Antiphospholipid syndrome (APS) may cause coronary thrombosis. This study was undertaken to determine the prevalence of silent myocardial disease in patients with APS, using late gadolinium enhancement (LGE) of cardiac magnetic resonance imaging (CMRI).
Twenty-seven consecutive patients with APS and 81 control subjects without known cardiovascular disease underwent CMRI. The prevalence of occult myocardial ischemic disease, as revealed by LGE, was compared between patients with APS and controls, and factors associated with myocardial disease were identified in patients with APS.
Myocardial ischemic disease, as characterized by LGE on CMRI, was present in 8 (29.6%) of 27 patients with APS, and imaging with LGE showed a typical pattern of myocardial infarction (MI) in 3 patients (11.1%). The myocardial scarring revealed on CMRI was not detected by electrocardiography or echocardiography. Although both patients with APS and control subjects shared a low risk of cardiovascular events, as calculated with the Framingham risk equation (mean ± SD 5.1 ± 8.2% and 6.5 ± 7.6%, respectively, for the absolute risk within the next 10 years; P = 0.932), the prevalence of myocardial ischemia was more than 7 times higher in patients with APS (P = 0.0006 versus controls). No association was found between myocardial disease in patients with APS and classic coronary risk factors. The presence of myocardial scarring tended to be more closely associated with specific features of APS, such as duration of the disease, presence of livedo, and positivity for anti–β2-glycoprotein I antibodies.
The finding of a significant and unexpectedly high prevalence of occult myocardial scarring in patients with APS indicates the usefulness of CMRI with LGE for the identification of silent myocardial disease in such patients.
The antiphospholipid syndrome (APS) is an autoimmune thrombophilic condition characterized by venous or arterial thromboses or pregnancy complications in the presence of antiphospholipid antibodies (aPL), namely, lupus anticoagulant (LAC) and/or anticardiolipin antibodies (aCL) and/or anti–β2-glycoprotein I (anti-β2GPI) antibodies (1). Regarding the clinical spectrum of APS, any combination of vascular occlusive events may occur in the same individual, with a highly variable time interval between each event (2–4).
Myocardial infarction (MI) is diagnosed in 5.5% of patients with APS and is the presenting manifestation in 2.8% of patients (2, 5). However, considering the low sensitivity of clinical symptoms and Q waves on electrocardiography (EKG) for the diagnosis of MI (6, 7), the true prevalence of MI may be higher. Late gadolinium enhancement (LGE) of cardiac images obtained by contrast-enhanced cardiac magnetic resonance imaging (CMRI) may detect and characterize myocardial scarring that may be missed by EKG (8), echocardiography (9), or nuclear imaging techniques (10). The primary objective of our study was to estimate the prevalence of occult MI, as characterized by LGE, in patients with APS who had no known history of coronary artery disease (CAD). The secondary objective was to identify the factors associated with occult MI in patients with APS.
PATIENTS AND METHODS
From December 2008 to August 2009, 61 patients with primary or secondary APS were screened at the Department of Internal Medicine of Bichat-Claude Bernard Hospital (Paris, France). The diagnosis of APS was based on a history of venous and/or arterial thromboses or recurrent miscarriages in the presence of aPL, in accordance with the Sapporo criteria (11). Patients were considered to have the secondary form of APS if they had concurrent systemic lupus erythematosus (SLE) as defined by the American College of Rheumatology criteria (12, 13). Forty-two patients met the criteria for APS as proposed in the International Consensus Statement in 2006 (1), and these patients were considered for inclusion.
The cutoff titers used to define positivity for aCL and anti-β2GPI antibodies in the patients' sera were 40 IgG phospholipid units or 40 IgM phospholipid units, and these titers had to be higher than the 99th percentile, as measured on 2 or more occasions at least 12 weeks apart by a standardized enzyme-linked immunosorbent assay, in accordance with published criteria (1). Seven patients did not undergo CMRI because of renal insufficiency along with a creatinine clearance rate of <30 ml/minute (n = 5), claustrophobia (n = 1), or the presence of an orthopedic prosthesis (representing a metallic hazard) (n = 1). Five patients declined to participate in the study. One patient with a history of MI and 2 with a history of acute coronary syndrome (ACS) underwent CMRI but were subsequently excluded from the analyses. Thus, 27 patients with APS and no known history of CAD were included in the study. All patients provided their informed consent to undergo CMRI and to participate in the observational study, in accordance with the Declaration of Helsinki.
