Population Pharmacokinetics of Enoxaparin During the Antenatal PeriodClinical Perspective
Background—The optimal dosing strategy of low-molecular-weight heparins for the treatment of antenatal venous thromboembolism is not known. The physiological changes associated with pregnancy alter the pharmacokinetic profile of low-molecular-weight heparins, which has led to controversy and subsequent variation in practice, when pregnant women with venous thromboembolism are treated with low-molecular-weight heparins. Our objective was to develop a robust pharmacokinetic model of enoxaparin during the antenatal period to address this problem.
Method and Results—Women prescribed antenatal enoxaparin were eligible to enroll in the study. Recruited women were reviewed monthly and had up to 3 anti-Xa activities (trough and 1 and 3 hours after dose) drawn at each clinic attendance. Compartmental pharmacokinetic modeling was conducted using nonlinear mixed-effects modeling. One hundred twenty-three patients contributed 795 anti-Xa activities for pharmacokinetic modeling purposes. Both enoxaparin clearance and volume of distribution were increased during pregnancy. Simulations of once- versus twice-daily enoxaparin administration demonstrated that both dosing regimens would reach target 3-hour plasma concentrations throughout the duration of the pregnancy. When trough anti-Xa activity was simulated, both once- and twice-daily regimens exhibited an increase in trough anti-Xa activity with the progression of pregnancy. This is explained by the significant increase in volume of distribution observed during pregnancy.
Conclusions—The half-life of enoxaparin is prolonged with the progression of pregnancy, and our work provides compelling evidence for prescribing once-daily enoxaparin for the treatment of antenatal venous thromboembolism. National and international guideline recommendations should be reconsidered.
Medication use during pregnancy is common, with estimates suggesting that >80% of women take a prescribed medication at some point during pregnancy.1 For some women, the prescription medication is essential for maintaining maternal (and fetal) health for conditions often unrelated to pregnancy, for example, epilepsy, asthma, diabetes mellitus, and HIV.2 The evidence informing the doses of medications prescribed during pregnancy is often poor, representing, at best, the findings of small, observational, antenatal studies, so women might receive suboptimal doses. Venous thromboembolism (VTE) is a case in point, where anticoagulant therapy is prescribed during the gravid period, but where the evidence base is poor. After the first successful reports involving the use of low-molecular-weight heparins (LMWHs) for the management of antenatal VTE >20 years ago,3,4 LMWHs have superseded unfractionated heparin as the agents of choice for this indication. Whilst extensive reviews have concluded that LMWHs are safe during pregnancy,5,6 many uncertainties remain on how best to manage women using an LMWH for antenatal VTE, especially whether LMWH should be prescribed once or twice a day.7 The physiological changes associated with the gravid state are known to alter the pharmacokinetics of LMWH; the glomerular filtration rate is reported to increase by up to 50% by 15 weeks’ gestation.8 This increase is maintained for the remainder of pregnancy and begins to normalize only as term approaches. Plasma volume also increases by up to 50% over the course of pregnancy.8 These changes directly affect both the clearance and volume of distribution of LMWH. This has led many to believe that because the glomerular filtration rate increases, so must LMWH clearance, resulting in the need for a higher dose or more frequent dosing. Current national and international guidelines capture this confusion by suggesting either a once- or twice-daily dose of LMWH for this indication. For example, the current United Kingdom Royal College of Obstetricians and Gynaecologists (RCOG) guidelines for the management of antenatal VTE suggest a twice-daily dose of enoxaparin or dalteparin,9 whereas the American College of Chest Physicians guidelines state that a once- or twice-daily dose could be used, acknowledging the uncertainty that exists around the optimal LMWH dosing regimen for this indication.10 The evidence for increasing the total daily dose or dosing frequency during pregnancy is based on small, historical, observational studies.11–27 Prescribing a twice-daily dosing regimen of LMWH has significant implications for women with respect to injection burden, particularly if the VTE event occurs early during pregnancy.
