Thrombosis Research 144 (2016) 62–68

Contents lists available at ScienceDirect

Thrombosis Research

journal homepage: www.elsevier.com/locate/thromres

Full Length Article

Low-molecular-weight heparin to prevent recurrent venous

thromboembolism in pregnancy: Rationale and design of the Highlow
study, a randomised trial of two doses
Suzanne M. Bleker a,⁎, Andrea Buchmüller b, Céline Chauleur c, Fionnuala Ní Áinle d, Jennifer Donnelly d,
Peter Verhamme e, Anne Flem Jacobsen f, Wessel Ganzevoort g, Martin Prins h, Jan Beyer-Westendorf i,
Maria DeSancho j, Stavros Konstantinides k, Ingrid Pabinger l, Marc Rodger m,n,
Hervé Decousus b, Saskia Middeldorp a
a
Department of Vascular Medicine, Academic Medical Center, Amsterdam, The Netherlands
b
INSERM, CIC 1408 – F Crin, INNOVTE; CHU Saint-Etienne, Hôpital Nord, Service Médecine Vasculaire et Thérapeutique, F-42055 Saint Etienne, France
c
INSERM, U1059, F-42023, CHU Saint-Etienne, Hôpital Nord, Service de Gynéco-Obstétrique, F-42055 Saint Etienne, France
d
Mater Misericordiae University Hospital and Rotunda Hospital, Dublin, Ireland; SPHERE Research Group, University College Dublin School of Medicine
e
University Hospital Leuven, Center for Molecular and Vascular Biology, Leuven, Belgium
f
Oslo University Hospital, Oslo, Norway
g
Department of Obstetrics and Gynaecology, Academic Medical Center, Amsterdam, The Netherlands
h
Department of Clinical Epidemiology, Maastricht University, Maastricht, The Netherlands
i
Center for Vascular Diseases, Division “Thrombosis Research”, University Hospital “Carl Gustav Carus”, Dresden, Germany
j
Weill Cornell Medical Center New York, New York, United States
k
Center for Thrombosis and Haemostasis, Johannes Gutenberg University of Mainz, Germany
l
Clinical Division of Haematology and Haemostaseology, Medical University of Vienna, Vienna, Austria
m
Departments of Medicine, Epidemiology and Community Medicine, University of Ottawa, Ottawa, Canada
n
Department of Obstetrics and Gynecology, University of Ottawa, Ottawa, Canada

a r t i c l e i n f o s u m m a r y

Article history: Background: Women with a history of venous thromboembolism (VTE) have a 2% to 10% absolute risk of VTE re-
Received 18 March 2016 currence during subsequent pregnancies. Therefore, current guidelines recommend that all pregnant women
Received in revised form 4 May 2016 with a history of VTE receive pharmacologic thromboprophylaxis. The optimal dose of low-molecular-weight
Accepted 1 June 2016 heparin (LMWH) for thromboprophylaxis is unknown. In the Highlow study (NCT 01828697; www.
Available online 3 June 2016
highlowstudy.org), we compare a fixed low dose of LMWH with an intermediate dose of LMWH for the preven-
tion of pregnancy-associated recurrent VTE. We present the rationale and design features of this study.
Keywords:
Prophylaxis
Methods: The Highlow study is an investigator-initiated, multicentre, international, open-label, randomised trial.
Pregnancy Pregnant women with a history of VTE and an indication for ante- and postpartum pharmacologic
Venous thrombosis thromboprophylaxis are included before 14 weeks of gestation. The primary efficacy outcome is symptomatic re-
Clinical trials current VTE during pregnancy and 6 weeks postpartum. The primary safety outcomes are clinically relevant
LMWH bleeding, blood transfusions before 6 weeks postpartum and mortality. Patients are closely monitored to detect
cutaneous reactions to LMWH and are followed for 3 months after delivery. A central independent adjudication
committee adjudicates all suspected outcome events.
Conclusion: The Highlow study is the first large randomised controlled trial in pregnancy that will provide high-
quality evidence on the optimal dose of LWMH thromboprophylaxis for the prevention of recurrent VTE in preg-
nant women with a history of VTE.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Venous thromboembolism (VTE), comprising deep vein thrombosis

