REVIEW

Safety and Efficacy of Enoxaparin in Pregnancy:A Systematic Review and Meta-Analysis

Barry Jacobson . Virendra Rambiritch . Dara Paek . Tobias Sayre .

Poobalan Naidoo . Jenny Shan . Rory Leisegang

Received: September 23, 2019 / Published online: October 31, 2019� The Author(s) 2019

ABSTRACT

Introduction: International guidelines supportthe use of low molecular weight heparins for thetreatment of thromboembolism and thrombo-prophylaxis during pregnancy. However, evi-dence of the benefit and harm associated withspecific low molecular weight heparins such as

enoxaparin is dated. No current systematicreview and meta-analysis describing the safetyand efficacy of enoxaparin for thromboem-bolism and thromboprophylaxis during preg-nancy exists.Methods: PubMed, Embase, and Cochranedatabases were searched on August 17, 2018 forclinical trials or observational studies in preg-nant women receiving enoxaparin; patientswith a prosthetic heart valve were excluded.Risk ratios (RR) with 95% confidence intervals(CI) were calculated using a random effectsmodel, and heterogeneity was measured usingthe I2 statistic.

Enhanced Digital Features To view enhanced digitalfeatures for this article go to https://doi.org/10.6084/m9.figshare.9933413.

Electronic supplementary material The onlineversion of this article (https://doi.org/10.1007/s12325-019-01124-z) contains supplementary material, which isavailable to authorized users.

B. JacobsonDepartment of Haematology and MolecularMedicine, University of Witwatersrand, 1 Jan SmutsAve, Johannesburg, South Africa

V. RambiritchDepartment of Pharmacology, University ofKwaZulu-Natal, University Rd, Westville, SouthAfrica

D. Paek � T. SayreDoctor Evidence, 301 Arizona Ave, Santa Monica,CA, USA

P. NaidooSanofi Affiliate, Medical, 2 Bond St, Grand Central,Midrand, South Africa

P. NaidooDepartment of Health Informatics, School of HealthProfessions, Rutgers, State University of New Jersey,57 US Highway 1, New Brunswick, NJ, USA

J. ShanSanofi Global, Medical, 54-56 Rue la Boetie, Paris,France

R. Leisegang (&)Department of Pharmaceutical Biosciences, UppsalaUniversity, 752 36 Uppsala, Swedene-mail: rory.leisegang@gmail.com

R. LeisegangFAMCRU, The Department of Paediatrics & ChildHealth, Stellenbosch University, Tygerberg, CapeTown, South Africa

Adv Ther (2020) 37:27–40

https://doi.org/10.1007/s12325-019-01124-z

Results: Of the 485 records identified in thesearch, 24 studies published clinical trials, andobservational studies were found dating back to2000. Only one observational cohort and onerandomized control trial focused on the use ofenoxaparin for thromboprophylaxis and there-fore efficacy was not assessed; the other studiesincluded women with recurrent pregnancy loss(15 studies), history of placental vascular com-plications (five studies), and recurrent in vitrofertilization failure (two studies) and weretherefore analyzed in terms of safety only.Bleeding events were non-significantly moreoften reported for enoxaparin compared tountreated controls (RR 1.35 [0.88–2.07]) but lessoften reported for enoxaparin versus aspirin (RR0.93 [0.62–1.39]); thromboembolic events,thrombocytopenia, and teratogenicity wererarely reported events; in patients with a historyof recurrent pregnancy loss, encouragingly therates of pregnancy loss were significantly lowerfor enoxaparin compared to untreated controls(RR 0.58 [0.34–0.96]) and enoxaparin ? aspirinversus aspirin alone (RR 0.42 [0.32–0.56]) as wellas observably lower for enoxaparin versusaspirin alone (RR 0.39 [0.15–1.01]), though sig-nificant heterogeneity was observed (I2[60).Conclusion: Literature on the efficacy andsafety of enoxaparin for thromboembolism andthromboprophylaxis remains scanty, andtherefore efficacy was not assessed; in terms ofsafety, when including other indications forenoxaparin in pregnancy, we found thatenoxaparin was associated with significantlylower complications than aspirin. Given differ-ences in study design and study heterogeneity,pregnancy loss results should be interpretedwith caution. Moreover, reports of throm-boembolic events, thrombocytopenia, andcongenital malformations were rare.Funding: Sanofi.

Keywords: Cardiology; Enoxaparin; Lowmolecular weight heparin (LMWH); Preg-nancy; Thromboembolism; Thromboprophy-laxis; Venous thromboembolic events (VTE)

Key Summary Points

Venous thromboembolic (VTE) disease is amajor cause of morbidity and mortalityduring pregnancy and the puerperium.Enoxaparin, a low molecular weightheparin (LMWH), is standard of care forthe management of VTE risk inpregnancy, after individualizedbenefit–risk assessment. This systematicreview and meta-analysis demonstrate thesafety of enoxaparin in pregnancy.

Though reports of thromboembolicevents, thrombocytopenia, andteratogenicity were rare, enoxaparin useduring pregnancy was associated withlower occurrence of pregnancy loss whencompared with controls and similar ratesof bleeding.

Both real-world studies and clinical trialswere assessed in this analysis, and, in lightof the adoption of antenatal LMWH inclinical practice, real-world data andcontinued safety reporting will helpfurther inform future treatment guidelinessurrounding this important topic.

INTRODUCTION

Pregnancy is a pro-thrombotic state and hasbeen associated with a fourfold increase in therisk of thromboembolic events [1, 2]. Venousthromboembolic events (VTE), which includedeep vein thrombosis (DVT) and pulmonaryembolism (PE), are one of the leading causes ofmaternal mortality in developed countries[3, 4]. Treatment options for prevention andtreatment of VTE are limited, because of thematernal and fetal risk of anticoagulation, andonly a limited number of drugs are consideredsafe and prescribed during pregnancy.