Clinical Perspective on p 1469
We developed a population pharmacokinetic model28 to better understand the specific pharmacokinetic changes in LMWH during pregnancy and, through pharmacokinetic simulation, determined the optimal dosing strategy to use during pregnancy for the treatment of antenatal VTE.
Pregnant women ≥18 years of age who were referred to the specialist thrombophilia clinic at King’s College Hospital were eligible for recruitment to the study. King’s College Hospital is a 900-bed teaching hospital based in South East London where ≈5400 babies are delivered each year. Women prescribed enoxaparin for all antenatal indications were eligible for recruitment to the study. Women <18 years of age, those with renal dysfunction (creatinine clearance <60 mL/min), and women with suspected nonadherence to enoxaparin were excluded. After providing informed written consent, the recruited subjects attended the thrombophilia clinic monthly during pregnancy for ongoing monitoring, recording of weight, and drawing of blood samples for analysis of renal and liver function and a full blood count. Additionally, to determine the antenatal pharmacokinetic profile of enoxaparin, women had up to 3 anti-Xa activities drawn at each clinic visit: trough and 1 and 3 hours after enoxaparin injection. The anti-Xa sampling strategy was based on d-optimal design to ensure that gestation-related changes in enoxaparin clearance and volume of distribution could be estimated.29 As afforded by the population pharmacokinetic method, women were prescribed differing doses of enoxaparin according to their specific indications, and clinicians were able to alter the dose of enoxaparin when clinically indicated during the course of the pregnancy. Recruited women were asked to record the time they injected the enoxaparin each day in a diary, which was collected and replaced at subsequent clinic visits.
Determination of Anti-Xa Activity
Anti-Xa activity was used to characterize the activity of enoxaparin in a subject’s plasma. For determination of anti-Xa activity, a 2.7-mL blood (9 vol) sample was collected in a 0.109M (3.2% trisodium citrate) Becton-Dickinson Vacutainer. After collection, the sample was centrifuged in a Rotina 420 R centrifuge (Hettich Zentrifugen), double spun for 7 minutes at 2500g, and frozen within 1 hour of sample collection. The samples were stored at −40°C until analyzed. Samples were thawed and analyzed in weekly batches with the STA-Rotachrom heparin colorimetric assay (Diagnostica Stago, France) on the STA-R evolution analyser (Diagnostica Stago) in the laboratory at King’s College Hospital. Results were reported as international units per milliliter. The lower limit of quantification for this assay was 0.01 IU/mL. The quality of anti-Xa activity determination was assessed by comparison with human plasma samples containing a predetermined level of LMWH (STA-Quality LMWH; Diagnostica Stago). Interassay precision (coefficient of variation) for quality-control samples ranged from 3.7% to 4.6%.
