Cardiac involvement is one of the main complications substantially contributing to the morbidity and mortality of patients suffering from systemic autoimmune diseases. All the anatomical heart structures can be affected, and multiple pathogenic mechanisms have been reported. Non-organ-specific autoantibodies have been implicated in immune complex formation and deposition as the initial triggers for inflammatory processes responsible for Libman–Sacks verrucous endocarditis, myocarditis and pericarditis. Anti-phospholipid antibodies have been associated with thrombotic events in coronary arteries, heart valve involvement and intra-myocardial vasculopathy in the context of primary and secondary anti-phospholipid syndrome. Antibodies-SSA/Ro and anti-SSB/La antigens play a major pathogenic role in affecting the heart conduction tissue leading to the electrocardiographic abnormalities of the neonatal lupus syndrome and have been closely associated with endocardial fibroelastosis.
Nowadays, cardiac involvement is one of the major concerns in the management of patients suffering from systemic autoimmune diseases. Such an involvement has been recognized since the beginning of 20th century, but in the last decades, newly recognized clinical entities have been detailed due to the introduction of very sensitive, non-invasive or semi-invasive cardiac imaging techniques [ 1 ].
Several autoantibodies, such as anti-phospholipid antibodies (aPL), anti-SSA/Ro antibodies and anti-endothelial cells antibodies, can mediate cardiac damage.
These autoantibodies can directly affect heart tissue or, alternatively, can trigger mechanisms able to cause heart damage: for example, aPL can contribute to cardiac damage enhancing atherosclerosis phenomena, causing thrombosis of coronary arteries or starting an immune-complexes-mediated reaction and deposition at the valves level.
Consequence of autoantibodies damage has been reported in several heart structures such as valves, myocardium, pericardium, conduction tissue and cardiac arteries in patients suffering from systemic lupus erythematosus (SLE), anti-phospholipid syndrome (APS), Sjogren syndrome and other autoimmune rheumatic diseases (ARD).
Cardiovascular disease has recently been acknowledged as a primary cause of morbidity and mortality in SLE as well as in APS, and numerous factors leading to accelerated atherosclerosis have been characterized.
Other cardiac manifestation, such as pericarditis, myocarditis, endocarditis and conduction disturbances, when present, are often mild and, usually, subclinical features are more prevalent than clinically apparent disease.
Heart involvement in neonatal lupus
Complete heart block (CHB) is the most serious manifestation of the neonatal lupus syndrome (NLS), a congenital syndrome in which maternal IgG anti-Ro/SS-A autoantibodies cross the placenta and injure an otherwise normally developing heart [ 2 ].
Fetal/neonatal disease is independent of maternal disease: in fact, mothers may have SLE, Sjogren syndrome or other autoimmune symptoms, or may be entirely asymptomatic [ 3 ].
CHB generally appears on an intact heart and must be differentiated from secondary heart block due to cardiac malformations, which is not associated with anti-SSA/Ro or anti-SSB/La antibodies.
CHB is often regarded as a model of passively acquired autoimmunity in which antibodies are necessary but insufficient to cause CHB, and fetal factors are likely contributory [ 4 ] some evidence against the role of anti-SSA/Ro and anti-SSB/La antibodies alone in the pathogenesis of CHB are listed below: (i) CHB is rare in adults with these antibodies; (ii) there is a discordance of CHB in monozygotic twins [ 5 ]; and (iii) there are few babies of Ro/La mothers affected by CHB [ 3 ].
Detection of CHB in the fetuses most commonly occurs in utero between 17 and 24 weeks of gestation [ 3 ]. CHB may be associated with fetal myocarditis [ 6 ] that can lead to fetal hydrops and stillbirth.
The degree of heart block includes all levels from first degree, discovered accidentally with electrocardiogram after the birth or in utero by a prolonged PR interval detected on echocardiogram, through third degree (complete) heart block, the most frequently recognized. Complete CHB, once established, is irreversible. It is clearly documented that incomplete blocks (including those improving in utero with dexamethasone) can progress after birth despite the clearance of the maternal autoantibodies from the neonatal circulation [ 7 ]. In Buyon's registry, among nine children with a CHB of first degree, four have shown block progression after birth and among others, two of four newborns with CHB of second degree progressed toward a third degree CHB.
