Abstract
Objective:
Systemic lupus erythematosus (SLE) patients have more pregnancy complications than healthy patients. Data regarding pregnancy outcomes in women with undifferentiated connective tissue disease (UCTD) are more limited, and existing studies are concentrated in Italy and predominantly in patients with a new diagnosis. We compared pregnancy outcomes for UCTD and SLE patients with established disease.
Methods:
Between 2008 and 2017, patients with UCTD and SLE at an academic medical center were recruited to a prospective pregnancy registry. UCTD was defined as a positive autoantibody plus connective tissue disease symptoms, not meeting criteria for another rheumatic diagnosis. SLE was defined by ACR or SLICC classification criteria or by physician diagnosis. Data were collected throughout pregnancy and postpartum. Comparator groups included UCTD, low activity SLE, and high activity SLE.
Results:
A total of 150 SLE and 51 UCTD pregnancies were analyzed. Disease activity was low in most patients, though more patients with SLE had severe activity during pregnancy (12% vs. 2%; p=0.05). The frequencies of prematurity and preeclampsia were significantly lower in UCTD than in high activity SLE patients (preterm: 17% vs. 45%, p=0.004 and preeclampsia: 6% vs. 34%, p=0.0008), though similar to low activity SLE patients. More SGA infants were born to SLE than UCTD patients (33% vs. 7%; p=0.0005), regardless of disease activity level.
Conclusion:
Pregnancies in women with UCTD managed by a rheumatologist have a high rate of pregnancy success and fewer risks than those in women with active SLE.
Keywords: SLE, UCTD, pregnancy, hydroxychloroquine, preterm birth, preeclampsia
It is well-established that, when compared to healthy women, women with systemic lupus erythematosus (SLE) have more complications in pregnancy, including higher rates of miscarriage, preterm delivery, preeclampsia, intrauterine growth restriction (IUGR), and small for gestational age (SGA) infants1–3. Women with SLE more often deliver by Cesarean section (C-section) or have labor induced1,3,4. The risk of pregnancy complications tends to be higher in patients with active lupus than in patients with low disease activity5,6.
Undifferentiated connective tissue disease (UCTD) is a term used to describe connective tissue diseases that do not meet criteria for another well-defined connective tissue disease7. Definitions of UCTD vary but generally include the presence of symptoms of a connective tissue disease as well as a positive antinuclear antibody (ANA) test8–10. UCTD is typically a milder condition with a more benign course than SLE, so the outcomes studies of pregnancy in SLE may not be applicable to patients with UCTD.
Previous studies have evaluated pregnancy outcomes in patients with UCTD9,11–14. Not all studies compared UCTD to other groups; however, the results of the comparative studies suggest that the risk of pregnancy complications in UCTD lies between those of healthy patients and that of patients with a defined rheumatic condition (Table 1). All of these studies were done in Italy and thus represent a more racially homogeneous population, distinct from those seen in other parts of the world. Additionally, half of these studies evaluated outcomes in patients without a prior diagnosis of UCTD who, during pregnancy, screened positive for the diagnosis as part of the study12–14. It is not clear whether these benign pregnancy outcomes would be found in women with symptoms significant enough to have sought rheumatologic care to obtain this diagnosis.
Table 1:
Prior studies of pregnancies in women with undifferentiated connective tissue disease.
| Author, Year Study Design | Groups | Pregnancy Loss | Mean GA at birth (wks) | Preterm | SGA or IUGR | Preeclampsia |
|---|---|---|---|---|---|---|
| Mosca, 2002 Cohort | UCTD (A), pregnant: n=25 (20 patients) | 12% | 36.7 | n=3 (12%) | IUGR: n=2 (twin gestation) | NR |
| Castellino, 2011 (1/2003–12/2008) Case-control** | UCTD (B), pregnant: n=55 pregnancies (50 patients) | n=3 (5.4%) | 38.6 | NR | IUGR: n=2 (3.6%) | n=2 (3.6%) |
| Spinillo, 2008 (3/2004–6/2007) Case-control | New dx UCTD (C), pregnant: n=41 | NR | 38.9* | NR | SGA: 12/40 (30%)* | n=3 (7.5%) |
| Healthy controls: n=82 | NR | 39.6 | NR | SGA: 11/80 (13.8%) | n=1 (1.3%) | |
| Spinillo, 2012 (5/2005–4/2010) Case-control | Defined CTD, pregnant: n=24 | n=3 (12.5%) | 37.8 | n=3 (12.5%) | IUGR: n=3 (12.5%) SGA: n=4 (16.67%) |
n=0/21 |
| New dx UCTD (D), pregnant: n=62 | n=2 (3.23%)* | 38.7* | Preterm, prior to 34 WGA n=2 (3.23%)* | IUGR: n=5 (8.06%)* SGA: 12 (19.35%)* |
n=7/60 (11.3%)* | |
| Insufficient criteria for dx, pregnant: n=57 | n=2 (3.51%) | 39.4 | n=0 | IUGR: n=0 SGA: n=3 (5.26%) |
n=0/55 | |
| Asx controls, pregnant: n=211 | n=2 (0.95%) | 39.4 | n=1 (0.47%) | IUGR: n=4 (1.90%) SGA: n=19 (9.00%) |
n=3/209 (142) | |
| Spinillo, 2016 (3/2009–6/2014) Case-control | Defined CTD, pregnant: n=68 | n=0 | 37.7 | n=5 (7.4%) | SGA: n=18 (26.5%) | n=20 (29.4%) |
| New dx UCTD (E), pregnant: n=131 | n=0 | 38.4* | Preterm, prior to 34 WGA n=6 (4.6%) | SGA: 23 (17.5%) | n=30 (22.9%) | |
| Insufficient criteria for dx, pregnant: N=150 | n=1 (0.6%) | 38.9 | n=22 (14.6%) | SGA: n=17 (11.3%) | n=9 (6%) | |
| Asx controls: n=597 | n=0 | 39 | n=17 (2.8%) | SGA: n=46 (7.7%) | n=19 (3.2%) | |
| Zucchi, 2019 (2000–2018) Cohort | UCTD (A), pregnant: n=100 (81 patients) | n=11 (11%) | 38.8 | n=9 (10%) | SGA: 10% | n=1 (1.1%) |
| Radin, 2020 (2010–2019) Cohort | UCTD (F), pregnant: n=224 (133 patients) | Sab: n=45 (20.1%) SB: n=2 (0.9%) |
36.2 | n=30 (16.9%) Late preterm: n=22 (12.4%) |
SGA: n= 21(11.9%) IUGR: n= 6 (2.7%) |
n=5 (2.2%) |
Definitions: A= Signs, symptoms suggestive of CTD lasting ≥1 year, not fulfilling criteria for a defined CTD + positive ANAs, on 2 separate occasions. B= Signs, symptoms suggestive of CTD lasting ≥1 year, not fulfilling criteria for a defined CTD + -positive ANAs, on 2 separate occasions (>1:80). C= Signs, symptoms suggestive of CTD lasting >6 months, not fulfilling criteria for defined CTD + positive ANA. D= Signs, symptoms suggestive of CTD lasting ≥3 years, not fulfilling criteria for a defined CTD + positive ANAs, on 2 separate occasions. E= signs, symptoms suggestive of CTD lasting ≥3 years, not fulfilling criteria for a defined CTD + positive ANA. F= UCTD diagnosis + positive ANA.