Pharmacokinetic Modeling and Simulation
Pharmacokinetic analysis was conducted using nonlinear mixed-effects modeling, an approach endorsed by both the US Food and Drugs Administration30 and the European Medicines Agency.31 Population pharmacokinetic modeling fits mathematical models to describe pharmacokinetic data that arise from >1 individual.32 The method does not require each subject to provide sufficient data to characterize his or her own pharmacokinetic profile because pharmacokinetic information is shared between individuals to develop the population pharmacokinetic profile.32 The method therefore allows the use of sparse sampling study designs, having an obvious advantage when applied in the clinic setting. A typical population pharmacokinetic model is integrated with both a covariate and a statistical model.33,34 The covariate model describes relationships between pharmacokinetic parameters and patient characteristics. The statistical model describes the variance in pharmacokinetic between and within individuals, as well as residual variance resulting from biological variability, measurement errors, and errors in the fit of the model to the data. Examples of the clinical application of population pharmacokinetic approach in an antenatal setting include the study of lumefantrine in women with uncomplicated multidrug-resistant Plasmodium falciparum malaria35 and the study of tenofovir in HIV-1–infected pregnant women.36
Developing the Antenatal Pharmacokinetic Enoxaparin Model
Initially, several structural base models were developed, for example, a 1- and 2-compartment model, and the model that best fit the data was selected. Goodness-of-fit plots, a statistical improvement in the fit of the model to the data using the objective function (minus twice the log-likelihood of the data), assessment of the precision of the parameter estimates, and residual variability were the criteria used to evaluate and choose the base model to take forward for full covariate analysis. The specific covariates evaluated as part of this analysis were those that had a mechanistic meaning: age, ethnicity, weight, creatinine, creatinine clearance, baseline lean body weight, and gestation. The covariate analysis involved a classic graphical approach, plotting individual estimates of the pharmacokinetic parameters random effects (ETAs) against each of the covariates being considered for inclusion in the model. Each selected covariate was then tested by univariate addition into the base model to confirm its relevance. A decrease in objective function value of at least 6.64 (P<0.01) was required to retain the covariate in the intermediate model. All significant covariates were then simultaneously added to the base model, and their continued relevance was evaluated with a stepwise backward elimination method in which each covariate was removed singularly from the model. An increase in the objective function value of >10.82 (P<0.001) was required to retain the covariate in the final model.
Both the base and the final models were evaluated with the use of a nonparametric bootstrap procedure (1000 replicates).37 The final model was additionally evaluated with a visual predictive check38 in which the 5th, 50th, and 95th prediction intervals, simulated from the posterior distribution of the final model parameter estimates, were overlaid with the 5th, 50th, and 95th percentiles from the observed data. A well-performing model would see the observed percentiles and simulated prediction intervals superimposed.
Finally, simulations from the final antenatal enoxaparin pharmacokinetic model developed were conducted to determine how enoxaparin exposure would change during pregnancy under different dosing regimens, that is, if women injected once-daily enoxaparin (according to the King’s College Hospital protocol) or twice-daily enoxaparin (according to current RCOG guidelines for enoxaparin; see Tables 1 and 2 for details of these dosing schedules). The simulations were based on women beginning enoxaparin on day 1 of their pregnancy and delivering at 40 weeks gestation and assumed that subjects injected at the same time each day, that there was no dose change throughout the pregnancy, and that there was 100% adherence to treatment.
The study received ethics approval from the Isle of White, Portsmouth, and South East Hampshire ethics committee (REC reference, 09/H0501/57; September 2009).
During a 28-month recruitment period (September 2009–December 2011), 158 women were referred to the thrombophilia clinic. Thirty-five of these women were not recruited or did not contribute data to this study for the following reasons: 1 patient had a previous allergy to enoxaparin and was therefore managed on tinzaparin during the index pregnancy; 8 women miscarried before providing informed consent; 6 women declined to participate; 3 women were not approached to take part in the study; 7 women prescribed enoxaparin attended clinic erratically and thus were not approached because any information that would be collected from these women was considered too unreliable; 4 women were moving out of the area and therefore were referred to their new local provider for ongoing obstetric and hematology care; the clinicians suspected nonadherence to enoxaparin for 3 women; 1 woman developed an allergy (rash) to enoxaparin; 1 women miscarried after providing informed consent; and the principal researcher could not obtain any blood samples from 1 woman. Therefore, 123 women consented to participate and contributed data for this study. Table 3 describes the demographic information of the women recruited and the associated indication and doses of enoxaparin prescribed.
Because of the hospital’s location in South East London, one third of the women were of African-Caribbean origin, one fifth were class I obese or greater, and two thirds were prescribed antenatal enoxaparin for VTE prophylaxis.
The recruited women contributed 795 anti-Xa activities for pharmacokinetic modeling purposes, of which 712 were drawn during the antenatal period. Figures 1 and 2 illustrate the time after dose and gestational stage at which these anti-Xa activities were drawn. These figures illustrate the breadth and depth of the anti-Xa activities drawn during the course of this study.