Such post-natal progression of CHB has been described in the past by others and today justifies performing electrocardiogram in all children born to anti-SSA/Ro positive mothers [ 8 ].
CHB carries a significant morbidity and mortality (15–30%) most often in utero or in the first few months of life: in fact, 67% of all the recognized cases require a pacemaker insertion before reaching adulthood [ 3 ]. Risk factors for permanent pacing are represented by very slow heart rates and symptoms like poor exercise tolerance, cardiomegaly, long QRS or QT durations, ectopy, syncope or structural or functional heart disease [ 9 ].
Cardiac damage may extend beyond the conduction system. There is a 10% incidence of late-onset dilated cardiomyopathy (CM) developing despite the early successful pacemaker implantation for the associated heart block [ 3 , 10 ]; in these patients, CM led to congestive heart failure and subsequently death or cardiac transplantation in four and eight infants, respectively. The risk to develop CM for children with CHB was valuated at 5–11% [ 10 ]: for this reason, children with CHB need a close follow-up of electrocardiogram and function of pacemaker, as well as ventricular function. Moreover, CM can be seen in the absence of heart block.
Recently, Nield et al . [ 11 ] have reported an endocardial fibroelastosis associated with CHB in 13 children born to anti-SSA/Ro or anti-SSB/La positive mothers. In all 13 cases (six in prenatal, seven in post-natal), a severe ventricular dysfunction was diagnosed: 11 of them died and two received cardiac transplantation.
The incidence of CHB in an offspring of a mother with anti-SSA/Ro antibodies is about 2%, while, if the mother already had a first affected child, the risk of CHB in a subsequent pregnancy rises to 18% [ 3 , 12 , 13 ]. Optimally, all pregnant women with anti-SSA/Ro or -SSB/La antibodies should have a serial fetal echocardiography done by an experienced paediatric cardiologist weekly from 16 to 26 weeks, and every other week until about 34 weeks to evaluate the fetal heart during the period of presumed vulnerability [ 14 ]; the development of a non-invasive Doppler technique to measure the mechanical PR interval may, in this way, allow an earlier diagnosis and treatment opportunities [ 15 ]. The early diagnosis of the CHB and its potential complications (pericardial effusion, myocarditis) usually can avoid the deterioration of the fetal cardiac function. When imaging techniques show in utero the presence of incomplete CHB, the suggested treatment of mothers is based on fluorinated steroids (dexametasone or betametasone). These cross the placenta and are available to the fetus in an active form [ 16 ], inhibiting the immune process in the fetal heart. Fluorinated steroids may also improve the survival of fetuses with a complete CHB, even if there is no evidence of a durable recovery [ 17 ]. According to some authors, betametasone should be preferred because of the reported risks associated with high dexamethasone doses to the fetal brain [ 18 , 19 ]. We have recently evaluated the neuropsycological development in 13 children born with CHB, of which 11 were exposed to high doses of dexamethasone during fetal life and two were not: even if some bias due to the small cohort, interestingly, a clinically borderline case of learning disabilities was found in a child not exposed to dexamethasone [ 20 ].
The prophylactic therapy of the high-risk mother (those with anti-SSA/Ro antibodies and a previous child with CHB) with fluorinated steroids actually is not recommended because of the suspected neurological toxicity of dexamethasone [ 18 , 19 ] and the high rate of adverse obstetrical events (including spontaneous abortions, stillbirth, severe intrauterine growth restriction and adrenal suppression) in patients treated with steroids to prevent CHB [ 21 ].
In addition to the classical complete and incomplete CHB, a high frequency of transient fetal first degree CHB has been recently reported, that in most of the cases is spontaneously reverted before or shortly after birth [ 22 ].
Other possible cardiac manifestations, such as sinus bradycardia or prolonged QTc interval in otherwise healthy children of anti-SSA/Ro positive mothers, have not been confirmed and are currently a matter of debate [ 23 , 24 ].
