Over the past 2 decades, the antiphospholipid syndrome (APS) has become a major research font in rheumatology. Antiphospholipid antibodies (aPL) are now recognized as representing the most frequent acquired risk factor for thrombophilia and as being a treatable cause of pregnancy loss (1). With the widespread use of tests to detect aPL, rheumatologists and obstetricians frequently have to make decisions regarding the consequences of positive test results in women who wish to conceive or are pregnant. We review herein studies on the management of pregnancy in women with aPL and provide guidelines for treatment, despite our conclusion that the many shortcomings of currently available data hamper evidence-based recommendations.
The obstetric APS
In the early 1980s, retrospective studies of patients with systemic lupus erythematosus (SLE) established associations between aPL and thrombosis (2), fetal loss, (3) and thrombocytopenia (4). This clinico-serologic entity was termed APS (5). Soon, APS in individuals without another autoimmune disease was described as primary APS (6). Criteria for the diagnosis of APS were proposed in 1987 (7), with (recurrent) fetal loss being the obstetric criterion. Terms such as recurrent pregnancy loss (8) and recurrent spontaneous abortion (9) were also used. This confusing nomenclature refers to the traditional classification of pregnancy loss, with losses prior to 20 weeks of gestation classified as abortions, and death in utero thereafter classified as stillbirth or fetal loss.
From a pathophysiologic point of view, this classification is inaccurate. Recent advances in reproductive biology and observations in normal early pregnancies allow pregnancy loss to be categorized into at least 3 developmental periods (10). The preembryonic period lasts from conception through to the beginning of the fifth menstrual week. This is followed by the embryonic period, which lasts from 5 weeks through the ninth menstrual week. The fetal period begins at 10 menstrual weeks (i.e., 70 days from conception) and extends until the time of delivery. In unselected obstetric patients, 10–15% of recognized pregnancies end in loss, with >85% of losses occurring during the (pre)embryonic periods (11, 12). The rate of fetal loss after 14 weeks of gestation is 2% (11). Chromosomal abnormalities of the conceptus account for more than half of sporadic (pre)embryonic losses, and, in many cases, a visible embryo never forms. In contrast, genetic abnormalities of the conceptus are less common in women with at least 3 consecutive pregnancy losses (12). In women with ≥3 spontaneous abortions, a normal embryonic ultrasound beyond 6–8 weeks of gestation provides a favorable prognosis with a ≥80% chance for a live birth (13). The preliminary classification criteria for definite APS published in 1999 (14) (Table 1) include this refined classification of pregnancy loss and recognize that a preterm live birth accompanied by severe preeclampsia or severe placental insufficiency is comparable with a loss late in pregnancy.
Table 1. Preliminary classification criteria for APS*
Definite antiphospholipid syndrome (APS) is considered to be present if at least 1 of the clinical and 1 of the laboratory criteria are met.
a.One or more clinical episodes of arterial, venous, or small-vessel thrombosis in any tissue or organ, AND
b.Thrombosis confirmed by imaging or Doppler studies or histopathology, with the exception of superficial venous thrombosis, AND
c.For histopathologic confirmation, thrombosis should be present without significant evidence of inflammation in the vessel wall.
a.One or more unexplained deaths of a morphologically normal fetus at or beyond the 10th week of gestation, with normal fetal morphology documented by ultrasound or by direct examination of the fetus, OR
b.One or more premature births of a morphologically normal neonate at or before the 34th week of gestation because of severe preeclampsia or severe placental insufficiency OR
c.Three or more unexplained consecutive spontaneous abortions before the 10th week of gestation, with maternal anatomic or hormonal abnormalities and paternal and maternal chromosomal causes excluded.
a.Anticardiolipin antibody of IgG and/or IgM isotype in blood, present in medium or high titer, on at least 2 occasions at least 6 weeks apart, measured by standard enzyme-linked immunosorbent assay for β2-glycoprotein I–dependent anticardiolipin antibodies, OR
b.Lupus anticoagulant present in plasma, on 2 or more occasions at least 6 weeks apart, detected according to the guidelines of the International Society on Thrombosis and Hemostasis (17).
The serologic criteria for APS are the persistent presence of lupus anticoagulant (LAC) or medium-to- high levels of IgG or IgM anticardiolipin antibodies (aCL) (7, 14). LAC refers to antibodies that prolong in vitro clotting times of plasma by interference with assemblage of components of the coagulation cascade on a phospholipid template (15). International guidelines for the identification of LAC have been published (16, 17). Apart from defining suitable test samples, the guidelines indicate that sensitive phospholipid-dependent clotting tests should be used for screening, that mixing studies using normal plasma are essential to rule out coagulation factor deficiencies as a cause for prolonged clotting times, and that extra phospholipids should be added to the test system to show neutralization of the abnormality. It is advised to screen with more than 1 test, because no coagulation assay is fully sensitive for LAC.
