Abstract
Objective. Few data exist to guide the frequency and type of monitoring in systemic lupus erythematosus (SLE) pregnancies. A systematic literature review was performed to address this gap in the literature.
Methods. A systematic review of original articles (1975–2015) was performed using Medline, Embase, and Cochrane Library. We included search terms for SLE, pregnancy, and monitoring. We also hand-searched reference lists, review articles, and grey literature for additional relevant articles.
Results. The search yielded a total of 1106 articles. After removing 117 duplicates, 929 articles that were evidently unrelated to our topic based on title and/or abstract, and 7 that were in a language other than English or French, 53 articles were included for full-text review. Following a more in-depth review, 15 were excluded: 6 did not use any measure of SLE activity and 6 did not specifically address SLE monitoring in pregnancy; 1 case series, 1 review, and 1 metaanalysis were removed. Among the 38 included studies, presence of active disease, antiphospholipid (aPL) antibodies positivity, and abnormal uterine and umbilical artery Doppler studies predicted poor pregnancy outcomes. No studies evaluated an evidence-based approach to the frequency of monitoring.
Conclusion. Few existing studies address monitoring for optimal care during SLE pregnancies. The available data imply roles for aPL antibodies measurement (prior to pregnancy and/or during the first trimester), uterine and umbilical artery Doppler studies in the second trimester, and following disease activity. Optimal frequency of monitoring is not addressed in the existing literature.
Systemic lupus erythematosus (SLE) predominantly affects women during their reproductive years, occurring in about 1/1000 women aged between 15 and 45 years1. SLE is associated with substantial maternal and fetal morbidity during pregnancy. Compared with the non-SLE population, SLE has been shown to be associated with an increased risk of preterm birth, cesarean delivery, preeclampsia, low birth weight, intrauterine growth restriction (IUGR), congenital heart block (CHB), and intrauterine and neonatal death2,3,4,5. Preterm birth is the most common adverse pregnancy outcome in women with SLE, with incidence ranging from 15% to 50% (as opposed to 10% in unaffected women), with increased risk in women with lupus nephritis or high disease activity. In the general population, preterm birth is the leading cause of neonatal death and the second most common cause of death (after pneumonia) in children younger than 5 years6.
SLE has a waxing and waning course and it has been shown that if women conceive during a period of disease quiescence, this will minimize the risk of flare in pregnancy, but will not eliminate it, with rates of flare still ranging between 20% and 40% of pregnancies that are conceived during a period of remission7,8,9,10. In addition to a greater frequency of pregnancy complications, the SLE disease course itself can be negatively affected by pregnancy, with a greater number of women developing an SLE flare in the peripartum period11,12,13.
SLE pregnancies are considered high risk, being associated with higher maternal and fetal morbidity. Although the majority of SLE pregnancies end with live births, active disease and major organ involvement can affect the outcomes in both mother and fetus. In addition, major fetal issues such as IUGR and neonatal SLE syndromes make monitoring imperative in SLE pregnancies14.
Finally, some symptoms of SLE can be silent, such as renal flare or thrombocytopenia, emphasizing the need for closer monitoring in pregnancy; these silent flares can still increase the chance of obstetrical complications15.
Quality indicators pertaining to reproductive health in SLE have been developed but are limited only to the recommendation that anti-SSA/SSB and antiphospholipid (aPL) antibodies be documented in the chart, prior to conception16. Optimal quality indicators remain undefined, and topics to be addressed include not only laboratory monitoring but also newer imaging modalities for monitoring high-risk pregnancies, such as umbilical and uterine artery Doppler studies. These are believed to be useful in monitoring pregnancies at high risk of placental insufficiency, which include SLE pregnancies17. Because the issue of how best to follow disease activity, and how often to monitor for disease flares, has not been systematically addressed in the literature, current practice is heterogeneous and not necessarily evidence-based18.
To address this important knowledge gap, a Canadian SLE working group was established, funded by the Canadian Institutes of Health Research and endorsed by the Canadian Rheumatology Association. One of the aims of this group was to determine what investigations are needed to optimally monitor pregnancy in SLE, in the Canadian context (which includes universal healthcare access).
