Laboratory and clinical features of pregnant women with antiphospholipid syndrome and neonatal outcome

Authors


Abstract

Objective

To evaluate the relationship between the antiphospholipid profile and clinical characteristics of pregnant women with antiphospholipid syndrome (APS) and neonatal outcome.

Methods

We retrospectively considered 109 treated pregnancies of 93 patients with primary APS and reviewed the medical records of their 111 infants. Neonatal outcome was assessed using the following variables: weeks of gestational age at delivery, percentiles of birth weight, Apgar score at 5 minutes, need for cardiopulmonary resuscitation in the delivery room, time in the neonatal intensive care unit, infections, and other neonatal complications. Univariate statistical analysis was performed to evaluate the relationship between APS maternal features and neonatal outcome parameters.

Results

When maternal APS features and variables of infant outcome were analyzed, it was evident that lupus anticoagulant (LAC), triple antiphospholipid positivity, and history of vascular thrombosis were significantly associated with some parameters of a poor infant outcome. History of pregnancy morbidity alone was, instead, significantly correlated to the variables of favorable neonatal outcome.

Conclusion

There seems to be more than one kind of pregnant woman with APS. Even when treated with a second-line therapy plan, mothers with LAC and/or triple antiphospholipid positivity and/or previous thromboembolism seem to have a high probability of poor neonatal outcome, whereas those with pregnancy morbidity alone, treated with conventional drugs, seem to have a high probability of favorable outcome.

INTRODUCTION

The outcome of infants born to mothers with primary antiphospholipid syndrome (APS) has been the object of several studies (1–6). Disorders related to prematurity are common (1–4), but neonatal APS manifestations are quite rare (5, 6). Maternal APS features predicting neonatal outcome have not as yet been clearly established. We recently reported (7, 8) that different subsets of obstetric patients with APS can be identified on the basis of their clinical and laboratory characteristics. It was found, in fact, that women with previous thromboembolism and/or triple antiphospholipid positivity (lupus anticoagulant [LAC], plus IgG anticardiolipin [aCL], plus IgG anti–β2-glycoprotein I [anti–β2-GPI] antibodies) are at high risk for other unsuccessful pregnancies and thrombosis, whereas those with single (LAC or aCL or anti–β2-GPI) or double positivity (aCL plus anti–β2-GPI) and without thrombosis are at a low risk for complications during their next pregnancy (8).

It seemed important to evaluate the relationship between the laboratory profile and clinical features of mothers with APS and the outcome of their infants. Several parameters of neonatal outcome of a large group of newborns were therefore analyzed in relationship with the antiphospholipid profile and clinical characteristics of their mothers diagnosed with primary APS and prescribed antithrombotic therapy during their pregnancies.

PATIENTS AND METHODS

Study population.

We retrospectively considered the pregnancies of women who attended our Center and were diagnosed with primary APS on the basis of the International Consensus Statement classification criteria (9). These included history of thrombosis (venous, arterial, or small vessel) and/or pregnancy morbidity (≥1 fetal deaths occurring at or beyond the tenth week, and/or premature delivery occurring before the thirty-fourth week due to severe preeclampsia or placental insufficiency, and/or ≥3 miscarriages occurring before the tenth week of pregnancy). In accordance with the laboratory criteria of the Consensus Statement, LAC and/or IgG/IgM aCL (medium/high titers) and/or IgG/IgM anti–β2-GPI (medium/high titers) antibodies were detected and confirmed at ≥12 weeks apart to verify persistence. Exclusion criteria were clinical and laboratory data confirming any other definite autoimmune systemic disease. All of the pregnant patients were treated with conventional protocol treatments, including heparin ± low-dose aspirin (10, 11). A second-line therapy plan (conventional treatment plus plasma exchange or immunoabsorption) was considered when some pregnancy complications occurred, including intrauterine growth restriction or a significant (>20% of the basal value) diminution in platelet count during the first or second trimester of pregnancy in the absence of antibodies to platelets and/or to platelet factor 4–heparin complex (12). Informed consent statements were obtained from all the eligible subjects to consult their and their infants' charts.

