Abstract
Objectives:
The increased relative risk of heart failure (HF) from systemic lupus erythematosus (SLE) is greatest at younger ages, but the etiology remains unclear. We identified risk factors for HF in children and adults with SLE and evaluated associations between SLE manifestations and HF.
Methods:
Incident SLE cases without preceding HF were identified using Clinformatics DataMart® (OptumInsight, Eden Prairie, MN) US claims data (2000–2015), and categorized by age of SLE onset (children 5–17, young adults 18–24, adults 25–44 years old). The primary outcome was the first HF ICD-9-CM diagnosis code (428.x), categorized as early-onset (< 6 months) or delayed-onset. Multivariable logistic regression was used to identify factors associated with early or delayed-onset HF. Cox proportional hazards regression was used to identify time-dependent associations between the onset of SLE manifestations and incident HF.
Results:
There were 523 (2.3%) HF cases among 1,466 children, 2,163 young adults and 19,349 adults age 25–44 with SLE. HF in children and young adults was early-onset in 50% and 60% of cases, respectively, compared to 35% of cases in adults 25–44 years old. There was a temporal association between incident myopericarditis and valvular disease diagnoses and early-onset HF, whereas nephritis and hypertension were more strongly associated with delayed-onset HF. Black race remained independently associated with a 1.5-fold increased HF risk at any time.
Conclusion:
Hypertension remains an important traditional CV risk factor across all ages and should be managed aggressively even in younger patients with SLE. Cardiac dysfunction due to acute cardiac manifestations of SLE may contribute to the very high relative incidence of early HF diagnoses among younger SLE patients. Therefore, future prospective studies will need to address heterogeneity in the types and severity of heart failure in order to determine etiology and which patients should be monitored.
Keywords: systemic lupus erythematosus, heart failure, cardiac disease, pediatric lupus
1.0. Introduction
Systemic lupus erythematosus (SLE) is a chronic autoimmune condition associated with cardiac manifestations of the disease as well as cardiovascular (CV) comorbidities of chronic inflammation or its treatment. While there has been long-standing recognition of the increased risk of CV events from premature atherosclerosis, only in more recent years has there been literature emerging on the increased risk of heart failure (HF) in SLE.[1] HF is a major public health problem that affects 6.2 million Americans over the age of 20 and is associated with significant morbidity, mortality, and health care expenditures.[2] A recent epidemiologic study demonstrated that SLE is associated with a 5-fold increased incidence of HF compared to the general population and an adjusted odds ratio of 3.2 not explained by CV risk factors. Although the absolute risk of HF increased with age, the highest relative risk of HF was observed in the youngest age group, with 65-fold and 50-fold increases among males and females age 20–24, respectively.[3] However, there are no epidemiologic data regarding HF in pediatric-onset SLE. Furthermore, there are no data on the timing and risk factors for HF in children and young adults with SLE who do not yet have cardiac disease, and therefore it is unclear when or who to monitor for this potential complication.
The potential etiologies of HF in this population include accelerated atherosclerosis, hypertensive chronic kidney disease, subclinical myocardial dysfunction due to chronic inflammation, and acute cardiac manifestations of SLE such as myocarditis, pericarditis or valvular disease.[4,5] Although direct myocardial injury from lupus carditis is thought to be too rare to explain HF rates in adults with SLE, pediatric-onset SLE may be associated with higher rates of acute myopericardial involvement at SLE presentation,[6] as well as fewer traditional CV risk factors. Furthermore, as most acute cardiac manifestations occur within the first 6 months of SLE diagnosis,[6,7] HF due to direct cardiac involvement from SLE may be expected to present early in the disease course, whereas HF due to atherosclerosis or other chronic comorbidities would be expected to present later. Therefore, we sought to understand the relative contribution of acute cardiac and non-cardiac manifestations of SLE to the risk and timing of HF in children and adults with SLE. The objectives were to: 1) estimate the association between acute cardiac manifestations of SLE (myopericardial or valvular involvement) and incident HF diagnoses in children and adults with SLE, 2) identify other risk factors for early or delayed-onset HF, and 3) determine whether the association between cardiac or non-cardiac organ involvement and HF differs by age of SLE onset.
2.0. Methods
2.1. Study Design.
This was a retrospective cohort study. An exemption of this study was approved by the Institutional Review Board at The Children’s Hospital of Philadelphia (IRB #17–013682).
2.2. Data Source and Subjects.
We identified children (age 5–17), young adults age 18–24, and adults age 25–44 with an incident diagnosis of SLE using the Clinformatics® DataMart (OptumInsight, Eden Prairie, MN) de-identified U.S. administrative data from 2000 through 2015. OptumInsight data includes commercial health insurance and Medicare Advantage (C and D) claims for a nationally representative sample of approximately 20% of US residents.
The diagnosis of SLE was determined by the presence of at least 3 hospital or physician claims with an ICD-9-CM diagnosis code for SLE (710.0), each > 30 days apart.[8,9] To be considered an incident case, subjects were also required to have at least 12 months of continuous insurance eligibility in the database with no SLE codes in any position preceding the first SLE diagnosis code, as previously described.[10] Subjects with a diagnosis of HF preceding SLE diagnosis were excluded. Age was determined at the time of SLE diagnosis. For pediatric-onset SLE, the lower age limit was set to exclude monogenic or neonatal lupus. Given that both the highest relative risk of HF and the highest standardized mortality rates have been described in transition-age young adults under age 25,[3,11] adult-onset SLE was further categorized into young adults age 18–24, and adults age 25–44, inclusive. The upper age limit was selected to focus on the risk of cardiovascular disease at a young age.
2.3. Study Measures.
The primary outcome was the first occurrence of a primary or secondary ICD-9-CM diagnosis code for heart failure (428.x), which has been previously validated for adults with a positive predictive value of 83% and negative predictive value of 87%.[12,13] As most acute cardiac manifestations of SLE (myocarditis, pericarditis, endocarditis, valvular insufficiency) occur within the first 6 months of SLE diagnosis,[6,7] HF diagnoses occurring within the first 6 months of the first SLE code were considered early-onset, and HF diagnoses occurring after that date were considered delayed-onset. Outcomes were assessed until the end of enrollment or October 1, 2015, whichever occurred earlier.
