Portal de Periódicos da CAPES

Prevention of pre-eclampsia by low-molecular-weight heparin in addition to aspirin: a meta-analysis

2015; Wiley; Volume: 47; Issue: 5 Linguagem: Inglês

10.1002/uog.15789

1469-0705

S. Roberge, S. Demers, K. H. Nicolaides, Marc Bureau, Stéphane Côté, Emmanuel Bujold,

Maternal and fetal healthcare

To estimate the impact of adding low-molecular-weight heparin (LMWH) or unfractionated heparin to low-dose aspirin started ≤ 16 weeks' gestation on the prevalence of pre-eclampsia (PE) and the delivery of a small-for-gestational-age (SGA) neonate. A systematic review and meta-analysis of randomized controlled trials (RCTs) was performed by searching the medical databases PubMed, EMBASE, Web of Science and Cochrane Central. Pregnant women randomized to receive LMWH or unfractionated heparin in addition to low-dose aspirin were compared with those who received low-dose aspirin alone. Outcome measures were PE, severe PE, early-onset PE and SGA. Pooled relative risks (RRs) with 95% CI were calculated using a random-effects model. Eight RCTs met the inclusion criteria; the indication for recruitment was previous recurrent miscarriage in five studies (three included women with thrombophilia) and a history of severe or early-onset PE in three studies (including women with thrombophilia in one). LMWH was administered in seven studies and unfractionated heparin in one. In women with a history of PE, treatment with LMWH and aspirin, compared with aspirin alone, was associated with a significant reduction in development of PE (three trials (n = 379); RR, 0.54 (95% CI, 0.31–0.92); P = 0.03) and in delivery of SGA neonates (two trials (n = 363); RR, 0.54 (95% CI, 0.32–0.91); P = 0.02). These outcomes were not significantly reduced in women with recurrent miscarriage who received LMWH and aspirin, compared with aspirin alone. The small number of studies precluded sensitivity analyses and the evaluation of publication biases. Blinding to the allocation treatment was absent in all RCTs. Based on limited evidence, the addition of LMWH to low-dose aspirin could reduce the prevalence of PE and SGA in women with a history of PE. This observation should be the basis of a well-conducted future trial rather than a recommendation for immediate clinical application. Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd. Estimar el impacto de añadir heparina de bajo peso molecular (HBPM) o heparina no fraccionada (HNF) a aspirina en dosis bajas iniciado antes de la semana 16 de gestación, sobre la prevalencia de preeclampsia (PE) y el parto de recién nacidos pequeños para la edad gestacional (PEG). Se llevó a cabo una revisión sistemática y metaanálisis de ensayos controlados aleatorios (ECA) mediante la búsqueda en las bases de datos médicas PubMed, EMBASE, Web of Science y Cochrane Central. Se comparó un grupo aleatorio de mujeres embarazadas a quienes se suministró HBPM o heparina no fraccionada además de dosis bajas de aspirina, con uno que solo recibió dosis bajas de aspirina. Las medidas de resultado fueron PE, PE severa, PE de aparición temprana y PEG. Se calcularon los riesgos relativos (RR) combinados con un IC del 95% mediante un modelo de efectos aleatorios. Ocho de los ECA cumplieron los criterios de inclusión; la indicación de reclutamiento fue el aborto previo recurrente en cinco estudios (tres incluían mujeres con trombofilia) y un historial de PE severa o de aparición temprana en tres estudios (uno de ellos con mujeres con trombofilia). La HBPM se administró en siete estudios y la heparina no fraccionada en uno. En las mujeres con antecedentes de PE, el tratamiento con HBPM y aspirina, en comparación con aquellas en tratamiento con sólo aspirina, se asoció a una reducción significativa en el desarrollo de PE (tres ensayos (n = 379); RR, 0,54 (IC 95%, 0,31–0,92); P = 0,03) y al parto de recién nacidos PEG (dos ensayos (n = 363); RR, 0,54 (IC 95%, 0,32–0,91); P = 0,02). Estos resultados no fueron significativamente menores en mujeres con abortos recurrentes que recibieron HBPM y aspirina, en comparación con las que tan solo recibieron aspirina. El pequeño número de estudios impidió hacer un análisis de