Standardization of the aCL enzyme-linked immunosorbent assay (ELISA), which proved difficult (18–20), began with the introduction of standards based on patient sera, such as the widely used Kingston Antiphospholipid Study (KAPS) samples (18). Use of these samples enables the definition of “standard” units for ELISA results. Cutoff levels for IgG aCL (in IgG phospholipid [GPL] units) and IgM aCL (in IgM phospholipid [MPL] units) have been presented in guidelines issued by the Association of Clinical Pathologists (9), with negative results defined as <5 GPL units and <3 MPL units, low-positive results defined as values <15 GPL units and <6 MPL units, medium levels defined as 15–80 GPL units and 6–50 MPL units, and high levels defined as >80 GPL units or >50 MPL units. However, variations in the performance of the assay itself and the patient-derived standards (21) cause wide variations in (semiquantitative) results (21–25) and make generalization of a specific, clinically relevant cutoff (26) hazardous. The fact that many important studies on the treatment of obstetric APS (27–32) evoked letters to the editor (33–39) disputing their reported cutoffs in the aCL ELISA illustrates that consensus is urgently needed.
Treatment of pregnancy in women with aPL
Studies on the management of aPL-related pregnancies comprise 3 groups of women: 1) those without a poor obstetric history, SLE, or previous thrombosis (low-risk pregnancies), 2) those with recurrent early pregnancy loss or (at least) one fetal loss in the absence of SLE or a history of thrombosis, and 3) those with a high frequency of fetal loss, SLE, previous thrombosis, or a combination of these. In contrast to women in the second and third categories, patients in the first category (low risk) did not receive pharmacologic treatment. We will critically review these studies, emphasizing whether or not patients met the classification criteria for APS (14).
Four studies described the natural course of pregnancy related to aPL in women regarded as being at low risk for pregnancy complications (28, 40–42). Patients were defined as “a normal pregnancy population” (28), “women with no more than 2 consecutive abortions and no stillbirth” (41, 40), and “absence of recurrent fetal wastage” (although women with fetal loss after 24 weeks of gestation were included) (42). Because all studies used results from a single blood sample to categorize women as positive or negative for aPL, by definition no patient met the serologic criteria for APS (7, 14). Other breaches of the criteria were testing for IgG aCL only (42) and the absence of a neutralization procedure (to confirm that the addition of extra phospholipids corrects abnormal coagulation times) in LAC tests (28, 40, 41). The studies included 16–95 women with aPL and 294–915 women negative for aPL (28, 40–42) (Table 2). The timing of blood sampling for aPL tests varied from the first antenatal visit (28) to 9 weeks (42), 13 weeks (40), and 16 weeks (41) of gestation. This timing is relevant, because the later in pregnancy that this sample is obtained, the higher the chance that early losses are not taken into account when the outcome in women with aPL and those without aPL is compared.
Table 2. Studies in which live-birth rates related to aPL in low-risk pregnancies were reported*
Despite these shortcomings and differences, all studies found lower live-birth rates in aPL-positive women (range 62–84%) than in aPL-negative women (range 90–98%) (28, 40–42) (Table 2). An association between the presence of aPL and severe preeclampsia was found in 2 studies (28, 42), but not in another study (40). The Yasuda et al study (42) found a negative association and the Lynch et al study (40) found a positive association between aPL and neonatal birth weight.
All investigators advised against routine screening of women with a low risk for pregnancy complications, because the frequency of aPL is low, and pharmacologic treatment is not indicated in women in whom aPL are found (28, 40–42). The latter conclusion is based on observations of uncomplicated pregnancies in untreated healthy women with aPL (even those with high titers ) and is supported by results from a small randomized study in which the pregnancy outcome in aPL-positive women at low risk for pregnancy complications was similar for those treated with aspirin (81 mg per day) and those who received standard care (43).
Women with recurrent early pregnancy loss or at least one fetal loss in the absence of SLE or previous thrombosis.