Unfractionated heparin (UFH) and lowmolecular weight heparin (LMWH) do not crossthe placenta and pose low risk of fetal exposure

28 Adv Ther (2020) 37:27–40

[5, 6]. LMWHs have potential benefits overUFH, such as higher bioavailability, longer half-life, and more predictable response. These fac-tors together allow for once or twice dailyadministration with minimal laboratory moni-toring, which makes LMWH ideal for outpatientuse [5]. International guidelines now supportthe use of LMWH over UFH and other oralanticoagulant agents for the treatment of VTEand thromboprophylaxis in pregnancy follow-ing appropriate risk factor assessment [7–13].However, recommendations are largely basedon consensus or expert opinion, and furtherhigh-quality research is still needed to guideclinical decisions.

Systematic reviews and meta-analyses whichhave examined antenatal LMWH for VTE interms of the benefit and harm associated withspecific LMWHs such as enoxaparin are dated[14]. The aim of this review was to update theevidence surrounding the safety and efficacy ofenoxaparin in pregnancy, focusing on thethromboprophylaxis efficacy and safety, butincluding other indicators in terms of safety.

METHODS

A systematic literature review was conductedfollowing a standardized review protocol usingthe PICOTSS framework, which outlined thepopulation, interventions, comparators, out-comes, timing, setting, and study designs ofinterest. This review followed standardmethodology for conducting systematic reviewsas per guidelines provided by the Centre forReviews and Dissemination [15], NationalInstitute for Health and Care Excellence [16],and the Cochrane Handbook for SystematicReviews of Interventions [17]. Results werereported according to the Preferred ReportingItems for Systematic Reviews and Meta-Analyses(PRISMA) guidelines [18].

This review is based on previously conductedstudies and does not contain any studies withhuman participants or animals performed byany of the authors.

Literature Search

A comprehensive search was conducted inPubMed, Embase, and the Cochrane Register ofControlled Trials for publications available upto August 17, 2018. Conference proceedingsfrom 2015 to 2018 were also searched throughEmbase. The search strategies are provided inTable S1 of the Electronic SupplementaryMaterial. Electronic literature searches wereperformed by medical librarians from DoctorEvidence, LLC using a proprietary softwareplatform (Doctor Evidence: Library Manage-ment System. Santa Monica, CA: Doctor Evi-dence, LLC) [19].

Study Selection

Given the particular concerns around maternaland fetal safety, we explored whether anyantenatal use of enoxaparin has been associatedwith increased risks of adverse outcomes. Stud-ies were included in the review if they met thefollowing criteria: (1) study population ofpregnant women; (2) clinical trials or observa-tional studies examining antenatal use ofenoxaparin compared to another non-enoxa-parin group including placebo; (3) outcomesincluded incidence of thromboembolic events,bleeding, thrombocytopenia, pregnancy loss,and congenital malformations. To summarizethe available literature on the potential benefitsand harms of enoxaparin use during pregnancy,women requiring thromboprophylaxis or otherpregnancy-related indications were included.Patients with mechanical heart valves wereexcluded. The study eligibility criteria are pro-vided in Table S2 (Electronic SupplementaryMaterial).

The initial screening of references retrievedin the search was performed by two indepen-dent medical librarians to identify potentiallyrelevant studies based on the titles andabstracts. The full texts of studies were thenreviewed, and detailed reasons for inclusion orexclusion were documented. The initial agree-ment was above 85%. Any disagreements wereresolved by discussion or an independent third-party reviewer.

Adv Ther (2020) 37:27–40 29

Data Extraction and Quality Assessment

Data extraction included study-level character-istics, treatment information, and the prespec-ified safety/efficacy outcomes of interest. Dataextraction was carried out by two clinical dataassociates using the Digital Outcome Conver-sion (DOC) Data Version 2.0 Software platform(Doctor Evidence, LLC, Santa Monica, CA,USA). All characteristic and outcome terms werecollected as reported in each paper, and syn-onyms were bound using the DOC OntologySystem (DOC Data version 2.0. Santa Monica,CA: Doctor Evidence, LLC).

Quality assessments were conducted by twoindependent reviewers, and any disagreementbetween the reviewers was resolved by discus-sion. Data were stored andmanaged inMicrosoftExcel. The Cochrane Collaboration tool forassessing the risk of bias in randomized trials wasused to assess clinical trials [20]. This instrumenthas been used to evaluate seven domains of bias:random sequence generation (selection bias),allocation concealment (selectionbias), blindingof participants and personnel (performancebias), blinding of outcome assessment (detectionbias), incomplete outcome data (attrition bias),selective reporting (reporting bias), and othersources of bias (other bias). The quality of obser-vational cohort studies was assessed using theNewcastle–Ottawa Scale (NOS) [21]. The NOSassigns up to amaximum of 9 points (or stars) forthe least risk of bias in three categories: (1)Selection of study groups (maximum 4 points);(2) Comparability of study groups (maximum 2points); and (3) Outcome ascertainment (maxi-mum 3 points). For the second item in the out-come category, ‘‘2. Was follow-up long enoughfor outcomes to occur?’’, a follow-up of 40 weeks(9 months) ormorewas considered adequate. Forthe third item in the outcome category, ‘‘3. Ade-quacy of follow-up of cohorts’’, a follow-up rateof at least 90% or a description provided of thoselost was considered adequate to assign a point.

