Electrocardiographic QT Intervals in Infants Exposed to Hydroxychloroquine Throughout Gestation

Introduction

As the world faces the pandemic of the novel severe acute respiratory syndrome coronavirus 2 and its resulting illness (coronavirus disease 2019 [COVID-19]), the reach for therapeutics assumes the highest of priorities. Two antimalarials, hydroxychloroquine and chloroquine, have surfaced as promising candidates. By raising endosomal pH, these weak bases have been shown in vitro to decrease severe acute respiratory syndrome coronavirus 2 viral replication.¹ It has been suggested that endosome maturation is blocked at intermediate stages of endocytosis, which would then result in decreased transport of virions to the ultimate releasing site.² Since the race against time has precluded mature placebo-controlled trials, favorable results from China support efficacy of chloroquine against COVID-19–associated pneumonia³ and a very limited French study using hydroxychloroquine reported a significant reduction in viral carriage.⁴ Although such positive results were not reproduced in a limited study comparing hydroxychloroquine to placebo,⁵ the United States has embraced massive testing of hydroxychloroquine for both prophylaxis of high-risk patients and treatment (eg, see https://www.clinicaltrials.gov; Unique identifier: NCT04308668). As the use of hydroxychloroquine continues to expand, 2 relevant points surface: that of toxicity and that of inclusion or exclusion of pregnant women. Concerns regarding cardiac toxicity have surfaced based on the Food and Drug Administration–approved package insert for hydroxychloroquine, which states that the drug may prolong the corrected QT interval (QTc). Accordingly, data are needed to address cardiac safety in the neonate exposed to hydroxychloroquine.

This is especially true since there is known transplacental passage of hydroxychloroquine,⁶ and the terminal elimination half-life of hydroxychloroquine is long.⁷ After absorption, the half-life of hydroxychloroquine is ≈40 days owing to the large volume of distribution in the blood. Moreover, hydroxychloroquine can distribute to aqueous cellular and intercellular compartments, resulting in long mean residence times.⁸

Physicians caring for patients with systemic lupus erythematosus are intimately familiar with hydroxychloroquine. It is virtually the most frequently prescribed long-term medication used to treat this disease, a practice driven by extensive literature supporting the prevention of flares and reduction of mortality.^8,9 Predictably, these clinically favorable effects have led to the strong recommendation by the American College of Rheumatology to maintain hydroxychloroquine during pregnancy in these patients.¹⁰ Furthermore, based on promising retrospective and case-control studies, maternal use of hydroxychloroquine may extend to the prevention of congenital heart block (CHB) associated with anti-Sjögren’s Syndrome A/Ro (anti-SSA/Ro) autoantibodies.^11,12 The overall safety of hydroxychloroquine during pregnancy is supported by the absence of congenital malformations in over 400 pregnancies.^13,14

To provide further data on the fetal cardiac effects of exposure to hydroxychloroquine, we leveraged a recently completed open-label study in which hydroxychloroquine was prospectively evaluated to reduce the recurrence rate of CHB.¹⁵ Here, reported are the results of available ECGs obtained within four months of postnatal life of fetuses exposed to hydroxychloroquine from 10 weeks of gestation until delivery. Supporting maternal compliance and unambiguous fetal exposure, levels of hydroxychloroquine during each trimester and delivery as well as in cord blood are presented.