In those women whose pregnancy and clinical histories put them in the second category (those with recurrent early fetal loss or at least 1 fetal loss, without SLE or previous thrombosis), 3 observational studies evaluated the outcome in women treated with aspirin alone (44), those who received a combination of heparin and aspirin (45), and those who declined pharmacologic treatment and received standard care (46). In 9 prospective studies (12, 29–32, 47–50), 2 treatment strategies were compared. All but 2 of these studies (47,50) were randomized studies. The treatments included placebo (12, 29, 32), aspirin alone (12, 30–32, 47, 49), aspirin plus prednisone (29, 48, 49), and heparin plus aspirin (30, 31, 47, 48, 50) (Table 3). Evaluation of pregnancies started with a positive pregnancy test (12, 29, 32, 44, 47, 48, 50) or the presence of fetal heart activity (30, 31, 45, 46, 49). In contrast to the former studies, the latter studies could not score blighted ovum or absent fetal heart beat as pregnancy loss.
Table 3. Studies on live-birth rates related to pharmacologic treatment in women with aPL and recurrent early pregnancy loss or at least 1 fetal loss in the absence of SLE or previous thrombosis*
First author (reference)
No. of women
Live-birth rate (%) according to pharmacologic treatment
Most women in these studies met the criteria for obstetric APS (14), because they had at least 3 consecutive miscarriages (12, 30–32, 45–47, 50), at least 2 unexplained pregnancy losses before 12 weeks of gestation or at least 1 pregnancy loss after 12 weeks (44, 48, 49), or at least 2 consecutive fetal losses before 32 weeks of gestation (29). All studies, except one that tested for IgG aCL only (12), included tests for LAC as well as both IgG aCL and IgM aCL. All women tested repeatedly as aPL positive. In 4 studies (29, 32, 44, 48), however, LAC was not tested according to accepted guidelines (16, 17), because phospholipid dependency of coagulation abnormalities was not demonstrated. All but 2 studies (12, 44) used KAPS standards (18) in the aCL ELISA. The cutoffs used for inclusion in the studies ranged from 5 GPL units (30, 32, 45) to 31 GPL units (48) and from 3 MPL units (30, 45) to 21 MPL units (47, 50). Three studies (47, 48, 50) emphasized exclusion of women with low aCL levels, and another (49) emphasized inclusion of women with low aCL levels.
The studies differed widely regarding the proportion in which LAC and aCL contributed to aPL positivity in included women. Two studies (47, 50) excluded women with LAC, and another did not test for LAC (12). In studies that tested women for both LAC and aCL, the prevalence of LAC varied from 6% (49) to 91% (30), whereas the frequency of LAC in the absence of aCL was 6% (49), 22% (32, 44), 47% (31), 55% (45), and 82% (30). We will review these studies in relation to treatment (Table 3).
Prednisone and aspirin.
Three randomized studies included a regimen of prednisone and aspirin (29, 48, 49). The pregnancy outcome of women receiving this treatment was compared with that of women receiving a combination of heparin and aspirin (48), aspirin alone (49), or placebo (29). The daily dose of aspirin was 80–100 mg. Each study used a different dosage of prednisone, as follows: a standard daily dose of 40 mg throughout pregnancy, starting when pregnancy was confirmed (48); a daily dose of 20 mg starting when viability was confirmed, with dose adjustments throughout pregnancy based on results of aPL tests (daily dose 10–40 mg) (49); and a daily dose of 0.8 mg/kg of body weight (maximum 60 mg) for 4 weeks, followed by 0.5 mg/kg of body weight per day until delivery (29). In the study by Cowchock et al (48), subcutaneous injections of unfractionated heparin (10,000 units twice daily) were begun when the presence of a viable pregnancy was confirmed; downward dose adjustments were used to obtain a mid-interval activated partial thromboplastin time (APTT) that was within the normal range or was similar to prolonged baseline values (mean dosage 17,000 IU/day).
Live-birth rates were 75% for both treatment groups in the study by Cowchock et al (48) and 100% in that by Silver et al (49). Laskin et al (29) reported live births in 60% of treated pregnancies and in 52% of pregnancies in which placebo was given. All studies (29, 48, 49) found that use of prednisone was associated with significantly more maternal morbidity and more preterm deliveries (often associated with preterm rupture of membranes and preeclampsia). In one study (49), lower neonatal birth weights were observed in patients who were receiving prednisone.
Heparin and aspirin.
Three randomized studies compared pregnancy outcome in women receiving a combination of heparin and aspirin with the outcome in women receiving aspirin alone (30, 31, 47), and one study evaluated pregnancy outcome in women treated with aspirin combined with different doses of heparin (50). A study comparing treatment with heparin and aspirin with treatment with prednisone and aspirin was discussed above (48).