Neonatal outcome was assessed on the basis of the following parameters: weeks of gestational age at delivery, percentiles of birth weight, 5-minute Apgar score, need for cardiopulmonary resuscitation in the delivery room, time spent in the neonatal intensive care unit (NICU), thrombotic events, infections, other neonatal complications (e.g., respiratory distress syndrome, bronchopulmonary dysplasia, pneumothorax, anemia, retinopathy, etc.), and perinatal mortality.

Antiphospholipid antibody assays.

We determined aCL and anti–β2-GPI antibodies using enzyme-linked immunosorbent assay following the minimal requirements proposed by the European Forum on Antiphospholipid Antibodies (13, 14). According to internationally accepted recommendations, LAC was detected using dilute Russell's viper venom and dilute activated partial thromboplastin times as screening tests (15).

Statistical analysis.

Statistical analysis was performed using the Statistical Package for the Social Sciences, version 14.0 (SPSS, Chicago, IL). The association between categorical variables of neonatal outcome and laboratory and clinical maternal features was estimated using the Pearson's chi-square or Fisher's tests; the odds ratios with 95% confidence intervals were calculated. The Mann-Whitney or Kruskal-Wallis tests were employed to analyze neonatal outcome variables measured along a continuous scale. The correlation coefficients among continuous neonatal parameters in the women with negative outcome variables were calculated using Spearman's rho. P values less than 0.05 were considered statistically significant.

RESULTS

We followed 155 pregnancies of 134 women diagnosed with primary APS from August 1991 to September 2008. A total of 128 women gave birth to 147 live infants during that time. Pregnancy loss of an unknown origin occurred in 8 of these patients (5.1%), including fetal loss in 7 and early miscarriage in 1. The medical records of 111 neonates born to 93 of these women were reviewed because those of the other 36 infants were untraceable or unavailable.

Maternal APS features.

The mean ± SD age of the 93 women was 33 ± 4.1 years (range 21–42 years). Of these patients, 74 (79.6%) had a history of pregnancy morbidity alone: 57 (77.0%) of fetal loss, 1 (1.4%) of premature birth, and 16 (21.6%) of early miscarriage. Previous thrombosis alone was found in 9 (9.7%) of the women: 4 (4.3%) venous and 5 (5.4%) arterial. Previous pregnancy morbidity together with a history of thromboembolism was reported in 10 (10.8%) of the patients. Seventy-nine (84.9%) were positive to aCL (59 to the IgG isotype, 11 to IgM, and 9 to both), 67 (72.7%) to anti–β2-GPI (47 to the IgG isotype, 13 to IgM, and 7 to both), and 14 (15.1%) to LAC. Single antiphospholipid positivity (LAC or IgG/IgM aCL or IgG/IgM anti–β2-GPI) was found in 38 patients (41.1%), double positivity consisting only of IgG/IgM aCL plus IgG/IgM anti–β2-GPI in 43 patients (46.2%), and triple positivity (IgG/IgM aCL plus IgG/IgM anti–β2-GPI plus LAC) in 12 patients (12.9%). Of the 83 patients (89.2%) prescribed conventional therapy (i.e., prophylactic or therapeutic heparin ± low-dose aspirin), 74 (89.2%) presented with pregnancy morbidity alone, 6 (7.2%) with vascular thrombosis alone, and 3 (3.6%) presented with both. Plasma exchange or immunoabsorption was added to conventional treatment in 10 women (10.8%): 7 (70%) with vascular thrombosis and pregnancy morbidity and 3 (30%) with vascular thrombosis alone. The relationship between clinical features and antiphospholipid antibodies in pregnant women with APS is outlined in Table 1.

Table 1. The relationship between clinical features and aPL profiles in pregnant women with APS*
 NIgG/IgM aCLIgG/IgM anti–β2-GPILACaPL positivity
SingleDoubleTriple
  • *

    Values are the number (percentage) unless otherwise indicated. aPL = antiphospholipid antibody; APS = antiphospholipid syndrome; aCL = anticardiolipin antibodies; anti–β2-GPI = anti–β2-glycoprotein I antibodies; LAC = lupus anticoagulant.

Thrombosis98 (88.9)4 (44.4)3 (33.3)6 (66.7)0 (0)3 (33.3)
Pregnancy morbidity7462 (83.8)53 (71.6)3 (4.1)31 (41.9)42 (56.8)1 (1.4)
Both criteria109 (90.0)10 (100.0)8 (80.0)1 (10.0)1 (10.0)8 (80.0)

Neonatal outcome parameters.

