Subclinical changes

Subclinical changes in heart have been observed despite the absence of evident cardiac symptoms in RA. Up to half of RA patients with no diagnosis and clinical manifestation of CVD showed cardiac fibrosis and inflammation [15, 16]. Systolic and diastolic left ventricle function was found to be reduced in up to half of RA patients with no cardiac manifestations [17, 18]. Echocardiographic studies also showed asymptomatic pericarditis and valvular involvement [19]. Another study revealed dysfunction of coronary microcirculation in 1/3 of cardiac asymptomatic RA patients via measurement of myocardial flow reserve [15]. Coronary plaques measured by computed tomography (CT) angiography were found out in higher prevalence and severity in RA patients [20] and the risk of having a silent (unrecognized) heart attack was observed to be doubled in rate [21].

CAD

Epidemiological studies showed that 51% of long-standing RA patients have a Framingham risk score over 20%, meaning increased frequency of CV ischemic events [22]. Therefore, an expert committee of the European Leage Against Rheumatism recommends that CV risk scores (e.g., Framingham) should be multiplied by 1.5 in RA patients to draw attention to the increased risk in this patient group [23].

Patients with a diagnosis of RA have 1.5–2 times higher risk of developing CAD compared to the general population [21, 24] and this risk is equivalent to that in type 2 diabetes patients [25, 26]. Similarly, Lindhardsen and colleagues [27] reported that the increased acute MI risk was similar in magnitude that was expected in diabetes mellitus patients. A cohort study of 1,135 patients with RA and acute coronary syndrome showed that sudden cardiac death, ST-elevated MI (STEMI), elevated levels of troponin, inpatient complications were more common in RA patients, implying that acute coronary syndrome is expected to be much more severe than in general population [28]. In another study the incidence of heart attack was found equal to 10 years older non-RA patients which was 70% higher compared to normal population of the same age [27]. Considering that the history of having a heart attack is 3 times higher in individuals diagnosed with RA for the first time, compared to the normal population [21], the CV risk should be considered to have increased long before the diagnosis was made.

Interestingly, CAD presents a silent clinic in most patients with RA [21]. Chest pain during an acute coronary event is less expected in patients with RA than in the general population [29]. Even though the exact cause is unknown, absence of clinical angina pectoris may be due to not being physically active enough to cause angina, attributing pain to RA itself or altered pain perception by RA medications.

HF

Ischemic heart disease, inflammatory mediators, medications and amyloidosis can lead to the development of HF in RA [30]. The most frequently blamed factors in the development of HF in RA patients can be listed as follows: (1) ischemic cardiomyopathy due to increased CAD; (2) treatment-related cardiomyotoxicity; (3) granulomatous (less frequently necrotizing or non-specific) myocarditis secondary to inflammation and (4) infiltrative cardiomyopathy due to AA-type amyloidosis secondary to RA [31]. Patients with RA have 2-fold increased risk of HF [32]. This risk is more pronounced in patients who are seropositive for rheumatoid factor (RF).

Clinical presentation of HF is distinct in RA patients. Even though having a better cardiac function and lower blood pressure, higher mortality rates are expected [33]. RA patients rarely show expected sign and symptoms of HF and respond poorly to treatment with a worse prognosis [33].

Preserved ejection fraction (EF) is often expected (> 50%) in RA patients with HF and clinical findings of CAD are expected to a lesser extent [34] because HF seen in RA is often related to diastolic dysfunction, which may be related to systemic inflammation. The progression of HF in RA patients may be due to elevated levels of C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), RF, anti-citrullinated protein antibody (ACPA) [21]. Therefore, disease activity score-28 (DAS-28) can be used as a measure in the estimation of HF in RA patients [35].

Non-atherosclerotic heart disease

Pericarditis is among the most common cardiac complications of RA and it may develop even before the disease is diagnosed. Even tough 20–50% of RA patients show pericardial involvement on echocardiography (ECG), less than 15% are clinically symptomatic with chest pain or dyspnea and less than 10% have hemodynamic abnormalities [36]. Autopsy studies showed that 20–40% of seropositive patients for RF had pericarditis [37]. Restrictive pericarditis with tamponade is extremely uncommon.

Myocarditis is a rare complication of RA. It is mostly seen in active disease with other extraarticular manifestations [38] and may be granulomatous or interstitial. Granulomatous type has higher specificity for RA. Endocardial involvement may result in mitral insufficiency while involvement of conduction system results in atrioventricular block.

Rheumatoid nodules can develop in the myocardium, pericardium, cardiac valves and can cause different manifestations such as syncope, arterial embolization, valvular insufficiency and death according to the site of formation. Furthermore, anti-rheumatic drugs may cause acute myocardial dysfunction [39].