A group of 81 control subjects without a history of cardiac disease, including ischemic cardiac disease, myocarditis, hypertrophic or infiltrative heart disease, or systemic disease with cardiac or pericardial involvement, underwent delayed-enhancement CMRI at the same Department of Medical Imaging during the same period of time. None of the control subjects had a history of venous and/or arterial thromboses or recurrent miscarriages suggestive of APS. Forty-one of these control subjects were included in a clinical trial that involved cardiovascular evaluation of adult patients with pseudohypoaldosteronism type 1 and their relatives, conducted at the Assistance Publique-Hôpitaux de Paris (NCT00646828). Forty subjects underwent CMRI for systematic cardiovascular screening because of diabetes (n = 10) or ischemic stroke (n = 10) or because they were asymptomatic relatives of young subjects who had died as a result of a sudden cardiac event (n = 20), in accordance with current clinical practice.
Definition and scoring of coronary risk factors.
All patients underwent a detailed medical history review at the time of CMRI. Hypertension was defined as a systolic blood pressure (BP) >140 mm Hg or diastolic BP >90 mm Hg, with consistency across ≥2 readings obtained on ≥2 visits, or a need for antihypertensive treatment according to the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure 7 criteria (14, 15). Hypercholesterolemia was defined as any indication for cholesterol-lowering drug treatment according to the Adult Treatment Panel III of the National Cholesterol Education Program guidelines (16). A diagnosis of diabetes mellitus was based on a persistent finding of fasting hyperglycemia and the need for antidiabetic drug therapy (17). A family history of CAD was defined as the occurrence of MI in first-degree relatives at the age of <55 years in men and <65 years in women. Smoking was considered significant when tobacco consumption exceeded 10 pack-years.
Clinical evidence of MI was based on either a history of MI or documentation of an MI in the medical records, or a finding of significant Q waves in at least 2 contiguous leads on EKG. The risk of cardiovascular events was calculated as the absolute risk within the next 10 years using the Framingham risk equation, which includes adjustments for age, sex, total cholesterol level, high-density lipoprotein cholesterol level, smoking history, and systolic BP (18).
EKG and echocardiography measurements.
Resting 12-lead EKG and transthoracic echocardiography were performed on the day of CMRI. A single reader, who was blinded to the CMRI and clinical data, interpreted the findings on EKG. Echocardiography, which was performed using a Vivid 7 Vantage machine (General Electric Medical Systems), was conducted by a single investigator (FH), who was blinded to the clinical data and the EKG and CMRI results. Myocardial contraction was analyzed in a 2-dimensional mode in the parasternal long axis, short axis, apical 4-chamber, apical 2-chamber, and apical long-axis views.
CMRI protocol and data interpretation.
All patients were placed in a supine position and assessed using a 1.5T CMRI system (Signa CV/i; GE Healthcare) with a 4- or 8-element phased-array surface coil. The CMRI studies consisted of cine steady-state free precession imaging (repetition time 3.4 msec, echo time 1.2 msec, in-plane spatial resolution 1.6 × 2 mm) of left ventricular function, and imaging with LGE (repetition time 4.8 msec, echo time 1.3 msec, inversion time 200–300 msec) for myocardial scarring. All images were acquired with EKG gating and with the patients kept in a breath-holding state. Cine imaging and LGE imaging were obtained in 8–14 matching short-axis planes (8-mm thick with 0-mm spacing) and 3 radial long-axis planes. A segmented inversion-recovery pulse sequence for LGE was used, starting 15 minutes after achieving a cumulative dose of 0.15 mmoles/kg of gadolinium DTPA (19). Parallel imaging techniques (array spatial sensitivity encoding technique, with an accelerating factor of 1.5–2) were used throughout some studies to shorten the duration of the patients' breath-hold.
Two independent and experienced readers (JMS and JPL), who were blinded to the clinical data, analyzed the CMR images separately. They visually judged the occurrence (absence versus presence), localization, and pattern of LGE imaging of myocardial scarring. The pattern and extent of LGE-detected myocardial scarring were assessed by using short- and long-axis views, and myocardial scarring was defined as present only if it was detectable by LGE on 2 orthogonal planes. Areas of LGE were allocated to the American Heart Association 17-segment model (20). Regions of LGE were defined to reflect myocardial fibrosis. On the basis of previous experience from clinical and experimental studies, regions of LGE were defined as indicative of either typical MI (involving the subendocardium) or atypical MI (subepicardial, patchy midwall, or diffuse, circumferential pattern). In cases of disagreement between the readers, a third data review was performed.