UCTD= undifferentiated connective tissue disease. Sab= spontaneous abortion. GA= gestational age. BW= birthweight SGA= small for gestational age. IUGR= intrauterine growth restriction. C/S= Cesarean section. CTD= connective tissue disease. ANA= antinuclear antibody. NR= not reported. n/a= not applicable. Dx= diagnosis. Asx= asympomatic. WGA= weeks gestational age. SB= stillbirth (intrauterine death after 20 weeks gestation).
Significant result
Control group= UCTD patients, not pregnant. Outcomes not reported here.
The objective of our study was to compare pregnancy outcomes of our cohort of patients with UCTD to pregnancy outcomes of patients with SLE. We hypothesized that the outcomes of UCTD patients would be more favorable than those of SLE patients, regardless of SLE activity in pregnancy. Our goal was to provide clinically useful information to women with UCTD who wish to pursue pregnancy.
Patients and Methods
Patients and Recruitment
Patients for this prospective cohort study were recruited into the Duke Autoimmunity in Pregnancy (DAP) registry (Duke University Hospital Institutional Review Board number: Pro00000756) from the Duke rheumatology clinic by providing written informed consent at their first clinic visit in pregnancy. Patients were recruited from 2008 to 2017 and may be included in the analysis more than once for distinct pregnancies.
Patients were included if they had a singleton pregnancy, a diagnosis of either UCTD or SLE, and were enrolled into the DAP registry prior to 33 weeks. In our cohort, we diagnosed UCTD if patients had at least one positive autoantibody and symptoms suggestive of an autoimmune disease without meeting criteria for another rheumatic condition. SLE was defined according to ACR criteria, SLICC criteria, or physician diagnosis19–21. Exclusion criteria included the patient having a primary diagnosis other than UCTD or SLE, including those with a diagnosis of primary antiphospholipid syndrome (APS). Women with SLE and APS were included in the study. Multiple gestation pregnancies (n=9) are reported below but were excluded from statistical analysis of pregnancy and neonatal outcomes.
Data Collection
Data collected at baseline included patient demographics and autoantibody profile. At each visit throughout pregnancy, data on medications, current lupus-related labs, and disease activity using the SLEPDAI22 and the physician global assessment (PGA) were recorded at each visit. The maximum disease activity during pregnancy was used to classify patients into low, moderate, and severe activity groups. A PGA of 0 to <1 was considered low activity, PGA 1 to <2 was moderate activity, and PGA ≥2 severe activity. High activity SLE was defined as moderate or severe activity. Neonatal and maternal outcomes were collected via chart review and patient report soon after delivery. These outcomes included pregnancy loss for women enrolled prior to 20 weeks; preterm delivery (<37 weeks); early preterm birth (<30 weeks); weeks of gestational age at delivery; method of delivery; small for gestational age (SGA; defined as birthweight <10th percentile for gestational age at birth, based on U.S. population reference percentiles of birth weight, stratified by infant sex); very low birthweight (<1500 g); infant APGAR (Appearance, Pulse, Grimace, Activity, Respiration) scores at one and five minutes; APGAR ≤6 at five minutes; neonatal intensive care unit (NICU) admissions; and complete heart block23.
Data Analysis
Outcomes were compared between pregnancies in women with UCTD versus SLE, as well as between pregnancies in which the mother had high activity SLE (PGA≥1) during pregnancy, low activity SLE during pregnancy, or UCTD. Due to significant differences in maternal race between women with SLE and UCTD and the known impact of non-Caucasian race on pregnancy outcomes, odd ratios were adjusted by maternal race. Differences in proportions were estimated by Fisher’s exact test. Differences in continuous variables were estimated by t-test or Wilcoxon rank-sum test, depending on the distribution of the data. To account for the lack of independence in pregnancy outcomes for multiple pregnancies in the same woman, a sensitivity analysis was performed excluding all women with multiple pregnancies in the registry. Data were analyzed using SAS 9.4 (Cary, NC).
Results
Patients
Data were collected on a total of 201 pregnancies, including 150 pregnancies in 134 women with SLE and 51 pregnancies in 47 women with UCTD (Table 2). In patients with SLE, there were 8 twin pregnancies and 1 triplet pregnancy. There were significant differences in the racial composition of the diagnostic groups with the majority of women with UCTD being Caucasian (67%) and half of women with SLE being Black (51%). The average maternal age at delivery and gestational age at study entry were not significantly different between women with UCTD and SLE.