The daily dose of aspirin used in these studies was 75 mg (30, 31) or 81 mg (47, 50). In some studies, aspirin was started before conception (47, 50), and in others aspirin was initiated at the time of a positive pregnancy test (30) or at the time of randomization (before 12 weeks) (31). Heparin was started at the time of a positive pregnancy test (47, 50), when fetal heart activity was recorded (30), or before 12 weeks of gestation (31). Three studies used unfractionated heparin (30, 47, 50), and another used a single, fixed dose (5,000 IU) of low molecular weight heparin (31). The daily dose of unfractionated heparin was either fixed (5,000 IU subcutaneously twice daily) (30) or adjusted to maintain an APTT value that was either 1.2–1.5 times the value at baseline (range 16,000–40,000 IU/day) (47, 50) or at the upper limits of the normal range (range 10,000–25,000 IU/day) (50).
Two (30, 47) of the 3 studies that compared a combination of heparin and aspirin with aspirin alone (30, 31, 47) found higher live-birth rates in women who received combination therapy. Live-birth rates in women treated with heparin plus aspirin were 71%, 78%, and 80%, and in those treated with aspirin alone the rates were 42% (30), 72%, and 44%, respectively (30, 31, 47). High and low doses of unfractionated heparin, given in combination with low-dose aspirin, were associated with similar live-birth rates (80% and 76%, respectively) (50). Rai et al (30) found that the superiority of combination therapy holds only for the first 13 weeks of gestation. Because this is the time when the first wave of trophoblast invasion is complete and placentation is established, it was speculated that heparin may protect trophoblast cells from aPL-induced damage. The same group of investigators affirmed a live-birth rate of 71% in a prospective observational study in which women received either fixed dosages of unfractionated heparin (5,000 IU twice daily) or enoxaparin (20 mg once daily) (45).
In most studies, the overall frequency of prematurity (birth at <37 weeks of gestation) was between 7% (31) and 13% (47), but a frequency of 24% was found in 2 studies from a single clinic (30, 45). None of the studies found an association between prematurity and the therapeutic regimen used. Also, there was no association between the therapeutic regimen used and the frequency of preeclampsia, a complication that was absent in some studies (31) but reached a prevalence of 2% (30) to >10% in other studies (47, 50). In 2 studies, preeclampsia and intrauterine growth restriction accounted for more than one-third of prematurity (30, 45).
Aspirin alone (supportive care only).
The live-birth rate in aspirin-treated aPL-positive women with recurrent pregnancy loss was relatively good in most studies (72% , 80% , 91% , and 100% ) but was strikingly low in another 3 studies (17% , 42% , and 44% ). With placebo, live-birth rates of 17% (12), 52% (29), and 85% (32) were reported. In the smallest study (12), live-birth rates were spuriously low, because blighted ovum and ectopic pregnancy accounted for the majority of losses. Pattison et al (32) emphasized that the high live-birth rate with placebo makes it questionable that pharmacologic treatment had additional value over and above that of supportive care alone. This opinion contrasts sharply with that of Rai et al (46), who reported a live-birth rate of only 10% in 20 aPL-positive women with recurrent pregnancy loss who declined pharmacologic treatment in their current pregnancies.
The chance of pregnancy-related thrombosis developing in healthy aPL-positive women with recurrent pregnancy loss appears to be low. Only one case of thrombosis was reported (49) among a total of 673 such women included in 12 studies (12, 29–32, 44–50). That case originated from the only study that included patients (n = 2) with previous thrombosis (49). It was not specified whether the current bilateral deep vein thrombosis was a recurrent event. The other studies indicated that pharmacologic treatment was given “throughout pregnancy,” but most did not specify when it was stopped. Two reports (47, 50) mentioned that patients continued receiving heparin treatment in the postpartum period, and in another 2 studies (30, 45) it was reported that treatment with heparin and aspirin was stopped at 34 weeks.
Women with a history of fetal loss, SLE, thrombosis, or a combination of these.
The pregnancy outcomes in women in the third category (those with a high frequency of fetal loss, SLE, thrombosis, or a combination) were reported in 3 retrospective studies in which patients received treatment according to the patient's and physician's judgment (51–53), one retrospective study in which all women received monotherapy with heparin (54), and one placebo-controlled study that evaluated the effects of high-dose intravenous immunoglobulins (IVIG) added to treatment with heparin and aspirin (55).
In the retrospective studies, several patients contributed >1 pregnancy (Table 4). Many patients in these studies had SLE (77% , 36% , 32% , 13% , not specified ), previous thrombotic events (56% , 41% , 30% , 29% , not specified ), or both. All patients had APS according to the 1987 criteria (7), but many did not fulfill the obstetric criteria (7, 14), because several studies included primigravidae (accounting for 11% , 17% , 36% , and an unspecified percentage  of included women) or included women with 1 or 2 first trimester losses or only successful pregnancies (52–54). All patients fulfilled the serologic criteria for APS, because all studies apparently used adequate tests for LAC and aCL and included patients with at least medium levels of aCL. In one study (51), an insensitive LAC test was used, and another study (52) excluded patients with IgM aCL only.