Continuous data are expressed as the mean ± SD, while discrete variables are presented as percentages. For the prevalence of LGE-detected MI in patients with APS and controls, results are expressed as the relative risk along with 95% confidence interval (95% CI). The nonparametric Mann-Whitney test (for continuous data) or Fisher's exact test (for discrete variables) was used to calculate differences between patients with APS and controls and between groups of APS patients (on images with or without LGE). P values less than 0.05 were considered significant. Because of the explorative nature of the analysis, no correction for multiplicity of tests was performed.
Baseline characteristics of the subjects.
Twenty-seven consecutive patients with APS (9 male and 18 female, mean ± SD age 42.9 ± 17.3 years) were studied (Table 1 and Figure 1). Patients had APS for a mean ± SD duration of 10.4 ± 8.7 years. Twenty-one patients (77.8%) had primary APS. Six patients (1 male and 5 female, mean ± SD age 47.5 ± 20.9 years) had APS associated with SLE, and at the time of CMRI referral, the SLE Disease Activity Index score was 0 for all of these patients, and all 6 were being treated with a low dose of steroids and hydroxychloroquine, while 2 of them had previously received immunosuppressive therapy for lupus nephritis. Eighteen patients (66.7%) had isolated venous thrombosis, including deep and superficial vein thrombosis of any site or pulmonary embolism. Seven patients (25.9%) had arterial thrombosis, including ischemic stroke (n = 5) and peripheral arterial ischemia (n = 2), which was associated with venous thrombosis in 4 cases. Two patients had isolated recurrent miscarriages, defined as more than 3 unexplained consecutive spontaneous abortions before the tenth week of gestation. The prevalence of concurrent cardiovascular risk factors, such as hypertension, hypercholesterolemia, tobacco use, and diabetes, was low in the patients with APS, as shown in Table 1.
Table 1. Demographic characteristics and medical histories of the study cohorts*
APS patients (n = 27)
Control subjects (n = 81)
Except where indicated otherwise, values are the number (%) of subjects. APS = antiphospholipid syndrome; BMI = body mass index; LDL = low-density lipoprotein; CAD = coronary artery disease; NA = not applicable.
Within the next 10 years, calculated using the Framingham risk equation.
Eighty-one control subjects were also studied. None of the control subjects had a known history of cardiac disease and all were asymptomatic. Although the percentage of subjects with diabetes tended to be higher among the control group than among patients with APS, no significant differences in the prevalence of classic coronary risk factors were observed between the groups (Table 1). The absolute risk of cardiovascular events occurring within the next 10 years, calculated with the Framingham risk equation, was a mean ± SD 5.1 ± 8.2% in patients with APS compared with 6.5 ± 7.6% in control subjects (P = 0.932). The clinical characteristics and disease features of the patients with APS are shown in more detail in Table 2.
Table 2. Clinical characteristics and disease features of the patients with antiphospholipid syndrome (APS), according to the presence or absence of myocardial infarction (MI) revealed on late gadolinium enhancement (LGE) of cardiac magnetic resonance imaging*
All patients (n = 27)
LGE evidence of MI
Present (n = 8)
Absent (n = 19)
Except where indicated otherwise, values are the number (%) of patients. No significant differences between patients with and those without LGE evidence of MI were observed. BMI = body mass index; LDL = low-density lipoprotein; HDL = high-density lipoprotein; CAD = coronary artery disease.
Among the patients with antiphospholipid antibodies (aPL), groups were based on positivity for lupus anticoagulant, anticardiolipin, and anti–β2-glycoprotein I (anti-β2GPI) antibodies: group 1 = positivity for 1 of the 3 aPL, group 2 = positivity for 2 of the 3 aPL, and group 3 = positivity for all 3 aPL.
Prevalence and pattern of myocardial scarring on LGE of CMRI in patients with APS.
There was a good interobserver reliability (κ = 0.61) and high intraobserver reliability (κ = 0.86) for the LGE detection of myocardial scarring on CMRI. Eight discrepant interpretations were resolved on a third data review.
Myocardial ischemic disease characterized by LGE on CMRI was present in 8 (29.6%) of 27 patients with APS and in 3 (3.7%) of 81 control subjects. Patients with APS had a relative risk of LGE-detected myocardial disease of 7.4 (95% CI 4.4–15.6; P = 0.0006), despite sharing similar clinical characteristics as those in controls at baseline, including the prevalence of cardiovascular risk factors (Table 1).