Analysis

Analyses were conducted for comparisons ofenoxaparin (alone and in combination with

aspirin) versus placebo/untreated controls,aspirin, and other anticoagulation comparatorgroups where there was sufficient data reported.Traditional pairwise meta-analyses were per-formed for the outcomes of bleeding and preg-nancy loss. Only qualitative analyses werefeasible for thromboembolic events, thrombo-cytopenia, and congenital malformations. Rel-ative risk/risk ratios and 95% confidenceintervals for outcomes were determined using arandom effects model. For studies that reportedno events for a single treatment arm, a conser-vative statistical continuity correction wasapplied by assigning the group a 0.5 eventvalue. Studies that reported no events in bothtreatment arms were excluded from the meta-analysis. Heterogeneity between studies wasestimated using the I2 statistic. If significantheterogeneity was observed (I2[ 50), differ-ences in study design, single vs. multicentersetting, and primary indication were examined.

RESULTS

A total of 485 records were identified in thesearch, of which 422 records were excludedduring title/abstract screening. On the basis of areview of the full texts of 63 studies, an addi-tional 39 studies were excluded, primarily forinappropriate comparison (i.e., no comparisonbetween different treatments) and for notreporting the outcomes of interest. Twenty-fourstudies met eligibility criteria and were includedfor data extraction. The PRISMA flow diagram ispresented in Fig. 1.

There were 18 randomized controlled trials(RCTs), two prospective non-randomized stud-ies, two prospective observational studies, andtwo retrospective observational studies, withpublication dates ranging between 2000 and2017. Study populations included women withrecurrent pregnancy loss (15 studies), history ofplacental vascular complications (five studies),recurrent in vitro fertilization failure (twostudies), and thromboprophylaxis (two studies).The characteristics of studies and reported out-comes are summarized in the Electronic Sup-plementary Material (Table S5).

30 Adv Ther (2020) 37:27–40

Twenty publications [22–41], pertaining to19 clinical trials [22–33, 35–41] were included inthe Cochrane risk of bias assessment. A sum-mary of the assessment for the randomized tri-als is presented (Table S3 in the ElectronicSupplementary Material). The risk of bias acrossthe included studies was generally low andlimited to the blinding of participants and per-sonnel (performance bias) and the blinding ofoutcome assessment (detection bias): 14 studieswere judged to be at low risk for the randomsequence generation item in the selection biasdomain [23–33, 38–40], three studies werejudged as unclear, as the random sequencegeneration methods were not clearly described[22, 35, 40], and four studies were consideredhigh risk as they were non-randomized[36, 37, 42, 43]. Regarding allocation conceal-ment, 14 studies were considered to be at lowrisk of bias [22–25, 27, 28, 30–32, 35, 36, 38, 40,41], four studies were considered unclear asmethods for allocation concealment were notprovided or not clearly described, one study wasconsidered as high risk as randomization wasnot carried out and the groups were made awareof the treatment they received upon enrollment[37]. Few studies presented adequate blindingmethods to reduce performance bias: only threestudies were considered as low risk for suffi-ciently packaging their treatments to disguisetheir identities to patients receiving them at allpoints in the study [28, 38, 41], eight studieswere considered unclear as blinding methodswere not provided, ten studies were open-labelfor participants and personnel and were thusconsidered to be at high risk for performancebias [22, 24, 25, 27, 30–33, 37, 40]. Similarly, fordetection bias, only five studies were consideredlow risk [28, 30, 38, 39, 41], six studies were ofunclear risk as the study was not described aseither open-label or blinded[23, 29, 35, 36, 42, 43], and nine studies wereconsidered high risk because of an open-labelstudy design [22, 24–27, 31, 33, 37, 40]. Allstudies were considered as low risk in thedomains of incomplete outcome data (attritionbias) and selective reporting (reporting bias).Most studies used an intent-to-treat analysis intheir primary outcome analysis, and wheredropouts were present, the numbers were few

and similar between groups. All studies reportedon their prespecified primary outcome, asmentioned in the publication. Finally, no othersources of bias other than the ones previouslydescribed were detected across the studies.

Four studies were included in the NOSquality assessment, all of which were cohortstudies [42–45]. A summary of the assessment ispresented in the Electronic SupplementaryMaterial (Table S4). One study had a score of 8and three studies scored the maximum score of9. All studies scored the maximum score of 4stars in the selection category, indicating thatthe data used was representative of the targetedcommunity. For the comparability category,one study scored only one star because theauthors did not describe any additional baselinecharacteristics that were well matched betweencohorts [42] and the other three studies scoredthe maximum score of 2 stars [43–45]. Finally,in the outcome category, all studies scored themaximum of 3 stars.

Thromboembolic Events: Efficacyand Safety

Only two studies looked at enoxaparin in thecontext of thromboprophylaxis; a pooled meta-analysis was therefore not possible.

In terms of safety of enoxaparin, a pooledmeta-analysis for thromboembolic events wasnot feasible because of the low number ofreported cases across treatment comparisons.Data reported in the studies are summarized inTable S6 (Electronic Supplementary Material).Two cases of DVT were reported in patients withrecurrent pregnancy loss who received enoxa-parin monotherapy compared to four cases inpatients who received no intervention (RR 0.5;95% CI 0.09–2.69) [22]. No thromboembolicevents were reported in patients receivingcombined therapy with enoxaparin plus aspirin.In one study that compared enoxaparin plusaspirin versus aspirin in women with a historyof preeclampsia, one case of superficial venousthrombosis was reported in patients whoreceived aspirin [34].