⁎ Corresponding author at: Department of Vascular Medicine, Academic Medical
(DVT) and pulmonary embolism (PE), is an important cause of short
Center, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands. and long term morbidity during pregnancy and postpartum periods.
E-mail address: [email protected] (S.M. Bleker). Moreover, PE is one of the leading causes of maternal mortality in

http://dx.doi.org/10.1016/j.thromres.2016.06.001
0049-3848/© 2016 Elsevier Ltd. All rights reserved.
 S.M. Bleker et al. / Thrombosis Research 144 (2016) 62–68 63

developed countries. In the United Kingdom for instance, between 2006 Table 2
and 2008, 0.79 deaths per 100,000 maternities (95% CI 0.49–1.25) were Inclusion- and exclusion criteria of the Highlow study.

attributed to VTE [1–3]. VTE occurs in 1 to 2 per 1000 pregnancies, and Inclusion Age ≥18 years
the risk is 5-fold higher in pregnant women compared to non-pregnant criteria Pregnancy confirmed by urinary pregnancy test, blood test or
women of the same age [2]. The ante- and postpartum incidences of VTE ultrasound examination
Gestational age b14 weeks since the first day of the last menstrual
are similar, but given the much longer duration of the antepartum peri-
period
od than the postpartum period, the daily absolute risk of VTE is highest Previous objectively confirmed VTEa, either:
postpartum [4,5]. The majority of postpartum VTE occur in the first Unprovoked, or
6 weeks after delivery, with event rates decreasing sharply thereafter In the presence of oral contraceptive or estrogen/progestogen
use, or
[6,7].
Related to pregnancy or the postpartum period, or
Women with a personal history of VTE have a 2% to 10% absolute risk In the presence of a minor provoking risk factorb
of developing recurrent VTE during a subsequent pregnancy in the ab- Exclusion Previous VTE related to a major provoking risk factorc as the sole
sence of pharmacologic thromboprophylaxis, with an odds ratio of criteria risk factor
24.8 (95% CI 17.1–36.0) compared to pregnant women without previ- Indication for treatment with a therapeutic dose of anticoagulant
therapy (e.g. acute VTE, atrial fibrillation, a mechanical heart valve,
ous VTE [2,8–10]. Circumstances under which the first VTE occurred in-
recurrent VTE for which an indefinite duration of anticoagulant
fluence the risk of recurrence. In two retrospective studies, women therapy is used prior to pregnancy)
whose first VTE was provoked by the use of oral hormonal contracep- Inability to provide informed consent
tives or was related to pregnancy had a higher risk of recurrent VTE dur- Any contraindication listed in the local labelling of LMWH
ing a subsequent pregnancy compared with women whose first VTE VTE: venous thromboembolism; LMWH: low-molecular-weight heparin.
a
was unprovoked or provoked by a non-hormonal transient risk factor, Patient with a history of extensive superficial thrombophlebitis that was treated as
although these differences did not reach statistical significance [9,10]. deep vein thrombosis (i.e. if it was close to the deep venous system), are also eligible
b
Long distance travel or minor trauma.
Similarly, in a large retrospective cohort of women with prior VTE, c
Surgery, major trauma or plaster cast immobilization in the 3 months prior to VTE.
those who had a history of pregnancy-associated VTE had a higher
risk of recurrence during subsequent pregnancies compared to those
with prior unprovoked VTE (4.5% versus 2.7% respectively; relative
risk 1.7; 95% CI 1.0–2.8) [11].
The American College of Chest Physicians (ACCP) guideline recom- pharmacologic thromboprophylaxis is recommended. In contrast,
mends that all pregnant women with a history of VTE receive postpar- women with a moderate or high risk of VTE recurrence (e.g. women
tum pharmacologic thromboprophylaxis [12]. Low-molecular-weight with prior hormone/pregnancy-associated VTE or recurrent unpro-
heparin (LMWH) is the preferred anticoagulant for VTE prophylaxis in voked VTE or VTE associated with a persistent risk factor such as paral-