sensibilidad y evaluar los sesgos de publicación. En ninguno de los ECA hubo asignación ciega del tratamiento. Basado en la evidencia limitada, la adición de HBPM a dosis bajas de aspirina podría reducir la prevalencia de PE y PEG en mujeres con un historial de PE. Esta observación debería constituir la base de un futuro ensayo bien realizado, en vez de una recomendación para la aplicación clínica inmediata. Pre-eclampsia (PE), which occurs in 2–8% of pregnancies, is associated with evidence of impaired placentation1, 2. Placental vascular lesions and incomplete transformation of uterine spiral arteries are also found in some pregnancies that are complicated by the delivery of small-for-gestational-age (SGA) neonates2. Meta-analyses of randomized controlled trials (RCTs) have reported that, in pregnancies at high risk of PE, the prophylactic use of low-dose aspirin started ≤ 16 weeks' gestation can reduce the prevalence of PE and SGA3-5. The role of heparin in prevention of these conditions is becoming increasingly apparent, mainly due to its antithrombotic and anti-inflammatory effects, similar to those of aspirin6. A Cochrane review concluded that prophylactic use of heparin in high-risk pregnancies may reduce the prevalence of PE and SGA; however, the study included women who did not receive low-dose aspirin7. A recent randomized trial in 292 pregnant women with thrombophilia, not included in the Cochrane review, reported that prophylactic use of low-molecular-weight heparin (LMWH) does not reduce the occurrence of placenta-mediated pregnancy complications; however, the relative risk (RR) for women who were concomitantly taking low-dose aspirin was 0.30 (95% CI, 0.1–1.1) compared with 1.20 (95% CI, 0.7–2.2) for those who were not taking aspirin8. As it has been demonstrated that aspirin is effective in early gestation, the objective of this meta-analysis was to determine if the combined treatment of LMWH and low-dose aspirin starting ≤ 16 weeks' gestation is superior to the administration of low-dose aspirin alone in the prevention of PE and SGA in women at risk9. Included in this meta-analysis were RCTs that compared the administration of a combination of any type of LMWH or unfractionated heparin and aspirin with the administration of aspirin alone, started ≤ 16 weeks' gestation in women at risk of hypertensive disorders. Studies that included other treatment or cotreatment were excluded. An electronic search was made of PubMed, EMBASE, Web of Science and Cochrane Central to identify studies published from 1945 to March 2015; references of systematic reviews were investigated to identify further articles. No language restriction was imposed. A combination of the following keywords and MeSH terms were included: 'heparin', 'bemiparin', 'certoparin', 'dalteparin', 'enoxaparin', 'nadroparin', 'parnaparin', 'reviparin', 'tinzaparin', 'LMWH', 'aspirin', 'acetylsalicylic acid', 'preeclampsia', 'pre-eclampsia', 'PE', 'toxaemia', 'toxemia' and 'eclampsia'. Screening of the titles of identified articles was performed by one reviewer (S.R.). The relevant selected abstracts were screened separately by two reviewers (S.R. and M.B.). Final selection and data extraction from the complete articles was performed independently by two reviewers (M.B. and E.B.) and any discrepancies were resolved by discussion. The quality of studies was evaluated using the Cochrane Handbook Criteria tool for judging risk of bias, and studies with high risk of bias were considered for sensitivity analysis10. The quality and integrity of this review were validated with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)11. The outcome measures were PE, early-onset PE and SGA. The definition of PE was that of the American College of Obstetricians and Gynecologists as systolic blood pressure (BP) ≥ 140 mmHg or diastolic BP ≥ 90 mmHg occurring > 20 weeks' gestation in a woman with previously normal BP plus proteinuria, defined as urinary excretion of ≥ 0.3 g protein in a 24-h urine specimen or 2+ protein on dipstick12. Considering the variations in the definition of PE around the world, similar definitions were accepted13. Early-onset PE was defined as PE ≤ 34 