- R Rai, research fellowa,
- H Cohen, senior lecturerb,
- M Dave, scientific officerb,
- L Regan, professora
- a Imperial College School of Medicine at St Mary's London W2 1PG Department of Obstetrics and Gynaecology
- b Department of Haematology
- Correspondence to: Professor Regan
- Accepted 12 November 1996
Objective: To determine whether treatment with low dose aspirin and heparin leads to a higher rate of live births than that achieved with low dose aspirin alone in women with a history of recurrent miscarriage associated with phospholipid antibodies (or antiphospholipid antibodies), lupus anticoagulant, and cardiolipin antibodies (or anticardiolipin antibodies).
Design: Randomised controlled trial.
Setting: Specialist clinic for recurrent miscarriages.
Subjects: 90 women (median age 33 (range 22-43)) with a history of recurrent miscarriage (median number 4 (range 3-15)) and persistently positive results for phospholipid antibodies.
Intervention: Either low dose aspirin (75 mg daily) or low dose aspirin and 5000 U of unfractionated heparin subcutaneously 12 hourly. All women started treatment with low dose aspirin when they had a positive urine pregnancy test. Women were randomly allocated an intervention when fetal heart activity was seen on ultrasonography. Treatment was stopped at the time of miscarriage or at 34 weeks' gestation.
Main outcome measures: Rate of live births with the two treatments.
Results: There was no significant difference in the two groups in age or the number and gestation of previous miscarriages. The rate of live births with low dose aspirin and heparin was 71% (32/45 pregnancies) and 42% (19/45 pregnancies) with low dose aspirin alone (odds ratio 3.37 (95% confidence interval 1.40 to 8.10)). More than 90% of miscarriages occurred in the first trimester. There was no difference in outcome between the two treatments in pregnancies that advanced beyond 13 weeks' gestation. Twelve of the 51 successful pregnancies (24%) were delivered before 37 weeks' gestation. Women randomly allocated aspirin and heparin had a median decrease in lumbar spine bone density of 5.4% (range -8.6% to 1.7%).
Conclusion: Treatment with aspirin and heparin leads to a significantly higher rate of live births in women with a history of recurrent miscarriage associated with phospholipid antibodies than that achieved with aspirin alone.
The prognosis for pregnancies in women with recurrent miscarriage associated with phospholipid antibodies is poor
This randomised controlled trial found that the prognosis improved with low dose aspirin and was further improved with the addition of low dose heparin to the aspirin
This combination may promote successful embryonic implantation in the early stages of pregnancy and protect against thrombosis of the uteroplacental vasculature after successful placentation
Most miscarriages occurred before 13 weeks' gestation
Nearly a quarter of the successful pregnancies were delivered prematurely (before 37 weeks' gestation), so close surveillance is necessary
Long term use of low dose heparin was associated with few complications
Phospholipid antibodies, lupus anticoagulant, and cardiolipin antibodies are associated with recurrent miscarriage, thrombosis, and thrombocytopenia.1 Fifteen per cent of women with a history of recurrent miscarriage (three or more consecutive losses of pregnancy) have persistently positive results for phospholipid antibodies.2 These women have a rate of fetal loss of 90% when no specific treatment is given during pregnancy.3 4 Miscarriage has been attributed to thrombosis of the uteroplacental vasculature and placental infarction,5 6 but the extent of placental disease cannot account for pregnancy loss in all cases.6 Several treatments including corticosteroids,7 8 9 10 low dose aspirin,9 11 heparin8 11 12, and immunoglobulins13 have been used either as single agents or in combination to try to improve the rate of live births in women with phospholipid antibodies. However, available data are limited by the small number of patients in individual studies, which also have had varying entry criteria and treatment protocols, and by the lack of standardisation of laboratory assays used to detect phospholipid antibodies. This last point has recently been addressed.14 15 16
The use of corticosteroids in pregnancy is associated with significant maternal and fetal morbidity.8 9 Low dose aspirin or heparin are currently the favoured treatments. We determined whether the addition of low dose heparin to low dose aspirin results in a higher rate of live births than that achieved with low dose aspirin alone in women with a history of recurrent miscarriage associated with phospholipid antibodies.
Patients and methods
Local ethics committee approval was obtained and patients gave informed consent. Between April 1993 and July 1995, 90 patients were recruited from the recurrent miscarriage clinic at St Mary's Hospital in London.
Patients were eligible for this study if they had a history of three or more consecutive miscarriages and had positive results for phospholipid antibodies on at least two occasions more than eight weeks apart before becoming pregnant. Women with previous thromboembolism were excluded as they require thromboprophylaxis during pregnancy. Also excluded were those with systemic lupus erythematosus,17 a uterine abnormality (detected on ultrasound scanning), hypersecretion of luteinising hormone,18 and multiple pregnancy. Women were also excluded if either they or their partner had an abnormal karyotype.
