Antithrombotic treatment for recurrent pregnancy loss?


Ian A. Greer, Faculty of Health & Life Sciences, University of Liverpool, 3rd Floor, The Foundation Building, 765 Brownlow Hill, Liverpool L69 7ZX, UK.
Tel.: +44 0151 795 0422; fax: 44 151 795 5256.


Summary. Background: Recurrent pregnancy loss (RPL) is a major issue for women’s health. Acquired and heritable thrombophilias are associated with RPL, this association could reflect a general prothrombotic phenotype rather than a specific thrombophilia. Antithrombotic intervention has therefore been assessed for RPL. Results: Two large randomised trials with untreated control groups showed no benefit from antithrombotic treatment with LMWH and low dose aspirin in women with RPL. These trials had insufficient power to exclude an effect in women with underlying thrombophilia, ≥ 3 losses, or late losses. Conclusions: Antithrombotic intervention should not be recommended for unexplained RPL in general. There may be specific groups such as those with an heritable thrombophilia, or with three or more losses, or second trimester losses that might benefit and where further trials are required. Further there is a need to consider the benefits of LMWH on implantation such as in women undergoing assisted conception therapy.


Recurrent pregnancy loss (RPL) is a major issue for women’s health because it is estimated that 5% of women of reproductive age have two or more RPL and around 1% have three or more losses [1]. Although a substantial number of losses are associated with an underlying chromosomal problem in the embryo or fetus, these are difficult to identify in routine practice as products of conception may not be available for such testing and so a pragmatic approach is required. RPL is a feature of antiphospholipid syndrome (APS), which is associated with evidence of increased thrombin generation and thrombotic placental damage. This condition benefits from treatment with antithrombotic interventions – low dose aspirin and heparin [2,3]. Heritable forms of thrombophilia have also been associated with RPL [4]. These data come largely from retrospective case control studies. The degree of association between heritable thrombophilia and RPL varied considerably between reports. This is likely to reflect heterogeneity in study design and methodology, sample size, inclusion of high-risk groups, the number of thrombophilias assessed, definition of RPL, and differences in the genetic background of the population studied. However these associations persist on systematic review where thrombophilias such as Factor V Leiden and Prothrombin G20210A are associated with an increased relative risk of first trimester RPL, although the magnitude of risk is modest being around 2-fold. There is also an association with recurrent second trimester loss [4]. As both heritable and acquired thrombophilias have been associated with RPL this raises the hypothesis that unexplained RPL could reflect a prothrombotic phenotype in general rather than a specific thrombophilia. Such an effect could reflect not only a thrombotic problem at the placental bed, but also adverse effects of coagulation activation on trophoblast during implantation. If a prothrombotic phenotype, regardless of specific heritable thrombophilia, underlies RPL then it is logical to consider antithrombotic treatment for unexplained RPL. Further as heparins have potentially beneficial effects on trophoblast [5], an effect mediated through mechanisms other than an anticoagulant effect [5] is possible.