2.4. Covariates.
Baseline characteristics were assessed from 12 months before until 6 months after the first SLE code for delayed-onset HF, or up until the first HF code for early-onset HF cases. We included demographics (age, sex, race/ethnicity, US census region, highest household education); major SLE organ manifestations (nephritis,[9] seizure, cerebrovascular disease, psychiatric disorders,[14] myocarditis/pericarditis, endocarditis, and valvular insufficiency) as previously defined in this database;[6,10] antiphospholipid antibody syndrome (APLS) or venous thromboembolism (VTE);[6,15] traditional cardiovascular risk factors (hypertension,[16] diabetes,[17] hyperlipidemia (272.0–272.4), smoking status,[18] obesity (278.00–278.03), and coronary artery disease);[19] and medication use (prescription claims for hydroxychloroquine, glucocorticoids, and other conventional immunosuppressants, including azathioprine, methotrexate, mycophenolate, leflunomide, tacrolimus, sirolimus, and cyclophosphamide; J-codes for biologic therapies, including belimumab and rituximab).
As an additional proxy for disease activity, we characterized escalation of immunosuppression at the time of HF diagnosis, defined as the presence of at least one J-code for IV glucocorticoid administration, new biologic therapy or new conventional immunosuppressant within 30 days of the first HF diagnosis code.
2.5. Statistical Analysis
Baseline demographic and clinical characteristics were summarized using standard descriptive statistics. HF characteristics were compared across age categories using Fisher’s exact or chi-square tests as appropriate. Multivariable logistic regression with backward selection (p-value ≥ 0.2 for removal) was used to identify baseline factors associated with early-onset heart failure within 6 months of SLE diagnosis. A separate logistic regression model was used to identify baseline factors associated with delayed-onset heart failure > 6 months after SLE diagnosis. Age was retained in all models, and additional covariates considered confounders were forced into the model if coefficients of interest (SLE manifestations, including nephritis, myopericardial involvement, valvular disease, APLS/VTE) changed by more than 15%. Pre-specified interactions between initial SLE manifestations by age of SLE onset were assessed using likelihood ratio tests.
To account for heterogeneity in the onset of SLE manifestations over time, a secondary analysis using multivariable Cox proportional hazards regression was performed to determine longitudinal associations between new SLE manifestations and time to HF. Covariates identified to be time-dependent (nephritis, APLS/VTE, myopericardial involvement, valvular disease) were updated at 90-day intervals with a 90-day time lag to avoid potential reverse causality. Censoring occurred at the time of the first HF diagnosis or the end of the eligibility period for enrollment, whichever came first. There was no mortality data available, so all subjects were censored as alive. Schoenfeld residuals were used to evaluate the proportional hazards assumption. Age and sex were included in all models, and additional covariates were retained if p-values were < 0.2 or if coefficients of interest changed by more than 15%, as detailed above. Likelihood ratio tests were used to evaluate pre-specified interactions by age at SLE onset.
We also performed a sensitivity analysis using an expanded primary outcome definition with the addition of diagnosis codes for secondary cardiomyopathy, excluding alcoholic (425.0, 425.8, 425.9). Although these have not been demonstrated to improve the positive predictive value of heart failure definitions in the general adult population,[13] they have not previously been assessed in patients with underlying conditions such as SLE.
3.0. Results
We identified 22,978 incident SLE cases, of which 1,466 were diagnosed in childhood and 2,163 were diagnosed as young adults prior to age 25. Baseline demographic and disease characteristics are shown in Table 1. We observed 523 (2.3%) incident cases of heart failure during an average follow-up time of 3.2 years after the first SLE diagnosis (SD 2.6). The estimated annual incidence rate of HF was 0.8% per person-year. Pediatric-onset SLE cases had the lowest HF incidence estimate of 0.3% per person-year (95% CI [0.2% – 0.5%]), whereas the incidence among young adults age 18–24 (0.9% per person-year, 95% CI [0.7% – 0.12%]) was comparable to adults age 25–44 (Table 2). Median time to heart failure was 217 days (interquartile range (IQR) 4 – 561) in children, 97 days (IQR 4 – 405) in young adults, and 373 days (IQR 70 – 903) in adults age 25–44.
Table 1.
Baseline patient characteristics by age of SLE onset
| Age 5–17 | Age 18–24 | Age 25–44 | |
|---|---|---|---|
| N = 1,466 | N = 2,163 | N = 19,349 | |
| Demographics | |||
| Age of SLE onset, mean (SD) | 13.5 (3.3) | 21.1 (2.0) | 36.5 (5.4) |
| Observation time (years) | 3.4 (2.6) | 2.6 (1.9) | 3.3 (2.7) |
| Male sex, n (%) | 350 (24%) | 368 (17%) | 2782 (14%) |
| Race/ethnicity | |||
| White | 899 (61%) | 1386 (64%) | 11623 (60%) |
| Black | 175 (12%) | 236 (11%) | 2331 (12%) |
| Hispanic | 207 (14%) | 291 (13%) | 2815 (15%) |
| Asian | 78 (5%) | 77 (4%) | 850 (4%) |
| Unknown | 107 (7%) | 173 (8%) | 1730 (9%) |
| Region | |||
| Midwest | 316 (22%) | 432 (20%) | 3485 (18%) |
| Northeast | 218 (15%) | 312 (14%) | 2767 (14%) |
| South | 697 (47%) | 1068 (49%) | 9974 (52%) |
| Pacific | 229 (16%) | 349 (16%) | 3082 (16%) |
| Highest household education | |||
| High school or less | 348 (24%) | 524 (24%) | 5223 (27%) |
| More than high school | 1054 (72%) | 1538 (71%) | 13102 (68%) |
| Unknown | 64 (4%) | 101 (5%) | 1024 (5%) |
| Initial SLE Manifestations | |||
| Myocarditis/Pericarditis | 43 (3%) | 61 (3%) | 326 (2%) |
| Endocarditis/Valvular disease | 65 (4%) | 154 (7%) | 1863 (10%) |
| Nephritis | 182 (12%) | 164 (8%) | 1041 (5%) |
| Cerebrovascular event | 46 (3%) | 93 (4%) | 1200 (6%) |
| Seizure | 93 (6%) | 128 (6%) | 928 (5%) |
| Psychiatric disorder | 475 (32%) | 820 (38%) | 7971 (41%) |
| APLS/VTEa | 48 (3%) | 119 (6%) | 1430 (7%) |
| Traditional Cardiovascular Risk Factors | |||
| Hypertension | 112 (8%) | 165 (8%) | 4006 (21%) |
| Diabetes | 36 (2%) | 51 (2%) | 1359 (7%) |
| Obesity | 88 (6%) | 138 (6%) | 2633 (14%) |
| Hyperlipidemia | 116 (8%) | 339 (16%) | 6274 (32%) |
| Coronary artery disease | 2 (0%) | 15 (1%) | 245 (1%) |
| Current/prior smoker | 6 (0%) | 153 (7%) | 2315 (12%) |
| Initial SLE Medication Use | |||
| Hydroxychloroquine | 347 (24%) | 691 (32%) | 5675 (29%) |
| Glucocorticoid | 660 (45%) | 1087 (50%) | 9876 (51%) |
| Other immunosuppressant | 211 (14%) | 271 (13%) | 2392 (12%) |
All characteristics ascertained within baseline period from 12 months before SLE diagnosis up to 6 months after initial SLE diagnosis.