Arrhythmias

Conduction abnormalities can be seen in RA patients due to local ischemia, rheumatoid nodules, HF or amyloidosis. Even in the absence of structural hearth anomalies, risk of fatal arrhythmias is found to be increased in RA patients [40].

In terms of AF, the results of the studies are conflicting. Atrial tissue affected by systemic inflammation in the background of RA undergoes structural and electrical remodeling. Greater than 40% increase in the incidence of AF and a 5-fold higher thromboembolism rate are reported in a prospective cohort study of 18,247 RA patients [11]. On contrary, another cohort study involving 20,852 RA patients and 104,260 control patients reported no difference in AF risk when comorbidities and medications are adjusted [41]. CRP level, evidence of systemic inflammation, is shown to be correlated with autonomic dysfunction and QT corrected for heart rate (QTc) prolongation [40].

Non-cardiac vascular diseases

Hypertension (HT) is a common CV comorbidity in RA patients. RA patients have increased peripheral vascular resistance due to oxidative stress, endothelial dysfunction and genetic polymorphisms resulting in increased blood pressure. Some agents used in RA treatment have been shown to cause HT via different pathophysiological pathways. Non-steroidal anti-inflammatory drug (NSAID) therapy may cause HT by activating the renin-angiotensin-aldosterone system, glucocorticoid therapy may increase sodium absorption and vascular tone (prednisolone ≥ 10 mg/day or equivalent), and leflunomide may cause HT by increasing sympathetic activation [42].

The effect of RA on non-cardiac vascular diseases is not clearly known yet. Although there are reports showing increased rate of cerebrovascular disease in RA patients [43], there are also publications claiming the opposite [44].

Venous thromboembolism risk is found to be increased 2–3 times compared to the general population [44, 45]. The risk of peripheral vascular disease seems to be increased, especially in patients with extra-articular involvement [46–48]. Pulmonary HT can also be a rare complication in patients with RA. Rheumatoid vasculitis may rarely cause coronary arteritis but in the absence of other extra-articular involvement, it is unlikely to cause MI.

Impaired immunity and chronic inflammation

Impaired immunity and chronic inflammation seen in patients with RA may be responsible for the cardiac involvement and understanding them may shed light on the development of novel therapeutic targets for disease prevention. The cornerstone of the ischemic CV events is thought to be chronic inflammation. In patients with RA, elevated CRP, indicating systemic inflammation, is an independent prognostic marker of CV mortality [49]. Seven percent increase in CV risk was reported with each period of increased inflammation in the joints [50]. It has been shown that RA patients with increased disease activity have more unstable plaques at carotid artery by ultrasonography (USG) [51] and CT angiography [20].

Inflammation in the synovium is also seen in the vascular wall of RA patients [52] which explains why both the inflamed synovial membrane and the inflamed atherosclerotic plaques show similarity in terms of abnormal immune responses [53, 54]. An autopsy study comparing coronary artery pathology among RA vs. non-RA patients who died after having MI, showed that even though RA patients have less atherosclerosis compared to the control group, their plaques were more inflamed and unstable [55].

Unlike ischemic CV events, the role of chronic inflammation in non-ischemic CV events is less well known. Although atherosclerosis and chronic inflammation is mostly blamed so far, there may also be different mechanisms that can lead to cardiac dysfunction. It is discussed that diastolic dysfunction can be seen in RA patients even in the absence of detectable atherosclerosis. This situation gives rise to the hypothesis that there may be as yet unexplained mechanisms in the development of early cardiac dysfunction that precede the development of atherosclerosis.

It is thought that inflammation may cause myocardial fibrosis leading to systolic and diastolic dysfunction and arrhythmias. The correlation between RA disease activity and myocardial inflammation and fibrosis has been shown by cardiac magnetic resonance imaging (MRI) and positron emission tomography and CT (PET-CT) [15, 56]. Also, the prevalence of diastolic HF was increased in RA patients with increased disease activity and high CRP levels [57]. Another supporting data is that myocytes of RA patients with low disease activity showed normal structure and function [58, 59]. Inflammation is both cell and cytokine mediated.

Genetic susceptibility and epigenetics

Before overt synovitis that is clinically detectable, genetically susceptible host for RA has a preclinical, asymptomatic phase in which behaviors and environment play crucial role in delaying or potentially preventing RA onset [85].