pregnant women, as it does not cross the placenta and is therefore ysis) should receive thromboprophylaxis during the entire pregnancy
safe for the fetus [13]. The risk threshold for instituting antepartum (Table 1). Antepartum thromboprophylaxis should be commenced as
thromboprophylaxis is higher than for postpartum early as possible, as the risk of VTE recurrence is increased from the be-
thromboprophylaxis. The rational for this is the lower average daily ginning of pregnancy [14].
risk of antepartum VTE and the need to self-inject LMWH for several The optimal LMWH dose for pharmacologic thromboprophylaxis of
months compared to 6 weeks postpartum. Therefore, in women with pregnancy-associated recurrent VTE is unknown, as no randomised
a low risk of recurrence (e.g. women with a single prior VTE associated controlled trials (RCTs) have been performed. Therefore the ACCP
with a major transient risk factor such as surgery, use of a plaster cast or guideline suggests the use of either a prophylactic or intermediate
trauma), close antepartum clinical surveillance rather than (half therapeutic) dose of LMWH in this setting, without a preference
for one dose over the other [12]. Many centres prescribe a prophylactic
dose. However, numerous treatment failures have been reported in ret-
rospective studies and in the TIPPS trial, with an estimated recurrence
Table 1 risk of 5 to 8% using this strategy [9,10,15–18]. Of note, compliance
Summary of the 9th American College of Chest Physicians (ACCP) recommendations to was not assessed in these studies and the results are inconsistent with
prevent pregnancy-related venous thromboembolism (VTE) in women with prior VTE those from another study [19]. It has been postulated that an intermedi-
[12]. ate dose of LMWH could have superior efficacy compared to a prophy-
Antepartum and postpartum Postpartum prophylaxis No lactic dose of LMWH, but potentially at the cost of a higher bleeding
prophylaxis during 6 weeks pharmacological risk. Reassuringly, in a retrospective study in pregnant women receiving
prophylaxis therapeutic doses of LMWH, there was no increased risk of clinically rel-
Women with a single Women with a history of a General evant or severe postpartum bleeding compared with women who had
unprovoked episode of VTE single episode of VTE related pregnant delivered in the same hospital without LMWH use [20]. In another
to a major nonhormonal population
study, women receiving therapeutic LMWH during pregnancy were
transient risk factor#
Women with a single episode of found to have an increased risk of blood loss N500 mL and b1000 mL
VTE provoked by use of after vaginal delivery [21]. The use of therapeutic-intensity LMWH for
hormonal contraceptives, pharmacologic thromboprophylaxis during pregnancy is not widely ac-
pregnancy or the postpartum cepted in view of the anticipated elevated bleeding risk in the peripar-
period
tum period and because this strategy may preclude neuraxial
Women with a single episode of
VTE provoked by a minor anaesthesia.
nonhormonal transient risk There is an urgent need for evidence regarding the optimal strategy
factor£ in pregnant women who require pharmacologic thromboprophylaxis.
Women with a history of
To investigate the optimal LMWH dose for prevention of recurrent
multiple unprovoked
episodes of VTE VTE in pregnant patients with a history of VTE, we are currently
Women with a history of VTE conducting the Highlow study (NCT 01828697). The results of this
and a persistent risk factor RCT are very likely to impact current clinical practice and modify con-
VTE: venous thromboembolism; £ Long distance travel or minor trauma; # Surgery, major sensus guidelines. We summarize herein the design of this study, and
trauma or plaster cast immobilization in the 3 months prior to VTE. discuss the rationale for some of the unique study design features.
64 S.M. Bleker et al. / Thrombosis Research 144 (2016) 62–68

