Should we recommend universal aspirin for all pregnant women?
Prevention of preeclampsia and fetal growth restriction
LDA that is commenced after the first trimester has been demonstrated to reduce the risk of preeclampsia by 24% and sequelae, such as pre-term birth and fetal growth restriction, by 14% and 20%, respectively, when given to women who are at risk of preeclampsia. Based on such evidence, the administration of LDA is now recommended for women with 1 major risk factor or 2 moderate risk factors for preeclampsia, as per guidelines from the National Institute for Health and Care Excellence, UK. These recommendations are echoed by the US Preventative Services Task Force. The ACOG differs slightly in opinion and recommends LDA usage in women with a history of preeclampsia in >1 pregnancy or a history of preeclampsia that required delivery at <34 weeks gestation. The use of a maternal history approach for screening, based on 1 major risk factor (such as a previous preeclamptic pregnancy, chronic hypertension, diabetes mellitus, and systemic lupus erythematosus), which is similar to that proposed by the ACOG when set at a false-positive rate of 5%, can predict 37% and 28.9% or early-onset and late-onset preeclampsia. The use of the National Institute for Health and Care Excellence, UK, or US Preventative Task approach can improve these rates to 89.2% and 93.0% respectively.
LDA at a dosage of 60–150 mg has been tried and tested in pregnancy for decades and has robust safety data to support its use beyond the first trimester. The CLASP study (Collaborative Low-dose Aspirin Study in Pregnancy) was 1 of the original and largest randomized controlled trials to assess efficacy and safety of aspirin in the prevention of preeclampsia in at-risk pregnant women and concluded that, in >9000 subjects, LDA was generally safe for the fetus and newborn infant, with no evidence of an increased likelihood of maternal or fetal bleeding. The findings of the CLASP study have been supported through further studies that have concluded that there is no association between the use of LDA in the third trimester of pregnancy and neonatal intraventricular hemorrhage, neonatal bleeding, or antenatal closure of the ductus arteriosus. Use of aspirin beyond the first trimester of pregnancy is not associated with an increased risk of congenital anomalies, nor is it associated with increased maternal risk in terms of major postpartum bleeding, placental abruption, or adverse regional anesthetic outcome. There is an abundance of robust data to support the safety of LDA beyond the first trimester, most recently a systematic review from the US Preventative Services Task Force 2014 concluded that LDA usage was not associated with maternal or neonatal harm, with normal follow-up to an 18-year period.
In addition to being regarded as a safe drug in pregnancy, LDA is cheap in cost and is available widely. Bearing in mind the high costs to health services worldwide for treatment of preeclampsia and its sequelae, which encompass care for the mother and the neonate, it may be suggested that universal provision of aspirin for prevention of preeclampsia is more economically viable. The issue is knowing for which group of women is it more cost-effective to offer preeclampsia prevention, whether it is for all women or just for those with major risk factors or whether it is for those that test positive on a formalized screening test algorithm. This may explain the differing opinions in international guidelines on whom it is appropriate to offer LDA. A recent study has demonstrated that it is more cost-effective to treat all pregnant women universally with LDA or via a screened approach adopting the US Task Force policy. As addressed within this study, the introduction of screening tests have limitations in terms of their application and interpretation. Screening tests, such as the Fetal Medicine Foundation algorithm, incorporating first-trimester uterine artery Doppler assessment, maternal history, mean arterial blood pressure, and biomarker assessment, may introduce additional cost with resource implications, although a cost-effectiveness analysis compared with other measures has yet to be assessed.
Despite advances in the management of preeclampsia and a reduction in the rates of maternal death in the developed world, there remains disparity worldwide, with hypertensive disorders being 1 of the 3 major causes of maternal death worldwide and remaining the leading cause of death in regions such as Latin America and the Caribbean. Countries where resources are limited, despite the existence of good world health recommendations for preeclampsia recognition and management, may find it a challenge to afford tools to diagnose and manage the disease or to facilitate safe delivery and neonatal care. Where access to tools for screening, diagnosis, and treatment of preeclampsia are limited, 1 potential option is the universal provision of LDA, which may provide a cheaper alternative and could potentially reduce mortality rates in such countries.
It has been suggested that the most cost-effective approach to the reduction of preeclampsia is likely to be the provision of an effective, affordable, and safe intervention that is applied to all mothers without previous testing to assess levels of risk. Universal aspirin for all may be the answer to this in terms of efficacy because it is regarded as a safe and cheap drug that may serve major benefit in pregnancy on a worldwide scale.
Should we recommend universal aspirin to all pregnant women?
Efficacy in low-risk populations
There are multiple randomized controlled trials and metaanalyses that support a risk reduction in preeclampsia and sequelae in women who have existing risk factors for the disease. In terms of such evidence, studies vary in terms of aspirin dosing, starting times in pregnancy, characteristics of populations included, combinations of other antiplatelet agents, types of analyses conducted, and inclusion of studies that are of variable quality. Hence, metaanalyses provide variable risk reduction rates for preeclampsia. In terms of risk reduction of preeclampsia in low-risk women, this has not been assessed within an appropriately powered randomized controlled trial, which would likely require >5000 participants to assess preeclampsia as an outcome. One may suggest that an evidence-based approach is required on the efficacy of LDA in terms of preeclampsia risk reduction before proceeding with a universal aspirin policy.