Antiphospholipid antibody syndrome / venous thromboembolism
Table 2.
Incidence rate of heart failure by age of SLE onset
| Age of SLE onset | Person-years | Failures | IR | 95% CI |
|---|---|---|---|---|
| 5–17 | 1657914 | 14 | 0.003 | [0.002, 0.005] |
| 18–24 | 1847425 | 45 | 0.009 | [0.007, 0.012] |
| 25–44 | 21102439 | 464 | 0.008 | [0.007, 0.009] |
| Total | 24607777 | 523 | 0.008 | [0.007, 0.008] |
Incidence rate per person-year of follow-up after SLE diagnosis
3.1. Early-onset Heart Failure
In children and young adults, the proportion of HF diagnoses presenting within the first 6 months of SLE diagnosis was 50% (7/14) and 60% (27/45), respectively. Myocarditis/pericarditis preceded 21% of cases in children and 27% of cases in young adults. Similarly, endocarditis or valvular insufficiency preceded 36% and 31% of cases in children and young adults, respectively. 43% of early-onset HF cases in children and 31% in young adults with early-onset HF diagnoses were treated with IV glucocorticoids or initiated a new immunosuppressant within 30 days of the first HF diagnosis code. In contrast, only 35% of HF diagnoses in adults age 25–44 presented early within 6 months of SLE diagnosis. Compared to children and young adults, HF in adults age 25–44 was less frequently associated with a preceding diagnosis of myocarditis/pericarditis (10%, p-value <0.01) and also less likely to be treated with increased immunosuppression (20%, p-value 0.05).
After adjustment for other baseline characteristics, young adults 18–24 years of age had the highest risk of early-onset HF (OR 1.7, 95% CI [1.1 – 2.8] compared to adults age 25–44) (Table 3). Myopericardial and valvular disease were identified as the strongest disease-related risk factors for early-onset HF (OR 4.0, 95% CI [2.3 – 6.9] and OR 2.2, 95% CI [1.5 – 3.2], respectively). Black race and other markers of increased SLE disease severity, including nephritis, cerebrovascular events, APLS and glucocorticoid use, were also associated with a higher odds of early-onset HF (Table 3). In contrast, hydroxychloroquine use was associated with a 0.3-fold decreased odds of early-onset HF (95% CI [0.2 – 0.5]). With respect to traditional CV risk factors, hypertension was most strongly associated with an increased odds of early-onset HF (OR 2.3, 95% CI [1.6 – 3.2]).
Table 3.
Baseline characteristics associated with early-onset heart failure
| Unadjusted | Adjusted | |||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p-value | OR | 95% CI | p-value | |
| Demographics | ||||||
| Age of SLE onset | ||||||
| 25–44 years old | - | - | - | - | - | - |
| 18–24 years old | 1.5 | [1.0, 2.2] | 0.06 | 1.7 | [1.1, 2.8] | 0.02 |
| 5–17 years old | 0.6 | [0.3, 1.2] | 0.14 | 0.5 | [0.2, 1.3] | 0.14 |
| Male sex | 1.3 | [0.9, 1.8] | 0.18 | - | - | - |
| Race/ethnicity | ||||||
| White | - | - | - | - | - | - |
| Black | 2.0 | [1.4, 2.9] | <0.01 | 1.5 | [1.0, 2.2] | 0.03 |
| Hispanic | 0.8 | [0.5, 1.2] | 0.28 | 0.7 | [0.4, 1.2] | 0.24 |
| Asian | 0.8 | [0.3, 1.7] | 0.51 | 0.9 | [0.4, 2.1] | 0.82 |
| Region | ||||||
| Midwest | - | - | - | - | - | - |
| Northeast | 0.5 | [0.3, 0.9] | 0.02 | 0.4 | [0.2, 0.8] | 0.01 |
| South | 1.0 | [0.7, 1.4] | 0.87 | 0.9 | [0.6, 1.3] | 0.44 |
| Pacific | 0.7 | [0.4, 1.2] | 0.18 | 0.7 | [0.4, 1.2] | 0.23 |
| Higher household education† | 0.7 | [0.5, 1.0] | 0.03 | |||
| Initial SLE Characteristics | ||||||
| Myopericarditis | 7.6 | [4.9, 11.8] | <0.01 | 4.0 | [2.3, 6.9] | <0.01 |
| Endocarditis/valve insufficiency | 3.3 | [2.4, 4.6] | <0.01 | 2.2 | [1.5, 3.2] | <0.01 |
| APLS/VTE* | 3.1 | [2.1, 4.4] | <0.01 | 1.9 | [1.3, 3.0] | <0.01 |
| Nephritis | 3.8 | [2.6, 5.5] | <0.01 | 2.0 | [1.3, 3.0] | <0.01 |
| Cerebrovascular event | 1.7 | [1.1, 2.8] | 0.02 | |||
| Seizure | 1.9 | [1.2, 3.1] | 0.01 | |||
| Pyschiatric diagnosis | 0.8 | [0.6, 1.1] | 0.26 | 0.7 | [0.5, 1.0] | 0.04 |
| Traditional Cardiovascular Risk factors | ||||||
| Hypertension | 3.7 | [2.8, 4.9] | <0.01 | 2.3 | [1.6, 3.2] | <0.01 |
| Diabetes | 2.3 | [1.5, 3.5] | <0.01 | 1.5 | [1.0, 2.4] | 0.08 |
| Obesity | 1.4 | [0.9, 2.0] | 0.11 | |||
| Hyperlipidemia | 1.4 | [1.1, 1.9] | 0.02 | |||
| Current/prior tobacco use | 1.5 | [1.0, 2.2] | 0.05 | |||
| Coronary artery disease | 5.0 | [2.6, 9.5] | <0.01 | |||
| Initial Medication Use | ||||||
| Hydroxychloroquine | 0.3 | [0.2, 0.5] | <0.01 | 0.3 | [0.2, 0.5] | <0.01 |
| Glucocorticoids | 1.0 | [0.8, 1.4] | 0.76 | |||
| Other immunosuppressants | 0.9 | [0.6, 1.5] | 0.79 | |||
Univariable and multivariable logistic regression models estimating associations between baseline characteristics in SLE patients age < 45 and early-onset heart failure (n=198) occurring ≤ 6 months after SLE diagnosis. Covariates were assessed until the first HF diagnosis or 6 months after SLE diagnosis, whichever occurred first.