Pharmacokinetic Model Development
Several base models were initially explored (ie, 1- and 2-compartment models). A 1-compartment model with between-subject variability on enoxaparin clearance and volume of distribution with a combined (additive and proportional) error model best fit the data. The online-only Data Supplement provides information on parameter estimates for key base models explored (Table I in the online-only Data Supplement) and how data below the limit of quantification were handled. Table 4 gives parameter estimates on the final chosen base model with the associated bootstrap results.
A full covariate analysis was then conducted to evaluate the following selected covariates: weight, creatinine, creatinine clearance, baseline lean body weight, and gestation (see Table 5). All the covariates evaluated met the criteria for inclusion in the final model. At this stage, it made no physiological sense to include both creatinine and creatinine clearance as covariates on clearance. Therefore, creatinine was removed because of the smaller drop in the objective function when creatinine was added to the base model compared with when creatinine clearance was added to the base model. All remaining covariates were simultaneously added to the base model, and then backward elimination of each covariate was executed to ensure the continued relevance of a particular covariate in the presence of the other covariates. Table 6 lists the results from this analysis and shows that creatinine clearance was removed at this stage.
The final enoxaparin pharmacokinetic estimates for typical values of clearance (CL) and volume of distribution (Vd) can be represented mathematically as follows: CL=POPCL×GESTCL×(Wt/80)1.06 and Vd=POPVd×GESTV×(LBW/42)1.25, where during pregnancy GESTCL=1 and GESTV=(GEST/MGEST)0.357 and after delivery GESTCL=(GEST/MGEST)−4 and GESTV=(GEST/MGEST)−5.49. POPCL is the population estimate for enoxaparin clearance; POPVd refers to the population estimate for enoxaparin volume of distribution; GESTCL and GESTV represent the change in CL and Vd over gestation and after delivery, with GEST representing the specific gestational stage (weeks) of the woman; Wt is the subject’s actual body weight (in kilograms); LBW is the subject’s lean body weight (in kilograms; before pregnancy, calculated according to a mechanistically derived lean body weight equation41); and MGEST is the gestational week at which the woman delivers.
Table 7 lists the typical values of clearance and volume of distribution from the final model developed, along with the associated bootstrap results. Finally, a visual predictive check was conducted (see Figure 3). The fact that the observed and predicted anti-Xa activities are superimposed in the visual predictive check illustrates that the final model developed was a robust antenatal enoxaparin pharmacokinetic model.
Having established the antenatal enoxaparin pharmacokinetic model, the next step was to simulate the important “what if” clinical question: How does anti-Xa activity change with gestation when the same dose is injected through the course of pregnancy? Simulations from the final model were conducted, comparing anti-Xa activity of women injecting enoxaparin according to RCOG antenatal dosing guidelines with those injecting according to King’s once-daily dosing guidelines (Figure 4A and 4B).
Figure 4A and 4B demonstrates that with both regimens, the current RCOG 3-hour target of 0.5 to 1.2 would be achieved without a need for a change in dose for the duration of pregnancy. The trough activity is more informative. The simulations show an increase in anti-Xa trough activity with the progression of pregnancy. This occurs as a result of an increase in the volume of distribution, prolonging the half-life of enoxaparin, and demonstrates that it would be appropriate to prescribe a once-daily dose of LMWH for the management of antenatal VTE.
This study is the largest pharmacokinetic study undertaken in an antenatal setting of any LMWH to date and demonstrates that as pregnancy progresses, although 3-hour anti-Xa activity decreases, trough anti-Xa activity increases. This decrease in 3-hour anti-Xa activity and the increase in trough activity can be explained by an increase in the apparent volume of distribution and should be expected given the physiological changes known to occur during pregnancy. Although clearance increases during pregnancy, so does the volume of distribution; the elimination half-life is a function of both clearance and volume of distribution. Our work demonstrates that the increase in volume of distribution during pregnancy leads to a prolongation of enoxaparin half-life, so once-daily dosing is adequate for this group of patients.