In a systematic review of safety of LMWH in pregnancy the success rate for pregnancy in those receiving LMWH for RPL was in excess of 85% [6]. A comparison between different doses of LMWH in patients with thrombophilia and adverse pregnancy outcome including RPL suggested a possible beneficial effect from LMWH (Live birth rates: 84% with 40 mg enoxaparin vs. 78% with 80 mg Enoxaparin) when compared to historic controls [7], but with such an historic control group with previous poor outcomes there is the likelihood of bias towards a positive result. Thus benefit of LMWH could not be concluded. Gris et al. [8] studied low dose aspirin and LMWH in patients with a history of pregnancy loss and thrombophilia (FV Leiden, Prothrombin G20210A or protein S deficiency). They randomised 160 women with pregnancy loss after 10 weeks to low dose aspirin (100 mg) or enoxaparin 40 mg day−1 from 8 weeks in their subsequent pregnancy. This study reported that 86% of the LMWH group and 29% of the aspirin group had successful pregnancies (OR 15.5 95% CI 15–34). As there was no untreated control group it was impossible to determine whether aspirin was harmful or LMWH beneficial. Dolitzky et al. [9] reported a randomised trial of 104 women with ≥ 3 unexplained recurrent miscarriages to low dose aspirin (100 mg) or enoxaparin 40 mg with livebirth rates of 84% and 81.5% respectively, but again there was no untreated controls. More recently the Habenox trial [10] randomised 207 women with ≥ 3 consecutive first trimester (< 13 weeks) miscarriages, ≥ 2 second trimester (13–24 weeks) miscarriages, or one third trimester fetal loss and one first trimester miscarriage before 7 weeks to either aspirin 100 mg, enoxaparin 40 mg and placebo aspirin or enoxaparin 40 mg and aspirin 100 mg (the aspirin/placebo was double blind). The live birth rate in the aspirin 100 mg group (= 76) which was the referent group, was 61%, for the enoxaparin 40 mg and placebo aspirin group (= 68) the live birth rate was 71% RR 1.1, 95%CI 0.92–1.48, and in the enoxaparin 40 mg and aspirin 100 mg group (= 63) the live birth rate 65% RR 1.08, 95% CI 0.83–1.39. These were not significantly different from the referent group. The trial was ended prematurely because of slow recruitment with no significant difference in any outcome, complications or adverse events.

It should be noted that the expected success rate for pregnancy, without treatment, after one loss is > 75%. Coppens et al. [11] reported that in women who had had a first pregnancy loss, the live birth rate in the second pregnancy after an early first loss (≤ 12 weeks of gestation) was 77% (95% CI 62–87) for carriers of Factor V Leiden and Prothrombin 20210A, and 76% (95% CI 57–89) for non-carriers. Further, it is recognised that women with recurrent loss receiving specific antenatal counseling and psychological support have a pregnancy success rate of 86% [12].

As LMWH is considered safe in pregnancy [6], the association between RPL and thrombophilia, the success of antithrombotic treatment in acquired thrombophilia (APS) and the suggestion of a possible benefit as described above, led clinicians to prescribe antithrombotic interventions for these women before reliable evidence from adequately controlled trials was available. While there was biological plausibility for this approach there was a clear need for randomized and adequately controlled trials of antithrombotic interventions in women with RPL.

Randomised controlled trials of antithrombotics for RPL

There have been several recent randomised trials of LMWH in women with RPL and no identified thrombophilia. Two of these trials compared enoxaparin 20 mg day−1 vs. no treatment [13] or placebo [14] in women with at least three idiopathic recurrent miscarriages and reported higher rates of successful pregnancy outcome with LMWH. However both had significant methodological limitations [15] such as lack of prospective trial registration and unclear blinding and thus may be associated with significant bias.

Kaandorp et al. [16] reported the Aspirin Combined with Low-Molecular-Weight Heparin and Aspirin Alone in Women with Recurrent Miscarriage (ALIFE) study (ISRCTN58496168)). This trial compared low dose aspirin combined with a LMWH, nadroparin, and low dose aspirin alone, with aspirin placebo in women with unexplained RPL. Of the 364 women with recurrent miscarriage, who were attempting to conceive, 299 became pregnant. The intervention commenced at 6 weeks gestation following the confirmation of a viable pregnancy. The trial was halted prematurely when the Data Safety Monitoring Board advised discontinuation because of futility. Low dose aspirin combined with the LMWH nadroparin, and aspirin alone, did not improve live birth rates compared to placebo in these women. The live birth rates in women who became pregnant are shown in Table 1.

