Abstract
Objective. Antiphospholipid antibodies (aPL) play an active role in the pathogenesis of the antiphospholipid syndrome (APS). Primary prevention in APS may be aimed at decreasing existing elevated aPL levels, or preventing high aPL titers and/or lupus anticoagulant (LAC) from developing in the first place. Hydroxychloroquine (HCQ) has been shown in retrospective studies to decrease aPL titers in laboratory studies, and to decrease thrombosis risk in patients with systemic lupus erythematosus (SLE). We investigated an association between HCQ use and persistent aPL and/or LAC in SLE.
Methods. We identified all patients over 21 years old with SLE from an urban tertiary care center who had aPL and LAC measured on at least 2 occasions at least 12 weeks apart. We defined the presence of persistent LAC+ and/or at least 1 aPL ≥ 40 U [immunoglobulin A (IgA), IgG, or IgM] as the main outcome variable.
Results. Among 90 patients included in the study, 17 (19%) had persistent LAC+ and/or at least 1 aPL ≥ 40 U. HCQ use was associated with significantly lower odds of having persistent LAC+ and/or aPL ≥ 40 U (OR 0.21, 95% CI 0.05, 0.79, p = 0.02), adjusted for age, ethnicity, and sex.
Conclusion. This is the first study to show that HCQ use is associated with lower odds of having persistently positive LAC and/or aPL. Data from this study provide a basis for the design of future prospective studies investigating the role of HCQ in primary and secondary prevention of APS.
According to the “2-hit hypothesis,” the presence of antiphospholipid antibodies (aPL) is necessary to create a prothrombotic state (first hit). However, aPL alone are not sufficient, and may persist for a long time before the second hit results in the actual thrombotic event1,2. Therefore, primary thrombosis prevention may be aimed at decreasing existing elevated aPL, or preventing high aPL titers and/or lupus anticoagulant (LAC) from developing3.
Hydroxychloroquine (HCQ) has been shown to decrease aPL titers in laboratory studies4,5. However, only 1 published study to date evaluated the association between HCQ and aPL in a secondary analysis, with a negative result6. We investigated whether patients with SLE treated with HCQ were less likely to develop or to maintain persistently positive aPL and/or LAC.
MATERIALS AND METHODS
We included all adult patients with SLE by American College of Rheumatology (ACR) criteria7 who had LAC, anticardiolipin (aCL), anti-ß2 glycoprotein I (anti-ß2-GPI), and antiphosphatidylserine antibodies measured at least twice, at least 12 weeks apart, between January 2006 and May 2012 at Montefiore Medical Center (MMC), a large urban tertiary care center in Bronx, New York.
Patients were considered to be taking a medication (immunosuppressives, aspirin, HCQ, prednisone, or anticoagulant) if they ever took this medication, similar to previous retrospective studies6,8. Race and ethnicity were analyzed as African American/non-African American and Hispanic/non-Hispanic, respectively, based on self-report. Over 90% of non-Hispanics were African American, reflecting the overall racial/ethnic distribution in our center.
Enzyme immunoassay kits (Bio-Rad Laboratories, Hercules, CA, USA) were used to test aPL. Moderate to high aPL positivity (aPL+) was defined as at least 1 aPL [immunoglobulin G (IgG), IgM, or IgA] ≥ 40 units (moderate/high)9. LAC was reported as positive or negative (LAC+/LAC−) by the MMC laboratory in accord with the guidelines of the International Society on Thrombosis and Haemostasis10.
Because of the retrospective study design, we did not obtain informed consent from the patients, as no identifying information was stored or used in the data analysis. This project was approved by the Institutional Review Board at Albert Einstein College of Medicine/MMC.
Statistical analysis was performed using Stata 12.0 (StataCorp, College Station, TX, USA). No adjustments were made for multiple comparisons in this exploratory study.
RESULTS
The frequencies of aPL and/or LAC among 90 patients included in our study are shown in Table 1. The number of patients who converted from aPL and/or LAC-positive to negative, or from negative to positive, was small.
The frequencies of aPL/LAC positivity in the entire cohort (n = 90). All data are n (%).
