Cardiomyopathy in ankylosing spondylitis



Based on clinical observations, the incidence of cardiac involvement in ankylosing spondylitis (AS) appears to have reduced over recent decades. The reason for this is unclear, but may be attributed to earlier recognition of AS and better control of the disease. When faced with unexplained cardiomyopathy in an AS patient, the challenging question facing cardiologists and rheumatologists alike is whether the cardiomyopathy, in the absence of other identifiable causes, is an extraarticular manifestation of AS and, if so, what are the underlying pathogenesis and the effect on the cardiomyopathy following treatment of the underlying inflammatory process in AS.

Extraarticular manifestations of AS are not uncommon. In 30–40% of AS patients, anterior uveitis may occur, being the most common extraarticular manifestation of AS (1). While involvement of the ascending aorta, aortic valves, and conduction system have been recognized in AS, myocardial manifestations remain less well defined.

Cardiac involvement appears to occur more commonly in patients with a longer duration of disease. For example, aortic incompetence was noted in 3.5% of AS patients with 15 years of disease duration and in 10% with 30 years of disease (2), while cardiac conduction disturbance was noted in 2.7% in those with 15 years of disease duration and in 8.5% after 30 years (1). These data suggest that the underlying inflammation in AS, if untreated, may affect various cardiac tissues, with more involvement after longer periods of inflammation. Although to our knowledge there has been no study to show a decline in the prevalence of these conditions with better control of inflammation, the observation in our AS cohort shows a very low incidence of AS-related valvular heart disease in recent years (3 of 503 AS patients), probably related to improved management of AS.

The association between AS and cardiomyopathy is not well defined. Cases of cardiomyopathy in AS in the past were mainly reports of left ventricular dysfunction defined by echocardiography (2–4). In an echocardiographic study of asymptomatic juvenile- and adult-onset AS patients with longstanding disease, it was found that up to 25% of juvenile-onset AS patients and 32.2% of adult-onset AS patients had evidence of cardiomyopathy (5).

Here we present 4 cases, 2 cases each from Singapore General Hospital (SGH) and Toronto Western Hospital (TWH), of idiopathic cardiomyopathy in AS and a review of the literature. The 2 patients from Singapore are ethnic Chinese, whereas the 2 cases from Canada are white. Case selection excluded circumstances in which the cardiomyopathy had a defined etiology such as coronary artery disease or valvular disease. Two patients were noted to have improvement of their ejection fractions following treatment of their AS, suggesting that cardiomyopathy in AS may be reversible with effective treatment of the underlying inflammatory process. Early detection of cardiomyopathy therefore may lead to better prognosis.

Case 1 (SGH)

A 52-year-old man first presented in 1992 with acute anterior uveitis, and in 2001 started experiencing neck stiffness, followed by knee and ankle arthritis. He was diagnosed with AS based on characteristic inflammatory back pain, evidence of bilateral sacroiliitis on pelvic radiographs, and positive HLA–B27. Despite treatment with salazopyrin and diclofenac, the patient continued to have recurrent flares of peripheral arthritis.

In 2003, he was admitted with congestive cardiac failure. A 2-dimensional (2-D) echocardiogram revealed evidence of dilated cardiomyopathy with a severely impaired left ventricular ejection fraction (LVEF) of 11% (Table 1). A dipyridamole myocardial perfusion scan showed no evidence of ischemia, but confirmed the impaired LVEF. No cardiac angiography was performed. He was started on angiotensin-converting enzyme (ACE) inhibitors, beta-blockers, and calcium-channel blockers for his heart failure. The patient was a nonsmoker and nondrinker.

Table 1. Summary of 4 AS patients with dilated CMP*
CaseSex/raceAge at onset of AS symptoms, yearsAge at diagnosis of CMP, yearsPeripheral arthritisAS treatmentCMP treatmentOutcome
  • *

    AS = ankylosing spondylitis; CMP = cardiomyopathy; MTX = methotrexate; SSZ = sulfasalazine; ACE = angiotensin-converting enzyme; CCB = calcium-channel blocker; LVEF = left ventricular ejection fraction; CHF = congestive heart failure; AF = atrial fibrillation.