weeks of gestation. The definition of SGA was a neonatal birth weight below the 10th, 5th or 3rd percentile (the first available in this order) for gestational age at delivery, or an equivalent. Pooled RRs were computed for dichotomous outcomes, using Review Manager 5.2.3 software. Global RRs were calculated according to the DerSimonian and Laird random-effects model in case of significant heterogeneity and variance, and fixed-effects model in case of homogeneity between studies. Heterogeneity between studies was analyzed using the Higgins I2 statistic14, 15. The distribution of trials was examined using funnel plots to assess publication bias16. It was planned to conduct sensitivity analyses to investigate robustness of the findings and to assess heterogeneity between studies, with comparison of the inclusion criteria, blinding, study quality and trial size. The search of databases identified 1807 potentially eligible studies and 250 were selected for further evaluation. Eight (885 participants) met the selection criteria and were included in the analysis (Figure 1). Gestational age at treatment randomization varied among the trials but all women were randomized between the time of a positive pregnancy test and 14 weeks' gestation. Three RCTs used dalteparin17-20, two used enoxaparin21, 22, two did not specify the type of LMWH23, 24 and one used unfractionated heparin25. Three studies recruited women with a history of PE, one of which included only women with thrombophilia, and five studies recruited women with at least two consecutive miscarriages, three of which included only women with thrombophilia. The dose of aspirin varied between 75 and 100 mg/day. The characteristics of each trial, including the number of patients, inclusion criteria, treatment and outcome measures are summarized in Table 1. Due to the wide heterogeneity between inclusion criteria for the two subgroups of women, we stratified our results according to the entry criteria and used a random-effects model. We observed that, in women with a history of PE, the addition of LMWH to low-dose aspirin reduced the risk of PE (three trials (n = 379); RR, 0.54 (95% CI, 0.31–0.92); P = 0.03) and SGA (two trials (n = 363); RR, 0.54 (95% CI, 0.32–0.91); P = 0.02), whereas the risks for these outcomes were not significantly reduced in the subgroup of women with recurrent miscarriage (PE: two trials (n = 211); RR, 0.57 (95% CI, 0.08–4.35); P = 0.59; SGA: three trials (n = 337); RR, 0.73 (95% CI, 0.18–2.99); P = 0.66 (Figures 2 and 3)). Moreover, LMWH with low-dose aspirin appears to show a trend in the reduction of early-onset PE (two trials (n = 155); RR, 0.14 (95% CI, 0.02–1.10); P = 0.06) but not late-onset PE (two trials (n = 155); RR, 1.20 (95% CI, 0.53–2.72); P = 0.65) in women with a history of PE. The Higgins I2 statistic demonstrated no heterogeneity between studies for PE and early-onset PE (I2 = 0%) and low heterogeneity for SGA (I2 = 22%). The small number of included RCTs precluded appropriate analysis of the funnel plot and other sensitivity analysis and therefore we cannot exclude selection biases. According to the Cochrane Handbook Criteria tool for judging risk of bias, the majority of included studies were judged to have low or unclear risk of bias, except for the blinding of the participants as no study used a placebo in the control group (Figure 4)10. The findings of this systematic review and meta-analysis suggest that, in women who developed PE in a previous pregnancy, the combination of LMWH and low-dose aspirin started in early pregnancy is superior to low-dose aspirin alone for the prevention of early-onset PE and the delivery of a SGA neonate. In women with recurrent miscarriage, there appeared to be no benefit, in terms of occurrence of PE or SGA, of adding LMWH compared with treatment by aspirin alone. A trial including 135 participants that evaluated LMWH without low-dose aspirin vs no treatment did not find a significant reduction in adverse perinatal outcome (risk difference, 2.2 (95% CI, −11.6 to 16.0); P = 0.76)26. Similarly, as mentioned earlier, the TIPPS trial including 292 participants showed that LMWH without aspirin had no