Adv Ther (2020) 37:27–40 31

Bleeding

Results of the pooled meta-analyses for bleedingare presented in Table 1 and Fig. 2. The inci-dence of bleeding events with enoxaparinmonotherapy was 35% higher compared toplacebo/no treatment controls. When com-pared to aspirin, risk of bleeding was 7% lowerwith enoxaparin monotherapy and 5% lowerwith enoxaparin plus aspirin. These results werenot statistically significant, and heterogeneitybetween the studies was low. Enoxaparin wascompared to tinzaparin in one study in womenwith recurrent pregnancy loss [45]. No bleedingevents were reported in either treatment group.In another study in women with recurrentpregnancy loss who also tested positive forantiphospholipid antibodies, three bleedingevents were reported in patients receivingenoxaparin plus aspirin compared to two eventsin those who received UFH plus aspirin [37].Bleeding event data are summarized in Table S7(Electronic Supplementary Material).

Pregnancy Loss

Results of the pooled meta-analyses for preg-nancy loss are presented in Table 1 and Fig. 3.

Risk of pregnancy loss was reduced by 58%when enoxaparin monotherapy was comparedto control groups, and significant heterogeneitywas observed (Fig. 3a). Though there was a largenumerical difference, reduction in pregnancyloss for enoxaparin monotherapy compared toaspirin did not achieve statistical significance(Fig. 3b). A sensitivity analysis removing onestudy with no events in the enoxaparin groupdid not change the result. Pregnancy loss wassignificantly reduced by 42% with enoxaparinplus aspirin compared to aspirin alone(p\ 0.0001) (Fig. 3c). In a sensitivity analysisbased on the study design, a lower risk ratio wasfound for pregnancy loss in the observationaldata than the RCT data. Enoxaparin plus aspirinwas still statistically favored over aspirin andthe study design did not appear to influencethis result. A non-statistically significant reduc-tion of 66% was found when enoxaparin plusaspirin was compared to UFH plus aspirin(Table 1). Pregnancy loss reported in the studiesis summarized in Table S8 (Electronic Supple-mentary Material).

24 records included in evidence synthesis

24 records accepted (based on full text)

39 references excluded in full-text screening:Comparison: 13Outcomes: 13Study design: 4Intervention: 3Duplicate publication: 3Population: 3

63 records accepted (based on title and abstract)

485 records screened (based on title and abstract)

422 records excluded in title-abstract screening:Study design: 253Not a clinical study: 54 Population: 43In vitro study: 20Not a treatment study: 15 Intervention: 14Duplicate publication: 8Outcomes: 4Comparison: 4Animal study: 3 Publication date cutoff: 3 Not English: 1

485 records identified through MEDLINE, Embase and Cochrane

Elig

ibili

tyIn

clud

edSc

reen

ing

Iden

tific

atio

n

Fig. 1 PRISMA diagram

32 Adv Ther (2020) 37:27–40

Thrombocytopenia

Few cases of thrombocytopenia occurred in thestudies (Table S9 in the Electronic Supplemen-tary Material). In two RCTs in women withrecurrent pregnancy loss, ten events werereported for enoxaparin and three events forplacebo [22, 38]. Thrombocytopenia withenoxaparin plus aspirin was reported in onestudy, which reported two cases for enoxaparinplus aspirin and four cases for aspirin, in womenwith a history of preeclampsia [34].

Congenital Malformations

Eight cases of congenital malformations werereported for enoxaparin, two cases for aspirin,and five cases for placebo in three RCTs inwomen with recurrent pregnancy loss[24, 26, 38]. In an observational study in

women with a history of obstetric and/orthromboembolic complications, single cases ofcongenital malformations were reported inpatients receiving enoxaparin plus aspirin inthe first trimester or aspirin only. There were nocases observed with second trimester enoxa-parin plus aspirin (Table S10 in the ElectronicSupplementary Material).

Maternal Mortality

Maternal mortality was reported in one studyand no deaths were reported in either ofenoxaparin plus aspirin and aspirin groupsduring the study.

Allergic Reaction

Allergy or skin reactions were reported in threestudies. One of the studies reported no allergy,and the remaining two studies reported aroundone to four more cases of allergic reaction whencompared to the standard care or aspiringroups.

Study Withdrawal

Study withdrawal was reported in eight studies.However, five reported zero discontinuations inthe remaining three studies, and each of themreported a unique treatment comparisoninvolving enoxaparin.

DISCUSSION

The findings of this systematic review and meta-analysis were that antenatal enoxaparin use wasgenerally safe, but that enoxaparin efficacyspecifically for antenatal VTE could not beexamined as a result of only two studies meet-ing the inclusion criteria. Given that the use ofenoxaparin in pregnancy remains off-label,careful individual benefit–risk assessment andinformed consent are still required before itsuse. In terms of efficacy. These findings arealigned with those of the review conducted byGreer et al. [14] in LMWH in general.

Table 1 Pooled relative risks of bleeding events andpregnancy loss

Outcome No. ofstudies

RR (95% CI)Random effects;I2

Bleeding

Enoxaparin vs

placebo/controls

5 1.35 (0.88–2.07);

I2 = 16.05

Enoxaparin vs aspirin 3 0.93 (0.62–1.39);

I2 = 0

Enoxaparin ? aspirin vs

aspirin

4 0.95 (0.71–1.28);

I2 = 11.28

Pregnancy loss

Enoxaparin vs

placebo/controls

6 0.58 (0.34–0.96);

I2 = 65.80

Enoxaparin vs aspirin 4 0.39 (0.15–1.01);

I2 = 78.24

Enoxaparin ? aspirin vs

aspirin

8 0.42 (0.32–0.56);

I2 = 14.10

Enoxaparin ? aspirin vs

UFH ? aspirin

2 0.66 (0.33–1.34);

I2 = 0

Adv Ther (2020) 37:27–40 33

34 Adv Ther (2020) 37:27–40

In terms of safety, the incidence of throm-boembolic events was rare, with only twoantenatal cases of DVT with enoxaparin repor-ted in a single study. More bleeding eventsoccurred with enoxaparin than placebo/un-treated controls, but the risk of bleeding waslower compared to aspirin. Risk of pregnancyloss was lower with enoxaparin compared toplacebo/untreated controls, aspirin, and UFHplus aspirin. Statistically significant reductionin risk of pregnancy loss was shown for enoxa-parin plus aspirin compared to aspirin alone.Few studies reported thrombocytopenia andcongenital malformations occurring withenoxaparin, the majority reporting only one ortwo events.