Women started taking low dose aspirin (75 mg daily) as soon as they had a positive pregnancy test (urinary human chorionic gonadotrophin concentration >50 IU/ml; Clearview, Unipath, Bedford). Vaginal ultrasound scans were performed from five weeks of amenorrhoea. When fetal heart activity was seen on ultrasonography women were randomly allocated either to continuing low dose aspirin or to self administered subcutaneous calcium heparin (5000 U 12 hourly; Calciparine, Sanofi Winthrop, Surrey) in addition to low dose aspirin. From 24 weeks' gestation pregnancies were monitored by serial ultrasonography and Doppler studies of the umbilical artery circulation. Treatment was stopped at the time of miscarriage or at 34 weeks' gestation.
Women randomly allocated heparin had dual energy x ray absorbitometry bone densitometry (Lunar Corporation, Madison, Wisconsin) of the lumbar spine (L2-L4) performed at 12 weeks' gestation and postnatally.
Lupus anticoagulant–A coagulation screen consisting of the prothrombin time, activated partial thromboplastin time, and thrombin time was performed on all patients using an Organon Teknika XC Plus and standard methods. The activated partial thromboplastin time was also measured in test plasma samples diluted in the ratio of 80 to 20 parts with normal plasma. The dilute Russell's viper venom time was measured in duplicate on a semiautomated coagulometer (KC4; Heinrich Amelung, Lehbrinksweg, Lemgo) using a kit (Unicorn Diagnostics, London). Patient samples with a dilute Russell's viper time ratio of ≥1.1 (test:normal) were retested with a platelet neutralisation procedure consisting of washed, freeze-thawed platelets. A decrease of 10% or more in the ratio was considered to be positive for lupus anticoagulant.15
Anticardiolipin antibodies–Samples for IgG and IgM cardiolipin antibodies were assayed in duplicate using a standardised enzyme linked immunosorbent assay (ELISA).16 Results were expressed in either GPL or MPLs. One such unit is defined as the binding activity of 1 μg/ml of affinity purified IgG or IgM standard.19 A positive concentration of cardiolipin antibodies was considered to be IgG ≥5 GPL or IgM ≥3MPL.16
Patients were randomly assigned in equal proportion to the two treatment groups by means of a computer generated random number list (Systat 5.2.1; Macintosh). The randomisation list was kept by an independent member of staff not involved in the trial. All patients remained in their originally allocated treatment group and the outcome of all pregnancies was analysed.
Data available at the start of the study showed that the rate of live births among women treated with low dose aspirin alone is 30%.10 We considered that an improvement to 60% with the addition of heparin would be clinically important. A power calculation at the start of the study indicated that between 80 and 90 women would have to be recruited to achieve a study power of 80% at a significance level of 0.05 (two tailed) to prove the hypothesis correct: that the use of low dose aspirin and heparin leads to a higher rate of live births than that achieved with low dose aspirin alone.
The primary outcome measure of this trial–live birth or miscarriage–was compared between the two groups using Fisher's exact test. Secondary quantitative outcome comparisons were made using the Mann-Whitney U test.
Forty five women were randomly allocated low dose aspirin alone and 45 low dose aspirin and heparin. Table 1) shows their demographic details and phospholipid antibody status. The median concentration of IgG cardiolipin antibodies was 12.5 GPL (range 5.2-80 GPL). One woman with positive results for IgM cardiolipin antibodies had a concentration of 15 MPL. No woman withdrew from the trial.
The outcome of the 90 pregnancies is shown in figure 1) and table 2). Treatment with low dose aspirin in combination with heparin led to a significantly higher rate of live births (71%) than that achieved with low dose aspirin alone (42%; odds ratio 3.37 (95% confidence interval 1.40 to 8.10)). The birth weights and gestation at delivery were similar in the two groups (table 3)).
Twelve of the 51 women with successful pregnancies (24%) delivered prematurely (<37 weeks' gestation). This was due to preterm labour in two out of 19 women treated with low dose aspirin (11%) and in five out of 32 women treated with low dose aspirin and heparin (16%; P=0.70). Intrauterine growth retardation (estimated fetal weight below the 10th centile for gestational age) was the indication for delivery in one woman treated with low dose aspirin (5%) and for three women treated with low dose aspirin and heparin (9%; P=0.99). One woman treated with low dose aspirin alone was delivered at 36 weeks' gestation for pre-eclampsia. No woman developed a thromboembolic complication during pregnancy or the puerperium.
Most miscarriages in the two groups occurred in the first trimester (table 4)). In pregnancies that progressed beyond 13 weeks' gestation there was no difference in the outcome of those treated with low dose aspirin alone compared with those treated with low dose aspirin and heparin.
All babies were examined by a paediatrician shortly after delivery. No congenital abnormalities were detected. Three babies were admitted to the neonatal unit because of prematurity. Only one baby, who was delivered by caesarean section at 32 weeks' gestation for intrauterine growth retardation to a mother treated with low dose aspirin alone, required ventilatory support for a week. The two other babies, delivered at 33 and 36 weeks' gestation, were admitted to the neonatal unit for help with feeding. One baby had a parietal lobe infarction two days after delivery (diagnosed on magnetic resonance imaging). This baby was born at 42 weeks' gestation to a mother who was treated with low dose aspirin alone. The mother had high concentrations of IgG cardiolipin antibodies (median 80 GPL) and was negative for lupus anticoagulant. The baby tested negative for both IgG and IgM cardiolipin antibodies.