In patients with APS, LGE displayed a typical pattern of MI in 3 (11.1%) of 27 patients, by showing scarring of the subendocardium with partial (n = 2) or complete (n = 1) transmural spreading (Figure 2). By assigning myocardial segments (MS) to coronary artery territories, the findings revealed by LGE were ascribed to a monotroncular CAD pattern in all 3 patients with APS in whom a typical pattern of MI was observed. LGE of CMRI revealed scarring in the right coronary artery territory (MS 9) in 1 patient, in the left circumflex coronary artery territory (MS 6 and 12) in 1 patient, and in the left anterior descending coronary artery territory (MS 8) in 1 patient. These 3 patients were asymptomatic and had no evidence of MI in the medical records or on EKG or echocardiography. LGE patterns of linear subepicardial scarring (Figure 3A) and patchy, midwall scarring (Figure 3B) were seen in 4 patients and 1 patient, respectively. The findings on CMRI were normal in all 6 patients with APS associated with SLE. The patterns of LGE-detected myocardial scarring that were observed in the 3 control subjects were subepicardial, midwall, and subendocardial, respectively.
One patient with APS experienced sudden cardiac death within 2 months after undergoing CMRI. The patient was asymptomatic, had no evidence of a prior MI, and was at low risk of CAD (absolute cardiovascular risk of 8.9% within the next 10 years). CMRI with LGE in this patient had shown a subendocardial pattern with complete transmural spreading.
Biologic, EKG, and echocardiography findings in patients with APS.
Levels of serum creatinine (mean ± SD 69 ± 16.1 μmoles/liter), pro–B-type natriuretic protein (138.2 ± 324.8 pg/ml), and proteinuria/creatininuria (13.8 ± 16.4 mg/mmoles) were all within the normal range in patients with APS. All patients with APS had normal EKG findings, including the 8 patients with myocardial ischemic disease revealed by LGE on CMRI (designated the APS LGE+ group). Echocardiography showed thickening of the mitral or aortic valves in 4 patients with APS, moderate left ventricle hypertrophy in 5 patients, mild pericardial effusion in 1 patient, and elevation of the systolic pulmonary artery pressure between 40 mm Hg and 50 mm Hg in 3 patients. No segmental wall motion abnormalities could be found in the patients in the APS LGE+ group. Six (75%) of 8 patients in the APS LGE+ group and 13 (68.4%) of 19 in the APS LGE− group (those without LGE evidence of myocardial disease) had normal findings on echocardiography (P > 0.999).
Factors associated with LGE-detected myocardial ischemic disease in patients with APS.
The clinical characteristics of the patients with APS according to the presence or absence of LGE evidence of MI are shown in Table 2. We found no statistically significant differences in terms of age, sex ratio, race, and frequency of coronary risk factors, including history of hypertension, diabetes, hypercholesterolemia, and smoking, family history of CAD, body mass index, and ratio of homocysteine to creatinine between the APS LGE+ and APS LGE− groups. APS LGE+ and APS LGE− patients had a mean ± SD absolute cardiovascular risk within the next 10 years of 9.1 ± 12.9% and 3.4 ± 4.8%, respectively (P = 0.287).
APS LGE+ patients had APS for a longer duration compared with those in the APS LGE− group (mean ± SD duration 13.1 ± 8.5 years versus 9.2 ± 8.8 years), but the difference did not reach statistical significance (P = 0.148). Interestingly, the prevalence of livedo was 7-fold higher in APS LGE+ patients than in APS LGE− patients (P = 0.064). The site of thrombosis—whether arterial or venous—did not differ between the groups. The percentage of patients with anti-β2GPI antibodies in the serum (i.e, IgG levels higher than the 99th percentile on 2 or more occasions at least 12 weeks apart) was much higher in the APS LGE+ group (62.5%) than in the APS LGE− group (26.3%); however, the difference between the groups failed to reach statistical significance (P = 0.101). Moreover, we did not find any significant differences in the immunologic profile of APS between the groups. The presence of myocardial ischemic disease, revealed by LGE on CMRI, was not associated with SLE, nor was it associated with use of medications, including oral anticoagulants (P = 0.201), aspirin (P > 0.999), statins (P > 0.999), or oral contraceptives (P = 0.592).