Table 2.
Demographic and baseline characteristics for 150 pregnancies in 134 SLE patients and 51 pregnancies in 47 UCTD patients. Data include 141 singleton gestations, 8 twin gestations, and 1 triplet gestation in the SLE group. All pregnancies in the UCTD group were singleton.
| Disease Group: | UCTD | SLE | p-value* | High SLE | Low SLE |
|---|---|---|---|---|---|
|
| |||||
| Total patients | 47 | 134** | 58 | 81 | |
| Total pregnancies | 51 | 150 | 61 | 89 | |
|
| |||||
| Age (years), mean (SD) | 31.4 (5.5) | 30.0 (5.5) | 0.1 | 28.9 (5.5) | 30.8 (5.4) |
| Age (years), range | 17–42 | 19–45 | 20–42 | 19–45 | |
| Gestational age at entry (weeks), mean (SD) | 15.0 (7.7) | 12.6 (6.6) | 0.05 | 13.5 (6.2) | 12.0 (6.8) |
|
| |||||
| Race, n (%) | |||||
| Black | 10 (20%) | 76 (51%) | 0.0001 | 33 (54%) | 43 (48%) |
| Caucasian | 34 (67%) | 64 (43%) | 0.003 | 26 (43%) | 38 (43%) |
| Asian | 3 (6%) | 5 (3%) | 1 (2%) | 4 (4%) | |
| Other | 3 (6%) | 5 (3%) | 1 (2%) | 4 (4%) | |
| Unknown | 1(2%) | 0 (0%) | 0 (0%) | 0 (0%) | |
| Hispanic ethnicity (n=196) | 3 (6%) | 6 (4%) | 0.7 | 4 (7%) | 2 (2%) |
|
| |||||
| History of lupus nephritis | 0 (0%) | 42 (28%) | <0.0001 | 23 (38%) | 19 (21%) |
|
| |||||
| Auto-antibodies (ever) | |||||
| ANA (n=189) | 44 (92%) | 139 (99%) | 0.04 | 54 (100%) | 85 (98%) |
| Anti-Ro (n=196) | 18 (37%) | 75 (51%) | 0.1 | 25 (42%) | 50 (57%) |
| Anti-La (n=195) | 9 (19%) | 23 (16%) | 0.7 | 12 (20%) | 11 (13%) |
| Anti-RNP (n=194) | 5 (10%) | 64 (44%) | <0.0001 | 29 (48%) | 35 (41%) |
| Anti-Sm (n=193) | 0 (0%) | 47 (32%) | <0.0001 | 25 (42%) | 22 (26%) |
|
| |||||
| APLA*** (ever) (n=171) | 3 (8%) | 21 (16%) | 0.3 | 10 (18%) | 11 (14%) |
| Antiphospholipid syndrome | 0 (0) | 5 (3%) | 0.3 | 3 (5%) | 2 (2%) |
|
| |||||
| Disease Activity | |||||
| At First Visit | |||||
| Low | 45 (88%) | 110 (73%) | 0.1 | 21 (34%) | 89 (100%) |
| Moderate | 5 (10%) | 30 (20%) | 30 (49%) | 0 (0%) | |
| Severe | 1 (2%) | 10 (7%) | 10 (16%) | 0 (0%) | |
| Highest During Pregnancy | |||||
| Low | 40 (78%) | 89 (59%) | 0.02 | -- | 89 (100%) |
| Moderate | 10 (20%) | 43 (29%) | 43 (70%) | -- | |
| Severe | 1 (2%) | 18 (12%) | 18 (30%) | -- | |
|
| |||||
| Inflammatory Markers During Pregnancy | |||||
| Low C3 &/or C4 (n=187) | 2/42 (5%) | 50/145 (34%) | <0.0001 | 26/59 (44%) | 24/86 (28%) |
| dsDNA+ (n=187) | 0/42 (0%) | 35/145 (24%) | <0.0001 | 14/59 (24%) | 21/86 (24%) |
| Low Platelets (n=193) | 2/45 (4%) | 6/148 (4%) | 1.0 | 5/60 (8%) | 1/88 (1%) |
| Proteinuria (n=192) | 0/44 (0%) | 13/148 (9%) | 0.04 | 13/60 (22%) | 0/88 (0%) |
| Proteinuria prior to 20 weeks (n=155) | 0/30 (0%) | 10/125 (8%) | 0.2 | 10/50 (20%) | 0/88 (0%) |
|
| |||||
| Medications, n (%) | |||||
| Immunomodulatory | |||||
| Hydroxychloroquine | 33 (65%) | 125 (83%) | 0.009 | 52 (85%) | 73 (82%) |
| Azathioprine | 0 (0%) | 30 (20%) | 0.0001 | 20 (33%) | 10 (11%) |
| Steroids§ | 9 (18%) | 70 (47%) | 0.0002 | 46 (75%) | 24 (27%) |
| None | 15 (29%) | 17 (11%) | 0.003 | 4 (7%) | 13 (15%) |
|
| |||||
| Anti-hypertensive | 5 (10%) | 25 (17%) | 0.3 | 14 (23%) | 11 (12%) |
| Aspirin | 25 (49%) | 112 (75%) | 0.001 | 45 (74%) | 67 (75%) |
| Anticoagulation | 2 (4%) | 19 (13%) | 0.1 | 10 (16%) | 9 (10%) |
Abbreviations: SD= standard deviation.
The p-value compares UCTD and SLE pregnancies with statistical significance set at p≤ 0.05.
5 patients had both low and high SLE pregnancies.
APLA= antiphospholipid antibody (positive [≥40 IU] anti-cardiolipin IgG, IgM; positive [≥40 IU] anti-beta 2 glycoprotein IgG, IgM; positive lupus anticoagulant test).
Oral (including dexamethasone), intramuscular, intravenous, or intra-articular steroids.