Some bachelor level education or more
Antiphospholipid antibody syndrome / venous thromboembolism
3.2. Delayed-onset Heart Failure
Hypertension and baseline coronary artery disease were the strongest risk factors for delayed-onset HF (OR 3.1, 95% CI [2.4 – 4.1] and OR 2.5, 95% CI [1.5 – 4.1], respectively). Of note, there were no cases of baseline coronary artery disease in children or young adults with HF. Nephritis remained a significant disease-related risk factor for delayed-onset HF. Myopericardial involvement or valvular disease at baseline were also independently associated with delayed-onset HF, although the magnitudes of the associations were lower compared to that of early-onset HF (Table 4). As observed with early-onset HF, black race remained associated with a 1.5-fold increased odds of delayed-onset HF (95% CI [1.1 – 2.0]). In contrast to early-onset HF, younger age of SLE onset was associated with a significantly lower risk of delayed-onset HF (Table 4). The effects of initial organ involvement on HF risk did not differ significantly by age of SLE onset in either the early or delayed HF models.
Table 4.
Baseline characteristics associated with delayed-onset heart failure
| Unadjusted | Adjusted | |||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p-value | OR | 95% CI | p-value | |
| Demographics | ||||||
| Age of SLE onset | ||||||
| 25–44 years old | - | - | - | - | - | - |
| 18–24 years old | 0.5 | [0.3, 0.9] | 0.01 | 0.7 | [0.4, 1.1] | 0.15 |
| 5–17 years old | 0.3 | [0.1, 0.6] | <0.01 | 0.3 | [0.1, 0.7] | 0.01 |
| Male sex | 1.3 | [1.0, 1.7] | 0.07 | |||
| Race/ethnicity | ||||||
| White | - | - | - | - | - | - |
| Black | 2.1 | [1.6, 2.8] | <0.01 | 1.5 | [1.1, 2.0] | 0.01 |
| Hispanic | 1.0 | [0.7, 1.4] | 0.95 | 0.9 | [0.6, 1.3] | 0.47 |
| Asian | 0.9 | [0.5, 1.6] | 0.61 | 0.9 | [0.5, 1.7] | 0.70 |
| Region | ||||||
| Midwest | - | - | - | - | - | - |
| Northeast | 1.4 | [0.9, 2.1] | 0.14 | 1.5 | [0.9, 2.3] | 0.10 |
| South | 1.7 | [1.2, 2.4] | <0.01 | 1.5 | [1.0, 2.1] | 0.03 |
| Pacific | 1.1 | [0.7, 1.7] | 0.76 | 1.3 | [0.8, 2.1] | 0.23 |
| Higher household education† | 0.7 | [0.5, 0.9] | <0.01 | 0.8 | [0.6, 1.0] | 0.11 |
| Initial SLE Characteristics | ||||||
| Myopericarditis | 3.9 | [2.5, 6.3] | <0.01 | 2.0 | [1.2, 3.3] | 0.01 |
| Endocarditis/valve insufficiency | 2.6 | [1.9, 3.4] | <0.01 | 1.3 | [1.0, 1.8] | 0.07 |
| aPLS/VTE* | 2.2 | [1.6, 3.0] | <0.01 | 1.4 | [1.0, 1.9] | 0.07 |
| Nephritis | 4.3 | [3.3, 5.7] | <0.01 | 2.3 | [1.7, 3.1] | <0.01 |
| Cerebrovascular event | 2.7 | [2.0, 3.7] | <0.01 | 1.5 | [1.0, 2.1] | 0.03 |
| Seizure | 1.8 | [1.2, 2.6] | 0.01 | |||
| Pyschiatric diagnosis | 1.3 | [1.0, 1.6] | 0.02 | |||
| Traditional Cardiovascular Risk factors | ||||||
| Hypertension | 5.1 | [4.1, 6.4] | <0.01 | 3.1 | [2.4, 4.1] | <0.01 |
| Diabetes | 3.1 | [2.3, 4.2] | <0.01 | 1.4 | [1.0, 2.0] | 0.03 |
| Obesity | 2.0 | [1.5, 2.6] | <0.01 | |||
| Hyperlipidemia | 1.9 | [1.5, 2.3] | <0.01 | |||
| Current/prior tobacco use | 1.6 | [1.2, 2.2] | <0.01 | |||
| Coronary artery disease | 7.8 | [5, 12.1] | <0.01 | 2.5 | [1.5, 4.1] | <0.01 |
| Initial Medication Use | ||||||
| Hydroxychloroquine | 1.0 | [0.8, 1.3] | 0.82 | |||
| Glucocorticoids | 1.9 | [1.5, 2.3] | <0.01 | 1.5 | [1.2, 2.0] | <0.01 |
| Other immunosuppressants | 2.2 | [1.7, 2.8] | <0.01 | |||
Univariable and multivariable logistic regression models estimating associations between baseline characteristics in SLE patients age < 45 and delayed-onset heart failure (n=325) occurring > 6 months after SLE diagnosis. Baseline characteristics were assessed from 12 months before SLE diagnosis up to the index date (6 months after SLE diagnosis). Subjects with early-onset HF at SLE diagnosis were excluded from this analysis.