We are not the first to describe an increase in trough anti-Xa activity as pregnancy progresses. Lebaudy and colleagues42 in an evaluation of enoxaparin pharmacokinetics during the antenatal period reported an increase in the trough activity of enoxaparin as pregnancy progressed. They reported an increase in the clearance and volume of distribution for enoxaparin of 48% and 49%, respectively. Furthermore, the Casele and colleagues15 early noncompartmental pharmacokinetics study of prophylactic enoxaparin in 13 women reported a decrease in the maximum anti-Xa activity at 30 to 33 weeks’ gestation compared with 12 to 15 weeks’ gestation; however, the calculated mean resident time of enoxaparin rose from 12 to 15 weeks’ to 30 to 33 weeks’ gestation from 287 to 325 minutes. This finding can again be explained by the increase in volume of distribution of enoxaparin.
Surveys of antenatal anticoagulation for the management of VTE from the United Kingdom and Ireland have already reported that once-daily LMWH is commonly used by many clinicians, presumably for patient practicality and acceptability reasons. Voke and colleagues43 reported once-daily LMWH use in 66% of the total population of 126 women being treated for antenatal VTE, whereas Knight44 reported once-daily LMWH use in 49% of 143 women being treated for antenatal pulmonary embolism. The results from our study provide evidence to support the use of once-daily LMWH in this setting for this indication and would explain why few complications have been reported historically when a once-daily dose has been used. We acknowledge that some clinicians may feel apprehensive about using a once-daily LMWH regimen during the first few weeks after diagnosis, particularly in those women who suffer from extensive thrombosis or who suffer their event early in their pregnancy. In such situations, an initial twice-daily strategy followed by once-daily LMWH might be justifiable.
Although studies monitoring peak anti-Xa activity continue to suggest that the dose of LMWH should be increased during pregnancy, to maintain the peak anti-Xa activity,45 our work clearly shows that this recommendation is misplaced. We believe that monitoring 3-hour/peak anti-Xa activity is not appropriate in the antenatal setting because the result will reflect the changes seen in both clearance and volume of distribution, rendering any interpretation, and therefore subsequent dose change, difficult. Although we agree with current guidelines that most women do not require routine anti-Xa monitoring, if anti-Xa activity is monitored, for example, when clinicians are concerned about a lack of response or for women at the extremes of weight, monitoring the trough concentration would appear more informative.
Our findings have wider implications for the use of LMWH for other obstetric indications, particularly in women with mechanical heart valves. The current American College of Chest Physicians guidelines suggest maintaining a 4-hour peak anti-Xa activity between 1 and 1.2 IU/mL. We suggest that trough activity monitoring would be more informative because increasing the dose of LMWH to maintain 4-hour peak anti-Xa activity with the progression of pregnancy may result in overanticoagulation and the patient experiencing a bleed.
The limitations of our work should be considered. The simulation work is based on the assumption that women inject the enoxaparin at the same time each day and that adherence was complete (ie, 100%). Other work conducted by our group suggests that this is unlikely.46 One would anticipate, however, that injecting LMWH once a day would have only a positive impact on women’s adherence during pregnancy.
This study has developed a robust pharmacokinetic antenatal enoxaparin model, the simulations from which demonstrate that although clearance increases, so does volume of distribution, leading to a prolongation of the half-life of enoxaparin. Our study provides compelling evidence for the use of once-daily LMWH for the management of antenatal VTE without a need for a change in LMWH dose. Current national and international guidelines should be reconsidered in light of our findings.
Source of Funding
We acknowledge financial support from the Department of Health via the National Institute for Health Research Comprehensive Biomedical Research Center Award to Guy’s & St. Thomas’ NHS Foundation Trust in partnership with King’s College London and King’s College Hospital NHS Foundation Trust.