1Male/Chinese3445YesMTX, SSZACE inhibitor, CCB, beta-blockerImprovement of LVEF (11% to 35%)
2Male/Chinese2732YesInfliximabACE inhibitor, CCB, beta-blockerImprovement of LVEF (10% to 44%)
3Male/white2530NoNoNoStable dilated CMP, LVEF 50–59%
4Male/white2062NoNoACE inhibitor, beta-blocker, digoxinRecurrent CHF with AF

Following the addition of methotrexate in 2004 for persistent active peripheral arthritis, he began to notice significant improvement to his disease and exercise tolerance, and he had no recurrence of heart failure. His latest 2-D echocardiogram in 2009 showed significant improvement of his LVEF to 35%. Of note, the left ventricle remained moderately dilated with a severely dilated left atrial cavity but without ventricular hypertrophy. The aortic root was mildly dilated (left ventricular outflow tract annulus 2.3 cm) with trivial aortic valvular regurgitation. Currently, his AS is stable with methotrexate and sulfasalazine.

Case 2 (SGH)

A 37-year-old man was diagnosed with AS in 2000, when he presented with chronic back pain and stiffness and peripheral arthritis. At presentation, he had pelvic radiographs showing ankylosis of the sacroiliac joints and positive HLA–B27. His peripheral arthritis remained very active despite treatment with nonsteroidal antiinflammatory drugs (NSAIDs), sulfasalazine, and methotrexate.

In January 2005, the patient had presented with palpitations and congestive heart failure (CHF). His left ventricle was dilated with a depressed LVEF of 35% and moderate global hypokinesia. A dipyridamole myocardial perfusion scan (MIBI) showed a small area of previous inferior infarction with no residual ischemia. His LVEF deteriorated in 2006 to 10% with significant diastolic dysfunction.

By 2007, the patient was immobilized by severe axial and peripheral disease. His functional capacity was also impaired by the cardiomyopathy, which resulted in poor effort tolerance. Medical treatment options were few because an episode of gastrointestinal bleeding prohibited the use of NSAIDs. He was experiencing New York Heart Association class IV heart failure. Temporary relief on his joint pain was achieved with intermittent intraarticular steroid injections.

In 2008, a colonoscopy was performed for recurrent episodes of gastrointestinal bleeding and a biopsy sample confirmed the diagnosis of Crohn's disease. Infliximab was instituted in June 2008 for treatment of his AS and Crohn's disease. His cardiac function was monitored closely. Following 3 injections of infliximab, his axial and peripheral arthritis improved significantly and he was able to ambulate. His Crohn's disease stabilized and he recorded healthy weight gain. His exercise tolerance improved significantly and a repeat echocardiography demonstrated improvement of his LVEF from a low of 10% to 44% and normalization of his left ventricular diastolic function (Figure 1). There were no features suggestive of aortic valve disease or cardiac amyloidosis. Currently, he continues on maintenance infliximab infusions and weekly methotrexate (22.5 mg) and has remained well, with no recurrence of the CHF.

Figure 1.

Effect of anti–tumor necrosis factor (anti-TNF) on left ventricular ejection fraction (LVEF).

Case 3 (TWH)

A 35-year-old man was diagnosed with AS in 2004 after an episode of acute anterior uveitis. Prior to the development of uveitis, he had experienced 4 years of mild inflammatory back pain, and pelvic radiographs showed bilateral sacroiliitis and positive HLA–B27. There were no extraaxial or extraskeletal manifestations. His spinal symptoms were relieved with celecoxib, which he took only on demand.

Following recurrent episodes of palpitation with chest discomfort, he returned for further evaluation. There was no decrease in exercise tolerance, nor were there any abnormalities on his electrocardiogram and cardiac enzymes.

A 2-D echocardiogram in June 2005 showed a moderately dilated left ventricular chamber size (left ventricular end diastolic diameter of 6.4 cm and end systolic diameter of 4.7 cm [normal range 3.55–5.7]) with mild decrease in the global left ventricular systolic function (LVEF 50–59%). The right ventricular chamber size was reported to be mildly enlarged with low normal systolic function. There were no abnormalities in the aorta and aortic valves and no pericardial effusion was detected.

