At the Heart of the Matter - Speckle Tracking Echocardiography in Lupus Mothers and Their Offspring

Women with SLE have a high risk of placenta-mediated adverse pregnancy outcomes (APO), including gestational hypertension, preeclampsia, eclampsia, small for gestational age (SGA) neonate, placental abruption, and stillbirth. It is hypothesized that high levels of circulating anti-angiogenic factors, placental microparticles, and other soluble mediators cause endothelial dysfunction and lead to cardiovascular diseases (CVD) in women with APO. In women with SLE, CVD is the leading cause of death, accounting for almost half of all deaths.Until now, studies have been unable to identify the cause of increased CVD in SLE women, despite accounting for SLE disease activity, damage, and drugs. Given the huge burden of CVD in SLE, it is unfathomable that no one has, to date, comprehensively examined the relationship between APO and CVD risk in SLE women.

Importantly, the offspring of mothers with placenta-mediated complications also have an increased CVD risk and in SLE offspring, in-utero exposure to maternal autoantibodies might further contribute to increased CVD risk.

Hydroxychloroquine (HCQ), the cornerstone drug in SLE treatment, may prevent SLE flare during pregnancy and reduces the risk of APO. However, HCQ may pose cardiotoxicity. HCQ dosing is done by weight but surprisingly, drug level monitoring is almost non-existent in Canada, although it is increasingly championed by key opinion leaders in the US and Europe. Lack of drug level monitoring is particularly concerning during pregnancy, given potential risks to mother and baby for adverse outcomes related to under or potentially overdosing. The incidence and prevalence of HCQ cardiotoxicity are not known and it is likely to be an under-diagnosed complication as patients are not systematically screened. Furthermore, data are urgently needed to address cardiac safety in neonates exposed to HCQ.

To assess cardiac function (i.e., myocardial contraction and relaxation), the investigators will use global longitudinal strain (GLS) measured by speckle tracking echocardiography (STE). GLS has superior prognostic value to conventional measures such as the left ventricle ejection fraction (LVEF) for predicting major adverse cardiac events and provides incremental prognostic information beyond the traditional CVD risk stratification tools. Moreover, GLS assessed by STE has most consistently detected early myocardial changes during therapy and appears to be the best measure of early drug cardiotoxicity, predicting future reduction in LVEF or heart failure.

The investigators will test the hypothesis that SLE women with APO have higher frequency of impaired myocardial strain versus SLE women without APO. The investigators will also evaluate the hypothesis that exposure to maternal autoantibodies during pregnancy (ie anti-Ro/La and aPL) is associated with impaired myocardial strain in neonates (versus unexposed neonates). Another hypothesis is that maternal and cord HCQ levels correlate with impaired myocardial strain in mothers and neonates respectively, potentially following a J-curve relationship (i.e. lower HCQ levels failing to be cardio-protective against maternal autoantibodies and higher HCQ levels leading to cardiotoxicity).

Primary aims: Using STE, the investigators will: 1) Assess the impact of placenta-mediated complications on maternal and neonatal myocardial strain, 2) Determine the effect of maternal autoantibodies on neonatal myocardial strain, 3) Evaluate if maternal and cord HCQ levels are associated with maternal and neonatal myocardial strain.

Primary deliverables: The investigators will develop: 1) Innovative risk stratification approaches for CVD in SLE women, 2) New tools for identifying subclinical cardiac involvement in offspring ex-posed in utero to APO, maternal autoantibodies, and/or HCQ, 3) HCQ cardiotoxicity screening strategies in SLE women and offspring, and 4) Novel therapeutic drug monitoring approaches for efficacy and safety of HCQ in pregnancy. Furthermore, the knowledge gained by this study will allow an investigator to further assess the effects of APO in other high-risk pregnancies.