effect on adverse outcome, whereas LMWH with aspirin was associated with a non-significant reduction in adverse outcomes8. However, in a trial by Rey et al.27 evaluating LMWH vs no treatment in 116 participants, but in which about 90% of women randomized to LMWH also received low-dose aspirin, a significant reduction in PE (odds ratio (OR), 0.15 (95% CI, 0.03–0.70)) and severe PE (OR, 0.11 (95% CI, 0.01–0.90)) was reported. Mello et al.28 reported a beneficial effect of LMWH in reducing severe PE (RR, 0.25 (95% CI, 0.08–0.86)) in a trial including 85 participants, but it was not specified whether or not the participants were taking aspirin. Finally, Schleussner et al.29 did not observe a reduction in intrauterine growth restriction, placental insufficiency or PE with administration of LMWH alone, started between 5 and 8 weeks' gestation, in a trial of 449 women with recurrent miscarriage. On the basis of the results of the above trials and our meta-analysis, it could be suggested that, in high-risk pregnancies, the prophylactic use of LMWH is beneficial in reducing adverse pregnancy outcome only when there is concomitant therapy with low-dose aspirin. The main limitation to our analysis relates to the small number of studies and patients fulfilling the inclusion criteria. Consequently, the possibility of bias in results could not be assessed by funnel plot and sensitivity analysis, and therefore we cannot exclude the possibility of publication bias. Other limitations include heterogeneity in inclusion criteria, dose of aspirin and outcome measures. Half (4/8) of the studies included women with thrombophilia. Finally, our study did not allow for evaluation of the cost benefits or side effects of LMWH, or its impact on perinatal mortality and long-term maternal or infant outcomes. There is good evidence that, in women with a history of previous PE, the prophylactic use of low-dose aspirin starting ≤ 16 weeks' gestation can substantially reduce the rate of preterm PE, SGA and perinatal death9. There is some evidence that, to maximize the beneficial effect of aspirin, it would be preferable to use 100–160 mg rather than 75–81 mg, and that the drug should be administered at bed time30-32. The limited evidence presented in this study that the use of LMWH may have a beneficial effect in addition to that of low-dose aspirin should be the basis of a well-conducted future trial rather than a recommendation for immediate clinical application. We observed that the benefits of adding LMWH to aspirin are potentially limited to the reduction of early-onset PE and SGA, two disorders typically related to impaired placentation, in comparison with late-onset PE that does not have the same etiology33. Any strategy aiming at the prevention of early-onset PE, a disease affecting only a very small number of pregnant women, should rely on an effective method of screening that would identify the high-risk group in the first trimester of pregnancy. A large population-based study has reported that only about 15% of pregnancies that develop preterm PE have a history of previous PE34. Moreover, the prevalence of early-onset PE has been reported to be as low as 2% in women with previous history of PE but increasing to 14% when such history was combined with an abnormal first-trimester uterine artery Doppler assessment35. Therefore, future trials evaluating the impact of LMWH should focus on women at high risk of early-onset placenta-mediated complications of pregnancy based on such previous history or using a combined screening with adequate positive predictive values. It has been reported that a combination of maternal characteristics and medical history, uterine artery Doppler, mean arterial pressure and serum placental growth factor at 11–13 weeks' gestation can identify more than 90% of cases of early-onset PE36. Larger clinical trials, including women at risk of PE, should consider combining LMWH with aspirin for its prevention. Emmanuel Bujold, Suzanne Demers and Stéphanie Roberge are supported by the Fonds de la recherche du Québec – Santé (FRQ-S). Emmanuel Bujold holds the Jeanne et Jean-Louis Lévesque Perinatal Research Chair at Université Laval.