Table 4. Studies on live-birth rates in treated women with aPL and fetal loss, SLE, thrombosis, or a combination of these*
First author (reference)
No. of women
No. of pregnancies
Live-birth rate (%) according to pharmacologic treatment
Low-dose aspirin + heparin
Low-dose aspirin + heparin + IVIG
aPL = antiphospholipid antibodies; SLE = systemic lupus erythematosus; any = treatment according to patient's and physician's judgment (no pharmacologic treatment, low-dose aspirin, prednisone, heparin, high-dose intravenous gamma globulin [IVIG], or a combination of these).
In the studies in which pharmacologic treatment was administered according to the patient's and physician's judgment, therapy in individual patients included no treatment (51), monotherapy with low-dose aspirin (51, 53), prednisone (51, 52) or heparin (52), a combination of aspirin and prednisone (51–53), a combination of aspirin, prednisone, and heparin (52), a combination of heparin and prednisone (53), a combination of aspirin and heparin (52), and a combination of IVIG, aspirin, and heparin (55). The daily dose of aspirin was 75–80 mg (51–53) and that of prednisone was 10–30 mg (51), 40 mg (with adjustments upon evaluation of LAC) (52), or was not specified (53). Treatment with heparin was either low molecular weight heparin in a dose aiming for target values of 0.15–0.2 IU/ml of anti–factor Xa activity preinjection (53) or unfractionated heparin in a daily dose of 10,000–20,000 IU divided into 2 or 3 injections, with dosage adjustments to 15,000–20,000 IU/day at the beginning of the second trimester (52) or an unadjusted dose of 10,000–15,000 IU divided into 2 doses (53). Branch et al (55) started with a dosage of 15,000 IU of unfractionated heparin per day and increased the dosage to 20,000 IU/day in the second trimester. Rosove et al (54) used unfractionated heparin with a daily dose divided into 2 subcutaneous injections, with dose adjustments to obtain a mid-interval APTT of 1.5–2 times the control value (daily dose 10,000–36,000 IU).
The gestational age at which pharmacologic treatment started was 7.5 weeks (range 2–25 weeks) (52), 9 weeks (range 0–21 weeks) (51), and 10 weeks (range 6–18 weeks) (54) and was not specified in 2 studies (53, 55). Lima et al (53) indicated that patients with a history of thrombosis were switched from warfarin to heparin preconception or after a positive pregnancy test, and that patients continued to receive heparin for 6 weeks after delivery or were converted back to warfarin. Rosove et al (54) reported that they discontinued treatment of patients with heparin briefly at the time of delivery and then resumed and continued heparin for variable periods thereafter.
The study by Lockshin et al (51) clearly showed that prior fetal death (unfortunately, this term was not defined by the authors) and aCL levels of ≥40 GPL units contributed independently, in an additive manner, to current fetal loss. Their results did not demonstrate a beneficial effect of aspirin on pregnancy outcome and suggested that prednisone may worsen fetal outcome.
Analysis of the study by Branch et al (52) revealed that with treatment, 63% of pregnancies resulted in a surviving newborn (73% when spontaneous abortions were excluded). Of these infants, 92% were born before 37 weeks of gestation, and 37% were born before 32 weeks of gestation. Fetal or neonatal deaths attributable to complications of prematurity (related to preeclampsia, fetal distress, or both) occurred in 19 (23%) of 82 pregnancies. There was no statistically significant difference among the treatment groups in terms of the rate of surviving neonates. A comparison between the patient's first untreated pregnancy following a fetal death and the first treated pregnancy (n = 32) showed that a statistically significantly higher percentage of surviving neonates was associated with treatment (6% in untreated pregnancies versus 53% in treated pregnancies). Factors that were not related to outcome included a diagnosis of SLE, a history of preeclampsia or thrombosis, the IgG aCL level, or the number of previous fetal deaths. The overall rates of preeclampsia (51%) and severe preeclampsia (27%) were similar in all treatment groups. Fetal distress occurred in 53% of pregnancies resulting in live births, and 31% of the newborns were small for gestational age. The rates of both complications were similar in all treatment groups. In 3 patients, 4 thrombotic events occurred (overall thrombotic rate 5%). Two events occurred in a single patient who was receiving heparin and was found to have protein C deficiency. All events occurred within 2 weeks postpartum.