The major risk gene for RA, human leukocyte antigen-DR isotype1 (HLA-DRB1), promotes the selection and survival of auto-reactive CD4⁺ T cells which predisposes to disease. HLA-DRB1 alleles are also known to increase the risk of MI and other forms of non-RA associated heart disease [86, 87]. HLA-DRB1 positivity predisposes to expansion of CD28^null T cells in RA and CAD [88].

From epigenetic perspective, non-coding RNAs, including microRNAs (miRNAs), long noncoding RNAs (lncRNAs) and circular RNAs (circRNAs), are known to play important roles in the regulation of inflammation and autoimmunity [89]. Epigenetic mechanisms may be responsible from enhanced inflammation and joint destruction in RA patients. Correlatively, non-coding RNAs are being studied as significant novel regulators of CV risk factors and cell functions which make them therapeutic target candidates [90].

Autonomic nervous system dysfunction

It was shown years ago that RA patients have signs of autonomic dysfunction affecting CV nervous system. Chronic inflammation can lead to imbalances in autonomic nervous system (ANS) which then causes a vicious circle between inflammation and ANS [91].

Chronic inflammation in RA increases sympathetic activity while suppressing the parasympathetic function [92]. In a rat collagen-induced arthritis model study it is hypothesized that increased sympathetic drive of heart, so called “inflammatory reflex”, may cause CV problems [93]. The results of the study showed that even after the relief of arthritis attack and decline of sympathetic activity, the autonomic dysfunction in heart persists [93].

It has been found that that long term exposure of cardiomyoctes to high concentrations of epinephrine may lead to phosphorylation of beta-1 adrenergic receptors (β1-ARs) by G protein coupled receptor kinase-2 (GRK2), also known as β-AR kinase (βARK), resulting in receptor desensitization and endocytosis [94]. Chronic GRK2 hyperactivity may cause impairment in the contractility of cardiomyocytes leading to CV dysfunction [95]. Based on this hypothesis, it is proposed that GRK2 inhibitors can be used in the prevention of cardiac events and may modify disease activity in RA patients [96]. There is no doubt that further clarification is needed to understand the role of GRK2 to determine if GRK2 inhibition may be a new treatment target of cardiac therapy.

Altered metabolic profile

With the effect of chronic inflammation in RA, lipoprotein structure and function changes [97], serum amyloid A load carried by high-density lipoprotein (HDL) increases and apolipoprotein A-I decreases [98]. Furthermore, inflammation reduces the ability of HDL molecule to accept cholesterol from macrophages, so called “cholesterol efflux capacity” [99, 100] which turns the HDL effect from anti-atherogenic to pro-atherogenic [98].

Although the levels of total cholesterol, low-density lipoprotein (LDL) and HDL levels were all found to be decreased in RA patients, higher decrease in HDL cholesterol is observed. This increases CV risk by increasing atherosclerosis index (total cholesterol/HDL cholesterol) [101]. Lipid profile in RA patients is characterized by a higher suppression of HDL than LDL and total cholesterol during periods of high inflammation. This is called the “lipid paradox” and is associated with more severe systemic inflammation [102].

RA induced insulin resistance aggravates systemic inflammatory response and is associated with seropositivity [103, 104], pro-inflammatory cytokines, disease activity and glucocorticoid usage [105, 106]. Serum retinol binding protein 4 (RBP4) is found to be elevated in diabetes patients and it’s level is also detected higher in RA patients compared to controls and is relevant to the severity of RA [107]. Visceral adiposity seen in RA patients was found to be associated with insulin resistance, HT, metabolic syndrome and increased inflammatory burden [108].

Uncontrolled inflammation and lack of physical activity leads to decreased muscle mass resulting in low body mass index (BMI). It has been shown that RA patients have altered body composition (high fat mass, low muscle mass) and decreased physical function even when the disease is in the remission phase meaning that metabolic deficits sustain [109, 110].

Moreover, RA patients are reported to have reduced oxidative metabolism in both T cells and skeletal muscle via metabolic reprogramming, resulting in increased invasiveness of immune cells, disease activity and disability [111, 112]. It has been revealed that altered naive CD4⁺ T cell metabolism is associated with pro-inflammatory function and characterized by reduced glycolysis and low oxidative ATP production [112, 113] which adversely affects cardiorespiratory fitness.

Glucocorticoids

In addition to providing a beneficial anti-inflammatory affect, glucocorticoids may aggravate heart disease by increasing CV risk factors. Their adverse metabolic effects may outweigh their anti-inflammatory benefits. Blood pressure and glucose levels should be regularly monitored before and during treatment.

It has been reported that even 1-year treatment with corticosteroids in newly diagnosed RA patients, significantly increases the risk of non-ischemic HF [137]. The risk of heart disease in patients using high-dose (> 7.5 mg/day prednisolone) steroids is twice that of patients not using steroids [128].