1.1. Study objective and hypothesis 1.2.4. LMWH regimen

LMWH is injected subcutaneously once daily. Nadroparin is the pre-
We aim to compare a fixed low dose of LMWH with an intermediate ferred type of LMWH, but different types of LMWH are allowed in the
dose of LMWH in the prevention of recurrent VTE in pregnant women study to reflect heterogeneity in current clinical practice. Table 3A de-
with a history of VTE and an indication for ante- and postpartum picts the dosing schemes for all available types of LMWH. The fixed
thromboprophylaxis. We hypothesize that an intermediate dose is su- low dose regimen is based on weight at randomisation and will not be
perior to a fixed low dose of LMWH in preventing recurrent VTE, with changed throughout pregnancy or the 6 weeks postpartum. In the inter-
a comparable safety profile in terms of clinically relevant bleeding mediate dose regimen, the patient's weight will be monitored at every
complications. follow-up visit and if necessary the dose will be changed accordingly.
Prior to delivery, women are instructed to stop LMWH when con-
tractions start or when membranes rupture. If delivery is planned, the
1.2. Study design last dose of LMWH is given at least 24 h prior to delivery. In the fixed
low dose group, neuraxial anaesthesia is allowed if the interval after
1.2.1. Overview of study organization the last LMWH dose is more than 12 h. In the intermediate dose
The Highlow study is an investigator-initiated, multicentre, interna- group, an interval of 24 h between the last injection and neuraxial an-
tional, randomised, open-label, superiority study for efficacy. Patients aesthesia is required. LMWH is restarted 12 to 24 h after delivery at
receive either a fixed low dose or an intermediate dose of LMWH during the discretion of the obstetrician.
pregnancy and 6 weeks after delivery. The use of LMWH is open-label, and the medication is prescribed by
The protocol was reviewed and approved by the regulatory author- the treating physician and supplied by pharmacies in the standard set-
ity of the Netherlands and the ethics committee of the Academic ting of patient care or in accordance with national regulatory
Medical Center in Amsterdam. In each participating country, the proto- requirements.
col is or has been subsequently reviewed by the local regulatory author-
ity and for each participating centre by the local institutional review 1.2.5. Efficacy outcome variables
board or ethics committee. Informed consent is obtained from eligible The primary efficacy outcome is symptomatic confirmed recurrent
patients prior to randomisation. A central independent adjudication VTE, defined as the composite of recurrent DVT and PE during pregnan-
committee (CIAC) whose members are unaware of the treatment allo- cy and 6 weeks postpartum (Table 4). The secondary efficacy outcomes
cation will adjudicate all suspected episodes of recurrent VTE, major are 1) symptomatic confirmed recurrent VTE, defined as the composite
bleeding events, clinically relevant non-major bleeding events, cases of recurrent DVT and PE up to 3 months postpartum and 2) symptomat-
of suspected type 1 allergy to LMWH injections, cases of suspected hep- ic confirmed superficial thrombophlebitis up to 3 months postpartum.
arin-induced thrombocytopenia (HIT), and deaths. An independent
data monitoring committee (DMC) monitors patient safety and out- 1.2.6. Safety outcomes
comes at regular intervals during the study, and makes recommenda- The primary safety outcomes are major bleeding, the composite of
tions to the coordinating investigators. Monitoring is performed via an major bleeding and clinically relevant non-major bleeding, early post-
interdepartmental monitoring system. partum haemorrhage (within 24 h postpartum), late postpartum haem-
orrhage (within 6 weeks postpartum), blood transfusion within 24 h
1.2.2. Patient population and eligibility postpartum, blood transfusion within 6 weeks postpartum, and mortal-
Pregnant women of 18 years or older with a history of VTE and an in- ity. The definitions of major bleeding and clinically relevant non-major
dication for ante- and postpartum thromboprophylaxis (Table 1) are el- bleeding are based on the criteria of the International Society of
igible for the study. The inclusion- and exclusion criteria are listed in Thrombosis and Hemostasis [22] and are provided in Table 5. The
Table 2. Patients enter the study as soon as a home test confirms preg- CIAC will adjudicate major bleeding, clinically relevant non-major
nancy, up to 14 weeks after the last menstrual period. bleeding, and postpartum haemorrhage (defined as more than 500 mL).
Women previously enrolled in the Highlow study are allowed to The secondary safety outcomes are minor bleeding, bruises, mild
participate during subsequent pregnancies, if the 6 weeks of postpar- skin complications (e.g. itching, swelling, pain), severe skin complica-
tum thromboprophylaxis have been completed (after a full-term preg- tions (e.g. local erythema, edema, vesicles or bullae), type 1 allergic re-
nancy, miscarriage, active termination or stillbirth). actions to LMWH, the medical necessity to switch to another LMWH
type, confirmed HIT and congenital anomalies or birth defects. The
CIAC will adjudicate type 1 allergic reactions to LMWH and HIT.
1.2.3. Stratification and randomisation
Once the patient has signed the informed consent form, the investi- 1.2.7. Duration of study treatment and follow-up
gator provides information to a secure web-based randomisation pro- All patients have specified scheduled contacts; 2 weeks after starting
gram (ALEA version 2.2), which randomly assigns the patient to either treatment (in the outpatient clinic or by telephone), at 20 weeks of
the fixed low dose or intermediate dose of LMWH. Randomly permuted pregnancy (in the outpatient clinic or by telephone), at 30 weeks of
blocks with maximum block size of 6 are applied, stratifying for centre. pregnancy (in the outpatient clinic or by telephone), 24 h to 1 week

Table 3A
Dosing schemes for all LMWH types in the Highlow study.