In terms of safety, anticoagulant treatmentwith enoxaparin is expected to have higherbleeding rates when compared to placebo. Arecent systematic review and meta-analysisconducted by Sirico et al. [46] included datafrom 22,162 pregnant women, of whom 1320(6%) were administered LMWH. AlthoughLMWH-treated patients had a higher risk ofpostpartum hemorrhage, no difference in meanblood loss or risk of blood transfusion at deliv-ery was reported. The authors attributed thehigher risk for postpartum hemorrhage toLMWH usage in spite of there being no differ-ence in mean blood loss or in risk of bloodtransfusion at delivery but noted ‘‘the smallnumber of women included in the secondaryanalysis (784 for mean blood loss at delivery,883 for blood transfusion) compared to thewomen included in the primary analysis.’’Moreover, the inclusion of retrospective cohortdata and the lack of adjustments for con-founders do impact the generalizability andconclusions of this study. The increased bleed-ing in enoxaparin-treated patients compared toplacebo observed in our review, together withthe findings in the review by Sirico et al. [46],supports the practice of assessing risk/benefit

when prescribing enoxaparin in pregnancy;caution is required when using enoxaparin inpregnancy and is a trade-off between maternalbleeding and prevention of VTE.

In terms of safety, enoxaparin was associatedwith a reduction in pregnancy loss in thisanalysis, which is an unexpected result, as asignificant benefit of LMWH for prevention ofrecurrent pregnancy loss has not been observedin other systematic reviews and meta-analyses[47, 48]. Observational data and RCTs wereincluded in our review, and this finding must beinterpreted cautiously as a safety signal ratherthan an efficacy signal. Adequately poweredrobust clinical studies are required to demon-strate the efficacy of enoxaparin to preventrecurrent pregnancy loss in patients withacquired or inherited thrombophilia. One suchongoing study is ALIFE-2, a robust RCT, whichis investigating the efficacy of enoxaparin inpreventing pregnancy loss in women withinherited thrombophilia and recurrent miscar-riage [49]. There is evidence that enoxaparindoes not benefit patients with pregnancy lossunrelated to thrombophilia and the practice oftreating women with enoxaparin for pregnancyloss unrelated to thrombophilia may be futile,and pose an unnecessary risk to the mother[38].

Fetal safety is of particular concern for alldrugs administered during pregnancy. Consid-ering that many women who receive enoxa-parin treatment during pregnancy are already athigh risk for poor pregnancy outcome, we didnot find an incremental fetal risk that wasattributable to enoxaparin. The risk of congen-ital malformations was low and similar to thosereported by Shlomo et al. [50] in a population-based, retrospective cohort study of maternaland infant hospital records between 1998 and2009. Of the 109,473 singleton pregnanciesreported on, 418 and 572 were exposed toenoxaparin during the first and third trimesters,respectively. Importantly, exposure to enoxa-parin during the first trimester of pregnancy wasnot associated with an increased risk of majorcongenital malformations [adjusted odds ratio(aOR) 1.1, 95% CI 0.8–1.6], while exposureduring the third trimester was not associatedwith an increased risk of low birth weight, low

bFig. 2 Relative risk of bleeding events with enoxaparincompared to a placebo/no treatment or b aspirin. c Relativerisk of bleeding events with enoxaparin ? aspirin com-pared to aspirin

Adv Ther (2020) 37:27–40 35

36 Adv Ther (2020) 37:27–40

Apgar score, or risk of perinatal mortality. Theauthors concluded that ‘‘exposure to enoxa-parin during pregnancy was not associated withan increased risk of major malformations ingeneral or according to organ systems.Nonetheless, the risk for specific malformationscannot be ruled out.’’

In our opinion, the low number of reportedcongenital malformations with enoxaparin inthe literature has biological plausibility, giventhat enoxaparin has not been shown to crossthe placenta in clinical studies that havedirectly evaluated whether enoxaparin crossesthe placenta in vivo [51, 52]. Forestier et al. [51]studied anti-Xa and anti-IIa activities in mater-nal and fetal blood after enoxaparin adminis-tration in five cases of medical abortion duringthe second trimester (week 23 of gestation); 3 hafter the subcutaneous injection of 7500 anti-Xa IU of enoxaparin, anti-Xa and anti-IIaactivities in the five fetuses were not modifiedcompared to controls. Dimitrakakis et al. [52]studied the fetal and maternal effects ofenoxaparin 40 mg (4000 IU) in 14 pregnantwomen and 10 controls due for termination oftheir pregnancies because of the presence of amajor congenital malformation; in the 14patients receiving drug, no anti-IIa or anti-Xaactivities were observed 3 h post-dose in ultra-sound-guided fetal blood samples.

Ideally, the safety and efficacy of enoxaparinin pregnancy should be determined in an ade-quately powered, randomized, double-blindcontrolled trial. This study would, however,require an exceptionally large number ofpatients given the low number of reportedevents. It may also be challenging to conduct astudy because of the large sample size, andethical justification for such a trial given thatenoxaparin has become standard of care inter-nationally and one may argue that clinicalequipoise does not exist. A more pragmaticapproach would be to supplement the currentbody of scientific evidence with data from reg-istry studies and pharmacovigilance reports.