Methods

Description of Parent Study

The authors declare that all supporting data are available within the article. In brief, the PATCH study (Preventive Approach to Congenital Heart Block With Hydroxychloroquine) is an open-label single-arm phase II trial to assess whether hydroxychloroquine is effective for the prevention of CHB recurrence.¹⁵ Treatment with hydroxychloroquine 400 mg was required by completion of 10 weeks of gestation, and the dose was maintained throughout pregnancy. If the mother was already on hydroxychloroquine, the dose remained at 400 mg, and if the mother was taking 200 mg, the dose had to be escalated to 400 mg by 10 weeks. The trial was funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development and by philanthropic foundations. The protocol was approved by the NYU School of Medicine Institutional Review Board, and all study participants gave written informed consent. The trial was overseen by a data and safety monitoring board of independent experts which convened every six months by teleconference. Patients were eligible if they had anti-SSA/Ro with or without anti-SSB/La antibodies irrespective of rheumatologic maternal diagnosis and a previous pregnancy with any of the following complications: fetal 2° or 3° CHB, serious cardiac injury defined as autopsy evidence of a mononuclear infiltrate in the endocardium, myocardium, and pericardium or the presence of severe endocardial fibroelastosis associated with cardiac dysfunction seen on fetal echocardiography.¹⁵

Maternal blood was obtained at baseline, second trimester, and birth to determine antibody levels by ELISA using native Ro60, recombinant Ro52, and recombinant La48 in the lab of Dr Buyon, as previously described.¹⁶ To assess maternal compliance, hydroxychloroquine levels were measured when possible at baseline, second trimester, third trimester, delivery, and in cord blood in Paris, France, as previously described.^17,18 Values reported at <100 ng/mL, which implies that the measurement of hydroxychloroquine is below the level of detection, were assigned a specific value of 50 ng/mL in the calculation of mean values.

The primary end point of the main study was recurrence of 2° or 3° block in utero or at birth. The cardiac secondary end points consisted of prolonged fetal Doppler mechanical PR interval (>150 ms) documented on the fetal echocardiogram (AV interval) or PR interval on ECG done at or after birth. Evaluation of QT/QTc intervals (see below) were not included in the study outcomes. Infants had an ECG or echocardiogram performed at birth or up to 1 year after birth.

Included in the current substudy were all neonates in which an ECG result and hydroxychloroquine levels were available. Several additional neonates who were screen failures based on maternal dosing past 10 weeks were also included as the appropriate information was still collected. Available ECGs (birth to four months of age) with sufficient quality to evaluate the QTc were read by the pediatric cardiologist (Dr Friedman). The ECGs of affected newborns who met the primary outcome of advanced block were not included in this safety study so that the results only reflect those infants with no clinical cardiac disease. Using the Bazett formula to correct for heart rate, QTc intervals were calculated and compared to age-matched normal values.¹⁹ For the median (2nd percentile–98th percentile), values for QTc were 413 (378–448) ms in males and 420 (379–462) ms in females. QTc intervals were recorded in the absence of knowledge of the hydroxychloroquine levels. Values exceeding 448 ms for males and 462 ms for females were considered abnormal. We are aware of the common usage of Schwartz to calculate the QTc but chose the recent simplified method as our historical control.²⁰

Statistical Analysis

The Pearson correlation coefficient was used to estimate the association between blood hydroxychloroquine levels and the neonatal QTc intervals. In addition, hydroxychloroquine levels were divided into quartiles and neonatal QTc levels were compared between the highest and lowest hydroxychloroquine quartiles with the 2 sample t test and Mann-Whitney test. Since results were nearly identical, only those based on the former are reported. For all analyses, 2 hydroxychloroquine levels were used: (1) hydroxychloroquine levels in cord blood or close to birth and (2) the average of the maternal hydroxychloroquine levels obtained at each trimester and delivery plus cord levels. A 2-sided P value of <0.05 was considered statistically significant. All data analyses were performed utilizing GraphPad Prism (GraphPad Software, Inc, La Jolla, CA).