Fixed low dose Intermediate dose

Weight Nadroparin Enoxaparin Dalteparin Tinzaparin Weight Nadroparin Enoxaparin Dalteparin Tinzaparin

In kg In lbs In kg In lbs

b100 b220 2850 IU 4000 IU 5000 IU 3500 IU b50 b110 3800 IU 6000 IU 7500 IU 4500 IU
50 to b70 110 to b154 5700 IU 8000 IU 10,000 IU 7000 IU
≥100 ≥220 3800 IU 6000 IU 7500 IU 4500 IU 70 to b100 154 to b220 7600 IU 10,000 IU 12,500 IU 10,000 IU
≥100 ≥220 9500 IU 12,000 IU 15,000 IU 12,000 IU

All doses are administered once daily.

LMWH: low-molecular-weight heparin; kg: kilograms; lbs.: pounds; IU: International Units; mg: milligrams.
 S.M. Bleker et al. / Thrombosis Research 144 (2016) 62–68 65

Fig. 1. Flow chart Highlow study: from randomisation to end of follow-up.

after delivery (in the outpatient clinic or by telephone), 6 weeks after treatment group to which they were assigned. The valid-for-safety-
delivery (by telephone) and 3 months after delivery (by telephone). analysis population will consist of all patients who were randomised
During these contacts efficacy and safety of LMWH will be evaluated. and received at least one dose of study treatment. The per-protocol
In parallel, patients are followed at the outpatient clinic by a midwife (PP) population will consist of all randomised patients without any
or gynaecologist. In case of a suspected efficacy or safety outcome, ap- major deviation from the protocol. All efficacy analyses will be per-
propriate physical examination, laboratory or diagnostic testing is per- formed on the ITT population. Additionally, the primary efficacy out-
formed. Fig. 1 depicts the flowchart from randomisation until end of come will be analysed in the PP population.
follow-up.
In the event that a pregnancy results in a miscarriage or stillbirth, the 2. Rationale for some aspects of the Highlow study
patient will continue the use of LMWH until 6 weeks after termination,
and will be followed up as usual. The Highlow study is the first large RCT in pregnancy that will pro-
vide high-quality evidence on the optimal prophylactic dose of LMWH
1.2.8. Laboratory tests in pregnancy in women with a history of VTE. At present, only two
At baseline, creatinine level, platelet count and D-dimer are collect- RCTs with major methodologic weaknesses have evaluated the safety
ed. Two weeks after randomisation, the platelet count is determined and efficacy of thromboprophylaxis (compared with placebo or no
in order to detect a possible HIT. Anti-Xa peak levels (optional) and treatment) in pregnant women with a history of VTE, containing very
platelet count are determined 2 weeks after randomisation, at small sample sizes of 40 and 16 patients respectively [24,25]. Our
20 weeks and at 30 weeks of pregnancy. study has several unique features that deserve explanation and that
may help others design future studies on thromboprophylaxis in preg-
1.2.9. Baseline ultrasonography nant patients.
If the patient has a history of DVT, it is recommended that an ultra-
sound examination of the affected leg be performed at baseline, if this 2.1. Rationale for open-label design
has not yet been performed after initial treatment of the prior DVT.
Knowing whether there is any residual thrombosis in the leg, will be A double-blind design with labelling of the investigational medicinal
helpful in interpreting a new ultrasound examination in case the patient product (IMP) would have been the ideal design for this study, but the
presents with a suspicion of a recurrent VTE during the study. However, associated costs make it impossible for investigator-initiated studies to
a baseline ultrasound examination is not obligatory. implement such design. By law, IMPs should be available to subjects
by the sponsor free of charge, but in several countries, including the
1.2.10. Sample size and statistical analysis Netherlands, an exception is made for registered medicines even if
There is uncertainty about the actual incidence of recurrent VTE they are administered in a trial for another indication. Furthermore, in
among pregnant women receiving pharmacologic thromboprophylaxis. current practice LMWH is widely used in pregnant women in the dos-
Hence, an approach with a fixed sample size could lead to severe under- ages that are compared in the Highlow study, based on the ACCP guide-
powering or undue lengthening of the study. The sample size calcula- line recommendations. Following these principles, the Highlow study
tion in this study is based on the required number of events. Assuming uses a pragmatic open-label design, and we believe that our study will
a 65% relative risk reduction with the intermediate dose, a total of 29 be representative of how LMWH is likely to be used in clinical practice.
events would provide a power of 80% to demonstrate that an intermedi- As the CIAC will adjudicate all primary outcome events blindly, we trust
ate dose is superior to a low dose (two-sided α = 0.05). Similar risk re-
ductions have been achieved with current versus sub-standard
anticoagulant treatment to prevent recurrent VTE in patients after elec- Table 3B
Ratios of the intermediate dosages and the fixed low dosages
tive hip arthroplasty [23]. The efficacy analysis will be based on inten-
tion-to-treat (ITT) and the outcome is a symptomatic, objectively Nadroparin Enoxaparin Dalteparin Tinzaparin
diagnosed recurrent VTE. The expected loss to follow-up is close to Low dose group
zero, hence no further sample size adjustment was made. Based on b100 kg/b220 lbs Ref Ref Ref Ref
the available literature an incidence of recurrent VTE of 4 to 5% in the
Intermediate dose group
low dose group is expected, leading to a proposed sample size of 859 b50 kg/b110 lbs ×1.3 ×1.5 ×1.5 ×1.3
to 1074 women. However, this might be adjusted upward and down- 50 to b70 kg/110 to b154 lbs ×2 ×2 ×2 ×2
ward based on the overall number of events observed during the prima- 70 to b100 kg/154 to b220 lbs ×2.7 ×2.5 ×2.5 ×2.9
ry analysis period in the study. The ITT population will consist of all ≥100 kg/≥220 lbs ×3.3 ×3 ×3 ×4.3