SLE/UCTD Features:
The frequency of anti-Ro and anti-La antibodies was similar between women with UCTD and SLE, but anti-RNP and anti-Sm antibodies were both significantly more common in women with SLE. Antiphospholipid antibodies were not common within the cohort, and the frequency was not different for women with UCTD and SLE. Five pregnancies in five women with SLE met criteria for antiphospholipid syndrome. In SLE patients, 28% had a history of lupus nephritis.
Disease activity:
Most patients had low disease activity at the first clinic visit and throughout pregnancy. However, severe activity during pregnancy occurred in 2% of women with UCTD and 12% of women with SLE (p=0.05). More patients with SLE than UCTD had abnormalities in complement and dsDNA during pregnancy. Proteinuria prior to 20 weeks gestation was only found in patients with active SLE.
Medications:
The large majority of women with UCTD and SLE took immunomodulatory medications during pregnancy, most commonly hydroxychloroquine. Women with SLE were more likely to be prescribed azathioprine (none with UCTD compared to 20% with SLE, p=0.0001) and corticosteroids. Steroids of some form (oral, including dexamethasone; intramuscular; intra-articular; intravenous) were used by 47% of SLE patients and 18% of UCTD patients (p=0.0002). The frequency of corticosteroid use was significantly higher in women with active SLE compared to those with low activity SLE throughout pregnancy (high activity SLE 75% vs low activity SLE 27%; p<0.0001). Chronic daily steroid use of 5mg or less was uncommon and only prescribed in 4% of women with UCTD and 13% with SLE (Figure 1). Six patients, five with SLE (3%) and one with UCTD (2%), received dexamethasone to prevent development of complete heart block in the fetus as their only steroid exposure. More women with UCTD than SLE were not prescribed immunomodulatory medications.
Figure 1.
Maximum dose of oral steroids used in pregnancy among UCTD patients (left) and SLE patients (right). The majority of patients in both groups did not use oral steroids. A small percentage of patients in each group took dexamethasone as the only oral steroid for early neonatal heart block.
Low dose aspirin, which is recommended to decrease the risk for preeclampsia among high risk pregnancies, was more frequently prescribed to women with SLE (75%) than UCTD (49%, p=0.001). Anticoagulation was prescribed to women with APS or prior thrombosis in similar rates for women with UCTD and SLE. The frequency of antihypertensive therapy was not different between patients with UCTD and SLE.
Outcomes
Live birth:
The majority of the 192 singleton pregnancies resulted in a live birth with 86% of SLE and 92% of UCTD patients delivering a live infant (p=0.3) (Table 3). Pregnancy termination for medical indications, most commonly rheumatic disease activity, was most common in women with high activity SLE.
Table 3.
Pregnancy and neonatal outcomes in singleton pregnancies.
| Outcome | UCTD | All SLE | p-value* | High SLE | Low SLE |
|---|---|---|---|---|---|
|
| |||||
| Total pregnancies** | 51 | 141 | 59 | 82 | |
| Live birth | 47 (92%) | 121 (86%) | 0.3 | 49 (83%) | 72 (88%) |
| Losses*** | |||||
| Miscarriage§ | 2 (5%) | 12 (10%) | 0.5 | 2 (4%) | 10 (15%) |
| Stillbirth϶ | 1 (3%) | 3 (3%) | 1.0 | 3 (6%) | 0 (0%) |
| Termination (medically indicated) | 0 (0) | 5 (4%) | 0.3 | 5 (10%) | 0 (0%) |
| Termination (malformation) | 1 (3%) | 0 (0) | 0.2 | 0 (0%) | 0 (0%) |
|
| |||||
| Pregnancy Outcomes† | n=47 | n=121 | n=49 | n=72 | |
| Preterm (<37 weeks) | 8 (17%) | 34 (28%) | 0.2 | 22 (45%) | 12 (17%) |
| Early preterm (<30 weeks) | 1 (2%) | 5 (4%) | 1.0 | 5 (10%) | 0 (0%) |
| Weeks gestational age | |||||
| Mean (SD) | 37.7 (2.5) | 36.8 (3.0) | 0.06 | 35.4 (3.9) | 37.8 (1.7) |
| Range | 27–40 | 24–40 | 24–40 | 32–40 | |
| Preeclampsia or eclampsia‡ (n=170) | 3 (6%) | 25 (20%) | 0.02 | 17 (34%) | 8 (11%) |
| C-section (n=167) | 19 (40%) | 60 (50%) | 0.3 | 29 (60%) | 31 (43%) |
| Induction (n=167) | 18 (38%) | 77 (64%) | 0.003 | 35 (73%) | 42 (58%) |
| Medical Induction# | 10 (56%) | 53 (69%) | 0.3 | 29 (83%) | 24 (57%) |
| Induced for Dates# | 8 (44%) | 24 (21%) | 0.3 | 6 (17%) | 18 (43%) |
|
| |||||
| Neonatal outcomes† | |||||
| SGA¶ (n=161) | 3/45 (7%) | 38/116 (33%) | 0.0005 | 17/47 (36%) | 21/69 (30%) |
| Very low birthweight (<1500 g) | 1/45 (2%) | 8/116 (7%) | 0.4 | 6/47 (13%) | 2/69 (3%) |
|
| |||||
| APGAR 5 minutes, median (IQR) (n=117) | 9 (9–9) | 9 (8.5–9) | 1.0 | 9 (8–9) | 9 (9–9) |
| APGAR 5 minutes ≤6 | 1/37 (3%) | 5/80 (6%) | 0.7 | 4/32 (13%) | 1/48 (2%) |
| NICU admission (n=131) | 4/36 (11%) | 27/95 (28%) | 0.04 | 14 (37%) | 13 (23%) |
| NICU length of stay, median days (IQR) | 8 (2.5–29.5) | 5 (2–12) | 0.8 | 9 (4–16) | 3 (2–5) |
Abbreviations: SD= standard deviation. IQR= interquartile range
The p-value compares UCTD and SLE pregnancies.
