Background
Speckle-tracking echocardiography allows early detection of subclinical left ventricular systolic dysfunction (LVSD) in patients with rheumatoid arthritis (RA). In this prospective study, we assessed the prevalence and the prognostic role of subclinical LVSD detected by speckle-tracking echocardiography in RA patients.
Methods
Two-dimensional global longitudinal strain (GLS) and global circumferential strain (GCS) were measured in 209 RA patients without overt cardiac disease. LVSD was defined as low GLS (> −16.0%), low GCS (> −17.8%), or both. The primary end point was all-causes hospitalization; the coprimary end point was hospitalization for cardiovascular causes.
Results
The study population had a mean age of 58 ± 11 years; 67% were female, 52% had hypertension, and the RA duration was 14 ± 10 years. Low GLS was detected in 51 patients (24%), low GCS in 42 patients (20%), and combined low GLS and GCS in 18 patients (9%). During a median follow-up time of 16 months (range, 10–21 months), a primary end point occurred in 50 patients (24%), and 25 patients were hospitalized for a cardiovascular event. Multiple Cox regression analyses revealed that combined low GLS and GCS was independently associated with the end point defined as all-causes hospitalization together with higher aortic stiffness. Examined individually, neither low GCS nor low GLS showed an independent association with this typology of clinical outcome. Conversely, both low GCS and low GLS (examined individually or as combined low GLS and GCS) emerged as strong independent prognosticators of cardiovascular events.
Conclusions
Subclinical LVSD defined as low GLS, GCS, or both is common in RA patients without overt cardiac disease and provides additional prognostic information in these individuals.
Highlights
- •
A significant proportion of asymptomatic rheumatoid arthritis patients without history of cardiac disease have subclinical left ventricular systolic dysfunction detected by speckle-tracking echocardiography and defined as low global longitudinal strain (GLS) and/or low global circumferential strain (GCS).
- •
In rheumatoid arthritis subjects analyzed in primary prevention, all-causes hospitalizations are independently related to the condition of combined low GLS and GCS.
- •
The combined low GLS and GCS status is characterized by older age, left ventricular diastolic dysfunction, and left ventricular hypertrophy, all of which are factors predisposing patients toward the development of overt heart failure.
- •
Low GCS, low GLS, and combined low GLS and GCS are strong independent predictors of cardiovascular events at mid-term follow-up.
Subclinical left ventricular (LV) systolic dysfunction (LVSD) has been found in a significant proportion of patients with rheumatoid arthritis (RA) analyzed in primary prevention. In these patients, predominantly in those with concentric remodeling/hypertrophy, LV ejection fraction (LVEF) fails to detect subclinical LVSD because of the compensatory cross-fiber shortening phenomenon. By contrast, subclinical LVSD is flawlessly documented in these patients by means of more accurate echocardiographic parameters such as midwall shortening (M-mode LV minor axis), mitral annular peak systolic velocity (tissue Doppler imaging), or myocardial strain (speckle-tracking echocardiography).
These parameters consider separately the circumferential or longitudinal components of LV systolic function, which are individually impaired in a substantial number of patients with RA. Furthermore, in about one-fourth of these patients, circumferential and longitudinal LVSD coexist.
As demonstrated in several populations at high risk for cardiovascular (CV) events, LVSD detected by these measurements has emerged as a strong prognosticator of poorer clinical outcomes. Remarkably, despite RA patients have a significantly higher mortality than the general population, mainly as a result of CV disease, the prognostic role of LVSD has never been investigated in these individuals.
Accordingly, we designed this prospective study to assess the following features by two-dimensional (2D) speckle-tracking echocardiography: (1) the prevalence of abnormally low global longitudinal strain (GLS), global circumferential strain (GCS), and combined low GLS and GCS in patients with asymptomatic RA patients without a history of cardiac disease; and (2) whether decreased low GLS, low GCS, or both is associated in these patients with increased hospitalizations, either total or CV related.
Methods
Study Population
The study participants were 209 noninstitutionalized adult patients with RA diagnosed according to the American College of Rheumatology criteria. The participants had no history of heart disease, and absence of valvular disease defined as more than mild mitral regurgitation, aortic regurgitation, or both, or stenosis of any degree at echocardiographic baseline evaluation. They were consecutively recruited from March 2014 to March 2015 and represented the whole RA population surveyed by the Division of Rheumatology of the Verona University, where they underwent echocardiographic, clinical, and laboratory evaluations. All patients gave written informed consent. The study was approved by the ethical committees in all participating centers, and conforms to the ethical guidelines of the Declaration of Helsinki as revised in 2013.