On followup, he remained asymptomatic and led a normal lifestyle, and did not require any medication for his mild back symptoms. He was started on ACE inhibitors and beta-blockers by his cardiologist but admitted to being noncompliant to therapy, as he was asymptomatic. A stress echocardiogram in 2006 revealed no evidence of myocardial ischemia but persistent global mild hypokinesia of the left ventricular chamber. A repeat 2-D echocardiogram in 2007 revealed a persistently dilated left ventricular chamber with an ejection fraction of 53% and global mild dyskinesia.

Cardiac computed tomography performed in 2007 demonstrated dilatation of the main, right, and left pulmonary arteries, but without any signs of pulmonary valvulopathy or aortopathy. There was no evidence of flow-limiting lesions in the coronary arteries. Both ventricles were noted to be dilated with mildly reduced systolic function without any evidence of intracardiac shunt or other focal abnormalities. Cardiac magnetic resonance imaging could not be performed due to claustrophobia.

Case 4 (TWH)

A 69-year-old man with longstanding, stable AS was diagnosed with nonischemic dilated cardiomyopathy (LVEF 37%) in 2003 following recurrent episodes of congestive cardiac failure. A cardiac catheterization performed in 2003 showed normal coronary arteries except for a 60–70% narrowing of the ostia of the obtuse marginal branch of the left circumflex artery. A technetium scan ruled out myocardial ischemia as the cause of the cardiomyopathy. The lack of recognized risk factors for cardiomyopathy and the absence of significant coronary disease raised the possibility of cardiomyopathy secondary to AS.

The patient was diagnosed with AS at age 20 years when he presented with chronic buttock pain. Radiographs of his cervical and lumbar spine and pelvis taken years later revealed evidence of diffused ankylosis consistent with AS. He reported that his disease “burned out” when he was in his 40s. Metrology measurements revealed significant restriction of his spinal movements, as evidenced by the following: cervical spine rotation of 10 degrees bilaterally, Schober index of 0.5 cm, tragus-to-wall distance of 26 cm, and chest expansion of 1.0 cm. He had never experienced any significant peripheral arthritis or any extraarticular manifestations of AS such as uveitis or psoriasis. No antiinflammatory treatment was required for his AS ever since his disease “burned out.”

Subsequent to the diagnosis of cardiomyopathy, the patient had developed atrial fibrillation and had undergone electrocardioversion in 2003 due to issues with anticoagulation. Following treatment with digoxin, ramipril, and carvedilol, his heart remained in sinus rhythm with significant improvement of his left ventricular function (LVEF 60% in 2005, LVEF 58% in 2006), but in November 2008, he developed recurrent atrial fibrillation and a year later, recurrent heart failure. His LVEF decreased to 39% despite treatment with candesartan and carvedilol.


Previous echocardiographic reports have evaluated ventricular dysfunction in AS (2–4, 6–8). Our cases combined with these prior studies provide supportive evidence that cardiomyopathy in AS is a clinical reality. The cardiomyopathy in the cases reported above had no alternative explanation, with the possible exception of patient 4, who had a minor single-vessel disease. None of our patients had a history of rheumatic fever or evidence of longstanding hypertension or excessive alcohol consumption, myocarditis, or diabetes mellitus, all of which can contribute to ventricular dysfunction and dilatation. Except for patient 3, all of the patients met the criteria for dilated cardiomyopathy as classified by the World Health Organization/International Society and Federation of Cardiology Task Force (9). This report defines dilated cardiomyopathy as a cardiac muscle disorder with impaired ventricular function and increased end-diastolic and systolic volumes. The single exception was our third case, but this case is included to illustrate the fact that dilated cardiomyopathy may be asymptomatic in the early stages and may subsequently progress to significant CHF.