MATERIALS AND METHODS
The systematic literature review was performed according to the Preferred Reporting Items for Systematic review and Meta-analysis Protocols (PRISMA-P) 2015 statement19 (Table 1).
PRISMA-P checklist for preferred reporting of systematic reviews.
The review was conducted through 3 search engines: Embase, Medline, and Cochrane. The search was performed on publications between 1975 and 2015, without any language restrictions. We used Medical Subject Heading and free text terms adapted for each database to identify original articles. We included search terms for SLE, pregnancy, and monitoring variables. All terms within each set were combined using the Boolean operator “OR” and then the 3 sets were combined using “AND.” This was supplemented by hand-searching reference lists, review articles, and grey literature for relevant articles not identified by the electronic searches, as well as including relevant articles published after completion of our review. Exclusion criteria then included any abstract in a language other than French or English, case reports, case series, and review articles.
Study selection
First, titles from the initial search were reviewed by 3 individuals (EGM, LB, and EV) to initially include any potential studies related to the study question, and to exclude any duplicates. Second, titles and abstracts were reviewed to identify relevant studies that met our inclusion criteria and to exclude case reports and studies unrelated to the systematic review. Third, 2 reviewers (EGM and LB) independently reviewed each full-text article for inclusion in the final set of articles, with a third reviewer (EV) settling discrepancies. Two reviewers (EGM and EV) summarized evidence on monitoring of SLE in pregnancy. None of the reviewers were blinded to the authors or journal titles.
Data extraction and quality assessment
Relevant data pertaining to monitoring were extracted from each article as well as general information such as country of the study, type of study, year of publication, and first author. Monitoring was divided into 4 categories: serological tests [which included anti-DNA, antiextractable nuclear antibody, IgG and IgM anticardiolipin antibodies (aCL), lupus anticoagulant (LAC), and complement levels]; measures of SLE activity using a validated scoring system [examples include the SLE Disease Activity Index (SLEDAI) and European Consensus Lupus Activity Measure scores; studies that did not use any type of scoring system to measure disease activity were not included]; obstetrical Doppler ultrasound monitoring (either uterine or umbilical artery Doppler studies); and other (monitoring variables not falling into one of the above categories). When applicable, we recorded whether the frequency of monitoring was addressed, as well as any associations with negative maternal or obstetrical outcomes. Maternal outcomes were classified as the development of preeclampsia, or worsening of SLE disease activity, including lupus nephritis. Obstetrical outcomes were classified as IUGR, spontaneous abortion (prior to 20 weeks of gestational age), small for gestational age (SGA) or low birth weight (SGA and low birth weight were defined within individual studies), preterm delivery (prior to 37 weeks gestational age), complete CHB, or intrauterine fetal demise. Relevant data were extracted, synthesized, and presented in tabular format.
Risk of bias in individual studies: the Newcastle-Ottawa scale
Final studies included in the systematic review were evaluated for quality of evidence and risk of bias using the Newcastle-Ottawa scale, which uses a star system whereby a study is judged on 3 measures: the selection of the study groups, the comparability of the groups, and how the exposure for case-control (or outcome of interest for cohort studies) was ascertained.
RESULTS
Study characteristics
A general description of the characteristics of each study is presented in Table 22,9,13,15,20–53 including first author, year of publication, study population, and study type, as well as monitoring variables studied. A description of the search process, including the reasons for excluded studies, is shown in Figure 1. In total, 1106 titles were evaluated from the initial search strategy, of which 117 were removed because they were duplicates, and 7 were removed because they were in a language other than French or English. Then, on initial review of the title and/or abstracts, 929 articles that did not address SLE monitoring in pregnancy and were removed, leaving 53 articles that were reviewed in depth. Upon full review of the 53 articles, it was found that 6 more did not address SLE monitoring in pregnancy, 6 did not use any scoring system for SLE disease activity, 1 was a review, 1 was a case series, and 1 was a metaanalysis. These were removed, leaving 38 original observational studies for review. Final studies were selected based on reporting of adverse obstetrical or maternal outcomes related to monitoring variables (Figure 1) in SLE.