Heart involvment in systemic lupus erythematosus
The heart is frequently involved in SLE: very sensitive methods of cardiovascular investigation have found the prevalence of cardiac involvement to be >50% [ 25 ].
In the past, cardiac manifestations were severe, often leading to death and they were frequently found in post-mortem examinations. Nowadays, cardiac manifestations are often mild and asymptomatic and they can be recognized by echocardiography and other non-invasive tests [ 26 ].
All three layers of the heart—pericardium, myocardium and endocardium—can be involved by lupus; this section will focus on pericarditis and myocarditis because heart valve abnormalities are common lesions in either SLE and APS and will be discussed in next section.
Pericarditis is the most studied cardiovascular manifestation of SLE, although often not evident clinically, and it is included in the American College of Rheumatology (ACR) classification criteria for SLE.
The pericardium can be involved by acute and chronic inflammatory changes; granular deposition of immunoglobulin and C3, demonstrated by direct immunofluorescence, support the role of immune complexes in the development of pericarditis.
The reported prevalence of pericardial abnormalities, detected by echocardiographic studies, ranges from 11 to 54% [ 27 ]: this variability is partially attributable to the methods used to document pericardial disease and whether symptomatic or asymptomatic cases are included. Clinical (symptomatic) pericarditis is estimated to occur in 25% of SLE patients at some point in the course of their disease. Asymptomatic pericardial effusion is clearly more common than clinical pericarditis: in fact, 40% of unselected patients with SLE have pericardial effusion, detected using echocardiography. Moreover, a combined autopsy series revealed pericardial involvement in 62% of patients with SLE [ 28 ].
Pericardial involvement appears more frequently at SLE onset or during SLE relapses, although it can occur at any time of the disease [ 26 ]. Pericarditis usually appears as an isolated attack or as recurrent episodes [ 29 ].
Signs and symptoms of acute pericarditis include a typical precordial or substernal chest pain, usually positional (aggravated by lying down), often with a pleuric quality, sometimes with dyspnoea; moreover, patients may have fever, tachycardia and decreased heart sounds; pericardial rubs can be heard but usually are rare, perhaps because they are present often for only a few hours and are missed. The diagnosis can be confirmed by ECG findings of elevated ST segments and peaked T waves (although slight T-wave changes or transient elevation of ST segments are most characteristic). Co-existent pleurisy, effusion or both are common [ 30 ].
Patients with pericardial effusion (as opposed to thickening) are more likely to have pericardial pain and active lupus elsewhere; when present, pericardial effusion are usually small and do not cause haemodynamic problems [ 31 ]. Echocardiography represents the standard method to investigate pericardial abnormalities and is able to demonstrate mild effusion or thickening of pericardial layers, therefore, should be performed periodically in SLE patients.
Complications of pericarditis, such as cardiac tamponade, constrictive pericarditis and purulent pericarditis are rare, and invasive procedures such as pericardiocentesis or pericardial window are rarely needed.
Non-steroidal anti-inflammatory drugs and/or corticosteroids are the first line of treatment in mild pericarditis. Intravenous bolus of corticosteroid is necessary in more severe cases or if tamponade is present, while in patients with recurring pericarditis, chronic suppression with methotrexate, azathioprine or mycophenolate mofetil may be effective.
Myocarditis is the most characteristic feature of myocardial involvement in SLE.
The clinical detection of myocarditis ranges from 3 to 15%, although it appears to be much more common in autopsy studies (mainly done in the 1950s and 60s), suggesting the largely subclinical nature of lupus-associated myocarditis [ 26 , 30 ]. A more recent post-mortem study, reflecting the era of corticosteroids treatment, found much lower frequencies, from 0 to 8% [ 32 ].
Immunofluorescence studies demonstrate fine granular immune complexes and complement deposition in the walls and perivascular tissues of myocardial blood vessels, supporting the hypothesis that lupus myocarditis is an immune complex-mediated disease. Some reports demonstrate an association between anti-SSA/Ro antibodies and myocarditis [ 33 ].