Some bachelor level education or more
Antiphospholipid antibody syndrome / venous thromboembolism
3.3. SLE Manifestations and Time to Heart Failure
When accounting for the development of new SLE manifestations over time, myopericardial and renal involvement in a preceding time interval were the disease-related factors most strongly associated with a greater risk of HF (HR 2.6, 95% CI [1.9 – 3.6] and HR 2.4 [1.8 – 3.1], respectively) (Table 5). There were no significant interactions by age of SLE onset. Consistent with findings from the logistic regression analysis, valvular disease (HR 1.8, 95% CI [1.4–2.2]) and APLS (HR 1.6, 95% CI [1.2 – 2.1]) were also significant time-dependent risk factors for HF (Table 5). There was a trend towards an association between hydroxychloroquine initiation at baseline and a decreased risk of heart failure (HR 0.8, p-value 0.13), albeit not statistically significant.
Table 5.
Adjusted hazard ratios (HR) for heart failure by time-updated SLE manifestations
| HR | 95% CI | p-value | |
|---|---|---|---|
| Age of SLE onset | |||
| 25–44 years old | - | - | - |
| 18–24 years old | 0.9 | [0.6, 1.5] | 0.79 |
| 5–17 years old | 0.3 | [0.2, 0.7] | 0.01 |
| Race/ethnicity | |||
| Black | 1.5 | [1.1, 1.9] | 0.00 |
| Hispanic | 1.0 | [0.7, 1.4] | 0.88 |
| Asian | 0.9 | [0.5, 1.7] | 0.78 |
| Higher education | 0.7 | [0.6, 0.9] | <0.01 |
| Myopericardial disease* | 2.6 | [1.9, 3.6] | <0.01 |
| Valvular disease* | 1.8 | [1.4, 2.2] | <0.01 |
| Nephritis* | 2.4 | [1.8, 3.1] | <0.01 |
| aPLS/VTE* | 1.6 | [1.2, 2.1] | <0.01 |
| Cerebrovascular disease | 1.5 | [1.1, 2.0] | 0.02 |
| Diabetes | 1.6 | [1.2, 2.2] | <0.01 |
| Hypertension | 2.6 | [2.1, 3.4] | <0.01 |
| Current/prior tobacco use | 1.2 | [0.9, 1.6] | 0.23 |
| Coronary artery disease | 2.1 | [1.4, 3.2] | <0.01 |
| Baseline glucocorticoid use | 1.6 | [1.2, 2.0] | <0.01 |
| Other immunosuppressant use | 1.2 | [0.9, 1.6] | 0.16 |
| Baseline hydroxychloroquine use | 0.8 | [0.6, 1.1] | 0.13 |
Multivariable Cox proportional hazards regression with time-varying covariates estimating associations between new SLE manifestations in a preceding time interval and incident heart failure in a subsequent interval.
Time-varying covariate with 90-day time lag
3.4. Sensitivity Analysis
Inclusion of secondary cardiomyopathy codes in the HF outcome definition identified a total of 19 (1.3%) HF cases in children with SLE, 48 (2.2%) in young adults, and 520 (2.7%) in adults 25–44 years of age, with a similar distribution of early versus delayed-onset cases. Inclusion of secondary cardiomyopathy codes in the primary outcome did not significantly alter the results of the logistic or Cox regression models.
4.0. Discussion
In this population-based study of risk factors for HF among children and adults with SLE, there are several important findings. First, we found that 50% or more of HF diagnoses in children and young adults presented near the time of initial SLE diagnosis and were associated with acute cardiac manifestations of SLE, which may partially explain previous observations that younger SLE patients have a disproportionately high relative risk of HF compared to the general population. Secondly, among both pediatric and adult-onset cases, we identified several lupus-related risk factors for the development of HF later in the disease course, including baseline cardiac or renal involvement and antiphospholipid antibody syndrome, as well as modifiable CV risk factors such as hypertension. Third, despite accounting for disease-related and CV risk factors, there remain significant racial disparities in HF risk. Lastly, early antimalarial use may be associated with a decreased rather than increased risk of HF. Our findings have important implications for screening and can inform potential etiologies of heart failure in this high-risk population.
Although the types and etiologies of HF in SLE are multifactorial, lupus disease activity may play a more important role in early HF incidence, especially among the youngest SLE patients without significant atherosclerotic burden. In a multinational inception cohort of 1249 adults with SLE, there were 24 cases of HF, of which only 5 were attributed to atherosclerosis, 12 to active SLE, and 7 to other causes.[20] None of the children or young adults under age 25 in our study had known coronary artery disease, and HF diagnoses in these younger age categories more often resulted in increased immunosuppression compared to those in adults age 25–44. Active SLE could contribute directly to HF in a number of ways, including myocarditis, constrictive pericarditis, chronic inflammation, valvular disease, microvascular dysfunction, and coronary thrombosis or vasculitis.[4,5,21–23] Although lupus myocarditis and pericarditis are thought to be rare causes of overt HF in adults with SLE,[21,24,25] younger age of SLE onset is associated with greater disease severity and higher rates of major organ involvement,[26,27] including myopericardial disease.[6] We observed a temporal association between incident myopericaridial disease and subsequent HF diagnoses, but the types and severity of cardiac dysfunction could not be determined from diagnostic codes. It is also unclear whether the observed association between myopericardial involvement and HF is mediated by direct myocardial injury, or if baseline cardiac involvement is simply an indicator of greater overall disease severity. Direct myocardial injury could potentially be evaluated in the future using newer imaging techniques such as cardiac MRI.[28] Associations between disease activity and HF have also been described in rheumatoid arthritis, which are potentially modifiable by antirheumatic treatments.[29] As a result, early monitoring and aggressive disease control among patients with greater SLE severity may be particularly important for decreasing HF rates in children and young adults with SLE.