Outcomes presented for singleton pregnancies only. n=8 twin and n=1 triplet pregnancy occurred in 9 women with SLE
The rates of losses were calculated including only pregnancies enrolled prior to 20 weeks gestation. No losses occurred in pregnancies that enrolled after 20 weeks gestation.
Miscarriage= fetal loss before 20 weeks gestation.
Stillbirth= fetal loss at or after 20 weeks gestation.
Pregnancy and neonatal outcomes are reported for live births.
Among pregnancies with a gestational age at outcome of at least 20 weeks
Percentage includes only those who were induced.
Birthweight less than tenth percentile for gestational age.
Preterm Birth:
There was no statistically significant difference in the frequency of preterm delivery or early preterm delivery between SLE and UCTD patients. When adjusting for race and stratifying by the degree of SLE activity in pregnancy, however, women with high activity SLE had a higher frequency of preterm birth compared to women with UCTD (OR 3.85; 95% CI: 1.41, 10.53) and delivered, on average, 2.3 weeks earlier (β −2.30; 95% CI: −3.39, −1.21) (Table 4).
Table 4:
Pregnancy outcomes adjusted for maternal race.
| Outcome | Unadjusted | Adjusted for Race | ||
|---|---|---|---|---|
| High SLE vs. UCTD | Low SLE vs. UCTD | High SLE vs. UCTD | Low SLE vs. UCTD | |
|
| ||||
| β (95% CI) | β (95% CI) | β (95% CI) | β (95% CI) | |
| Gestational Age | −2.32 (−3.39, −1.25) | 0.08 (−0.90, 1.06) | −2.30 (−3.39, −1.21) | 0.09 (−0.91, 1.10) |
|
| ||||
| OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| Preterm Birth | 3.97 (1.54, 10.23) | 0.98 (0.37, 2.60) | 3.85 (1.41, 10.53) | 0.92 (0.32, 2.64) |
| Preeclampsia | 7.73 (2.09, 28.55) | 1.88 (0.47, 7.46) | 5.74 (1.50, 21.88) | 1.33 (0.32, 5.53) |
| SGA | 7.93 (2.13, 29.51) | 6.13 (1.71, 22.00) | 6.92 (1.82, 26.31) | 5.36 (1.46, 19.69) |
| C-section | 2.25 (0.99, 5.11) | 1.11 (0.53, 2.35) | 2.30 (0.98, 5.39) | 1.14 (0.53, 2.49) |
Preeclampsia:
Preeclampsia was more common in patients with SLE (20%) than in patients with UCTD (6%) (p= 0.02). When adjusting for race, the odds of having preeclampsia was not significantly higher in patients with SLE compared to UCTD (OR: 2.84; 95% CI: 0.79, 10.26), but it was significantly higher for SLE patients with high activity lupus than for patients with UCTD (OR: 5.74; 95% CI: 1.50, 21.88).
C-section & Induction:
The difference in the C-section rate between SLE and UCTD patients was not significant, and this did not change when stratifying by disease activity level. While the frequency of induction was similar among women with UCTD and SLE overall, the rate in pregnancies in women with UCTD (40%) was lower than in the pregnancies in women with active SLE (73%; p=0.0009). The reasons for induction were different, however, with 44% of pregnancies in women with UCTD induced routinely when they reached term, while 83% of pregnancies in women with high activity lupus were medically induced, either for the health of the mother or the infant.
SGA:
Significantly more infants born to mothers with SLE were small for gestational age, but there was no difference in infants born weighing <1500g. When adjusted for race, SLE continued to be associated with a higher risk of SGA infants, irrespective of disease activity, compared to UCTD (high SLE activity vs. UCTD: OR 6.92; 95% CI: 1.82–26.31 and low activity SLE vs. UCTD: OR 5.36; 95% CI: 1.46–19.69).
NICU:
The frequency of NICU admission was significantly higher for pregnancies in women with SLE (28% versus 11%; p-0.04). For both SLE and UCTD patients, the primary driver for admission to the NICU was preterm delivery, with at least half of babies delivered early spending some time in the NICU (UCTD: 50% of preterm versus 6% of term infants in the NICU (p=0.05); SLE 56% of preterm versus 18% of term infants in the NICU (p=0.0007)).
Multiples:
There were nine multiple gestation pregnancies, all in patients with SLE. Eight of the pregnancies were twin pregnancies, and one was a triplet pregnancy. Other than one twin stillbirth at 21 weeks, the remainder resulted in live births. The average gestational age for the twin pregnancies was 34 weeks, with four delivering at 36 weeks and one at 37 weeks. The triplet pregnancy delivered at 34 weeks. Six of the eight live births (twin and triplet gestations) were delivered by C-section, three for maternal health, two for infant health (growth retardation in triplet pregnancy and non-reassuring fetal heart tones in a twin pregnancy), and one for maternal history of prior C-section. Of the five pregnancies for which preeclampsia outcome was known, three of the pregnancies were complicated by preeclampsia.
APS:
Of the five pregnancies in women with SLE and APS, two resulted in terminations for medical reasons, one in a preterm delivery with preeclampsia at 35 weeks gestation, one with an SGA infant at term without preeclampsia, and one with a normal sized infant at term without preeclampsia.
Neonatal Lupus:
Only one infant born to a woman with anti-Ro antibodies developed complete heart block (1.1% of pregnancies in women with anti-Ro antibodies). The complete heart block was discovered at 19 weeks gestational age and was not reversed with dexamethasone; the mother was not taking hydroxychloroquine. Five additional patients received dexamethasone for early neonatal cardiac complications related to the presence of an anti-Ro antibody (Table 5). Two patients with fetal first-degree heart and two patients with findings concerning for carditis had reversal of these abnormalities with dexamethasone. One fetus was noted to have ectopy, which resolved with dexamethasone administration and caffeine avoidance. Of the five patients with early fetal cardiac changes, two patients took HCQ throughout pregnancy, two patients started HCQ after entering the second trimester, and one patient did not take HCQ. Overall, 80% of women with anti-Ro antibody took hydroxychloroquine during pregnancy (SLE 85%, UCTD 56%, p=0.009). The frequency of cardiac changes suggestive of neonatal lupus was similar between pregnancies in women with SLE (6.6%) and UCTD (5.5%).