Both low dose aspirin and heparin were well tolerated. Of those taking heparin, none developed thrombocytopenia or had symptomatic complications apart from mild bruising localised to the injection site. These women had a median decrease in lumbar spine bone density of 5.4% (range -8.6% to 1.7%) between 12 weeks' gestation and postnatally (median 8 weeks; range 0-12). No woman had a vertebral fracture.
This trial has shown that treatment with aspirin and low dose heparin leads to a significantly higher rate of live births than that achieved with aspirin alone in pregnant women with a history of recurrent miscarriage associated with phospholipid antibodies.
The laboratory assays used to detect phospholipid antibodies in this study complied with recently issued international guidelines for the testing of lupus anticoagulant and cardiolipin antibodies.14 15 16 The phospholipid antibody status of the patients recruited to this study was similar to that in a prevalence survey of 500 consecutive women with recurrent miscarriage in that most patients were positive for lupus anticoagulant alone and there was little crossover between lupus anticoagulant and cardiolipin antibodies positivity.2
The trial was not blinded as it was not considered ethical to ask a pregnant woman to self administer twice daily subcutaneous injections of a placebo for 28 weeks. The most common type of miscarriage in women with phospholipid antibodies occurs in the first trimester after fetal heart activity has become established.3 4 Entry into the trial and randomisation therefore occurred only when fetal heart activity was seen on ultrasonography. Treatment was stopped at 34 weeks' gestation to minimise the duration of exposure to both aspirin and heparin. All women in the trial attended for supportive care, in the form of weekly ultrasound scans and psychological support, during the first trimester. This has been reported to have a significant beneficial effect on pregnancy outcome in women with a history of recurrent miscarriage.20
Phospholipid antibodies and miscarriage
Studies on human tissue and in mice suggest that phospholipid antibodies cause pregnancy loss by binding to phospholipids expressed on the invading trophoblast,21 22 thereby inhibiting successful embryonic implantation into the endometrium. Once placentation is established their thrombogenic action leads to decreased placental perfusion and subsequent infarction.5 6 We recently reported that non-pregnant women with a history of recurrent miscarriage in association with phospholipid antibodies are in a prothrombotic state.23 This thrombogenic potential may be exacerbated by the known hypercoagulability that occurs in pregnancy.
Low dose aspirin may improve pregnancy outcome in women with phospholipid antibodies by irreversibly blocking the action of cyclo-oxygenase in platelets, thereby inhibiting platelet thromboxane synthesis and preventing thrombosis of the placental vasculature.24 Heparin may act to reduce fetal loss by binding to phospholipid antibodies, thereby protecting the trophoblast phospholipids from attack25 and promoting successful implantation in early pregnancy, in addition to its anticoagulant action. This is supported by the finding that there was no difference in pregnancy outcome between the two treatment arms in the pregnancies that survived beyond 13 weeks' gestation. By this time the first wave of trophoblast invasion is complete and placentation established.
High complication rate despite treatment
One quarter of successful pregnancies were delivered prematurely. This confirms previous reports of a high incidence of pregnancy complications in women with phospholipid antibodies3 4 5 6 and emphasises the need for close antenatal surveillance. The finding that most miscarriages (>90%) occurred before 14 weeks' gestation confirms previous prospective observations in women with phospholipid antibodies who received no pharmacological treatment during pregnancy and who were followed up from the time that they had a positive pregnancy test.4 Babies born to mothers treated with low dose aspirin and heparin tended to be of greater weight than those born to mothers treated with low dose aspirin alone, but this may be confounded by the higher proportion of boys born to mothers given aspirin and heparin (table 3)).
As the long term use of heparin is associated with the development of osteopenia, bone density measurements were performed on women randomly allocated heparin. The median loss in lumbar spine bone density of 5.4% is equivalent to that lost after six months of lactation.26 Osteopenia associated with heparin is reversible when heparin treatment is stopped.27
In conclusion, the poor obstetric outlook for women with a history of recurrent miscarriage in association with phospholipid antibodies may be improved with low dose aspirin, but it is further and significantly improved with the combined use of low dose aspirin and low dose heparin. This combination may promote successful embryonic implantation in the early stages of pregnancy and protect against thrombosis of the uteroplacental vasculature after successful placentation. Future studies should be aimed at refining the protocol used in this trial to determine the benefits of preconceptual administration of heparin and whether it can be stopped after 13 weeks' gestation without adversely affecting the rate of live births.
We thank Elizabeth Thomas and Margaret Murphy for performing the bone density scans; Katy Clifford for her constant support and advice; and Julia Wilson, Sue Head, Andrea Thompson, and Tracey McGrath for their help with patient follow up. Finally, we thank the patients and their referring clinicians, without whose enthusiastic support this study would not have been possible.
Funding: Arthritis and Rheumatism Council
Conflict of interest: None.