Results

As detailed in the Table, there were 45 ECGs available for interpretation within the first 4 months of life in unaffected infants. Levels of hydroxychloroquine obtained throughout pregnancy, albeit not available in 5 (11%) of the cord blood, confirmed fetal exposure. Given that the terminal elimination half-life of hydroxychloroquine is very long,⁷ the unavailability of a cord blood hydroxychloroquine level did not preclude inclusion in this study. Overall, there was no correlation between cord blood levels of hydroxychloroquine and the QTc (R=0.02, P=0.86) or the average value of hydroxychloroquine levels obtained during each individual pregnancy and cord blood and the QTc (R=0.04, P=0.80), as shown in Figure 1A and Figure 1B. Likewise there was no correlation between the average of the maternal hydroxychloroquine levels obtained at each trimester and delivery plus cord levels and the QTc on the ECGs of the 31 infants evaluated on day of life 1 to 4 (R=0.08, P=0.63) or those of the 14 children older than 4 days (R=0.01, P=0.95).

• Download figure
• Download PowerPoint

Maternal values of hydroxychloroquine were sustained throughout pregnancy and delivery (Figure 2). Mean QTc values were nearly identical between those in the highest and lowest quartiles of cord blood hydroxychloroquine levels (P=0.57) and between the highest and lowest quartiles of average hydroxychloroquine levels during pregnancy (P=0.54; Figure 3A and 3B).

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

Among these 45 infants, only 5 had prolongation of the QTc (11% [95% CI, 4%–24%]): 2 marked and 3 marginal. One (no. 45) was an asymptomatic female born prematurely at 35 weeks 3 days gestation whose QTc on the first day of life was 484 ms. No follow-up ECG is presently available. Although cord blood hydroxychloroquine levels were not obtained in this case, the mother reported taking the drug throughout pregnancy and provided blood during the first and second trimesters which substantiated compliance with values >1000 ng/mL. The second case (no. 61) was a full-term asymptomatic male newborn whose QTc on the first day of life was 503 ms. A repeat ECG on day 2 of life revealed sustained prolongation at 492 ms. No further ECG data are available; however verbal contact with the mother confirmed no cardiac concerns. Cord blood hydroxychloroquine was 585 ng/mL with an average level of hydroxychloroquine at 879 ng/mL throughout pregnancy. Of relevance, with regard to these 2 infants (no. 45 and no. 61), their ECG tracings were more challenging to interpret due to baseline artifact, with prominent u waves, such that the QTc may have been incorrectly read as prolonged. In fact, review by the local pediatric cardiologist was normal. The third case (no. 89) was a full-term asymptomatic female whose QTc was 471 ms on the first day of life. Cord blood hydroxychloroquine was 811 ng/mL with an average level of hydroxychloroquine at 793 ng/mL throughout pregnancy. Although the mother reports no health issues with the child, no further ECGs are available. The 2 other cases (no. 82, no. 84), both males, had minimal QTc prolongation of 451 ms on day 1 of life and 450 ms on day 2 of life, respectively. Cord blood levels were unknown in no. 82 but maternal levels were close to 1000 ng/mL during the first, second, and third trimesters. The cord blood level was 533 ng/mL in case no. 84 with an average hydroxychloroquine level of 697 ng/mL throughout pregnancy. No arrhythmias occurred in any neonate that was not known to have heart block.

Discussion

The safety of hydroxychloroquine use in pregnancy is a critical issue especially in consideration of its now very extended and immediate application to prevent and treat high-risk and positively infected individuals with severe acute respiratory syndrome coronavirus 2. Gathering information to balance against potential antiviral efficacy is needed. Albeit limited in numbers, the data presented in this substudy revealed that, of fetuses with no evidence of CHB who were exposed to hydroxychloroquine, 89% of their neonatal ECGs exhibited normal QTcs.

Moreover, assurances of compliance with the medication strengthen the value of the data since many previous reports in patients with systemic lupus erythematosus indicate nonadherence to hydroxychloroquine in up to 31%.²¹ Relevant to antiviral efficacy, the blood levels observed throughout pregnancy and in the cord blood were largely equivalent to or even exceeded the mean serum level of hydroxychloroquine (serum 460 ng/mL which approximates whole blood 920 ng/mL per personal communication with Dr Costedoat-Chalumeau: mean ratio of serum/whole blood levels of hydroxychloroquine is 0.53±0.15) reported in the one open-label COVID-19 study which suggested efficacy.⁴