NSAIDs

Cyclooxygenase-2 (COX-2) inhibition increases blood pressure, promotes atherosclerotic plaque rupture and thrombosis by reducing prostacyclin levels [138]. The Vioxx Gastrointestinal Outcomes Research trial (VIGOR) study reported an increased risk of heart disease with use of rofecoxib, leading to withdrawal of the drug from the market [139]. In another study, no difference was found in the development of heart disease between patients using celecoxib and ibuprofen [140]. A meta-analysis of seven NSAIDs, 4 of which were coxib group, showed that naproxen had the lowest CV risk while other 6 NSAIDs were found similar in terms of CV risk [141].

Conventional DMARDs

MTX, the first line treatment, has been associated with lower CV risk. Recent meta-analysis reports that CV events decreased by 21% with MTX use [142]. Another meta-analysis also reports reduced risk of major CV events with MTX use [143]. Interestingly, in a recent study MTX failed to prevent CV events in non-RA patients with CVD [144]. This could be interpreted as cardio-protective effect of MTX can only be seen under inflammatory conditions. It is thought that MTX does not alter the lipid profile. It’s effect on insulin resistance, HT and atherosclerotic plaque formation is not yet known. Furthermore, MTX, itself, increases homocysteine production which has pro-atherogenic effects on endothelial cells and stimulates LDL oxidation.

Hydroxychloroquine is found cardio-protective in most studies [143, 145, 146] and may inhibit platelet aggregation and thrombogenic effects of anti-phospholipid antibodies [147]. It is known that hydrocychloroquine is known to reduce risk of diabetes [148], levels of LDL and total cholesterol/HDL ratio in RA patients [149]. However, anti-malarials may also cause cardiomyopathies in RA patients [150].

Treatment with sulfasalazine in RA patients has been associated with lower CV risk compared with patients who never used sulfasalazine, MTX or hydroxychloroquine [151]. Leflunamide is also associated with lower rate of MI in RA patients [146] but different from other conventional DMARDs it may aggravate HT by increasing blood pressure [152].

Biological DMARDs

Rituximab (a humanized chimeric anti-CD20 monoclonal antibody) was shown to improve vascular pathophysiology in RA patients [153]. Depletion of B cells via CD20 monoclonal antibody was shown to inhibit atherosclerosis in mice [75]. Antibody therapy against BAFF, important for survival of B cells, has failed to decrease atherosclerosis in mice which suggests a distinct mechanism of action [154]. Clinical trials showed that safety profile of rituximab was similar to anti-TNF-α treatments [155]. Abatacept (T-cell co-stimulation inhibitor) is shown to have no effect on the risk of developing HF in RA patients when compared to etanercept [156].

TNF inhibitors

TNF inhibitor therapy has been associated with a reduced risk of all heart events [157]. A recent clinical study reports that anti-TNF-α therapy effectively reduces the incidence of acute coronary syndrome in RA patients [158]. This effect is thought to be due to controlling inflammation rather than modifying CV risk factors. Myocardial inflammation on positron emission tomography (PET) imaging, which is correlated with disease activity, was shown to be reduced in patients using biological DMARDs compared to patients using conventional DMARDs [15, 16].

TNF inhibitors increase total and HDL cholesterol levels resulting in stable atherogenic ratio [159]. Since RA activity is inversely related to HDL level, it is thought that drugs regulating disease activity may improve HDL levels [160]. However, there is no clear information yet whether this effect on lipid profile is a class effect or whether specific TNF inhibitors provide a more favorable lipid profile. Conversly, anti-TNF-α therapies in RA patients may also be associated with elevated risk of developing HT [161].

Although most data [162–164] shows that anti-TNF-α treatment reduces HF risk and improves survival in RA patients, in elderly population it might worsen HF and decrease survival [165]. It’s beneficial effect might be more evident in women [163] and younger patients [162] with RA. Infliximab has been shown to increase heart function and preserve left ventricular function [166]. In contrast, Infliximab was shown to adversely affect the prognosis in patients with HF [167].

In summary, anti-TNF-α therapies reduce risk of CV events in RA patients with preserved cardiac function, but in patients with HF they may not improve cardiac function and even might worsen it. Further studies are needed to clarify the effects of anti-TNF-α therapies in RA patients with serious HF.

IL-6 inhibitors

Tociluzumab (humanized antibody targeting IL-6) has been found to be associated with increased HDL, LDL and triglyceride (TG) levels [168]. It was shown to be a safe alternative for etanercept in terms of development of major CV events [121]. Although tocilizumab is successful in suppressing disease activity, it may impair lipid homeostasis meaning to have uncertain CV outcomes [169, 170].