Table 5.
Outcomes of patients who were prescribed dexamethasone for early fetal cardiac changes.
| Patient number | Maternal Diagnosis | Taking HCQ? | Abnormal cardiac finding(s) | Gestational weeks at abnormal finding | Outcome |
|---|---|---|---|---|---|
|
| |||||
| 1 | UCTD | No | Borderline 1st degree heart block | 23 | PR interval normalized |
| 2 | SLE | Started at 16 WGA | Borderline 1st degree heart block | 27 | PR interval normalized |
| 3 | SLE | Yes | Mild MR, Trivial TR | 19 | MR and TR resolved |
| 4 | SLE | Yes | MR, TR Ectopy | 25 | MR and TR improved to “mild” |
| 5 | SLE | Started at first consult (19 WGA) | Ectopy | 19 | Ectopy resolved (patient also stopped caffeine) |
| 6 | SLE | No | Complete heart block | 19 | Complete heart block |
Sensitivity analysis for multiple pregnancies in the same woman:
The comparisons between UCTD and SLE, as well as high and low SLE, were not different when 9 pregnancies in 5 women with UCTD and 29 pregnancies in 14 women with SLE were excluded from the analysis.
Discussion
Most pregnancies in women with UCTD resulted in a healthy baby and healthy mother, with similar outcomes to women with mild to inactive SLE but with fewer pregnancy and neonatal complications than pregnancies in women with active SLE. As has been shown previously, among pregnancies in women with SLE, most of the poor pregnancy and neonatal outcomes were related to SLE activity; given the low incidence of high activity disease in women with UCTD, a similar association is not found in this cohort1,5,29. In this cohort, collected prospectively in a university clinic in the southeastern part of the United States, there were significantly more Black patients with SLE compared to patients with UCTD. Because extensive data has demonstrated higher pregnancy complications in Black women than Caucasian women, pregnancy and neonatal outcomes were adjusted by maternal race, still revealing significant differences in rates of preterm birth, preeclampsia, and small for gestational age between pregnancies with UCTD and high activity SLE.
Pregnancy loss is difficult to assess in prospective pregnancy registries as pregnancy losses that occur prior to a woman signing informed consent are not included. For this reason, the miscarriage rate in this report is likely an underestimate of the actual rate. Stillbirth, when assessed for pregnancies enrolled prior to 20 weeks gestation, is more reliable. The rate of stillbirth in this study is comparable to other SLE and UCTD studies and, among women with active SLE, is higher than the general population16,24–26.
Our study is notably different from prior studies of pregnancies in women with UCTD by including only women with symptomatic UCTD that led to a rheumatologic diagnosis, by being located within the United States, and by including a racially diverse population. Multiple prior studies have primarily included women who were diagnosed with UCTD during pregnancy through a screening survey administered in the obstetric clinic. As these women had not presented for rheumatologic evaluation, it is likely that they had a lower level of rheumatic symptoms and, thus, less systemic inflammation and fewer pregnancy complications than our patients. We identified rates of preterm delivery, preeclampsia, and SGA that were all within the range of those reported in prior studies, thought the rate of preterm birth in women with UCTD was 17% in this cohort, at the top of the range of prior UCTD pregnancy reports (range 3–17%)13,14,26,27. The rate of preterm birth in women with SLE in this cohort is comparable to previously published reports as well1,28–30.
The rate of SGA infants was significantly higher in patients with SLE compared to those with UCTD. Interestingly, this finding was irrespective of disease activity and persisted after adjusting for race. The frequency of SGA in this cohort of women with UCTD (7%) was within the range of prior reports (5% to 30%) and the 10% in the general population. Infants are most commonly born small for gestational age due to placental insufficiency, which has been associated with SLE pregnancies31. In a comparison of placental histology between women with SLE and healthy women, the placentas from women with SLE, with and without preeclampsia, had abnormalities resembling those found in preeclampsia without SLE, including maternal vasculitis, laminar decidual necrosis, maternal-fetal interface hemorrhage, and non-occlusive fetal thrombotic vasculopathy32. The low frequency of SGA in the UCTD population in this study suggest that the placental pathology seen in women with SLE may not be common in women with UCTD.
The infant outcomes for women with UCTD in this cohort were very reassuring with a low rate of NICU admission, normal APGAR scores, and no increase in SGA rate. The primary cause for NICU admission was preterm delivery, as is seen in the general population.
The rate of complete heart block in our cohort of pregnancies in women with anti-Ro antibodies was as expected with only one infant (1.1%) developing this complication. The majority of patients in both groups were taking HCQ. Four retrospective cohort studies and one prospective trial show that HCQ may reduce the risk of developing complete heart block, so the high rate of HCQ use in our cohort may be responsible for keeping the number of cases with complete heart block low33–37. The use of dexamethasone in the patients with early signs of heart block may have prevented those fetuses from having progression of heart block, though the literature on fluorinated steroids (dexamethasone, betamethasone) to prevent or reverse early heart block due to anti-Ro antibodies is conflicting without randomized studies available38.
The primary limitation of our study is the small number of pregnancies in the UCTD cohort and the absence of a healthy control group for comparison. Additionally, this data is derived from a single center and managed by an expert in reproductive rheumatology, which limits generalizability.