Only published studies were included, andtherefore the review may be limited by publi-cation bias. The long-term effect of enoxaparinexposure during pregnancy on neurodevelop-ment requires substantial follow-up and was notreported in the included studies. The gray lit-erature was not searched, and very rare adverseevents may not have been even reported. Stud-ies of pregnant patients with mechanical heartvalves were excluded because these patientsmay have altered enoxaparin pharmacokineticsand pharmacodynamics, when compared topregnant patients without mechanical heartvalves and may benefit from a separate system-atic review. Therefore, the findings of this sys-tematic review and meta-analysis are notgeneralizable to pregnant patients withmechanical valves. Also, given the rarity ofoutcome events, we included both RCT andobservational data, which may present certainlimitations for analysis.

CONCLUSION

In terms of safety, enoxaparin use in pregnancywas associated with a significantly loweroccurrence of pregnancy loss when comparedwith control groups and similar rates of bleed-ing when compared with aspirin; reports ofthromboembolic events, thrombocytopenia,and teratogenicity were rare. The efficacy ofLMWH for the treatment of VTE is alreadyestablished, but evidence for enoxaparinspecifically remains limited. In light of theadoption of antenatal LMWH into clinicalpractice, real-world data and continued phar-macovigilance are needed to further informtreatment guidelines on the clinical benefitsand risks associated with this form of therapy.

ACKNOWLEDGEMENTS

Funding. The systematic search and analysisassistance was provided by Doctor Evidence andfunded by Sanofi, including open access andpublishing fees.

bFig. 3 Relative risk of pregnancy loss with enoxaparincompared to a placebo/no treatment or b aspirin. c Relativerisk of pregnancy loss with enoxaparin ? aspirin com-pared to aspirin

Adv Ther (2020) 37:27–40 37

Medical Writing Assistance. Medical writ-ing support was provided by Jung Min Han andSaranya Chandrudu of Doctor Evidence withpublication support by Angelica Stamegna andfunded by Sanofi.

Authorship. All named authors meet theInternational Committee of Medical JournalEditors (ICMJE) criteria for authorship for thisarticle, take responsibility for the integrity ofthe work as a whole, and have given theirapproval for this version to be published.

Authorship Contributions. Design of Studyand Conceptualization: PN, JS, DP, TS, BJ, RL.Data Curation and Analysis: DP, TS. OriginalDraft Construction: PN, DP, TS, RL. DraftReview and Scientific Revisions: BJ, VR, RL, JS.Final draft approval: BJ, VR, DP, TS, JS, PN, RL.

Disclosures. Barry Jacobson reports havingreceived speaker fees/honoraria from Sanofi,Bayer, Shire, and Aspen, outside of the contentsof this work. Virendra Rambiritch reports pre-vious speaking engagements with Sanofi, out-side of the contents of this work. Tobias Sayrewas an employee of Doctor Evidence at the timeof this analysis. Dara Paek was an employee ofDoctor Evidence at the time of this analysis.Poobalan Naidoo is an employee of Sanofi.Jenny Shan is an employee of Sanofi. Rory Lei-segang has nothing to disclose.

Compliance with Ethics Guidelines. Thisreview is based on previously conducted studiesand does not contain any studies with humanparticipants or animals performed by any of theauthors.

Data Availability. Full datasets generatedduring and/or analyzed during the currentstudy are available from the correspondingauthor on reasonable request. Additional dataare available in the supplementary materials.

Open Access. This article is distributedunder the terms of the Creative CommonsAttribution-NonCommercial 4.0 InternationalLicense (http://creativecommons.org/licenses/by-nc/4.0/), which permits any

noncommercial use, distribution, and repro-duction in any medium, provided you giveappropriate credit to the original author(s) andthe source, provide a link to the CreativeCommons license, and indicate if changes weremade.

REFERENCES

1. Heit JA, Kobbervig CE, James AH, Petterson TM,Bailey KR, Melton LJ. Trends in the incidence ofvenous thromboembolism during pregnancy orpostpartum: a 30-year population-based study. AnnIntern Med. 2005;143(10):697–706.

2. Pomp E, Lenselink A, Rosendaal F, Doggen CJM.Pregnancy, the postpartum period and prothrom-botic defects: risk of venous thrombosis in theMEGA study. J Thromb Haemost. 2008;6(4):632–7.

3. Knight M BK, Tuffnell D, Jayakody H, editors onbehalf of MBRRACE-UK. Saving Lives, ImprovingMothers’ Care: Lessons learned to inform maternitycare from the UK and Ireland Confidential Enqui-ries into Maternal Deaths and Morbidity 2014–16.Oxford: National Perinatal Epidemiology Unit,University of Oxford; 2018.

4. Petersen EE, Davis NL, Goodman D, et al. Vitalsigns: pregnancy-related deaths, United States,2011–2015, and strategies for prevention, 13 States,2013–2017. Morb Mortal Wkly Rep.2019;68(18):423.

5. Hirsh J, Warkentin TE, Raschke R, Granger C,Ohman EM, Dalen JE. Heparin and low-molecular-weight heparin: mechanisms of action, pharma-cokinetics, dosing considerations, monitoring, effi-cacy, and safety. Chest. 1998;114(5):489S–510S.

6. Ostrea EM, Mantaring JB, Silvestre MA. Drugs thataffect the fetus and newborn infant via the placentaor breast milk. Pediatr Clin. 2004;51(3):539–79.