Signs and symptoms are similar to those of myocarditis due to other causes (dyspnoea, tachycardia, arrhythmias) and they can progress to ventricular dysfunction, dilated CM and heart failure. There are no typical findings on ECG, and cardiac enzymes may be normal.
Echocardiographic studies cannot definitely diagnose myocarditis, but global hypokinesis, in the absence of other known causes, is strongly suggestive. Large echo series have found frequencies of global hypokinesis between 5 and 20%. However, also segmental areas of hypokinesis can be indicative of the disease [ 34 ].
Recently, other non-invasive investigations such as magnetic resonance, are employed for diagnosing myocardial involvement in SLE: T2 values sensitively indicated myocardial relaxation abnormalities, even at preclinical stage [ 35 ].
However, up to now, endomyocardial biopsy remains the technique of choice in diagnosing myocarditis even if the procedure is invasive and subject to sampling error.
Myocarditis, although mild, has to be treated immediately with high-dose steroids; in the most severe forms is necessary to use intravenous pulse corticosteroid followed by high oral doses. The addition of immunosuppressant such as azathioprine, cyclophosphamide or intravenous immunoglobulines (IVIG) may be helpful in the treatment of myocarditis [ 36 ].
Efficacy of the therapy can be assessed by serial echocardiographic studies or right ventricular endomyocardial biopsies.
Recently, an association of SLE and giant cell myocarditis has been reported. Despite clinical similarities, lupus myocarditis and giant cell myocarditis are histologically distinct entities, and the latter has a much more unfavourable prognosis [ 37 ].
Heart involvement in anti-phospholipid syndrome
Heart valve abnormalities (vegetations and/or thickening) are the most frequent cardiac manifestations of APS. These alterations were known as Libman–Sacks endocarditis, a verrucous endocarditis of valve leaflets, papillary muscles and the mural endocardium, originally described in SLE patients [ 38 ]; later both clinical observations and experimental data showed a close linkage between aPL and cardiac valvulopathy and documented the responsibilities of antibodies in the valvulopathy genesis.
Heart valve lesions (vegetation, valve thickening and dysfunction) are frequently reported in patients with APS with and without SLE [ 39 ] and in those with aPL alone [ 40 ]. According to echocardiographic studies, it is not clear if patients with SLE have valve disease more or less than in patients with primary APS; in addition, there is a discrepancy in the prevalence of valvular disease in SLE patients with or without aPL.
Hojnik et al . [ 40 ] reviewed echocardiographic studies of primary APS patients: the four largest transthoracic echocardiography (TTE) studies reported 32–38% prevalence of valve lesions that most frequently involved left-sided valves, mitral more commonly, followed by aortic (whereas Libman–Sacks involves the tricuspid valve most often).
Using transosophageal echocardiography (TEE), which is more sensitive for detection of valve lesions, Turiel et al . [ 41 ] demonstrated valve abnormalities in 82% of primary APS patients and mitral valve thickening in 63%, confirming previous data. This study also suggested that mitral valve thickening correlated with anti-cardiolipin antibodies (aCL) titre and aCL titre >40 GPL is a risk factor for thromboembolism, occurring in 25% of patients [ 40 ].
Recently Erdogan et al . [ 42 ] demonstrated cardiac involvement in 84% of primary APS patients and mitral regurgitation in 77.4%; interestingly, valve lesions were present in all stroke patients, confirming that the presence of cardiac valves pathology may be considered a risk factor for epilepsy, stroke and other CNS involvements, particularly in patients with primary APS [ 43 ].
The difference in the populations examined, the problems linked with aPL tests performance and the different echocardiography techniques (TEE vs TTE) can account for the variable prevalence of valve lesions in the studies mentioned above.
Valve abnormalities associated with aPL are similar to those reported in SLE, varying from minimal thickening and/or vegetations to severe valve distortion and dysfunction.
Valvular disease, for the most part, is mild and asymptomatic; only rarely (4–6%) do aPL positive patients develop valve disease severe enough to require surgical treatment.
