atherosclerotic risk factors and medications, as well as FS, EF and blood pressure (table 1). By logistic regression analysis including age, sex, hypertension, hyperlipidaemia, smoking, cardioactive medication, statins, baseline CFR and FMD, the calculated logit propensity scores were similar between the two treatment groups (0.195 (0.093) for anakinra-treated patients vs 0.180 (0.093), for prednisolone-treated patients, p = 0.535). Thus, the two groups were adequately balanced for markers of atherosclerosis. Twenty RA patients complained of mild erythema at the injection site and none was withdrawn from the study because of adverse effects or inadequate response to treatment.

Patients’ physical activity, as assessed by DASI, was similarly improved in both treatment groups (from 3.9 (1.0) to 6.3 (1.1) METs, post-anakinra and from 4.1 (1.2) to 5.9 (0.9) METs, post-prednisolone, F for interaction with treatment = 3.4, p = 0.07).

At baseline, the DASI (METs) was related to DAS (r = −0.454, p = 0.004) and CRP levels (r = −0.510, p = 0.001). Furthermore, the percentage improvement in physical activity as assessed by DASI, was related to the percentage improvement of DAS (r = −0.612, p<0.001). No other significant associations between changes in physical activity and changes in vascular or biochemical parameters were observed.

Inter-relations between 2D speckle tracking parameters, biochemical markers and vascular function indices at baseline

At baseline, CFR was associated with longitudinal strain (r = 0.436, p = 0.013), longitudinal systolic strain rate (r = 0.487, p = 0.005) and longitudinal early diastolic strain rate (r = −0.367, p = 0.046), as well as with circumferential strain (r = 0.390, p = 0.030), circumferential systolic strain rate (r = 0.452, p = 0.011) and circumferential early diastolic strain rate (r = −0.345, p = 0.057). FMD was related to diastolic longitudinal and circumferential strain rate (r = 0.554, p = 0.032 and r = 0.547, p = 0.015, respectively).

Also in NT levels was associated with longitudinal strain (r = 0.36, p = 0.04), longitudinal systolic strain rate (r = 0.479, p = 0.03) and diastolic longitudinal strain rate (r = −0.384, p = 0.036).

Finally, MDA levels were related to diastolic longitudinal strain rate (r = 0.367, p = 0.04) and circumferential systolic strain rate (r = −0.323, p = 0.05).

No significant correlations of myocardial deformation parameters with CRP were observed.

2D speckle tracking derived parameters in patients versus controls

At baseline, patients with RA had impaired longitudinal and circumferential strain and systolic strain rate, compared to controls, as well as radial strain (p<0.05, table 2). Furthermore, longitudinal and circumferential early diastolic strain rate modestly deteriorated in patients versus controls (p = 0.04 and p = 0.045, table 2). Radial systolic and early diastolic strain rate was similar between RA patients and controls.

Table 2

Comparison of myocardial deformation indices between RA patients and controls

2D-speckle tracking derived parameters post-anakinra versus post-prednisolone

Baseline 2D-speckle tracking parameters were similar between the two treatment groups (table 3). Additionally the two treatment groups had similar age, sex, BMI, atherosclerotic risk factors, DAS and medications as well as FS, EF, systolic and diastolic blood pressure (table 1).

Table 3

Chronic effects of anakinra on vascular and LV deformation parameters versus prednisolone-treated patients

Compared to baseline, there was an improvement in longitudinal strain, systolic and early diastolic strain rate, as well as circumferential strain and systolic strain rate and radial strain after 30 days of anakinra treatment (p<0.05 for all comparisons, table 3). In the prednisolone-treated group, there were no statistically significant changes after treatment compared to baseline. Conversely, longitudinal strain, systolic and early diastolic strain rate, as well as circumferential strain and systolic strain rate were higher in patients after anakinra treatment than after prednisolone treatment (p<0.05 for all comparisons).


To our knowledge, this is the first study to report a significant deterioration in longitudinal and circumferential strain and systolic strain rate in RA patients with preserved EF compared to healthy controls. These findings suggest that LV systolic deformation at longitudinal and circumferential directions is impaired, despite the presence of normal EF in rheumatoid arthritis patients without overt heart failure. Furthermore, in the present study we have shown that endothelial function, as assessed by FMD and CFR, and nitro-oxidative stress, as assessed by NT and MDA, are related to LV myocardial deformation in RA patients. Finally, in our study LV myocardial deformation improved to a greater extent after chronic inhibition of IL-1 activity by anakinra than after treatment with corticosteroids.