The findings reported here allow us to counsel women with UCTD that pregnancy is frequently safe with low rates of pregnancy and neonatal complications. More than half of the women with UCTD in this cohort were maintained on medications throughout pregnancy, including HCQ and aspirin, which may have decreased their risk for poor pregnancy outcomes. As this was not a randomized trial, the impact of the medications on pregnancy outcomes is unclear; this study does not suggest that women with UCTD without rheumatologic management will have similar outcomes. Women with SLE had higher rates of pregnancy complications than women with UCTD, with particularly high rates of preterm birth and preeclampsia in women with moderate to severely active SLE during pregnancy. An important finding is that our patients with low activity SLE had similar pregnancy outcomes to those of UCTD patients, with the exception of an elevated rate of SGA. This study suggests that women with UCTD, as well as women with well-controlled SLE, can be reassured that they have a very high likelihood of having a healthy pregnancy and baby.
Significance and Innovations.
This report is the first analysis of pregnancy outcomes of a cohort of women with UCTD in the United States.
The findings herein demonstrate that pregnancy outcomes of women with UCTD are similar to those of otherwise healthy women.
The findings also demonstrate that women with well-controlled SLE at the start of pregnancy have good pregnancy outcomes.
This study suggests that women with UCTD and women with well-controlled SLE can be counseled that they have a very high likelihood of having a healthy pregnancy and baby.
Grants/Financial Support:
Amanda Eudy is supported by an NIH NCATs award (1KL2TR002554). The Duke Autoimmunity in Pregnancy Registry has been funded by Building Interdisciplinary Careers in Women’s Health (K12HD043446), an Arthritis Foundation Arthritis Investigator Award, and the Duke Health Private Diagnostic Clinic ENABLE award
Footnotes
Disclosures: Dr. Clowse serves as a consultant for Astra-Zeneca and UCB, Inc. She receives independent medical education funding from GlaxoSmithKline.
Publisher's Disclaimer: This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi:10.1002/ACR.24644
References
- 1.Petri M Hopkins Lupus Pregnancy Center: 1987 to 1996. Rheum Dis Clin North Am. 1997;23:1–13. [DOI] [PubMed] [Google Scholar]
- 2.Lateef A, Petri M. Systemic lupus erythematosus and pregnancy [review]. Rheum Dis Clin North Am. 2017;43:215–226. [DOI] [PubMed] [Google Scholar]
- 3.Pastore D, Costa M, Parpinelli M, Surita F. A critical review on obstetric follow-up of women affected by systemic lupus erythematosus [review]. Rev Bras Ginecol e Obs / RBGO Gynecol Obstet. 2018;40:209–224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Eudy AM, Jayasundara M, Haroun T, Neil L, James AH, Clowse MEB. Reasons for cesarean and medically indicated deliveries in pregnancies in women with systemic lupus erythematosus. Lupus. 2018;27:351–356. [DOI] [PubMed] [Google Scholar]
- 5.Clowse MEB, Magder L, Petri M. Cyclophosphamide for lupus during pregnancy. Lupus. 2005;14:593–597. [DOI] [PubMed] [Google Scholar]
- 6.Buyon JP, Kim MY, Guerra MM, et al. Predictors of pregnancy outcomes in patients with lupus: A cohort study. Ann Intern Med. 2015;163:153–163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kallenberg CGM. Overlapping syndromes, undifferentiated connective tissue disease, and other fibrosing conditions. Curr Opin Rheumatol. 1992;4:837–842. [PubMed] [Google Scholar]
- 8.Mosca M, Neri R. Undifferentiated connective diseases (UCTD): a review of the literature and a proposed preliminary classification criteria [review]. Clin Exp Rheumatol. 1999;17:615–620. [PubMed] [Google Scholar]
- 9.Castellino G, Capucci R, Bernardi S, Padovan M, Giacuzzo S, Pivato E, et al. Pregnancy in patients with undifferentiated connective tissue disease: A prospective case-control study. Lupus. 2011;20:1305–1311. [DOI] [PubMed] [Google Scholar]
- 10.Mosca M, Tani C, Vagnani S, Carli L, Bombardieri S. The diagnosis and classification of undifferentiated connective tissue diseases [review]. J Autoimmun. 2014;48–49:50–52. [DOI] [PubMed] [Google Scholar]
- 11.Mosca M, Neri R, Strigini F, Carmignani A, Totti D, Tavoni A, et al. Pregnancy outcome in patients with undifferentiated connective tissue disease: A preliminary study on 25 pregnancies. Lupus. 2002;11:304–307. [DOI] [PubMed] [Google Scholar]
- 12.Spinillo A, Beneventi F, Epis OM, Montanari L, Mammoliti D, Ramoni V, et al. The effect of newly diagnosed undifferentiated connective tissue disease on pregnancy outcome. Am J Obstet Gynecol. 2008;199:1–6. [DOI] [PubMed] [Google Scholar]
- 13.Spinillo A, Beneventi F, Ramoni V, Caporali R, Locatelli E, Simonetta M, et al. Prevalence and significance of previously undiagnosed rheumatic diseases in pregnancy. Ann Rheum Dis. 2012;71:918–923. [DOI] [PubMed] [Google Scholar]
- 14.Spinillo A, Beneventi F, Locatelli E, Ramoni V, Caporali R, Alpini C, et al. The impact of unrecognized autoimmune rheumatic diseases on the incidence of preeclampsia and fetal growth restriction: A longitudinal cohort study. BMC Pregnancy Childbirth. 2016;16:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.March of Dimes. Miscarriage. 2017. Available at: https://www.marchofdimes.org/complications/miscarriage.aspx#.