Flowchart of the article selection process.
Study characteristics.
Participant characteristics
The majority of studies evaluated monitoring of SLE during pregnancy and immediately postpartum. Only a few studies looked specifically at lupus nephritis or focused solely on pregnancies in mothers who were anti-Ro- and/or La-antibody positive. The study populations were multiethnic, with patients from across the world including North America, Europe, Latin America, Asia, and the Middle East.
Monitoring variables characteristics
Of 38 final articles evaluated, 22/38 (58%) addressed the value of measuring serology, including the measurement of aPL antibodies prior to pregnancy and/or during the first trimester, and 18/38 (47%) assessed monitoring for SLE flare with a validated scoring system, such as the SLEDAI (Figure 2). The utility of umbilical and/or uterine artery Doppler monitoring for predicting poor obstetrical outcomes was evaluated in 5/38 (13%) articles. Disease activity scoring systems modified to account for the physiologic changes of pregnancy were observed in 4 articles, such as the SLE in Pregnancy Disease Activity Index54 and the Lupus Activity Index in Pregnancy55. Of note, none of them have been formally validated.
Histogram of the frequency of monitoring variables evaluated in the studies included in the systematic literature review. Serology = anti-DNA, antiextractable nuclear antibody, IgG and IgM anticardiolipin antibodies, lupus anticoagulant, and/or complement levels. Disease activity measured using a validated scoring system (examples include the SLEDAI and ECLAM scores; studies that did not use any type of scoring system to measure disease activity were not included). Renal = serum creatinine, urinalysis, urine microscopy and/or urine creatinine-to-protein ratio. Other = variables evaluated by only 1 included study (e.g., antifactor X11, C1q, and anti-PRL antibodies). AFP: alpha fetal protein; Hypo T4: hypothyroidism; CBC: complete blood count; PLT: platelets; BP: blood pressure; SLEDAI: Systemic Lupus Erythematosus Disease Activity Index; ECLAM: European Consensus Lupus Activity Measure; PRL: prolactin.
Outcomes
Findings from each study are summarized and presented in Table 32,9,13,15,20–53; studies were evaluated for risk of bias using the Newcastle-Ottawa scale and the quality of data is presented in Table 42,9,13,15,20–53. Across the studies included for full review, the presence of active disease (measured by various scoring systems), aPL antibodies positivity, and abnormal uterine and umbilical artery Doppler studies predicted poor pregnancy outcomes. Low complement and thrombocytopenia at the beginning of pregnancy were also predictors of poor obstetrical outcomes. No studies that assessed the value of serological and disease activity monitoring evaluated an evidence-based approach to the frequency of monitoring. Studies that addressed uterine and umbilical artery Doppler monitoring discussed the frequency at which these tests were performed but did not address the evidence behind the practice.
Outcomes related to monitoring variables.
Quality assessment by Newcastle-Ottawa scale.
DISCUSSION
Summary of findings
In this systematic review examining monitoring in SLE pregnancies, 38 articles were included, after an initial screening of 1106 abstracts and 53 articles that were reviewed in depth. Final articles addressed the monitoring of SLE-related pregnancy and focused on 3 main types of monitoring (blood and urine tests, SLE disease activity monitoring, and uterine or umbilical Doppler studies), as well as other types of monitoring. There have been no randomized controlled studies on this topic and so all articles were observational studies. The majority of studies were cohort studies that investigated predictors of poor maternal and fetal outcomes related to measures of SLE activity (both disease-specific and otherwise) throughout pregnancy. Studies included SLE women with varying degrees of disease activity, with some including only women with inactive or mildly active SLE, which might have influenced their findings.
Possible recommendations
Based on our systematic review of the literature, there may be a role for monitoring aPL (LAC and aCL), dsDNA, complement, and anti-Ro and anti-La antibodies prior to conception and early in pregnancy. Complement and dsDNA are likely helpful measures when a flare is suspected because these tests were shown in numerous studies that we examined to be predictive of poor outcomes (20/37 or 54% of studies addressed the predictive value of 1 or more of these monitoring variables). Because patients with quiescent SLE (i.e., with no or low disease activity) have fewer negative outcomes in pregnancy, if these autoantibodies and blood tests are measured prior to conception, it might help physicians stratify pregnancy risk in pregnant women and/or women planning a pregnancy. However, dsDNA, for example, does not always correlate with disease flare and/or activity in all patients and so this is not universally true, but might still contribute useful information. In addition, complement levels normally increase during pregnancy, which might further complicate its role as a reliable disease activity marker27.