Methods:

1. Patients with SLE At the time of the Lupus in prEGnAnCY (LEGACY) Biobank baseline visit, the investigator will recruit 30 pregnant women with SLE, without a prior history of placenta-mediated APO, and without preexist-ing CVD, hypertension, and non-gestational diabetes, which will form the "SLE group". LEGACY includes all intrauterine singleton pregnancies, before the 16th gestational week in SLE women, 18-45 years, regardless of SLE disease manifestations, level of disease activity, and medication exposures. At baseline, the investigator collects detailed data on maternal demographics, obstetrical history, co-morbidities, lifestyles, SLE activity and damage measures, as well as medications, including detailed information on HCQ use. LEGACY also measures data on relevant clinical parameters (e.g. blood pressure, random urine protein/creatinine ratio, etc) and maternal autoantibodies (e.g. aPL, anti-Ro/La). In addition, samples (serum, plasma, DNA) are biobanked at the Research Institute of McGill University Health Centre. Patients are reassessed by the investigator in the 2nd (between 20-24 weeks) and 3rd (between 30-34 weeks) trimesters, as well as in the postpartum (8-12 weeks after delivery). At these visits, clinical information is updated and occurrence of placenta-mediated APO is ascertained according to accepted international definitions.
2. Patients without SLE The investigators will recruit 3 groups of non-SLE women with a singleton pregnancy, without preexisting CVD, hypertension, non-gestational diabetes, nor prior history of placenta-mediated APO and matched 1:1 on age and ethnicity to SLE women. The first group, the "baseline control group", will consist of 30 pregnant women at less than 16 gestational weeks recruited during their first trimester visit at the Royal Victoria Hospital (RVH) obstetric clinic. The investigators will also identify 30 women with placenta-mediated APO as well as 30 women without any placenta-mediated APO upon admission for surveillance, delivery, and/or during the immediate postpartum at the RVH birth centre. These women will respectively constitute the "control group with APO" and the "control group without APO". For all non-SLE subjects, the investigators will obtain data on maternal demographics, obstetrical history, co-morbidities, lifestyles, drugs, and clinical parameters, similarly as for LEGACY subjects. An obstetrical medicine specialist has easy access to potentially eligible non-SLE subjects and will facilitate recruitment. Despite the pandemic, the investigators do not anticipate issues in recruiting SLE and non-SLE pregnant patients as prenatal care and hospitalizations for delivery are essential medical services maintained even at the worst of the pandemic. Pregnant SLE women are highly motivated for participation in research projects (e.g. LEGACY participation rate is 97%). The investigators will accommodate subjects by scheduling assessments with routine pregnancy visits. Strict safety measures will be implemented to minimize the risk of nosocomial COVID-19.
3. HCQ levels in mothers and offspring As previously mentioned, a key team member has already developed and is in the process of validating a method for whole-blood HCQ quantitative analysis (de-tails at www.agilent.com). Once validated, this method will go through the regulatory process for its full clinical implantation and approval for therapeutic drug monitoring. In short, this novel methodology uses whole-blood samples prepared by protein precipitation extraction followed with Agilent Captiva EMR-Lipid clean-up, then analyzed by triple quadrupole liquid chromatography mass spectrometry. This method provided a reliable solution with excellent quantitation accuracy (100±5%) for HCQ levels. Maternal whole-blood samples will be collected at the time of the echocardiography, while cord blood samples will be obtained at the time of delivery.
4. Echocardiographic assessment Myocardial imaging, including GLS assessment, will be performed at less than 20 gestational weeks in the SLE and baseline control group. In the SLE group, cardiac imaging will be repeated both in mothers and offspring in the immediate postpartum, up to 72 hours after delivery. Similarly, in the control groups with and without APO, echocardiography will be performed in mothers and offspring until 72 hours postpartum.Transthoracic echocardiogram will be done by an investigator, who will be blinded to maternal SLE and APO diagnoses, as well as maternal autoantibodies status, using a "Philips-Epiq7c" machine. Images will be obtained according to established guidelines.19 Echocardiography will be performed at bedside in mothers and neonates, after delivery at the RVH birth centre. Of note, all pregnant SLE women enrolled within LEGACY deliver at the RVH. The investigator in charge will perform a comprehensive echocardiographic examination of mothers and offspring to determine conventional measures of ventricular function, including a complete 2-dimensional (2-D) and colour flow-Doppler assessment. The LEVF will be calculated using Simpson's method. An investigator in charge will perform a blind review of all traditional echocardiographic assessments. 3-dimensional (3-D) echocardiography of right ventricle (RV) and left ventricle (LV) volumes will also be acquired. GLS will be measured using a 2-D and 3-D STE software (i.e. TomTec Arena). For evaluation of strain, the endocardial border will be traced, at end-systole in the appropriate imaging plane. Peak GLS will be measured by the software using an average of 2 consecutive cardiac cycles. Measures will be repeated twice to calculate intra-observer intraclass correlation coefficient (ICC). An investigator will blindly review STE assessments to calculate the inter-observer ICC.