- 16.MacDorman MF, Gregory ECW. Fetal and perinatal mortality: United States, 2013. Natl Vital Stat Rep. 2015;64:1–24. [PubMed] [Google Scholar]
- 17.Martin JA, Hamilton BE, Osterman MJK, Driscoll AK, Matthews TJ. Births: Final Data for 2015. Natl Vital Stat Rep. 2017;66:1–70. [PubMed] [Google Scholar]
- 18.Ananth C V, Keyes KM, Wapner RJ. Pre-eclampsia rates in the United States, 1980–2010: age-period-cohort analysis. BMJ. 2013;347:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Tan EM, Cohen AS, Fries JF, et al. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 1982;25:1271–1277. [DOI] [PubMed] [Google Scholar]
- 20.Hochberg MC. Updating the American College of Rheumatology Revised Criteria for the Classification of Systemic Lupus Erythematosus [letter]. 1997;40:1725. [DOI] [PubMed] [Google Scholar]
- 21.Petri M, Orbai AM, Alarcon GS, Gordon C, Merrill JT, Fortin PR, et al. Derivation and Validation of Systemic Lupus International. Arthritis Rheum. 2012;64:2677–2686. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Buyon JP, Kalunian KC, Ramsey-Goldman R, Petri MA, Lockshin MD, Ruiz-Irastorza G, et al. Assessing disease activity in SLE patients during pregnancy. Lupus. 1999;8:677–684. [DOI] [PubMed] [Google Scholar]
- 23.Oken E, Kleinman KP, Rich-Edwards J, Gillman MW. A nearly continuous measure of birth weight for gestational age using a United States national reference. BMC Pediatr. 2003;3:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Clowse MEB, Magder LS, Witter F, Petri M. Early Risk Factors for Pregnancy Loss in Lupus. Obstet Gynecol. 2006;107:293–299. [DOI] [PubMed] [Google Scholar]
- 25.Smyth A, Oliveira GHM, Lahr BD, Bailey KR, Norby SM, Garovic VD. A systematic review and meta-analysis of pregnancy outcomes in patients with systemic lupus erythematosus and lupus nephritis. Clin J Am Soc Nephrol. 2010;5:2060–2068. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Radin M, Schreiber K, Cecchi I, Bortoluzzi A, Crisafulli F, de Freitas CM, et al. A multicentre study of 244 pregnancies in undifferentiated connective tissue disease: maternal/fetal outcomes and disease evolution [published online ahead of print, 2020 Jan 14]. Rheumatology. 2020:1–7. [DOI] [PubMed] [Google Scholar]
- 27.Zucchi D, Tani C, Monacci F, Elefante E, Carli L, Parma A, et al. Pregnancy and undifferentiated connective tissue disease: outcome and risk of flare in 100 pregnancies [published online ahead of print, 2019 Oct 8]. Rheumatology. 2019:1–5. [DOI] [PubMed] [Google Scholar]
- 28.Chakravarty EF, Colón I, Langen ES, Nix DA, El-Sayed YY, Genovese MC, et al. Factors that predict prematurity and preeclampsia in pregnancies that are complicated by systemic lupus erythematosus. Am J Obstet Gynecol. 2005;192:1897–1904. [DOI] [PubMed] [Google Scholar]
- 29.Ko HS, Ahn HY, Jang DG, Choi SK, Park YG, Park IY, et al. Pregnancy outcomes and appropriate timing of pregnancy in 183 pregnancies in Korean patients with SLE. Int J Med Sci. 2011;8:577–583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kroese SJ, Abheiden CNH, Blomjous BS, van Laar JM, Derksen RWHM, Bultink IEM, et al. Maternal and Perinatal Outcome in Women with Systemic Lupus Erythematosus: A Retrospective Bicenter Cohort Study. J Immunol Res. 2017;2017:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Hanly JG, Gladman DD, Rose TH, Laskin CA, Urowitz MB. Lupus pregnancy. A prospective study of placental changes. Arthritis Rheum. 1988;31:358–366. [DOI] [PubMed] [Google Scholar]
- 32.Marder W, Knight JS, Kaplan MJ, Somers EC, Zhang X, O’Dell AA, et al. Placental histology and neutrophil extracellular traps in lupus and pre-eclampsia pregnancies. Lupus Sci Med. 2016;3:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Izmirly PM, Kim MY, Llanos C, Le PU, Guerra MM, Askanase AD, et al. Evaluation of the risk of anti-SSA/Ro-SSB/La antibody-associated cardiac manifestations of neonatal lupus in fetuses of mothers with systemic lupus erythematosus exposed to hydroxychloroquine. Ann Rheum Dis. 2010;69:1827–1830. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Izmirly PM, Costedoat-Chalumeau N, Pisoni CN, Khamashta MA, Kim MY, Saxena A, et al. Maternal use of hydroxychloroquine is associated with a reduced risk of recurrent anti-SSA/ro-antibody-associated cardiac manifestations of neonatal lupus. Circulation. 2012;126:76–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Tunks RD, Clowse MEB, Miller SG, Brancazio LR, Barker PCA. Maternal autoantibody levels in congenital heart block and potential prophylaxis with antiinflammatory agents. Am J Obstet Gynecol. 2013;208:64.e1–64.e7. [DOI] [PubMed] [Google Scholar]
- 36.Martínez-Sánchez N, Pérez-Pinto S, Robles-Marhuenda Á, Arnalich-Fernández F, Cameán MM, Zalvide EH, et al. Obstetric and perinatal outcome in anti-Ro/SSA-positive pregnant women: a prospective cohort study. Immunol Res. 2017;65:487–494. [DOI] [PubMed] [Google Scholar]
- 37.Izmirly P, Kim M, Costedoat-chalumeau N, Friedman D, Saxena A, Copel J, et al. The prospective open label preventive approach to congenital heart block with hydroxychloroquine (PATCH) study demonstrates a reduction in the recurrence rate of advanced block. Arthritis Rheumatol. 2019;71 (suppl 10). https://acrabstracts.org/abstract/the-prospective-open-label-preventive-approach-to-congenital-heart-block-with-hydroxychloroquine-patch-study-demonstrates-a-reduction-in-the-recurrence-rate-of-advanced-block/. Accessed May 14, 2020. [Google Scholar]
- 38.Saxena A, Izmirly PM, Mendez B, Buyon JP, Friedman DM. Prevention and treatment in utero of autoimmune associated congenital heart block [review]. Cardiol Rev. 2014;22:263–267. [DOI] [PMC free article] [PubMed] [Google Scholar]