Table 1.   Summary of outcomes from the ALIFE trial for women who became pregnant
 Aspirin +nadroparinAspirin onlyPlaceboP value
No. of women9799103 
Live birth no. (%)67 (69.1)61 (61.6)69 (67.0)0.52
Relative risk (95% CI)1.03 (0.85–1.25)0.92 (0.75–1.13)1.00 

The cause of RPL is heterogeneous, and, women with three or more losses, or those with heritable thrombophilia, or those with no previously successful pregnancy may represent more homogeneous groups and could therefore have different responses to antithrombotic therapy. Further, a substantial number of losses are associated with an underlying chromosomal problem in the embryo or fetus, which would not be expected to benefit from antithrombotic therapy. The trial had insufficient power to adequately assess subgroups, nonetheless this may be of value in generating hypotheses. In the 16% of women in the trial with an underlying thrombophilia, an a priori planned analysis in those women found a relative risk for live birth of 1.31 (95%CI: 0.74–2.33) for the combined intervention compared to placebo, and 1.22 (95%CI: 0.69–2.16) for aspirin so emphasizing the need for a controlled trial of antithrombotic in women with a thrombophilia [17,18].

These data are supported by a further multicentre randomized controlled trial, the Scottish Pregnancy Intervention Trial (ISRCTN06774126) [19]. This trial compared intensive pregnancy surveillance alone with an antithrombotic intervention, LMWH (enoxaparin), combined with low dose aspirin, in 294 women with a history of two or more RPL. There was no reduction in pregnancy loss rate with antithrombotic intervention [odds ratio 0.91 (95%CI 0.52–1.59)] compared with surveillance alone (Table 2).

Table 2.   Patients and outcome in the SPIN trial
CharacteristicAnticoagulant group (= 147)Intensive surveillance group (= 147)
Age (years)31 (26–36)32 (27–36)
≥ 1 previous live birth67 (45.6%)66 (44.9%)
No. of previous losses2 (2–3)2 (2–3)
No. with > 2 losses66 (44.9%)60 (40.8%)
BMI (Kg m−2)25.4 (21.8–29.0)26.4 (23.1–32.4)
Current smoker29 (19.7%)25 (17.0%)
Gestation at randomization (weeks)6 (6–7)6 (6–6)
Multiple pregnancy2 (1.4%)3 (2.0%)
Pregnancy outcome32 losses in 143 subjects (four subjects were lost to follow-up)29 losses in 140 subjects (seven subjects were lost to follow-up)

These studies are consistent in their finding of the lack of benefit from the use of LMWH and low dose aspirin for unexplained RPL. There were no significant safety issues, consistent with previous reports [6], but as expected there were significant numbers of minor side effects including injection site bruising.

While these trials show that women with two or more RPL, in general, should not receive antithrombotics for prevention of RPL, they do not exclude such an effect in women with a specific heritable thrombophilia. Thus the results of similar trials in women with thrombophilia are required [18]. It is also possible that women with late miscarriages, occurring in the second trimester might benefit as placental bed thrombosis may be a more likely mechanism in late miscarriage in contrast to early losses, which may be more heterogeneous including chromosomal problems. Further, if the intervention were to act through protective effects on the trophoblast, then treatment at an earlier stage may be more likely to exert an effect. It is noteworthy that many assisted conception units now use LMWH pragmatically, following embryo transfer because of possible beneficial effects, that could improve success rates. This is based on biological plausibility extrapolated largely from in vitro studies on trophoblast [5]. Thus there is a need for randomised controlled trials of antithrombotic treatment in women undergoing assisted conception to examine the effect of intervention in advance of the implantation process. Initial studies in women with recurrent implantation failure undergoing IVF have suggested possible benefit from LMWH reatment. Qublan et al. [20] randomized 83 women with history of three or more previous IVF failures to enoxaparin 40 mg day−1, or placebo. Both treatments started on the day of embryo transfer and continued until delivery or fetal demise was diagnosed. Women who received LMWH had a significant increase in their live birth rate compared with placebo (23.8% vs. 2.8%, respectively; < 0.05).


These data show that antithrombotic intervention should not be recommended for unexplained RPL in general. However, there may be specific groups such as those with an heritable thrombophilia, or with three or more losses, or second trimester losses that might benefit and where further trials are required. Further there is a need to explore antithrombotics in women undergoing assisted conception to assess a possible effect on implantation.

Disclosure of Conflict of Interest

The author states that he has no conflict of interest.