Valvular disease was also common in both age groups and independently associated with HF risk. Our findings are consistent with a previous estimate of the association between valvular disease and a 3-fold increased odds of HF incidence in adults with SLE.[3] An echocardiographic study performed in adults with SLE also demonstrated that although valvular abnormalities can come and go without temporal association with other disease manifestations, they are still associated with a higher rate of HF.[30] However, valvular abnormalities in pediatric-onset SLE are frequently mild without clinically significant effects on ventricular function, so it remains unclear whether there is a direct causal relationship.[31] Given the strong association between antiphospholipid antibodies and valvular disease in SLE, further studies are needed to determine whether it is the pathogenicity of the antiphospholipid antibodies or valvular insufficiency itself that serve as the primary contributor to HF in this population.
In addition to disease manifestations, black race was an important independent risk factor for HF in our cohort. This was not explained by the disparities between publicly or privately insured populations, as all subjects in the cohort had continuous private insurance coverage. SLE and HF are diseases that both disproportionately affect black populations in the US.[2,8] In our study, black race was consistently associated with a 1.5-fold increased risk of HF even after adjustment for other proxies of disease severity and demographic characteristics. Black children and adults with SLE have greater disease severity, worse renal outcomes, and higher mortality compared to their white counterparts, which may be related to a combination of delayed presentations to care, socioeconomic disparities, and genetic factors.[32–34] Previous work in pediatric-onset SLE has also demonstrated that black race is associated with higher rates of acute cardiac manifestations at initial SLE presentation.[6,35] In addition, the disproportionately higher rates of comorbidities such as obesity and hypertension among black US residents is well-recognized and may further compound HF risk.[36,37] Therefore, a low threshold for cardiac evaluation and close attention to comorbidity screening should be consistently advocated for in this population.
While traditional CV risk factors do not fully explain increased rates of HF in this population, they still warrant attention as being potentially modifiable. In our study, hypertension was a major independent risk factor, especially for delayed-onset HF. This suggests that, in contrast to early-onset HF, delayed-onset HF is more likely to be from hypertensive heart disease. In a previous study of older adults with SLE, hypertension was identified as the strongest predictor of HF.[3] The prevalence of hypertension is significantly higher in patients with SLE compared to the general population,[38] which may be secondary to renal disease or chronic glucocorticoid treatment. Therefore, aggressive blood pressure screening and management are warranted. The American College of Rheumatology has published guidelines on blood pressure targets in adults with lupus nephritis, but additional studies are needed to determine ideal blood pressure targets in children and adolescents with SLE.[39] Of note, none of the subjects under age 25 with HF in this cohort had a preceding diagnosis of coronary artery disease, and in contrast to previous studies in older adults, obesity and hyperlipidemia were not significant risk factors for HF. Therefore, relative contributions of other disease-specific factors in younger SLE patients may be more relevant early in the disease course. Longer follow-up will be needed to assess cumulative effects of conventional CV risk factors and atherosclerotic disease on HF in children and young adults with SLE.
Lastly, early initiation of hydroxychloroquine was not associated with an increased risk of HF in this cohort and may even be protective, though this relationship should be interpreted cautiously in the context of this study. Although there are reports of restrictive cardiomyopathy associated with chronic hydroxychloroquine use,[40–42] this is a rare complication that requires prolonged cumulative exposure and would not explain HF incidence in this inception cohort. In general, hydroxychloroquine use in SLE is thought to be protective against atherosclerotic cardiovascular disease in association with lipid-lowering effects, improvement of disease control, and prevention of thrombosis.[43–45] However, the relationship between hydroxychloroquine and HF has not yet been examined. While our study was not specifically designed to evaluate the effect of medications on HF outcomes, our findings suggest that the potential role of hydroxychloroquine deserves further investigation.
To our knowledge, this is the first population-based study of risk factors for HF that specifically addresses the timing of and relationship between disease manifestations and the significantly increased relative risk of HF among children and young adults with SLE. In addition, we accounted for heterogeneity in the onset of SLE manifestations over time. There are also several important limitations to our approach. First, although the primary outcome definition is validated for the general adult population, it has not been validated in children or in patients with SLE. Therefore, it is possible that patients with some degree of cardiac impairment from active myopericardial disease were being overcoded as HF, or others with overt HF were undercoded as SLE with myopericarditis.The de-identified nature of the dataset precluded separate adjudication, but our estimates were reassuringly similar to previously published incidence rates using Systemized Nomenclature of Medicine—Clinical Terms codes (incidence 0.36%/year in males and 0.27% in females ages 20–24), and sensitivity analyses with different outcome definitions did not significantly alter our results. More importantly however, ICD-9 diagnostic codes do not distinguish between different types or severity of HF, and therefore we were unable to determine which risk factors were most relevant for each type of HF. Secondly, there are no direct measures of disease activity, and therefore we used several proxies such as new major organ manifestations and immunosuppressant prescriptions. Third, we were unable to account for mortality as a potential competing risk due to the lack of mortality data. Lastly, patients in this database were all commercially insured, which limits the generalizability of our results to uninsured or publicly insured patients who are at greater socioeconomic disadvantage.
In summary, the etiology of HF in SLE is likely multifactorial, and both cardiac and non-cardiac lupus disease manifestations may play an important role in the significantly increased relative risk of HF that has been observed among younger SLE patients. Additional studies are needed to better understand heterogeneity in the types and severity of cardiac dysfunction in SLE in order to determine the etiologies of HF and which patients should be monitored. In particular, those with high disease activity, known cardiac or renal involvement, and secondary antiphospholipid antibody syndrome may warrant closer cardiology follow-up. Ideal blood pressure targets also need to be defined in pediatric-onset SLE. In addition, there remain significant racial disparities in SLE-related and cardiovascular outcomes that have yet to be effectively addressed. Knowledge of the interplay between disease activity, socioeconomic factors, and common CV comorbidities on HF risk in both pediatric and adult-onset SLE will be essential to the development of adequate preventative measures.
Funding:
This work was supported by the National Heart Lung and Blood Institute at the National Institutes of Health [F32HL142176 to J.C.].
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflicts of Interest Statement: The authors have no financial disclosures.