Matched Control Group
A group of 52 healthy individuals free of any medical treatment was selected during the same recruitment period of the RA patients to assess the range of normal values of strain measurements and in particular to identify the conditions of low GLS and low GCS. These 52 individuals were statistically comparable with those enrolled into the study for age, sex, and body mass index according to the Gower distance procedure as previously described. The 52 healthy control participants were selected from a pool of 116 individuals by taking for every four close patients with RA the closest healthy control. Their clinical and echocardiographic characteristics are reported in Table 1 and Table 2 , respectively.
| Variables | Total study population ( N = 209) | Healthy subjects used as reference ( n = 52) | P | Normal STE ( n = 154) | Abnormal STE ( n = 55) | P |
|---|---|---|---|---|---|---|
| Age (years) | 58 ± 11 | 59 ± 16 | .85 | 57 ± 11 | 62 ± 11 | .002 |
| Female (%) | 67 | 65 | .78 | 71 | 58 | .07 |
| Body mass index (kg/m 2 ) | 26.4 ± 4.1 | 25.8 ± 3.7 | .75 | 25.8 ± 4.0 | 25.2 ± 4.6 | .49 |
| Obese (%) | 14 | 12 | .72 | 14 | 15 | .74 |
| History of hypertension (%) | 52 | 0 | <.001 | 49 | 59 | .17 |
| Systolic blood pressure (mm Hg) | 133 ± 17 | 120 ± 13 | <.001 | 132 ± 18 | 132 ± 13 | .91 |
| Systolic blood pressure (mm Hg) | 83 ± 9 | 79 ± 5 | <.001 | 83 ± 9 | 84 ± 8 | .23 |
| Dyslipidemia (%) | 65 | 0 | <.001 | 64 | 70 | .40 |
| Active smoker (%) | 42 | 0 | <.001 | 41 | 44 | .80 |
| Diabetes (%) | 9 | 0 | <.001 | 7 | 13 | .14 |
| GFR (mL/min/1.73 m 2 ) | 94 ± 22 | 92 ± 10 | .36 | 93 ± 21 | 96 ± 25 | .30 |
| Cholesterol LDL (mg/dL) | 122 [99–147] | 126 [99–150] | .45 | 121 [97–149] | 125 [95–140] | .55 |
| Triglycerides (mg/dL) | 120 [84–152] | 113 [71–145] | .69 | 115 [80–157] | 132 [75–160] | .07 |
| C reactive protein (mg/L) | 2.8 [0.8–5.3] | – | 2.9 [0.7–5.8] | 2.6 [0.9–4.9] | .20 | |
| Rheumatoid factor + (%) | 38 | – | 38 | 42 | .60 | |
| Cyclic citrullinated peptide + (%) | 39 | – | 39 | 39 | .99 | |
| Duration of RA (years) | 14 ± 10 | – | 14 ± 10 | 14 ± 12 | .94 | |
| Clinical disease activity index | 10 ± 8 | – | 10 ± 9 | 10 ± 9 | .84 | |
| High activity of disease (%) | 22 | – | 24 | 18 | .38 | |
| Pharmacologic treatment | ||||||
| Beta-blockers (%) | 19 | – | 21 | 12 | .16 | |
| ACEi/ARB (%) | 32 | – | 28 | 42 | .08 | |
| Diuretics (%) | 17 | – | 16 | 20 | .58 | |
| Calcium antagonists (%) | 9 | – | 9 | 10 | .77 | |
| Antiplatelet agents (%) | 13 | – | 12 | 18 | .32 | |
| Statins (%) | 26 | – | 23 | 31 | .23 | |
| NSAIDs (%) | 35 | – | 37 | 26 | .16 | |
| Methotrexate (%) | 46 | – | 44 | 52 | .33 | |
| Hydroxychloroquine (%) | 7 | – | 7 | 10 | .46 | |
| BYOAs (%) | 78 | – | 80 | 72 | .22 | |
| Corticosteroids (%) | 42 | – | 42 | 42 | .93 | |