In the study by Lima et al (53), 70% of treated pregnancies ended with a live birth. Among successful pregnancies, 43% of the infants were born before 37 weeks; 31% of the neonates were small for gestational age, and 5% had fetal distress. Preterm delivery was significantly more common in women receiving prednisone. There were 3 neonatal deaths (all at 26 weeks). The 18 unsuccessful pregnancies comprised 1 therapeutic abortion because of severe active lupus, 9 fetal deaths (median gestational age 21 weeks [range 12–27 weeks]), and 8 spontaneous abortions. Preeclampsia necessitated induction of labor in 11 pregnancies (9 live births and 2 with fetal deaths). Preeclampsia was not related to the use of prednisone. There were 7 pregnancy-related thrombotic events in 7 women. Three of these were in the postnatal period, and none of the women was receiving low molecular weight heparin. It was not reported whether these were first or recurrent events.
In the small retrospective study of heparin treatment by Rosove et al (54), pregnancy outcome was very good, with live birth in 14 of 15 pregnancies (93%) and 1 miscarriage at 12 weeks. Half of the live births occurred before 37 weeks of gestation. No case of preeclampsia or pregnancy-related thrombosis was reported.
The results of the only randomized study of patients in this category (55) were surprising. Based on previous reports (52, 53), the authors expected in current pregnancies a 20–30% risk of fetal or neonatal death related to prematurity, a 20–50% risk of preeclampsia (18–27% severe), a 50% risk of fetal distress or oligohydramnios necessitating delivery, a 30% risk of intrauterine growth restriction, and a ≥30% risk of premature delivery with the treatment that was given to all patients (heparin in combination with aspirin). They hoped to improve these values by the addition of IVIG (1 gm/kg for each of 2 consecutive days every 4 weeks through 36 weeks of gestation). After randomization, 9 patients received heparin, aspirin, and intravenous 5% albumin solution and 7 patients received a combination of heparin, aspirin, and IVIG. In both groups, all women were delivered of live-born infants after 32 weeks of gestation. In the IVIG group the rate of preterm delivery was significantly higher (100% versus 33% in the group that did not receive IVIG), with preeclampsia and oligohydramnios being the most frequent reasons. No significant differences were found for gestational age at delivery, mean birth weight, oligohydramnios, fetal distress, and preeclampsia (44% versus 11% in the IVIG group versus the control group). Deep vein thrombosis occurred in one woman in the IVIG group 4 weeks after delivery, while she was receiving 20,000 IU of unfractionated heparin. As possible explanations for the unexpected salutary maternal and fetal outcome, the authors mention early institution (at 4–6 weeks of gestation) of aspirin and heparin, an organized and focused approach to fetal surveillance, and a Type II error. It was hypothesized that the early initiation of aspirin and heparin treatment benefited the pregnancies, perhaps by improving placentation.
In healthy aPL-positive women with only successful pregnancies or <3 consecutive early losses, there is no indication for aPL testing or pharmacologic treatment during pregnancy. This statement is based on results from studies in low-risk pregnancies (28, 40–43) and epidemiologic evidence that the definition of recurrent miscarriage should be 3 or more consecutive pregnancy losses (56). We realize, however, that in current daily practice, many pregnant women, in the absence of a poor outcome of previous pregnancies, are offered pharmacologic treatment (mostly aspirin) if aPL are found, based on their own or their physician's choice.
Unfortunately, several studies on aPL-related pregnancy included primigravidae (51–53, 55) and women with <3 early pregnancy losses (29, 44, 48, 49), and this may have influenced their results. Adhering to strict criteria for obstetric APS is important to minimize the role of causes other than aPL (particularly abnormal karyotypes), because even among losses in treated aPL-positive women with ≥3 previous early losses, abnormal fetal karyotypes may be expected in up to 24% of conception products (45). Apart from using deviating criteria for obstetric APS, many studies did not adhere to the definition of aPL as described in the APS criteria (14). For aCL, the currently poor standardization of the aCL ELISA provided an easy argument to express doubts regarding whether randomized trials included women with at least medium levels of aCL (33–39).
Results from randomized studies led to discontinuation of the use of prednisone to improve pregnancy outcome in women with aPL. However, many study results are confusing and difficult to understand. For instance, it is unclear why in 2 studies (30, 47) live-birth rates in women being treated with aspirin alone were low (42% and 44%) compared with those in women receiving a combination of heparin and aspirin (71% and 80%), whereas in another study using comparable obstetric inclusion and exclusion criteria (31) the rates were similar in women receiving aspirin alone and those receiving a combination of heparin and aspirin (72% and 78%, respectively). This discrepancy can be explained neither by obvious differences in the heparin regimens used, which were (unfractionated) heparin in fixed doses (30) or adjusted doses (47) or low molecular weight heparin (31), nor by differences in serologic characteristics of the included women (LAC was present in 0% , 41% , and 91% ) or the notion that 2 studies (30, 31) included women with aCL levels that were lower than those in women in the third study (47). Furthermore, although inclusion of women with low aCL levels possibly explains the 80% live-birth rate with aspirin alone in one study (32), this explanation does not hold for other studies with rates of 91% (44) and 100% (49).