References
- 1.Chen SK, Barbhaiya M, Fischer MA, Guan H, Yoshida K, Feldman CH, et al. Heart failure risk in systemic lupus erythematosus compared to diabetes mellitus and general medicaid patients. Semin Arthritis Rheum. 2019. December;49(3):389–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Benjamin EJ, Muntner P, Alonso A, Bittencourt MS, Callaway CW, Carson AP, et al. Heart Disease and Stroke Statistics—2019 Update: A Report From the American Heart Association. Circulation. 2019. March 5;139(10):e56–528. [DOI] [PubMed] [Google Scholar]
- 3.Kim CH, Al-Kindi SG, Jandali B, Askari AD, Zacharias M, Oliveira GH. Incidence and risk of heart failure in systemic lupus erythematosus. Heart. 2017. February;103(3):227–33. [DOI] [PubMed] [Google Scholar]
- 4.Dhakal BP, Kim CH, Al-Kindi SG, Oliveira GH. Heart failure in systemic lupus erythematosus. Trends Cardiovasc Med. 2018. April;28(3):187–97. [DOI] [PubMed] [Google Scholar]
- 5.Torres A, Askari AD, Malemud CJ. Cardiovascular disease complications in systemic lupus erythematosus. Biomark Med. 2009. June;3(3):239–52. [DOI] [PubMed] [Google Scholar]
- 6.Chang JC, Xiao R, Mercer-Rosa L, Knight AM, Weiss PF. Child-onset systemic lupus erythematosus is associated with a higher incidence of myopericardial manifestations compared to adult-onset disease. Lupus. 2018. November;27(13):2146–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Oshiro AC, Derbes SJ, Stopa AR, Gedalia A. Anti-Ro/SS-A and anti-La/SS-B antibodies associated with cardiac involvement in childhood systemic lupus erythematosus. Ann Rheum Dis. 1997. April;56(4):272–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Feldman CH, Hiraki LT, Liu J, Fischer MA, Solomon DH, Alarcón GS, et al. Epidemiology and sociodemographics of systemic lupus erythematosus and lupus nephritis among US adults with Medicaid coverage, 2000–2004. Arthritis Rheum. 2013. March;65(3):753–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hiraki LT, Feldman CH, Liu J, Alarcón GS, Fischer MA, Winkelmayer WC, et al. Prevalence, incidence, and demographics of systemic lupus erythematosus and lupus nephritis from 2000 to 2004 among children in the US Medicaid beneficiary population. Arthritis Rheum. 2012. August;64(8):2669–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Chang JC, Mandell DS, Knight AM. High Health Care Utilization Preceding Diagnosis of Systemic Lupus Erythematosus in Youth. Arthritis Care Res. 2017. Accepted; [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bernatsky S, Boivin J-F, Joseph L, Manzi S, Ginzler E, Gladman DD, et al. Mortality in systemic lupus erythematosus. Arthritis Rheum. 2006. August;54(8):2550–7. [DOI] [PubMed] [Google Scholar]
- 12.Goff DC, Pandey DK, Chan FA, Ortiz C, Nichaman MZ. Congestive heart failure in the United States: is there more than meets the I(CD code)? The Corpus Christi Heart Project. Arch Intern Med. 2000. January 24;160(2):197–202. [DOI] [PubMed] [Google Scholar]
- 13.McCormick N, Lacaille D, Bhole V, Avina-Zubieta JA. Validity of Heart Failure Diagnoses in Administrative Databases: A Systematic Review and Meta-Analysis. Guo Y, editor. PLoS ONE. 2014. August 15;9(8):e104519. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Knight AM, Xie M, Mandell DS. Disparities in Psychiatric Diagnosis and Treatment for Youth with Systemic Lupus Erythematosus: Analysis of a National US Medicaid Sample. J Rheumatol. 2016. July 1;43(7):1427–33. [DOI] [PubMed] [Google Scholar]
- 15.Yusuf HR, Hooper WC, Grosse SD, Parker CS, Boulet SL, Ortel TL. Risk of venous thromboembolism occurrence among adults with selected autoimmune diseases: A study among a U.S. cohort of commercial insurance enrollees. Thromb Res. 2015. January;135(1):50–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Quan H, Khan N, Hemmelgarn BR, Tu K, Chen G, Campbell N, et al. Validation of a Case Definition to Define Hypertension Using Administrative Data. Hypertension. 2009. December 1;54(6):1423–8. [DOI] [PubMed] [Google Scholar]
- 17.Chen G, Khan N, Walker R, Quan H. Validating ICD coding algorithms for diabetes mellitus from administrative data. Diabetes Res Clin Pract. 2010. August;89(2):189–95. [DOI] [PubMed] [Google Scholar]
- 18.Desai RJ, Eddings W, Liao KP, Solomon DH, Kim SC. Disease-Modifying Antirheumatic Drug Use and the Risk of Incident Hyperlipidemia in Patients With Early Rheumatoid Arthritis: A Retrospective Cohort Study: Risk of Hyperlipidemia in Early RA Patients Taking DMARDs. Arthritis Care Res. 2015. April;67(4):457–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Robitaille C, Bancej C, Dai S, Tu K, Rasali D, Blais C, et al. Surveillance of ischemic heart disease should include physician billing claims: population-based evidence from administrative health data across seven Canadian provinces. BMC Cardiovasc Disord. 2013. December;13(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Urowitz MB, Gladman D, Ibañez D, Bae SC, Sanchez-Guerrero J, Gordon C, et al. Atherosclerotic vascular events in a multinational inception cohort of systemic lupus erythematosus. Arthritis Care Res. 2010. June;62(6):881–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Wijetunga M, Rockson S. Myocarditis in systemic lupus erythematosus. Am J Med. 2002. October;113(5):419–23. [DOI] [PubMed] [Google Scholar]
- 22.Cesari M Inflammatory Markers and Onset of Cardiovascular Events: Results From the Health ABC Study. Circulation. 2003. November 3;108(19):2317–22. [DOI] [PubMed] [Google Scholar]