| Variables | Total study population ( N = 209) | Healthy subjects used as reference ( n = 52) | P | Normal STE ( n = 154) | Abnormal STE ( n = 55) | P |
|---|---|---|---|---|---|---|
| Echocardiography | ||||||
| LV end-diastolic diameter (cm/m 2 ) | 2.6 ± 0.3 | 2.9 ± 0.5 | <.001 | 2.6 ± 0.4 | 2.7 ± 0.5 | .62 |
| LV relative wall thickness | 0.46 ± 0.07 | 0.37 ± 0.04 | <.001 | 0.45 ± 0.06 | 0.48 ± 0.08 | .04 |
| Concentric LV geometry (%) | 67 | 2 | <.001 | 62 | 78 | .04 |
| LV mass index (g/m 2.7 ) | 45.0 ± 10.6 | 36.0 ± 7.2 | <.001 | 44 ± 10 | 46 ± 12 | .48 |
| LV hypertrophy (%) | 36 | 2 | <.001 | 35 | 36 | .86 |
| LV ejection fraction (%) | 67 ± 6 | 67 ± 7 | .88 | 67 ± 6 | 65 ± 6 | .03 |
| Left atrial maximal volume (ml/m 2 ) | 19.5 ± 6.8 | 19.0 ± 5.4 | .30 | 19 ± 5 | 19 ± 5 | .93 |
| E/A ratio transmitral flow | 0.95 ± 0.33 | 1.04 ± 0.21 | .78 | 0.97 ± 0.34 | 0.89 ± 0.26 | .14 |
| E/E′ ratio | 6.4 ± 1.6 | 6.4 ± 1.7 | .69 | 6.3 ± 1.5 | 6.5 ± 1.8 | .44 |
| LV diastolic dysfunction (%) | 27 | 9 | .008 | 26 | 32 | .39 |
| Mitral annular calcification (%) | 53 | 4 | <.001 | 53 | 54 | .91 |
| Aortic valve calcification (%) | 59 | 4 | <.001 | 58 | 61 | .79 |
| Aortic stiffness index (%) | 6.4 ± 4.8 | 4.6 ± 3.7 | <.001 | 6.2 ± 4.0 | 6.7 ± 4.2 | .42 |
| Longitudinal LV strain (%) | ||||||
| Global longitudinal strain (%) | −18.4 ± −3.4 | −19.9 ± 2.6 | .005 | −19.8 ± −2.5 | −13.9 ± 1.8 | <.001 |
| Low global LV strain (>−16) (%) | 24 | 5 | <.001 | 0 | 93 | <.001 |
| LV walls | ||||||
| Anteroseptum | −16.7 ± 2.9 | −18.6 ± 3.5 | <.001 | −18.0 ± 3.0 | −12.4 ± 2.5 | <.001 |
| Inferoseptum | −18.0 ± 3.4 | −19.4 ± 3.8 | .01 | −19.4 ± 3.8 | −14.4 ± 3.2 | <.001 |
| Inferior | −18.2 ± 3.7 | −19.5 ± 3.9 | .02 | −19.3 ± 3.9 | −14.7 ± 3.1 | <.001 |
| Posterior | −17.1 ± 3.2 | −17.7 ± 3.3 | .08 | −18.3 ± 3.4 | −13.2 ± 3.5 | <.001 |
| Lateral | −17.7 ± 3.3 | −19.8 ± 3.8 | <.001 | −19.4 ± 3.3 | −12.5 ± 2.8 | <.001 |
| Anterior | −18.2 ± 3.7 | −19.8 ± 3.9 | .006 | −19.5 ± 3.9 | −14.0 ± 2.9 | <.001 |
| Circumferential LV strain (%) | ||||||
| Global circumferential strain (%) | −23.6 ± 6.9 | −24.7 ± 4.5 | .31 | −24.5 ± 6.9 | −20.8 ± 6.2 | <.001 |
| Low global LV strain (>−17.8) (%) | 20 | 5 | <.001 | 0 | 76 | <.001 |
| LV segments (midlevel) | ||||||
| Anteroseptum | −24.6 ± 8.0 | −25.8 ± 6.7 | .35 | −25.5 ± 8.3 | −21.7 ± 8.4 | .001 |
| Inferoseptum | −21.1 ± 9.2 | −21.5 ± 7.9 | .84 | −21.8 ± 8.1 | −18.8 ± 7.8 | .03 |
| Inferior | −19.9 ± 9.0 | −19.0 ± 8.1 | .24 | −20.9 ± 9.2 | −18.9 ± 8.2 | .17 |
| Posterior | −20.5 ± 8.1 | −19.3 ± 7.0 | .33 | −21.2 ± 8.2 | −18.2 ± 7.8 | .02 |
| Lateral | −22.1 ± 8.5 | −22.8 ± 6.4 | .58 | −23.1 ± 8.5 | −18.7 ± 7.6 | .001 |
| Anterior | −24.6 ± 8.5 | −25.1 ± 6.9 | .67 | −25.6 ± 8.3 | −21.1 ± 8.4 | .001 |
| Combined low longitudinal strain + low circumferential strain (%) | 9 | 0 | 0.03 | 0 | 34 | <.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