7. American College of Obstetricians and Gynecolo-gists. ACOG Practice Bulletin No. 196: throm-boembolism in Pregnancy. Obstet Gynecol.2018;132(1):e1–17. https://doi.org/10.1097/AOG.0000000000002706.

8. Bates SM, Greer IA, Middeldorp S, Veenstra DL,Prabulos A-M, Vandvik PO. VTE, thrombophilia,antithrombotic therapy, and pregnancy:antithrombotic therapy and prevention of throm-bosis: American College of Chest Physicians evi-dence-based clinical practice guidelines. Chest.2012;141(2):e691S–736S.

38 Adv Ther (2020) 37:27–40

9. Bates SM, Rajasekhar A, Middeldorp S, et al. Amer-ican Society of Hematology 2018 guidelines formanagement of venous thromboembolism: venousthromboembolism in the context of pregnancy.Blood Adv. 2018;2(22):3317–59. https://doi.org/10.1182/bloodadvances.2018024802.

10. Chan WS, Rey E, Kent NE, et al. Venous throm-boembolism and antithrombotic therapy in preg-nancy. J Obstet Gynaecol Can. 2014;36(6):527–53.

11. Royal College of Obstetricians and Gynaecologists.Reducing the risk of thrombosis and embolismduring pregnancy and the puerperium. Green-topGuideline. 2015(37a). London: College of Obstetri-cians and Gynaecologists.

12. Royal College of Obstetricians and Gynaecologists.Thromboembolic disease in pregnancy and thepuerperium: acute management. Green-top Guide-line. 2015(37b). London: College of Obstetriciansand Gynaecologists.

13. Regitz-Zagrosek V, Roos-Hesselink JW, Bauersachs J,et al. 2018 ESC Guidelines for the management ofcardiovascular diseases during pregnancy. Eur HeartJ. 2018;39(34):3165–241. https://doi.org/10.1093/eurheartj/ehy340.

14. Greer IA, Nelson-Piercy C. Low-molecular-weightheparins for thromboprophylaxis and treatment ofvenous thromboembolism in pregnancy: a system-atic review of safety and efficacy. Blood.2005;106(2):401–7.

15. Centre for Reviews and Dissemination. CRD’sguidance for undertaking reviews in health care.Layerthorpe: University of York, Centre for Reviewsand Dissemination; 2009.

16. NICE. Guide to the methods of technology apprai-sal. London: National Institute for Health andClinical Excellence (NICE); 2008.

17. Higgins J, Green S, editors. Cochrane handbook forsystematic reviews of interventions version 5.1. 0.The Cochrane Collaboration. 2011. 2015.

18. Moher D, Liberati A, Tetzlaff J, Altman DG. Pre-ferred reporting items for systematic reviews andmeta-analyses: the PRISMA statement. Ann InternMed. 2009;151(4):264–9.

19. Mazzone PJ, Silvestri GA, Patel S, et al. Screening forlung cancer: chest guideline and expert panelreport. Chest. 2018;153(4):954–85.

20. Higgins JP, Altman DG, Gøtzsche PC, et al. TheCochrane Collaboration’s tool for assessing risk ofbias in randomised trials. BMJ. 2011;343:d5928.

21. Wells G. The Newcastle-Ottawa Scale (NOS) forassessing the quality of nonrandomized studies inmeta-analysis. http://www.ohrica/programs/clinical_epidemiology/oxfordhtm. 2013. Accessed 30 Aug2018.

22. Badawy A, Khiary M, Sherif L, Hassan M, Ragab A,Abdelall I. Low-molecular weight heparin inpatients with recurrent early miscarriages ofunknown aetiology. J Obstet Gynaecol.2008;28(3):280–4.

23. Clark P, Walker ID, Langhorne P, et al. SPIN: theScottish Pregnancy Intervention Study: a multi-centre randomised controlled trial of low molecularweight heparin and low dose aspirin in womenwith recurrent miscarriage. Blood.2010;115(21):4162–7.

24. Dolitzky M, Inbal A, Segal Y, Weiss A, Brenner B,Carp H. A randomized study of thromboprophy-laxis in women with unexplained consecutiverecurrent miscarriages. Fertil Steril.2006;86(2):362–6.

25. Elmahashi MO, Elbareg AM, Essadi FM, Ashur BM,Adam I. Low dose aspirin and low-molecular-weight heparin in the treatment of pregnant Libyanwomen with recurrent miscarriage. BMC Res Notes.2014;7(1):23.

26. Fawzy M, Shokeir T, El-Tatongy M, Warda O, El-Refaiey A-AA, Mosbah A. Treatment options andpregnancy outcome in women with idiopathicrecurrent miscarriage: a randomized placebo-con-trolled study. Arch Gynecol Obstet.2008;278(1):33–8.

27. Fouda UM, Sayed AM, Abdou AMA, Ramadan DI,Fouda IM, Zaki MM. Enoxaparin versus unfrac-tionated heparin in the management of recurrentabortion secondary to antiphospholipid syndrome.Int J Gynecol Obstet. 2011;112(3):211–5.

28. Gates S, Brocklehurst P, Ayers S, Bowler U. Throm-boprophylaxis and pregnancy: two randomizedcontrolled pilot trials that used low-molecular-weight heparin. Am J Obstet Gynecol.2004;191(4):1296–303.

29. Giancotti A, Torre RL, Spagnuolo A, et al. Efficacy ofthree different antithrombotic regimens on preg-nancy outcome in pregnant women affected byrecurrent pregnancy loss. J Maternal Fetal NeonatalMed. 2012;25(7):1191–4.

30. Gris J-C, Chauleur C, Faillie J-L, et al. Enoxaparinfor the secondary prevention of placental vascularcomplications in women with abruptio placentae.Thromb Haemost. 2010;103(04):771–9.