Although most data suggest that treatment improves the rate of live births, it is questionable whether the frequencies of pregnancy complications such as prematurity, small-for-gestational-age newborns, fetal distress, preeclampsia, or cesarean delivery for complicated pregnancy courses (32, 45, 52, 53) were influenced by treatment. In all published series maternal–fetal monitoring was optimized by including frequent antenatal visits, continuity of the care providers, a liberal admissions policy, a multidisciplinary approach to the management of autoimmune diseases, obstetric ultrasound scans every 2–4 weeks, monthly Doppler velocimetry of the umbilical arteries starting at 16 weeks of gestation, and weekly cardiotocography starting at 24 weeks, with more frequent testing if clinically indicated (30, 32, 44, 49, 53, 54). Some studies additionally included weekly nonstress testing beginning at 26–30 weeks (44, 49). Previous studies established the impact of supportive care alone in women with unexplained recurrent miscarriage; supportive care only resulted in live-birth rates of >85% (57–59).
Taken together, these data support the hypothesis that recognition of aPL as a risk factor for adverse pregnancy outcome by itself contributed significantly to higher live-birth rates in aPL-positive women who received treatment during their pregnancies (32, 44, 51–55). One should realize that almost all randomized treatment trials in pregnancy outcome involved aPL-positive women who were otherwise healthy. We do not know whether the presence of SLE, a history of thrombosis, or both augment aPL-related risks for poor pregnancy outcome.
We recommend that healthy women with ≤2 (pre)embryonic losses in the absence of fetal loss should not be advised to undergo aPL tests, because there is no evidence for the beneficial effects of any pharmacologic treatment during pregnancy. There are no data indicating that such advice should be different when SLE is present. We recommend close maternal–fetal monitoring during all pregnancies of aPL-positive women.
Figure 1 depicts an algorithm for the pharmacologic treatment of women with aPL. During the pregnancies of women meeting the criteria for APS (14), we recommend combined treatment with heparin and aspirin (75–100 mg/day). We realize that with such advice we rely heavily on the results of a recent systematic review of therapeutic trials (60). The conclusion drawn from that meta-analysis (i.e., that aspirin alone is ineffective) was based on the outcomes of only 71 pregnancies (12, 32, 43) in women with dissimilar obstetric histories and for whom different definitions of aPL were used. Furthermore, the conclusion regarding superiority of a combination of heparin and aspirin was based on results of only 2 trials (30, 47) (a total of 140 pregnancies) that observed remarkably low live-birth rates (<45%) in women treated with aspirin alone and enrolled patients who were substantially different with regard to laboratory tests. We also realize that the meta-analysis could not include results of the randomized trial by Farquharson et al (31), which found similar outcomes with aspirin and combination therapy. We recognize that the basis to advise combination therapy is weak, and that other investigators legitimately may conclude that selected patients should be treated with aspirin alone or with thorough and careful expectant care.
Low-dose aspirin should be started preconception (44, 47, 49, 50) or at the time of a positive pregnancy test (12, 30–32, 45, 48, 55), and heparin therapy should be initiated at the time of a positive pregnancy test (47, 50) or when fetal heart activity is proven (30). The optimal timing is unknown. An argument in favor of withholding therapy until fetal cardiac activity is verified is the avoidance of administering unnecessary drugs to women who have a positive pregnancy test but no developing embryo.
The absence or presence of previous thrombotic events dictates the dosage of heparin that we propose (Tables 5 and 6). The lowest dosage will be given to healthy women who meet the obstetric criteria for APS. For women with a previous aPL-related arterial thrombotic event who are receiving long-term aspirin treatment (61), the obstetric history dictates additional use of heparin, although some experts would be more comfortable with the administration of heparin in any such case. In patients with previous venous thrombotic events, the heparin dosage should be individualized, taking into account whether or not there is an indication for long-term anticoagulation, the circumstances under which the previous events occurred, the number and severity of such events, comorbidity, and bleeding risk assessment. When patients have a history of thrombosis, we advise a switch from oral anticoagulants to heparin before conception or, at the latest, within 2 weeks of the first missed period, because oral anticoagulants cross the placenta, are teratogenic when given between 6 and 12 weeks of gestation, and may cause intracranial bleeding in the fetus (62). As pregnancy progresses the volume of distribution for heparin increases, and dose adjustments in proportion to weight gain or based on APTT or anti–factor Xa levels can be considered. In selected patients a switch from heparin to oral anticoagulants may be practical between 15 and 34 weeks of gestation (63).