The mean ± SD gestational age at delivery was 36 ± 3.0 weeks (range 24–40 weeks), the mean ± SD birth weight was 49.15 ± 23.5 percentiles (range 3–90 percentiles), and the mean ± SD 5-minute Apgar score was 9.3 ± 1.0 (range 5–10). Eighteen infants (16.2%) received cardiopulmonary resuscitation in the delivery room, and 32 (28.8%) were admitted to the NICU for a mean ± SD time of 21.4 ± 22.0 days (range 1–83 days). Infections, including 6 cases of sepsis, occurred in 9 newborns (8.1%), whereas respiratory distress syndrome (associated with patent ductus arteriosus in 7 cases, pneumothorax in 5, anemia in 2, patent foramen ovale in 1, bronchopulmonary dysplasia in 1, and retinopathy in 1) was diagnosed in 29 newborns (26.1%). Complications related to prematurity (2 respiratory distress syndrome and 1 sepsis) caused death in 3 neonates (2.7%) at 1, 8, and 243 days of age, delivered at 24, 25, and 30 weeks of gestation, respectively. No thrombotic events were observed in the perinatal period.

Association between maternal and neonatal variables.

Statistical analysis data are outlined in Tables 2, 3, 4, and 5. The results are subdivided into 4 parts: Tables 2 and 3 show the association between maternal antiphospholipid profiles and continuous and categorical neonatal outcome variables, respectively, whereas Tables 4 and 5 indicate the association between maternal clinical APS features/antithrombotic treatments and continuous and categorical neonatal outcome variables, respectively.

Table 2. The association between maternal aPL profiles and continuous neonatal outcome variables*
 IgG aCLIgM aCLIgG anti–β2-GPIIgM anti–β2-GPILACaPL positivity
No (n = 25)Yes (n = 68)No (n = 73)Yes (n = 20)No (n = 39)Yes (n = 54)No (n = 73)Yes (n = 20)No (n = 79)Yes (n = 14)Single (n = 38)Double (n = 43)Triple (n = 12)P
  • *

    Values are the mean unless otherwise indicated. Only significant results are indicated. NICU = neonatal intensive care unit. See Table 1 for additional definitions.

  • For triple aPL positivity.

  • Significant association between LAC and a poor neonatal outcome (P < 0.001).

Gestational weeks36.235.936.135.536.335.836.135.736.632.436.836.232.00.000
Weight percentiles53.347.350.146.048.749.549.846.851.038.852.450.331.50.017
5-minute Apgar score9.29.49.49.09.39.39.49.19.58.49.59.38.30.002
Stay in the NICU4.66.85.58.04.87.25.58.13.719.63.04.226.50.000
Table 3. The association between maternal aPL profiles and categorical neonatal outcome variables*
 IgG aCLIgM aCLIgG anti–β2-GPIIgM anti–β2-GPILACaPL positivity
No (n = 25)Yes (n = 68)No (n = 73)Yes (n = 20)No (n = 39)Yes (n = 54)No (n = 73)Yes (n = 20)No (n = 79)Yes (n = 14)OR (95% CI)Single (n = 38)Double (n = 43)Triple (n = 12)P
  • *

    Values are the percentage unless otherwise indicated. OR = odds ratio; 95% CI = 95% confidence interval. See Table 1 for additional definitions.

  • OR (95% CI) values were not significant.

  • Significant association with a poor neonatal outcome.

  • §

    Included respiratory distress syndrome, patent ductus arteriosus, pneumothorax, anemia, patent foramen oval, bronchopulmonary dysplasia, and retinopathy.