There is no direct evidence that treatment with corticosteroids or cytotoxic therapy can prevent valvular damage; however, the decline in prevalence of Libman–Sacks lesions at autopsy following the introduction of corticosteroids supports a possible indirect beneficial role.
The valvular abnormalities resulting from Libman–Sacks lesions may predispose patients to bacterial endocarditis, so prophylactic antibiotics should be used for dental or surgical procedures with an increased risk of transient bacteraemia.
Atherosclerosis and coronary artery disease
Epidemiological studies showed an increase of cardio and cerebrovascular events in patients suffering from systemic autoimmune diseases, and autoptic investigations pointed out that an accelerated atherosclerotic process is largely responsible for such manifestations [ 44–46 ]. These observations support a possible role of autoimmunity in the genesis of atherosclerosis that may have clinical or subclinical features. The clinical edge of this phenomenon is coronary artery diseases (CAD) (myocardial infarction, angina, sudden death), while early endothelial dysfunction, abnormalities of circulation or atherosclerotic plaques, detected by different imaging techniques, identify the subclinical atherosclerosis expression.
Both preclinical (carotid plaque) and clinical (myocardial infarction) atherosclerotic diseases are more prevalent in SLE patients than in the general population; clinically, atherothrombotic events, such as myocardial infarction (MI), have been recognized as risk factors for mortality [ 26 , 47–51 ]; there may be a bimodal distribution of mortality risk factors in lupus: an ‘early’ peak in mortality is caused by disease activity and severity itself, as well as infections, while a ‘late’ peak is related to CAD [ 52 ]. In post-mortem studies, significant atherosclerosis was observed in >50% of deceased SLE patients regardless of the actual cause of death [ 53 ].
CAD is described with a prevalence ranging from 6 to 10%, and, in SLE patients, the risk of developing any CAD is 4–8 times higher than in controls [ 47 , 48 , 50 , 51 ]. In young women with SLE, the risk of MI is increased 50-fold [ 54 ]. In various cohort studies, MI was the cause of death in 3–30% of SLE patients [ 55 ].
In SLE patients, the role of traditional and non-traditional risk factors for atherosclerosis is still debated. Some studies have shown that traditional cardiovascular risk factors are also more predictive in SLE patients than in age- and sex-matched healthy subjects [ 56 ]; particularly, older age at diagnosis, hypercholesterolaemia and hypertension were the three most common predictors of CAD [ 57 ]. Hyperlipidaemia in SLE has two major patterns. Patients with active lupus, especially children, have low high-density lipoprotein-C (HDL-C) and elevated very-low-density lipoprotein-C (VLDL-C) and triglyceride levels. Moreover, corticosteroids therapy seems to increase the serum concentration of cholesterol, lipoproteins and triglyceride, whereas hydroxychloroquine seems to reduce them in SLE patients.
Other non-traditional risk factors associated with the autoimmune-inflammatory pathogenesis of the disease or with immunosuppressive therapy must also be taken into account: among these SLE-related risk factors, besides cumulative dosage and/or length of corticosteroids therapy, disease duration, high score of activity or damage could contribute to the development of atherosclerotic plaque [ 26 ].
More recently, some novel risk factors for atherosclerosis have been proposed and reviewed [ 58 ]; they include inflammatory markers (C-reactive protein, fibrinogen, interleukin-6), co-stimulatory molecules (CD40/CD40L), adhesion molecules, aPL including anti-cardiolipin (aCL) and anti-β2 glycoprotein I (anti-β2GPI); anti-oxidized low-density lipoprotein (anti-oxLDL), anti-oxidized palmitoyl arachidonoyl phosphocholine (anti-oxPAPC) and anti-hsp antibodies, homocysteine, lipoprotein(a) and HDL.
According to different reports, traditional risk factors were not different in APS and in the general population [ 48 , 49 ]. Therefore, non-traditional risk factors such as antibodies seem to be involved in APS-associated atherogenesis.