Studies have shown an impaired coronary microcirculatory function, as assessed by CFR, in patients with RA,10 22 ankylosing spondylitis23 and Behçet’s disease.24 Myocardial microcirculatory dysfunction in RA is attributed to inflammation and enhanced nitro-oxidative stress10 25 and is improved by treatment with anti-tumour necrosis factor α (anti-TNFα) 25 or interleukin-1ra.10 The impairment of coronary microcirculation may compromise myocardial perfusion, and thus causes LV systolic and diastolic dysfunction. In the present study, we demonstrated that coronary microcirculatory function as expressed by CFR was related to longitudinal and circumferential systolic and diastolic deformation parameters. Also, endothelial function as estimated by FMD was related to circumferential and longitudinal diastolic deformation. Thus, the impairment of myocardial deformation in RA patients compared to controls, as observed in our study, may be a result of the endothelial and coronary microcirculatory dysfunction characterising these patients.

However, previous studies suggest that myocardial cell function is impaired10 26 in RA. Pathological processes contributing to myocardial dysfunction in RA include increased oxidative4 and nitrosative5 27 stress, production of IL-1 and IL-6.6 7 8 21 26 Products of nitro-oxidative stress and inflammatory cytokines have a direct negative inotropic action.10 21 26 Additionally, interstitial fibrosis caused by cytokine-induced fibroblast activity and collagen deposition in the heart muscle are present in RA28 and thus may play a part in the abnormal myocardial deformation observed in our study. In support of this hypothesis, we observed that nitro-oxidative stress as expressed by NT and MDA levels was related to longitudinal systolic and/or diastolic deformation parameters at baseline.

Thus, the inflammatory process and enhanced nitro-oxidative stress may have a direct effect on myocardial function and, therefore, may explain the impairment of longitudinal and circumferential myocardial deformation at systole and early diastole in the present study.

Adenosine-induced CFR is endothelial dependent.29 IL-1 enhances the production of C-reactive protein (CRP)6 7 8 and inflammatory cytokines with negative inotropic action.8 10 21 Furthermore, IL-1 promotes the release of superoxide anion,2 10 contributing to an enhanced nitro-oxidative stress. Through the above mechanisms, IL-1 impairs vascular reactivity, endothelial function and coronary flow2 3 and contributes to myocardial dysfunction. In our previous study,10 we have shown that administration of anakinra, an IL-1 receptor antagonist, reduced IL-6 and nitro-oxidative stress with a parallel improvement of FMD and CFR after acute and chronic treatment. The reduction of nitro-oxidative stress demonstrated the highest correlation with improvement of endothelial and coronary function.

In the present study we have extended our previous findings, as we have now shown an improvement in longitudinal and circumferential myocardial deformation after 30 days of treatment with anakinra in RA patients. Furthermore, the improvement of these parameters was greater in the anakinra group than in the prednisolone group after chronic treatment. This improvement in myocardial deformation after anakinra administration is in parallel with the improvement in NT, MDA, FMD and CFR after anakinra treatment reported in our previous study. These findings suggest that IL-1 inhibition may improve myocardial deformation to a greater extent than corticosteroids because of a greater reduction of nitro-oxidative stress and improvement in endothelial function. In our previous study,10 the reduction of nitro-oxidative stress by anakinra caused normalisation of endothelial function and coronary microcirculatory function as assessed by CFR. Thus, chronic treatment with anakinra may increase myocardial perfusion and consequently improve longitudinal and circumferential deformation, as observed in the present study. Furthermore, anakinra treatment may contribute to the improvement in myocardial deformation parameters by limiting the direct detrimental effects of nitro-oxidative stress on myocardial cells. Experimental studies have demonstrated that anakinra treatment reduces the adverse cardiac remodelling after myocardial infarction by inhibiting cell apoptosis.30 We may assume that, through these mechanisms, chronic treatment with anakinra may improve LV myocardial deformation parameters in RA patients, as shown in the present study.


In the present study we have shown a significant impairment of myocardial deformation parameters of RA patients compared with control group, in the presence of preserved ejection fraction. The early detection of abnormalities in myocardial deformation by speckle-tracking might provide a means of identifying patients at risk for progressive heart failure and thus, may be important in clinical practice when assessing the prognosis and optimising treatment. Moreover, administration of an IL-1 receptor antagonist improves myocardial deformation in patients with rheumatoid arthritis, probably through the reduction of nitro-oxidative stress and improvement in endothelial and coronary function. Anakinra treatment was more effective than corticosteroids in improvement of LV deformation parameters.

Study limitations

The study design does not permit us to explore the causality between the changes of the vascular and LV function after anakinra treatment. As RA patients with coronary artery disease were excluded, the effects of anakinra in the presence of coronary artery disease were not explored in our study.

Improved physical activity after anakinra, as reflected by the reduced DAS, may have contributed to the improvement in biochemical, vascular and LV deformation parameters in our study.31 The non-invasive assessment of endothelial function and coronary flow should also be acknowledged as a limitation.14



  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • See Editorial, p 1471

Linked Articles