The results of the bivariate comparisons between patients with persistently positive LAC and/or any aPL ≥ 40 U (n = 17), and patients with either transiently positive or persistently negative LAC and aPL (n = 73) are summarized in Table 2. The 2 groups were similar in age, sex, comorbidity scores, HCQ duration, and disease duration. HCQ use was lower in the persistent aPL/LAC-positive group than in the comparison group, 11 (65%) and 65 (89%), respectively (p = 0.02). The median duration of HCQ use was 49 months (IQR 31, 61) in the aPL/LAC-positive group, and 36 months (IQR 19, 56) in the comparison group (p = 0.6). The minimum duration on HCQ was at least 1 month.
Baseline characteristics of the study patients.
The results of the logistic regression adjusted for age, ethnicity, and sex are shown in Table 3. HCQ use was associated with significantly lower odds of having persistent LAC+ and/or aPL antibodies ≥ 40 U (IgA, IgG, or IgM; OR 0.21, 95% CI 0.05, 0.79, p = 0.02). We did not observe an association between use of other immunosuppressives or prednisone and persistent LAC positivity and/or aPL antibodies ≥ 40 U. Adding these variables to the above model did not change the association between HCQ and LAC/aPL. Age and sex were not independently associated with persistent LAC/aPL.
Logistic regression for persistent LAC+ and/or aPL ≥ 40 (IgA, IgG, or IgM) and for persistent positive LAC and/or Sapporo criteria aPL (aCL or ß2-GPI IgG or IgM) ≥ 40 U.
Similarly, persistently positive LAC and/or moderate/high Sapporo criteria aPL (aCL IgG or IgM, or anti-ß2-GPI IgG or IgM)9 were associated with HCQ use (OR 0.24, 95% CI 0.06, 0.94, p = 0.04) and Hispanic ethnicity (OR 4.2, 95% CI 1.0, 16.7, p = 0.04; Table 3).
We performed additional analyses to explore a possibility of “by indication” bias with respect to HCQ use in our study, i.e., if HCQ is prescribed for milder SLE, the differences observed in our study may be confounded by SLE severity and multiple comorbidities. When we compared patients taking HCQ (HCQ+) with patients not taking HCQ (HCQ−), no differences were observed with respect to demographics, disease duration, Charlson Comorbidity Score11, or medications. When we limited our analysis to a subgroup of patients with SLE who were not taking immunosuppressives, presuming less severe SLE (n = 49), HCQ was associated with an OR of 0.17 (95% CI 0.03, 0.88, p = 0.03) of LAC+ and/or Sapporo criteria aPL ≥ 40 U, suggesting that HCQ was independently associated with LAC/aPL positivity.
DISCUSSION
HCQ treatment is currently recommended for patients with SLE who have persistent moderate-high aPL or LAC positivity, for primary prevention (grade 1B to 2B recommendation) based on the other beneficial effects of HCQ in SLE12,13 and thrombosis14. However, ours is the first study to show that HCQ use may be associated with lower odds of having persistently positive LAC and/or aPL in SLE, and therefore may be beneficial in primary prevention.
We also showed that Hispanic ethnicity was associated with higher odds of persistent Sapporo criteria aPL and/or LAC compared with non-Hispanics (predominantly African Americans). While it was previously reported that LAC was more prevalent in whites15, the association between Hispanic ethnicity and the higher odds of positive aPL and/or LAC has not been previously reported. It is likely that aPL are affected by multiple demographic and SLE-specific factors. Therefore, further studies are needed to investigate the cumulative effect of protective factors and risk factors on developing and maintaining LAC+ and/or elevated aPL.
Our exploratory study has several limitations related to its retrospective design and the sample size, including differential selection, differential and nondifferential misclassification, disease activity, and medication dosage, duration, and compliance. However, we included only patients who satisfied ACR criteria for SLE, and performed several sensitivity analyses, with consistent results. Most importantly, the causal relationship between HCQ use and aPL/LAC positivity could not be established in this cross-sectional study. However, while our study was not designed or powered to evaluate whether HCQ decreased aPL levels, the information we found may be used to design prospective studies to evaluate this important question.
Acknowledgment
We thank Miriam Gordon, PhD, for valuable assistance in editing the manu script.
Footnotes
-
Supported in part by an Empire Clinical Research Investigator Program Career Development Award to Dr. Broder. Work in Dr. Putterman’s laboratory is supported by National Institutes of Health grants RO1 AR048692 and DK090319.
- Accepted for publication June 6, 2012.