Our study shows that the prevalence of unrecognized myocardial scarring detected by CMRI is unexpectedly high in patients with APS. Importantly, despite the low cardiovascular risk observed in the study patients and the lack of clinical suspicion of CAD in most patients, LGE displayed a typical pattern of MI in 11% of the patients with APS. In our study, the prevalence of LGE in APS patients is comparable with the prevalence in patients with well-known cardiovascular risk, such as patients with diabetes (21), patients with end-stage renal disease (22), or patients with clinical signs of CAD (8). Furthermore, our study shows that the presence of myocardial scarring as revealed by LGE in patients with APS is not associated with classic CAD risk factors.
In a previous European observational study, MI was disclosed in up to 5.5% of patients with APS, using conventional assessment, i.e., segmental wall motion abnormalities on echocardiography and EKG data (2). LGE imaging can detect myocardial scars that represent “footprints” of prior subclinical coronary events. Importantly, such scars may remain unnoticed when using current methods for evaluating myocardial ischemia, including nuclear scintigraphy (8, 10, 21), and detection of such scars is of prognostic value, beyond the parameters of left ventricular function, coronarography, segmental wall motion abnormalities, or patient risk factor score results (8, 21, 23).
The pattern revealed by LGE was carefully analyzed in patients with APS. In 3 patients (11%), LGE showed a pattern typical of myocardial scarring in monotroncular CAD. Unfortunately, coronary arteriography was not performed in these 3 patients. However, a similar CMRI pattern was also observed in 3 patients with APS whose symptom history included MI (Figure 1) or ACS (results not shown) and whose findings on coronarography were considered normal, suggesting a thrombotic process with clot lysis. Multiple areas of subepicardial or patchy scarring on LGE, which does not correspond to the pattern expected in coronary ischemic damage, were present in 5 other patients with APS. Focal LGE is a nonspecific measure of segmental fibrosis that was recently described in patients with other autoimmune diseases, such as SLE, systemic sclerosis, and systemic vasculitis (24–26). However, evidence has indicated that the non-CAD pattern with small focal myocyte necrosis spots as revealed by LGE may develop from ischemic myocardial damage subsequent to coronary microembolization (27, 28). Moreover, evidence of microthrombotic disease without clinically prominent expression in APS has been well described in the kidneys of patients with APS (29). Myocardial damage in APS may be caused by intravascular thrombosis of minor coronary artery side branches or distal embolization from transient thrombus of major coronary vessels, rather than “conventional” CAD. Such a hypothesis would fit with the lack of cardiac symptoms, significant Q waves, and segmental wall motion abnormalities in APS patients with LGE-detected myocardial ischemic disease.
We were not able to demonstrate any statistically significant association between specific features of APS and LGE evidence of myocardial scarring. However, the duration of APS, presence of livedo, and levels of anti-β2GPI antibodies in the serum tended to be related to the presence of myocardial ischemia on LGE. Interestingly, livedo in APS is related to arteriolar occlusion, and the association of LGE evidence of myocardial scarring with the presence of livedo suggests a similar process in the heart.
Because our study was monocentric, selection bias may have led to sampling of high-risk patients with APS, in whom there is a high prevalence of subclinical coronary disease. However, the patients with APS in our cohort exhibited few demographic or clinical confounders, with a low pretest likelihood of CAD. Moreover, our patients with APS shared most of the same clinical characteristics, including age, race, frequencies of venous versus arterial thrombosis, and immunologic profile (i.e., frequency of patients with aCL and LAC), as have been reported in patients with APS in the largest cohort studied so far (2). Selection bias may also have existed in the composition of the control group. However, we did not observe demographic differences between the patients with APS and controls. Moreover, the 3.7% prevalence of LGE-detected myocardial ischemic disease in control subjects is consistent with the known prevalence reported in apparently healthy subjects (30).
The small size of our cohort may have limited the power of the study to detect differences between the APS LGE+ and APS LGE− subsets, particularly regarding classic risk factors for CAD that may also raise the specific thrombotic risk associated with APS. Prospective studies with prolonged followup are needed in order to address both the prognostic value of myocardial scars on CMRI and the specific therapeutic needs of patients with APS. Thus, CMRI, particularly with LGE, is useful in identifying a large subset of patients with APS who experience subclinical coronary disease.
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 published. Dr. Papo 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. Sacré, Brihaye, Hyafil, Lidove, Papo.
Acquisition of data. Sacré, Brihaye, Serfaty, Escoubet, Zennaro, Lidove, Laissy.
Analysis and interpretation of data. Sacré, Serfaty, Lidove, Laissy, Papo.
We thank C. Astoul and M. C. Vanderhaegen for assistance in organizing the patient study, and Dr. Florence Tubach for help with manuscript preparation.