JAK inhibitors

Despite worsening the plasma lipid profile, tofacitinib or baricitinib were not associated with increased CV risk in RA patients. It is known that statins can reverse dyslipidemia in tofacitinib treated RA patients [170].

Diagnosis and management of ischemic heart diseases

The main point for early diagnosis of CAD in patients with RA is high awareness. The diagnostic algorithm should include follow-up in accordance with international guidelines. In addition to the cardiac treatments recommended by the guidelines DMARDs should be planned to control the risk factors and RA activity. The efficacy of acetylsalicylic acid (ASA) as prophylaxis in patients with RA has not been demonstrated [171]. The use of low-dose ASA (75–150 mg/day) for secondary prevention in these patients was similar to that in the general population. The addition of non-ASA NSAIDs to ASA-treated patients may decrease the antiplatelet efficacy of ASA and increase CAD activation. Therefore, the use of non-selective NSAIDs (including selective COX-2 receptor blockers) at a low dose for as short duration as possible is recommended for CAD prophylaxis [171]. The management of CAD (acute or chronic coronary syndrome) and treatment goals in patients with RA are similar to those in patients without RA [115]. In these patients, especially the use of potent statins (atorvastatin or rosuvastatin) reduces the development of new CVD and suppresses the inflammatory response. Again, effective inflammation control with DMARDs in this patient group reduces the risk of CVD development [172].

Diagnosis and management of non-ischemic heart diseases

Pericardial involvement is mostly silent (symptom rate < 10%) and usually develops towards the end of the first decade after diagnosis but may be observed between 1–30 years depending on the course and severity of RA [173]. The disease usually presents with mild-to-moderate exudative pericardial effusion with symptoms of pericarditis and has a mild course. Advanced age, comorbid CVD and persistent fever increase the likelihood of a progressive and fatal disease course. Therefore, the management of the disease varies according to the symptoms and amount of pericardial fluid. For this purpose, echocardiographic examination is the gold standard for diagnosis and follow-up. Asymptomatic patients with pericardial fluid limited to a few millimeters do not require additional treatment or regimen modifications. Adding NSAIDs to the treatment regimen according to symptom status is sufficient. Invasive treatment is indicated in cases of tamponade or constrictive pericarditis. The agents used for treatment in the presence of pericarditis in RA patients are shown in Table 1 [174].

The evaluation of RA patients in the asymptomatic phase with conventional or advanced cardiac imaging modalities [such as 2 dimensional (2D)/3D ECG and strain ECG] allows the detection of subclinical left ventricular dysfunction. In cases of symptomatic HF developing in patients with RA, treatments recommended by current guidelines should be initiated immediately and RA medications that may cause cardiomyopathy should be discontinued. Prednisolone (or equivalent) at doses of ≥ 7.5 mg/day, NSAIDs, antimalarial drugs, especially choloroquine/hydroxychloroquine, and TNF-α inhibitors have been shown to cause cardiomyopathy in association with RA treatment. The recommended agents for RA patients with HF are shown in Table 2 [175].

Myocarditis is a rare complication that usually occurs during periods of disease exacerbation in patients with extensive extra-articular involvement [176]. The most common form is granulomatous, which may present with symptoms of mitral regurgitation or atrioventricular block. More rarely, there may be a nonspecific or necrotizing form with local involvement [32]. Although echocardiographic evaluation is initially recommended for diagnosis, cardiac MRI, 18-fluorodeoxyglucose (FDG) PET and cardiac biopsy are recommended in cases with rapid progression that are unresponsive to treatment. Treatment recommendations for myocarditis attacks in RA patients are shown in Table 3 [177].

The use of anti-TNF in this group of patients is controversial and it is recommended that decisions be made according to the patient’s clinical condition. Although it is known to increase mortality, especially in patients with advanced HF with functional capacity class 3–4, use of anti-TNF was shown to suppress myocardial inflammation and reduce secondary myocardial fibrosis in mild cases [15, 178].

Long-term use of glucocorticoid therapy in RA patients increases the likelihood of AF development, whereas hydroxychloroquine and anti-TNF medications decrease AF development [179]. Another point to be considered in these patients is that the association of RA with amiadoran treatment given for rhythm control increases the risk of pulmonary fibrosis [180].

In terms of HT, regular blood pressure checks are recommended for all patients even if they are asymptomatic but especially for those receiving NSAIDs, glucocorticoids or leflunamide treatments.