Adv Ther (2020) 37:27–40 39

31. Gris J-C, Chauleur C, Molinari N, et al. Addition ofenoxaparin to aspirin for the secondary preventionof placental vascular complications in women withsevere pre-eclampsia. Thromb Haemost.2011;105(06):1053–61.

32. Gris J-C, Mercier E, Quere I, et al. Low-molecular-weight heparin versus low-dose aspirin in womenwith one fetal loss and a constitutional throm-bophilic disorder. Blood. 2004;103(10):3695–9.

33. Groom KM, McCowan LM, Mackay LK, et al.Enoxaparin for the prevention of preeclampsia andintrauterine growth restriction in women with ahistory: a randomized trial. Am J Obstet Gynecol.2017;216(3):296.e1–14.

34. Haddad B, Winer N, Chitrit Y, et al. Enoxaparin andaspirin compared with aspirin alone to preventplacenta-mediated pregnancy complications.Obstet Gynecol. 2016;128(5):1053–63.

35. Khan E, Basharat A, Jamil M, Ayub S, Khan M.Preventive role of low-molecular-weight heparin inunexplained recurrent pregnancy loss. S Afr J ObstetGynaecol. 2017;23(1):17–9.

36. Mohamed KAA, Saad AS. Enoxaparin and aspirintherapy for recurrent pregnancy loss due to anti-phospholipid syndrome (APS). Middle East FertilSoc J. 2014;19(3):176–82.

37. Noble LS, Kutteh WH, Lashey N, Franklin RD,Herrada J. Antiphospholipid antibodies associatedwith recurrent pregnancy loss: prospective, multi-center, controlled pilot study comparing treatmentwith low-molecular-weight heparin versus unfrac-tionated heparin. Fertil Steril. 2005;83(3):684–90.

38. Pasquier E, de Saint Martin L, Bohec C, et al.Enoxaparin for prevention of unexplained recur-rent miscarriage: a multicenter randomized double-blind placebo-controlled trial. Blood.2015;125(14):2200–5.

39. Qublan H, Amarin Z, Dabbas M, et al. Low-molec-ular-weight heparin in the treatment of recurrentIVF–ET failure and thrombophilia: a prospectiverandomized placebo-controlled trial. Human Fertil.2008;11(4):246–53.

40. Urman B, Ata B, Yakin K, et al. Luteal phaseempirical low molecular weight heparin adminis-tration in patients with failed ICSI embryo transfercycles: a randomized open-labeled pilot trial. HumReprod. 2009;24(7):1640–7.

41. Visser J, Ulander V-M, Helmerhorst FM, et al.Thromboprophylaxis for recurrent miscarriage inwomen with or without thrombophilia. ThrombHaemost. 2011;105(02):295–301.

42. Bar J, Cohen-Sacher B, Hod M, Blickstein D, Lahav J,Merlob P. Low-molecular-weight heparin forthrombophilia in pregnant women. Int J GynecolObstet. 2000;69(3):209–13.

43. Sergio F, Maria Clara DA, et al. Prophylaxis ofrecurrent preeclampsia: low-molecular-weight hep-arin plus low-dose aspirin versus low-dose aspirinalone. Hypertens Pregnancy. 2006;25(2):115–27.

44. Merviel P, Cabry R, Lourdel E, et al. Comparison oftwo preventive treatments for patients with recur-rent miscarriages carrying a C677T methylenete-trahydrofolate reductase mutation: 5-yearexperience. J Int Med Res. 2017;45(6):1720–30.

45. Yuksel H, Kayatas S, Boza AT, Api M, Ertekin AA,Cam C. Low molecular weight heparin use inunexplained recurrent miscarriage. Pak J Med Sci.2014;30(6):1232.

46. Sirico A, Saccone G, Maruotti GM, et al. Lowmolecular weight heparin use during pregnancyand risk of postpartum hemorrhage: a systematicreview and meta-analysis. J Maternal Fetal NeonatalMed. 2019;32(11):1893–1900.

47. de Jong PG, Kaandorp S, Di Nisio M, Goddijn M,Middeldorp S. Aspirin and/or heparin for womenwith unexplained recurrent miscarriage with orwithout inherited thrombophilia. Cochrane Data-base Syst Rev. 2014. https://doi.org/10.1002/14651858.cd004734.pub4.

48. Zhang T, Ye X, Zhu T, et al. Antithrombotic treat-ment for recurrent miscarriage: Bayesian networkmeta-analysis and systematic review. Medicine(Baltimore). 2015;94(45):e1732. https://doi.org/10.1097/MD.0000000000001732.

49. de Jong PG, Quenby S, Bloemenkamp KW, et al.ALIFE2 study: low-molecular-weight heparin forwomen with recurrent miscarriage and inheritedthrombophilia-study protocol for a randomizedcontrolled trial. Trials. 2015;16(1):208.

50. Shlomo M, Gorodischer R, Daniel S, et al. The fetalsafety of enoxaparin use during pregnancy: a pop-ulation-based retrospective cohort study. Drug Saf.2017;40(11):1147–55.

51. Forestier F, Daffos F, Capella-Pavlovsky M. Lowmolecular weight heparin (PK 10169) does not crossthe placenta during the second trimester of preg-nancy study by direct fetal blood sampling underultrasound. Thromb Res. 1984;34(6):557–60.

52. Dimitrakakis C, Papageorgiou P, Papageorgiou I,Antzaklis A, Sakarelou N, Michalas S. Absence oftransplacental passage of the low molecular weightheparin enoxaparin. Pathophysiol HaemostThromb. 2000;30(5):243–8.

40 Adv Ther (2020) 37:27–40