With relatively rare events such as dilated cardiomyopathy, it is a challenge to determine if the occurrence is seen with greater frequency than would be expected on the basis of chance alone. To obtain an estimate of the frequency, we compared the prevalence of cardiomyopathy in our AS patients with the crude prevalence among the general population. On the basis of 500 AS patients at TWH and 250 AS patients at SGH, the prevalence of dilated cardiomyopathy can be estimated at 400–800 per 100,000 population. In contrast, population studies have estimated the prevalence of idiopathic dilated cardiomyopathy at 8.3–36.5 per 100,000 (10–12). The prevalence of cardiomyopathy in AS patients may be an underestimate, given the underrecognition of this condition in AS patients who may attribute any decreased exercise tolerance to the underlying AS.

The underlying pathogenesis for ventricular dysfunction in AS is unclear. The landmark pathologic study in such patients was published by Brewerton et al in 1987 (13). In that study, histopathologic and necropsy studies in 28 AS patients with cardiomyopathy without ischemic or valvular heart disease or hypertension revealed a mild but diffuse increase in interstitial connective tissue (reticulin) (14). No inflammation was observed. Using echocardiography, this diffuse fibrosis typically manifests as diastolic dysfunction and abnormal left ventricular filling (10), seen in several echocardiographic studies of apparently normal AS subjects (3, 6–8). Systolic dysfunction and ventricular cavity dilatation could not be explained by this diffuse myocardial connective tissue, except in advanced stages of fibrosis or in relation to an underlying valvular pathology. If diffuse fibrosis was the cause of the cardiomyopathy, it would be difficult to account for the significant improvement in the LVEF of 2 of our 4 patients.

In contrast, pathologic studies of the aortic root and valves have presented a clearer picture. In the first detailed study of aortic root and valvular disease in AS patients by Bulkley and Roberts in 1973 (15), aortitis was confirmed with the presence of plasma cells and lymphocytes at the vasa vasorum with extension below the aortic root to the base of the mitral valve and into the interventricular septum. Their study was based on autopsy findings in 8 patients who had AS and aortic regurgitation severe enough to cause heart failure and eventual death, despite aortic valve replacement surgery in 5 of the patients. The consequence of chronic inflammation was infiltration by dense scar tissue in the adventitia of the valves and cusps, along with the root and sinuses and fibrous proliferation in the intima. The finding of fibrosis in valvulopathy concurred with the histopathologic findings in cardiomyopathy reported by Brewerton et al (9). By extrapolating the inflammatory process to cardiomyopathy, it can be speculated that in our series, patients 1 and 2 had partial improvement of their ventricular function following control of an underlying inflammatory process related to spondylarthritis.

In a genetic study of 65 patients with idiopathic dilated cardiomyopathy, HLA–B27, the gene that is found in 90% of AS patients, occurred with increased frequency in 14.5% of patients with idiopathic dilated cardiomyopathy compared to the 3.3% of the control haplotypes (P < 0.001) (16). It is unclear how HLA–B27 could cause cardiomyopathy, but it is known that myocardial inflammation may be driven by tumor necrosis factor α (TNFα), the cytokine that has been studied most extensively in CHF. Circulating TNFα has been shown to be in the blood of patients with CHF, with an increase in levels as patients became symptomatic (17). It has been implicated in the causation of heart failure in animal studies (18), and thus it was postulated that the antiinflammatory effect of TNFα antagonist therapy may ameliorate the progressive course of CHF. While there is well-documented evidence for the beneficial effect of TNFα blockers in AS, the use of these agents in patients with cardiac dysfunction has remained controversial (19). Despite current expert consensus contraindicating the use of anti-TNFα agents in patients with moderate to severe heart failure, to date there are no compelling data that TNFα blockers worsen CHF (20). As shown in patient 2, treatment of severe AS with anti-TNF may be associated with a concomitant improvement in the cardiomyopathy.

Cardiomyopathy is usually detected when it is relatively advanced, since most AS patients are young and have ample cardiac reserves. Furthermore, systolic dysfunction may occur following diastolic dysfunction as a result of diffuse myocardial fibrosis. Consequently, cases of cardiomyopathy in AS patients often present late and pose a diagnostic dilemma to the managing physician as to the etiology of the cardiomyopathy, principally due to the fact that ischemic heart disease is more prevalent in older patients. Furthermore, it is not known if cardiomyopathy correlates with disease duration or its severity, in contrast to current concepts of conduction abnormalities and aortic incompetence (1, 2, 4). Although there are no published mortality data with the specific cardiovascular problems in AS, cardiovascular causes accounted for the cause of mortality in up to 50% for AS patients (8, 21).