We thank all the women who agreed to participate in the study, the specialist haematology registrars who rotated through the thrombophilia clinic while the study was underway, and the laboratory staff who processed the anti-Xa activity assays.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIRCULATIONAHA.113.003198/-/DC1.
- Received April 12, 2013.
- Accepted August 5, 2013.
- © 2013 American Heart Association, Inc.
- Greer IA,
- Nelson-Piercy C
- 9.↵Royal College of Obstetricians and Gynaecologists. Green-top Guidelines, Number 28: Thromboembolic Disease in Pregnancy and the Puerperium: Acute Management. November 2007. http://www.rcog.org.uk/files/rcog-corp/GTG37b_230611.pdf. Accessed September 3, 2013.
- Huxtable LM,
- Tafreshi MJ,
- Ondreyco SM
- 30.↵UF Food and Drug Administration. Innovation or stagnation: challenge and opportunity on the critical path to new medical products report, 2004. http://www.fda.gov/ScienceResearch/SpecialTopics/CriticalPathInitiative/CriticalPathOpportunitiesReports/ucm077262.htm. Accessed August 27, 2011.
- 31.↵European Medicines Agency, Evaluation of Medicines for Human Use. Innovative drug development approaches: final report from the EMEA/CHMP Think Tank Group on Innovative Drug Development, 2007. http://www.ema.europa.eu/docs/en_GB/document_library/Other/2009/10/WC500004913.pdf. Accessed August 27, 2011.
- Hirt D,
- Urien S,
- Ekouévi DK,
- Rey E,
- Arrivé E,
- Blanche S,
- Amani-Bosse C,
- Nerrienet E,
- Gray G,
- Kone M,
- Leang SK,
- McIntyre J,
- Dabis F,
- Tréluyer JM
- Tarning J,
- McGready R,
- Lindegardh N,
- Ashley EA,
- Pimanpanarak M,
- Kamanikom B,
- Annerberg A,
- Day NP,
- Stepniewska K,
- Singhasivanon P,
- White NJ,
- Nosten F
- Davison AC,
- Hinkley DV
- Beal S,
- Sheiner LB,
- Boeckmann A,
- Bauer RJ
- 40.↵R Development Core Team. R: a language and environment for statistical computing. Vienna, Austria: A Foundation for Statistical Computing; 2006. (http://www.R-project.org).
- Lebaudy C,
- Hulot JS,
- Amoura Z,
- Costedoat-Chalumeau N,
- Serreau R,
- Ankri A,
- Conard J,
- Cornet A,
- Dommergues M,
- Piette JC,
- Lechat P
Low-molecular-weight heparins (LMWHs) are the agents of choice for the treatment of antenatal venous thromboembolism. However, the optimal dosing of LMWH for the treatment of antenatal venous thromboembolism remains controversial, with many guidelines suggesting that a twice-daily dose of LMWH is required. Observational studies from clinical practice suggest that many clinicians do not adhere to the guidelines, with up to 50% already prescribing once-daily LMWH. This study applied the method of population pharmacokinetic modeling to characterize the pharmacokinetics of enoxaparin during pregnancy. Our study found that although the clearance of enoxaparin increases, so does the enoxaparin volume of distribution, so much so that the half-life of enoxaparin is actually prolonged with the progression of pregnancy. This study is not the first to report this, and the results provide compelling evidence for the use of once-daily LMWH for the majority of cases. The results explain why complications have not been observed when clinicians have prescribed LMWH once daily during pregnancy. Our study should provide confidence to those clinicians already prescribing once-daily LMWH for the management of venous thromboembolism, and we believe the current guidelines should be reconsidered in light of our findings. There will always be instances when clinicians may be apprehensive about using once-daily LMWH from the outset, for example, in patients with massive pulmonary embolism. In such situations, a twice-daily dose might be used until the patient’s condition has settled, after which a switch to once-daily LMWH would be appropriate.