Some pregnant women may be nonresponsive to aspirin. Such “aspirin resistance” may be due to multifactorial reasons such as (1) genetic polymorphisms, (2) aspirin adherence, (3) spontaneous regeneration of cyclooxygenase-2 and (4) increased platelet turnover in pregnancy. Aspirin resistance is well-established in cardiovascular research, and the reported prevalence ranges from 5–65%, varying with the assay used and the populations studied. A cohort of women is seen in pregnancy who, despite taking aspirin, still develop preeclampsia; this may represent the aspirin-resistant group. A clearly defined assessment that combines clinical and biochemical parameters in pregnancy to determine the prevalence of aspirin nonresponsiveness is currently the focus of some research groups. Studies suggest that LDA is as effective as high-dose aspirin in terms of cyclooxygenase suppression. However, enteric coating, which most preparations have that are used in pregnancy, can inhibit the cyclooxygenase inhibitory aspirin effect. Pregnant women differ from standard adults taking aspirin in that they have increased platelet turnover and production within the bone marrow, with additional sequestration of platelets in the placenta, which causes more immature platelets to be released from the placenta and increases a tendency for platelet aggregation. Hence, treating women with universal aspirin in pregnancy may only target the aspirin-responders, which may make up a much lower proportion of the population than previously anticipated.
Aspirin compliance and patient acceptability
Women typically are advised not to take any medications during pregnancy; however, up to 80% of women actually may require medications. In terms of compliance with medications in pregnancy, evidence suggests that, in the coexistence of chronic illness or new illness, compliance is high at 90–95% but varies based on drug type and preparation used. In terms of assessment of compliance, methods include assessment of “self-reporting,” such as questionnaires, in addition to prescription logs in the pharmacy; however, these tend to over-estimate compliance, and there may be a variation in the time taken and if the course of medication has been completed. In short, there is currently no robust method to assess aspirin compliance, and there are no studies unique to pregnancy. Hence, we do not know true compliance with aspirin specifically in pregnancy, notably compliance in a low-risk population where efficacy is not known. One must also consider how acceptable it is for women to take LDA in pregnancy if they have no predisposing risk factors for preeclampsia, which is another area that requires assessment.
Aspirin safety in pregnancy beyond the first trimester is widely known. LDA commenced before the completion of the first trimester has been demonstrated to be associated with a small increase in the incidence of gastroschisis (approximately 5/100,000 vs 12/100,000 deliveries). In terms of maternal risks, LDA is associated with a 10% increase in gastrointestinal symptoms, hence a specialized enteric-coated preparation is recommended. LDA uncommonly can cause hypersensitivity reactions and is contraindicated in those women who are aspirin-allergic or have aspirin-sensitive asthma. LDA is not associated with elevated rates of placental abruption, major postpartum hemorrhage, or intracranial neonatal hemorrhage, although it has been demonstrated to triple rates of vaginal bleeding in pregnancy (1.3% vs 3.9%; P =.004), as demonstrated in the recent Effects of Aspirin in Gestation and Reproduction randomized controlled trial that set out to determine whether the use of LDA (81 mg once daily) that was commenced preconceptually until 36 weeks gestation improved live birth rates in women who had had 1 previous pregnancy loss. Although LDA was not associated with an increased risk of miscarriage, it may not be acceptable to women with no predisposing risk factors for preeclampsia to experience additional concern and require further assessment of bleeding in pregnancy over a small theoretic preeclampsia risk reduction. This may also have further resource implications to health services if women attend as an emergency for assessment of vaginal bleeding in pregnancy. The recommended dosage of aspirin is unknown, although studies appear to prescribe it at a dosage of 60–150 mg once daily, depending on the preparations manufactured in respective countries. A once-daily dosage appears to be acceptable, and the effects of LDA do not appear to be half-life dependent but more so on the individual rate of platelet turnover. Doses of aspirin <81 mg, as used in the United States, are sufficient in maintaining COX-1 inhibition in nonpregnant subjects and hence appears to be sufficient in the absence of aspirin resistance. LDA appears to be most effective if commenced before placental implantation, typically at <16 weeks gestation. The optimal policy would be to dispense LDA at the lowest effective dose, which has optimal efficacy and which is freely and easily acceptable within a population. One must weigh the potential side-effects and acceptability of universal LDA for pregnant women and the risks of potential preeclampsia development against a screen-and-treat-all policy, which may involve additional investigations that are not acceptable to women.
The pathologic explanation behind preeclampsia is not fully understood. The rate of disease reduction is unknown with LDA use. Hence, one must consider whether universal LDA for the prevention of a disease, which we do not fully understand and do not know the efficacy of the proposed agent, will benefit the population. It may not be acceptable to women who have no risk factors to take a drug that may increase their risk of vaginal bleeding. Additionally, women may not be aspirin responders or may not be compliant with medication. Further research is required to support such a policy.
The authors report no conflict of interest.