Cardiopulmonary resuscitation17.615.515.418.514.018.016.415.312.735.23.7 (1.2–11.9)8.816.641.60.023
Infections2.910.37.111.16.09.88.27.63.135.216.5 (3.6–75.7)4.43.741.60.000
Other disorders§41.138.939.240.734.044.241.134.630.888.210.8 (3.4–34.9)35.531.491.60.000
Death0.03.92.33.70.04.83.50.00.017.6P = 0.0030.00.023.00.000
Table 4. The association between maternal clinical APS features/antithrombotic treatments and continuous neonatal outcome variables*
 Previous thrombosisPrevious pregnancy morbidityBoth criteriaPregnancy treatments
No (n = 84)Yes (n = 9)PNo (n = 19)Yes (n = 74)PNo (n = 83)Yes (n = 10)PConventional (n = 82)Second line (n = 11)P
  • *

    Values are the mean unless otherwise stated. APS = antiphospholipid syndrome; ns = not significant; NICU = neonatal intensive care unit.

  • Significant association with a poor neonatal outcome.

  • Significant association with a favorable neonatal outcome.

Gestational weeks36.234.00.0333.336.7< 0.00136.432.4< 0.00136.631.4< 0.001
Weight percentiles49.744.0ns38.351.90.0251.032.20.0151.828.9< 0.001
5-minute Apgar score9.39.1ns8.79.5< 0.0019.48.1< 0.0019.57.8< 0.001
Stay in the NICU5.015.10.0117.33.2< 0.0014.619.70.034.022.00.002
Table 5. The association between maternal clinical APS features/antithrombotic treatments and categorical neonatal outcome variables*
 Previous thrombosisPrevious pregnancy morbidityBoth criteriaPregnancy treatments
No (n = 84)Yes (n = 9)OR (95% CI)No (n = 19)Yes (n = 74)OR (95% CI)No (n = 83)Yes (n = 10)OR (95% CI)Conventional (n = 82)Second line (n = 11)OR (95% CI)
  • *

    Values are the percentage unless otherwise indicated. APS = antiphospholipid syndrome; OR = odds ratio; 95% CI = 95% confidence interval; ns = not significant.

  • Significant association with a favorable neonatal outcome.

  • Significant association with a poor neonatal outcome.

  • §

    Included respiratory distress syndrome, patent ductus arteriosus, pneumothorax, anemia, patent foramen oval, bronchopulmonary dysplasia, and retinopathy.

Cardiopulmonary resuscitation15.225.0ns39.110.20.22 (0.1–0.7)12.054.58.8 (2.3–33.3)12.246.26.1 (1.8–21.4)
Infections6.125.05.87 (1.2–28.0)26.13.40.9 (0.02–0.41)6.027.35.87 (1.2–28.0)4.138.514.7 (3.3–65.8)
Other disorders§36.466.7ns82.628.40.1 (0.1–0.4)33.0100.010.03 (2.4–41.1)32.792.313.9 (3.5–55.3)
Death2.08.3ns13.00.0P = 0.0081.018.2P = 0.0260.023.1P = 0.001

The analysis of the association between neonatal outcome variables and maternal APS features uncovered several important results. The mean gestational age at delivery was significantly shorter in the patients with LAC and/or triple antiphospholipid positivity, a history of thrombosis alone or associated with pregnancy morbidity, and in those treated with second-line therapy, but it was significantly longer in those with pregnancy morbidity alone. The mean percentile birth weight value was significantly lower in the infants of patients with triple positivity, thrombosis, and pregnancy morbidity and in those on second-line treatment, but it was significantly higher in the newborns of women with pregnancy morbidity alone. The mean 5-minute Apgar score was significantly lower in the infants of patients with LAC and/or triple positivity, thrombosis associated with pregnancy morbidity, and in those on second-line therapy, but it was significantly higher in the neonates of mothers with pregnancy morbidity alone. The frequency of cardiopulmonary resuscitation in the delivery room was significantly higher in the infants of patients with LAC and/or triple positivity, thrombosis associated with pregnancy morbidity, and in those treated with second-line therapy, but it was significantly lower in mothers with pregnancy morbidity alone. The mean time spent in the NICU was significantly longer in the infants of patients with LAC and/or triple positivity, thrombosis alone, thrombosis associated with pregnancy morbidity, and in those on second-line therapy, but it was significantly shorter in the neonates of patients with pregnancy morbidity alone.