Rarely, infiltrative cardiomyopathy secondary to cardiac amyloidosis has been reported in AS patients, often resulting in severe heart failure (22, 23). Amyloid A–related amyloidosis, a systemic disease that occurs in 5% of patients with poorly controlled chronic inflammatory disease such as rheumatoid arthritis, AS, and familial Mediterranean fever (22), more commonly affect the kidneys than the heart with amyloid infiltration (24). Although cardiac amyloidosis may occur, significant deposition resulting in cardiomyopathy is rare. When it does occur, the prognosis is poor, with 1-year, 3-year, and 5-year survival rates of 40%, 25%, and 17%, respectively, from a series of 16 Chinese patients with cardiac amyloidosis (25). Biopsy-proven cardiac amyloidosis was confirmed in a necropsy study of a 52-year-old male AS patient as reported by Child et al (26). Prior to his death, the patient was found to have echocardiographic features of cardiac amyloidosis, which included nondilated ventricles with concentric left ventricular thickening, right ventricular thickening, prominent valves, dilated atria, and thickening of the interatrial septum (27). A recent study showed that late gadolinium enhancement cardiac magnetic resonance imaging could play an important role in the detection of cardiac amyloidosis, given its high sensitivity and ability to detect interstitial expansion from amyloid deposition (28). It is of interest to note that among the 503 AS patients in the Toronto cohort, there is only a single case of renal amyloidosis, so this event appears to be much less common than in older AS series.

Chronic inflammation and immune dysregulation that are seen in autoimmune disorders such as rheumatoid arthritis and systemic lupus erythematosus are thought to result in accelerated atherosclerosis (29, 30). Given the inflammatory nature of AS, it is likely that accelerated atherosclerosis is also more prevalent in patients with AS. However, the link between inflammation, atherosclerosis, and AS is controversial (31), even as a cross-sectional study has shown a higher prevalence ratio of 1.5 compared with controls for atherosclerosis in patients with AS (32). Ultrasonographic studies measuring flow-mediated dilatation and endothelium-independent dilatation of the brachial artery showed significantly lower readings in AS patients compared to healthy controls (33), indicating impairment of endothelial function, a precursor of atherosclerotic changes (34). Evidence of subclinical atherosclerosis and arterial stiffness in AS patients was further supported by a case–control study by Peters et al, who found a significant increase in the intima-media thickness of the common carotid artery in 59 AS patients who were scheduled for etanercept therapy when compared with 30 healthy controls (P = 0.02) (35). Accelerated atherosclerosis and ischemic heart disease in AS patients may be multifactorial, with contributions from traditional risk factors such as hypertension, hyperlipidemia, hyperfibrinogenemia, thrombocytosis, and hypercoagulability; high C-reactive protein (CRP) levels; and poorer physical activity levels compared to the general population (36). In particular, high levels of CRP, which may be present in patients with active AS, are known markers for premature atheroma related to chronic inflammation (37, 38). Indeed, as an important etiology for cardiomyopathy, the presence of ischemic heart disease could have played a contributory role in the cardiomyopathy, unless the ischemic heart disease was mild, as was the case in patients 2 and 4.

In conclusion, clinically significant heart failure secondary to AS-related dilated cardiomyopathy can occur, albeit rarely, in AS patients, and may be related to the underlying disease activity. However, the pre–heart failure stage is generally underdiagnosed, especially if the patient is asymptomatic. As the incidence is greater in men, especially in adult males, it is often attributed to ischemic heart disease and occasionally to valvular heart disease. Early identification, especially in patients with traditional cardiac risk factors, may therefore have therapeutic and prognostic implications. The role of therapy with TNF blockers in AS patients with mild to moderate heart failure secondary to nonvalvular cardiomyopathy warrants further study.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Lui had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Lui, Thumboo.

Acquisition of data. Lui.

Analysis and interpretation of data. Lui, Thumboo, Inman.