Statistical analyses

1. Impact of placenta-mediated APO on maternal \& neonatal GLS (aim 1) In SLE mothers, the investigators will compare GLS in early pregnancy versus the postpartum. The investigators will further compare GLS in SLE with and without APO, both in early pregnancy and in the postpartum. The investigators will also determine if GLS is different in SLE versus controls in early pregnancy and in postpartum, stratifying for APO status. In offspring, the investigators will compare mean GLS in children born to mothers with APO versus those born to mothers without APO, stratifying for maternal SLE status. To compare the GLS between groups, the investigators will calculate the mean (with standard deviation) and median (with interquartile range). Comparisons between groups will be performed using either Student's t-test (for means) or Mann Whitney U test (for medians), depending on data distribution.
2. Maternal autoantibodies \& neonatal GLS (aim 2) In SLE offspring, the investigators will determine if GLS differs among those exposed to maternal autoantibodies (assessing respectively anti-Ro/La and aPL) versus unexposed offspring. The investigators will explore stratification by quartiles of HCQ cord blood levels. Comparisons between groups will be performed using either Student's t-test (for means) or Mann Whitney U test (for medians), depending on data distribution.
3. Association of maternal and cord blood HCQ levels with maternal and neonatal GLS (aim 3) In SLE mothers, the investigators will perform a univariate linear regression analysis with GLS as the dependent variable and maternal HCQ levels as the independent variable. The investigators will assess the possibility of a J-curve relationship by adding a quadratic term to the regression analysis and determine goodness of fit for both models (with and without quadratic term). In a sensitivity analysis, the investigators will explore the effect of cumulative HCQ drug exposure on GLS. In offspring, the investigators will perform a univariate linear regression analysis with GLS as the dependent variable and cord HCQ levels as the independent variable. The investigators will also assess the possibility of a J-curve relationship by adding a quadratic term to the regression analysis and determine goodness of fit for both models (with and without quadratic term). In a sensitivity analysis, the investigators will restrict the analysis to neonates born to SLE mothers without anti-Ro/La and aPL antibodies. The investigators will also explore the impact of cumulative maternal HCQ drug exposure during pregnancy on neonatal GLS.
4. Power calculation Based on estimates of mean GLS and standard deviations from prior studies, the proposed sample size will allow the investigators to detect a difference in mean GLS of at least -2.2% when comparing any groups (including between SLE groups with and without APO, assuming an APO incidence of 30%), with 0.80 power and 2-sided alpha of 0.05.

This research will lead to a novel approach for early risk stratification of SLE women and offspring at high risk of CVD. Identifying those at greatest risk is critical to implement preventive strategies and optimal management of modifiable CVD risk factors. The findings will inform efforts to develop a SLE-specific CVD risk prediction tool accounting for pregnancy history, using data from the SLICC Inception Cohort. Determining that maternal autoantibodies are associated with subclinical cardiac disease in offspring will not only offer insights into mechanistic pathways but offer the opportunity to examine therapeutic agents, such as HCQ, and improve management to reduce long-term CVD risk. This study will highlight the importance of monitoring HCQ levels in pregnancy. As virtually all SLE pregnancies (and a large number of rheumatoid arthritis and other rheumatic disease pregnancies) are exposed to HCQ, this study will provide critical data to inform the conduct of future studies in several populations of pregnant women with rheumatic diseases. The novel findings will be highly relevant to other high-risk pregnancy populations. Myocardial strain might prove not only to be a sensitive marker of cardiac involvement, but also a potentially sensitive outcome measure, responsive to therapy. Thus, findings from this study could help the conduct of future clinical trials and the methodology could be applicable to the assessment of subclinical cardiac involvement in other rheumatic diseases (e.g. systemic sclerosis). The investigators' close links with the Canadian Network for Improved Outcomes in SLE (CaNIOS) and SLICC will allow the investigator to lead the way nationally and internationally in establishing further studies to improve reproductive and cardiovascular outcomes in SLE.