The frequency of infections was significantly higher in the neonates of women with LAC and/or triple positivity, thrombosis alone or associated with pregnancy morbidity, and in those on second-line therapy, but it was significantly lower in the newborns of women with pregnancy morbidity alone. The frequency of other neonatal complications considered in this study was significantly higher in the infants of patients with LAC and/or triple positivity, thrombosis associated with pregnancy morbidity, and in those on second-line therapy, but it was significantly lower in the infants of women with pregnancy morbidity alone. The frequency of neonatal death was significantly higher in the mothers with LAC and/or triple positivity, thrombosis associated with pregnancy morbidity, and on second-line therapy, but it was significantly lower (absent) in the patients with pregnancy morbidity alone.

The correlation coefficients among the continuous neonatal variables in the women with negative outcome features (triple antiphospholipid antibody positivity, previous vascular thrombosis, and second-line treatment) showed that there was a significant correlation between gestational weeks and the 5-minute Apgar score (P < 0.05) and between gestational weeks and the time infants spent in the NICU (P < 0.05) (Figure 1).

Figure 1.

Significant correlation coefficients among continuous neonatal variables in women with some features of negative outcomes (dotted bar; n = 9) compared with those without (lined bar; n = 102). NICU = neonatal intensive care unit.

DISCUSSION

To our knowledge, this is the first study showing a relationship between laboratory and clinical features of women with primary APS and several parameters of neonatal outcome. The study group was made up of 128 selected patients affected with different clinical features of APS (vascular thrombosis alone, pregnancy morbidity alone, or both together), diagnosed with a variety of antiphospholipid profiles, and administered conventional or second-line therapy during their pregnancies. We retrospectively reviewed the data of 111 of 147 newborns because those of the other 36 infants (24.5%) were untraceable or unavailable.

In agreement with other works (1–3) and a previous report (4), the neonatal outcome of the infants born to the women with APS was characterized by a high rate of complications, such as respiratory distress syndrome, infections including sepsis, bronchopulmonary dysplasia, pneumothorax, retinopathy, anemia, patent ductus arteriosus, and patent foramen ovale usually associated with prematurity (16).

The statistical analysis of the relationship between maternal APS features and the variables of neonatal outcome that were considered showed that there is a significant association between LAC/triple antiphospholipid positivity and poor infant outcome. In 12 (85.7%) of 14 cases, however, LAC was not alone but part of a triple antibody positivity profile that should be considered the real antiphospholipid marker of poor neonatal outcome, also associated with neonatal death. A history of vascular thrombosis, especially of the arterial type, alone or associated with pregnancy morbidity was, likewise, significantly associated with poor infant outcome. These findings are in agreement with those of previous studies (7, 8) reporting that triple antiphospholipid positivity and previous thrombosis were predictive of unsuccessful new pregnancies in women with APS. It is interesting that triple antiphospholipid positivity, previous vascular thrombosis, or second-line treatment were frequently all contemporaneously present in the same patients. It is quite possible that these variables are related to one another, but a multivariate analysis was not possible due to the low number of cases studied. An analysis of the correlation coefficients among the continuous neonatal variables in the women with negative outcome features showed that there was a significant correlation between gestational weeks and the 5-minute Apgar score, and between gestational weeks and the time the infants spent in NICU. In agreement with our previous reports (7, 8), this study indicates that women with pregnancy morbidity alone and treated with conventional therapy are likely to have a favorable neonatal outcome.

In conclusion, it seems that there is more than one subset of women with APS. Those patients with triple antiphospholipid positivity and/or previous thrombosis seem to have a high probability of poor neonatal outcome, even when they are treated with a second-line therapy plan during pregnancy. Those mothers with pregnancy morbidity alone treated with conventional drugs seem to have, instead, a high probability of favorable outcome. Prenatal counseling and well-established therapy recommendations for women with APS planning a pregnancy could be predisposed if these findings are confirmed by prospective, multicentric large-scale studies.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. A. Ruffatti had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. A. Ruffatti, Calligaro, Hoxha.

Acquisition of data. A. Ruffatti, Calligaro, Hoxha, Trevisanuto, A. T. Ruffatti, Gervasi, Cuffaro.

Analysis and interpretation of data. A. Ruffatti, Calligaro, Hoxha, Pengo, Punzi.

Acknowledgements

The authors are grateful to Dr. Elisa Salvan for the statistical calculations and to Mrs. Linda Inverso Moretti for editing the English version of this manuscript.

Ancillary