Previous studies have considered the cardiac safety in offspring exposed to maternal hydroxychloroquine. Costedoat-Chalumeau et al²² reported on the ECGs of 92 newborns at the third day of life, 47 of whom were exposed to maternal hydroxychloroquine (doses 200 mg to 400 mg) and 45 control newborns whose mothers did not receive hydroxychloroquine. There were no statistical differences in the mean±SD QTc intervals between the 2 groups (406±34 versus 407±32 ms, hydroxychloroquine-exposed and hydroxychloroquine-unexposed respectively, P=0.40), with all newborns having normal ECGs. However, this study did not systematically verify maternal compliance by measuring hydroxychloroquine maternal or cord blood levels. Further evidence for safety is addressed in an autopsy performed on one fetus in the PATCH study in which complete heart block was detected at 19 weeks despite hydroxychloroquine. Reassuringly, there were no histological findings consistent with hydroxychloroquine toxicity, such as myelin figures.²³

One potential limitation in the study reported here is that a control group of anti-SSA/Ro pregnancies unexposed to hydroxychloroquine was not available. Therefore, an effect of maternal autoantibodies on the neonatal QTc cannot be entirely excluded. The contribution of anti-SSA/Ro is considered in part by Costedoat-Chalumeau et al²⁴ in a study which included an evaluation of neonatal ECGs in 58 anti-SSA/Ro–exposed children compared with 85 anti-SSA/Ro–negative mothers with a known connective tissue disease. There were no differences in mean QTc±SD between these groups (mean QTc values of 397±27 ms anti-SSA/Ro positive and 395±25 ms anti-SSA/Ro negative, P=0.57). Although hydroxychloroquine levels were not obtained, in 60% of pregnancies in both groups, the mother took hydroxychloroquine. Data were not specifically provided on anti-SSA/Ro mothers who did not take hydroxychloroquine. A second study also affirms that our results are not confounded by maternal anti-SSA/Ro antibodies and are likely to be applicable to the general population of otherwise healthy pregnant women who may be treated with hydroxychloroquine.²⁵ Specifically, in a prospective study by Gerosa et al,²⁵ ECGs were done in infants 20 to 90 days of age. The mean QTc interval±SD was 414±17 in 46 anti-SSA/Ro exposed infants, 420±23 in 25 non–anti-SSA/Ro exposed infants born to mothers with an autoimmune connective tissue disease, and 416±16 in 200 infants of otherwise healthy mothers. A QTc interval >470 ms was observed in none of the anti-SSA/Ro–exposed infants. Hydroxychloroquine was used in about a fifth of these mothers.

Two other considerations are also relevant for clinicians to put these data in perspective. Study subjects were at increased risk of potential prolongation of QTc shortly after birth because they were exposed to high levels of hydroxychloroquine throughout pregnancy in contrast to the short course of drug being used in COVID-19 trials. Historical controls were used, and specifically referred to the entire newborn period (birth to one month of life) and not specifically to days 1, 2, or 3 of life in which the QT interval is more variable and often somewhat prolonged. These considerations, combined with some difficulties related to baseline artifact and u waves, may have resulted in an over-reading of QTc intervals.

Although the present study was limited to high titer anti-SSA/Ro positive mothers who were not ill with a severe potentially multisystem virus and not simultaneously taking other QT-prolonging drugs, the data should be applicable to otherwise healthy pregnant women being considered for preventive anti–COVID-19 strategies, especially since hydroxychloroquine will not likely be given in sustained doses. In aggregate, these data provide reassurances that the maternal use of hydroxychloroquine is associated with a low incidence of infant QTc prolongation. However, since 11% of the neonates (with levels unequivocally assuring exposure to hydroxychloroquine) did manifest QTc prolongation shortly after birth, although asymptomatic, consideration should be given to an early postnatal ECG with appropriate follow-up ECG monitoring as determined by the initial findings.

Footnotes