The available literature also suggests that laboratory testing should be combined with evidence of increased disease activity index as measured by a validated scoring system, such as the SLEDAI, which has also shown to be associated with poor maternal and obstetrical outcomes in numerous studies included in this systematic review (18/38 or 47%). However, no study has directly compared the predictive value of laboratory or clinical data alone as compared to SLEDAI (which combines clinical activity as well as laboratory data).
One expanding area of monitoring in SLE pregnancies relates to the role of uterine and umbilical Doppler studies. Absent end-diastolic velocities of the umbilical artery predict early pregnancy-induced hypertension/preeclampsia and fetal or neonatal death in non-SLE pregnancies56,57. Abnormalities demonstrating increased resistive indices, notching of the arteries, or in very severe cases, reversal of end-diastolic flow, are highly predictive of poor fetal outcomes in SLE pregnancies28,32. This type of monitoring is minimally invasive and could be considered in the second and third trimester if the expertise exists in the center where the patient’s pregnancy is followed. This monitoring could therefore be used at the time of a suspected flare, or at the discretion of the maternal-fetal medicine or obstetrical medicine specialist who is involved in the patient’s care.
Four studies looked at alternative or novel tests for predicting poor outcomes in pregnancy such as anti-C1q antibodies, antiprolactin antibodies, antifactor X11, and alpha-fetal protein. Given that these studies have not been reproduced, and many of these serological tests are not widely available, we would not recommend their routine use at this time for monitoring of SLE pregnancies.
Regarding other tests, it is reasonable to measure thyroid stimulating hormone levels early in pregnancy, because hypothyroidism was found to be associated with an increased risk of prematurity in pregnant women from the general population as well as pregnant women with SLE, and was associated with an increased incidence of complete CHB among the offspring of mothers with anti-Ro antibodies40,53.
Prior to our study and the meeting of the Canadian SLE Working Group, guidelines did not exist regarding what monitoring variables should be measured in SLE pregnancies and at what frequency. A systematic review of the literature revealed there is no evidence to guide the frequency of routine monitoring of complete blood count, chemistry profile including liver enzymes, and urinalysis (routine or microscopy) or urine protein to creatinine levels, during pregnancy with SLE. There is some evidence for a role for measuring anti-Ro and anti-La antibodies prior to conception and early in SLE-related pregnancy, to risk-stratify pregnancies regarding the development of complete CHB. Similarly, the presence of aPL is predictive of poor outcomes, particularly preeclampsia and fetal loss, and so may be reasonable to measure prior to conception or early in pregnancy. Monitoring SLE disease activity should be done, potentially with a validated scoring system such as the SLEDAI because it relies on assessment of different clinical aspects of the disease (i.e., clinical manifestations and specific blood tests). If a flare is suspected, measuring anti-dsDNA antibodies and complement levels may be helpful. Women with SLE should be considered for referral to an expert in maternal-fetal medicine for fetal monitoring with umbilical artery Doppler studies, given that Doppler abnormalities are predictive of adverse outcomes; among pregnant populations without SLE, this monitoring modality has improved fetal outcomes. This systematic review will inform guidelines for the type of and frequency of monitoring of SLE in pregnancy, which are being developed by the Canadian SLE Working Group.
Footnotes
EV received salary funding from the Fonds de Recherches Santé Québec (FRSQ). EGM received funding from an FRSQ Master’s award and the McGill Clinician Investigator Program. SK received funding through a Canadian Institutes of Health Research Dissemination Grant.
Full Release Article. For details see Reprints and Permissions at jrheum.org
- Accepted for publication January 25, 2018.
Free online via JRheum Full Release option
REFERENCES
In this issue