- 23.Faccini A, Kaski JC, Camici PG. Coronary microvascular dysfunction in chronic inflammatory rheumatoid diseases. Eur Heart J. 2016. June 14;37(23):1799–806. [DOI] [PubMed] [Google Scholar]
- 24.Comarmond C, Cacoub P. Myocarditis in auto-immune or auto-inflammatory diseases. Autoimmun Rev. 2017. August;16(8):811–6. [DOI] [PubMed] [Google Scholar]
- 25.Rosenbaum E, Krebs E, Cohen M, Tiliakos A, Derk C. The spectrum of clinical manifestations, outcome and treatment of pericardial tamponade in patients with systemic lupus erythematosus: a retrospective study and literature review. Lupus. 2009. June;18(7):608–12. [DOI] [PubMed] [Google Scholar]
- 26.Brunner HI, Gladman DD, Ibañez D, Urowitz MD, Silverman ED. Difference in disease features between childhood-onset and adult-onset systemic lupus erythematosus. Arthritis Rheum. 2008. February;58(2):556–62. [DOI] [PubMed] [Google Scholar]
- 27.Brunner HI, Silverman ED, To T, Bombardier C, Feldman BM. Risk factors for damage in childhood-onset systemic lupus erythematosus: cumulative disease activity and medication use predict disease damage. Arthritis Rheum. 2002. February;46(2):436–44. [DOI] [PubMed] [Google Scholar]
- 28.Guo Q, Wu L-M, Wang Z, Shen J-Y, Su X, Wang C-Q, et al. Early Detection of Silent Myocardial Impairment in Drug-Naive Patients With New-Onset Systemic Lupus Erythematosus: A Three-Center Prospective Study. Arthritis Rheumatol. 2018. December;70(12):2014–24. [DOI] [PubMed] [Google Scholar]
- 29.Myasoedova E, Crowson CS, Nicola PJ, Maradit-Kremers H, Davis JM, Roger VL, et al. The Influence of Rheumatoid Arthritis Disease Characteristics on Heart Failure. J Rheumatol. 2011. August;38(8):1601–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Roldan CA, Shively BK, Crawford MH. An Echocardiographic Study of Valvular Heart Disease Associated with Systemic Lupus Erythematosus. N Engl J Med. 1996. November 7;335(19):1424–30. [DOI] [PubMed] [Google Scholar]
- 31.Chang JC, White BR, Elias MD, Xiao R, Knight AM, Weiss PF, et al. Impact of Pediatric-Onset Systemic Lupus Erythematosus on Echocardiographic Measures of Diastolic Function at Diagnosis. In Chicago; 2018. [Google Scholar]
- 32.Anderson E, Nietert PJ, Kamen DL, Gilkeson GS. Ethnic disparities among patients with systemic lupus erythematosus in South Carolina. J Rheumatol. 2008. May;35(5):819–25. [PMC free article] [PubMed] [Google Scholar]
- 33.Nee R, Martinez-Osorio J, Yuan CM, Little DJ, Watson MA, Agodoa L, et al. Survival Disparity of African American Versus Non–African American Patients With ESRD Due to SLE. Am J Kidney Dis. 2015. October;66(4):630–7. [DOI] [PubMed] [Google Scholar]
- 34.Alarcón GS, McGwin G, Sanchez ML, Bastian HM, Fessler BJ, Friedman AW, et al. Systemic lupus erythematosus in three ethnic groups. XIV. Poverty, wealth, and their influence on disease activity. Arthritis Rheum. 2004. February 15;51(1):73–7. [DOI] [PubMed] [Google Scholar]
- 35.Dalby ST, Tang X, Daily JA, Sukumaran S, Collins RT, Bolin EH. Effect of pericardial effusion on outcomes in children admitted with systemic lupus erythematosus: a multicenter retrospective cohort study from the United States. Lupus. 2019. March;28(3):389–95. [DOI] [PubMed] [Google Scholar]
- 36.Hertz RP, Unger AN, Cornell JA, Saunders E. Racial Disparities in Hypertension Prevalence, Awareness, and Management. Arch Intern Med. 2005. October 10;165(18):2098. [DOI] [PubMed] [Google Scholar]
- 37.Taveras EM, Gillman MW, Kleinman KP, Rich-Edwards JW, Rifas-Shiman SL. Reducing Racial/Ethnic Disparities in Childhood Obesity: The Role of Early Life Risk Factors. JAMA Pediatr. 2013. August 1;167(8):731. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Sabio JM, Vargas-Hitos JA, Navarrete-Navarrete N, Mediavilla JD, Jiménez-Jáimez J, Díaz-Chamorro A, et al. Prevalence of and Factors Associated with Hypertension in Young and Old Women with Systemic Lupus Erythematosus. J Rheumatol. 2011. June;38(6):1026–32. [DOI] [PubMed] [Google Scholar]
- 39.Hahn BH, McMahon MA, Wilkinson A, Wallace WD, Daikh DI, FitzGerald JD, et al. American College of Rheumatology guidelines for screening, treatment, and management of lupus nephritis. Arthritis Care Res. 2012. June;64(6):797–808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Manohar VA, Edwards WD, Kylew Klarich, KG Moder. Restrictive Cardiomyopathy Secondary to Hydroxychloroquine Therapy. J Rheumatol. 2009. February;36(2):440–1. [DOI] [PubMed] [Google Scholar]
- 41.Muthukrishnan P, Roukoz H, Grafton G, Jessurun J, Colvin-Adams M. Hydroxychloroquine-Induced Cardiomyopathy: A Case Report. Circ Heart Fail. 2011. March;4(2). [DOI] [PubMed] [Google Scholar]
- 42.Tselios K, Gladman DD, Harvey P, Mak S, Chantal M, Butany J, et al. Hydroxychloroquine-Induced Cardiomyopathy in Systemic Lupus Erythematosus. J Clin Rheumatol Pract Rep Rheum Musculoskelet Dis. 2016. August;22(5):287–8. [DOI] [PubMed] [Google Scholar]
- 43.Fasano S, Pierro L, Pantano I, Iudici M, Valentini G. Longterm Hydroxychloroquine Therapy and Low-dose Aspirin May Have an Additive Effectiveness in the Primary Prevention of Cardiovascular Events in Patients with Systemic Lupus Erythematosus. J Rheumatol. 2017. July;44(7):1032–8. [DOI] [PubMed] [Google Scholar]
- 44.Cairoli E, Rebella M, Danese N, Garra V, Borba E. Hydroxychloroquine reduces low-density lipoprotein cholesterol levels in systemic lupus erythematosus: a longitudinal evaluation of the lipid-lowering effect. Lupus. 2012. October;21(11):1178–82. [DOI] [PubMed] [Google Scholar]
- 45.Petri M. Thrombosis and systemic lupus erythematosus: the Hopkins Lupus Cohort perspective. Scand J Rheumatol. 1996;25(4):191–3. [DOI] [PubMed] [Google Scholar]