Adapted from recommendations in ref. 62. APTT = activated partial thromboplastin time; LAC = lupus anticoagulant.
5,000 units subcutaneously every 12 hours
Subcutaneously every 12 hours adjusted to target an anti–factor Xa level of 0.1–0.3 units/ml
Subcutaneously every 12 hours to target a mid-interval APTT (or, if LAC is present, an anti–factor Xa level) into the therapeutic range
Low molecular weight heparin (LMWH)
Dalteparin, 2,500 or 5,000 units subcutaneously every 24 hours; OR enoxaparin 20 or 40 mg subcutaneously every 24 hours; OR nadroparin 2,850 units subcutaneously every 24 hours; OR any once-daily LMWH adjusted to target a peak anti–factor Xa level of 0.2–0.6 units/ml
Weight-adjusted, full treatment doses of dalteparin, 200 units/kg subcutaneously in 1 or 2 injections; OR enoxaparin 1 mg/kg subcutaneously every 12 hours or 1.5 mg/kg subcutaneously every 24 hours; OR nadroparin, 171 units/kg subcutaneously in 1 or 2 injections
Table 6. Advocated heparin regimens according to different clinical situations*
APS = antiphospholipid syndrome; aPL = antiphospholipid antibodies; LMWH = low molecular weight heparin; SLE = systemic lupus erythematosus (see Table 5 for other definitions).
Patient meets criteria for obstetric APS (obstetric indication), OR
Mini- or moderate-dose unfractionated heparin, OR
Previous thrombotic event and aPL, no long-term use of oral anticoagulation
Previous thrombotic event and receiving long-term oral anticoagulation
Adjusted-dose unfractionated heparin, OR prophylactic- or adjusted-dose LMWH
Patient meets criteria for obstetric APS, OR aPL-positive SLE patient
Mini-dose unfractionated heparin, OR prophylactic-dose LMWH
Previous thrombotic event
Mini- or moderate-dose unfractionated heparin, OR prophylactic-dose LMWH
Indication for long-term anticoagulation
Adjusted-dose unfractionated heparin, OR adjusted-dose LMWH; resume long-term anticoagulation
Low molecular weight heparin has the advantage over unfractionated heparin of a longer plasma half-life and a more predictable dose response and therefore the potential for once daily administration. Furthermore, low molecular weight heparin causes less heparin-induced thrombocytopenia and heparin-induced osteoporosis (62, 64). Given these advantages, we prefer low molecular weight heparin, although this approach is more expensive. With prolonged use of heparin we recommend the intake of extra oral calcium (1,000–2,000 mg/day) and vitamin D (400–800 IU) as prophylaxis for heparin-induced osteoporosis.
We advise thromboprophylaxis (heparin, low molecular weight heparin, or warfarin) during the postpartum period (6 weeks) in aPL-positive women with, based on the foregoing, an indication for treatment during pregnancy (Table 6). This recommendation is based on notions that thrombosis in patients with APS is often related to pregnancy (52), and that cesarean deliveries (particularly emergency cesarean deliveries) constitute a high risk for venous thromboembolic events (62). We recognize, however, that in prospective studies in healthy aPL-positive women very few thrombotic events were reported. Both heparin and warfarin are safe for nursing mothers.
There are no data showing that the approach to an aPL-positive woman with SLE should be different from what we discussed previously. However, although scientific proof is lacking, most experts will agree with the use of postpartum thromboprophylaxis in an aPL-positive patient with SLE. We fully recognize that chronic organ damage, immunosuppressive treatment, and active disease per se contribute to adverse pregnancy outcome in patients with SLE (53, 65).
For further progress in the management of obstetric APS, the wide availability and use of well-characterized and well-standardized aPL assays and calibrators are essential. Periodic interlaboratory comparisons and quality control should abolish doubts about clinically relevant cutoff values. Future clinical trials need indisputable definitions of relevant aPL and obstetric entry criteria.
To establish any benefit of pharmacologic therapy, a trial with these prerequisites should compare optimal fetal–maternal supportive care plus placebo with similar supportive care plus pharmacologic treatment. Furthermore, randomized trials in patients with previous thrombotic events and aPL-positive women in whom treatment failed are much desired. It is hoped that, by such efforts, the misty research field on management of obstetric APS will clear up within the near future.