In histological studies of human carotid samples, β2GPI was shown to co-localize with T CD4+ lymphocytes in the subendothelial region of atherosclerotic plaques, supporting a possible role of antibodies in the disease progression [ 59 ]. On the other hand, in vitro , aPL accelerate the process of plaques formation, enhancing the macrophages transformation into foam cells by oxLDL. In fact, anti-β2GPI, reducing paraxonase activity, accelerates the formation of oxLDL [ 60 ]. In addition, since β2GPI binds oxLDL, in the presence of anti-β2GPI, the oxLDL uptake by human macrophages is enhanced because it seems to occur through the Fcγ receptors (FcγR) of the complex β2GPI–anti-β2GPI rather than via the usually employed and less efficient scavenger receptor [ 61 ].
Finally, the immunization with human β2GPI, which stimulates autologus anti-β2GPI antibodies formation, can also accelerate early atherosclerosis appearance in LDL-receptor-deficient mice or in anti lipoprotein E (APO-E) knock-out mice, without alteration of the animals’ lipids profiles [ 62 ]. It is relevant that oral feeding of the animals with human or bovine β2GPI was effective in reducing atherosclerosis as compared with control fed animals [ 63 ].
In agreement with these findings, APS patients suffer from an increased rate of cardiovascular accidents: myocardial infarction appears at same stage of the disease in up to 5.5% and is the presenting manifestation in 2.8% of APS patients [ 39 ]. Veres and others [ 64 ] showed correlation between serum levels of aCL and anti-β2GPI antibodies and the incidence and severity of acute coronary syndrome, MI and stroke. Lupus anticoagulant (LA) and anti-β2GPI are the risk factors for myocardial infarction in the study of SLE patients cohorts [ 65 , 66 ].
Regarding clinical and diagnostic aspects of APS-associated atherosclerosis, early endothelial dysfunction and increased common carotid intimal–medial thickness (ccIMT) have been studied [ 48 ]. Soltész et al . [ 71 ] reported abnormal flow-mediated vasodilatation of the brachial artery and increased ccIMT in 46 patients with primary APS. Others found a correlation between aCL IgG antibody levels and ccIMT [ 46 , 47 ]. Earlier development of carotid plaques were reported in SLE-associated APS in comparison with primary APS [ 46 , 72 ].
Atherosclerosis treatment strategies in SLE and SLE-associated secondary APS include an aggressive control of all traditional risk factors including hyperlipidaemia, hypertension, smoking, obesity and diabetes mellitus, which should be performed by using both drug treatment and changes in lifestyle [ 73 ].
Prophylactic therapy include anti-platelet and, in APS cases, anti-coagulant agents, as well as statins, folic acid, B vitamins and, as described above, possibly hydroxycloroquine (HCQ) that exerts evident anti-atherogenic properties. Statin therapy significantly reduces the risk of CAD and also prevents endothelial dysfunction [ 74 ].
Aspirin has been used for a long time to prevent CAD in the general population. Daily aspirin reduces the risk of MI and reduces CAD-related mortality; recent studies suggested that SLE patients might also benefit from aspirin prophylaxis. Again, there is no evidence from clinical trials to support this proposal. In a decision analysis model, aspirin intake in 40-yr-old lupus patients was estimated to gain 3 months of quality-adjusted survival in APA-negative and 11 months in aPL-positive individuals [ 75 ]. In a cohort study, the use of aspirin resulted in a 70% reduction of cardiovascular mortality in SLE [ 76 ]. According to recent guidelines, SLE patients with previous history of MI, angina or stroke; aPL positive subjects; patients with hypertension, hyperlipidaemia or diabetes mellitus, or smokers should be prescribed aspirin if there are no contraindications [ 77 ].
Lupus patients with secondary APS often take anti-coagulants, such as warfarin. As aspirin treatment has not been shown to add any benefit over warfarin alone, the use of aspirin may not be necessary in warfarin-treated SLE patients [ 78 ].
As described above, the role of corticosteroids in lupus-associated atherogenesis is rather controversial, as these agents may themselves be directly atherogenic, but they may also indirectly prevent premature atherosclerosis by controlling disease activity [ 48 , 49 ].
The authors have declared no conflicts of interest.