patients who have been randomised. Patients will be analysed in the Ref = reference category.
66 S.M. Bleker et al. / Thrombosis Research 144 (2016) 62–68

that the lack of blinded treatment will have little impact on the evalua- Table 4
tion of efficacy and safety. Compliance will be assessed by history taking Diagnostic criteria of confirmed symptomatic recurrent venous thromboembolism or su-
perficial thrombophlebitis.
during follow-up visits and by collection of batch numbers and other
details of used medication boxes. Suspected (recurrent) DVT or If there were no previous DVT
superficial thrombophlebitis with investigations
one of the following findings Abnormal CUS
2.2. Randomisation b14 weeks of gestation An intraluminal filling defect on
venography
If there was a previous DVT investigation
Patients should be randomised before the 14th week of gestation in Abnormal CUS where compression
this study. We have carefully considered including patients at more ad- had been normal or, if non-compressible
vanced gestations, but as this would lead to the potential for selection of during screening, a substantial increase
low risk patients with associated implications for generalizability, a de- (4 mm or more) in diameter of the
thrombus during full compression,
cision was made to not include these patients. Furthermore, as the risk
An extension of an intraluminal filling
of VTE is already increased early in pregnancy, we believe that setting defect, or a new intraluminal filling
this cut-off enhances institution of prophylaxis as early as possible, fol- defect or an extension of
lowing the recommendations of the current international guidelines. non-visualization of veins in the
presence of a sudden cut-off on
venography
2.3. Rationale for doses of LMWH in the ‘intermediate dose group’ and in Suspected PE with one of the following A (new) intraluminal filling defect in
findings subsegmental or more proximal
obese patients in the ‘low dose group’
branches on spiral CT scan
A (new) intraluminal filling defect or
The ACCP guideline recommends the following daily doses of LMWH an extension of an existing defect or a
for prophylaxis in pregnancy: nadroparin 2850 International Units (IU), new sudden cut-off of vessels more than
enoxaparin 4000 IU, dalteparin 5000 IU or tinzaparin 4500 IU [12]. We 2.5 mm in diameter on the pulmonary
angiogram
use the same doses in the ‘fixed low dose’ group, except for tinzaparin
A (new) perfusion defect of at least
for which we chose 3500 IU instead of 4500 IU, mainly based on avail- 75% of a segment with a local normal
ability of syringes. The doses in the ‘intermediate dose group’ are ap- ventilation result (high-probability) on
proximately half of a therapeutic dose, and were chosen based on 1) VPLS
the examples given in the ACCP guideline, 2) the availability of pre-filled Inconclusive spiral CT, pulmonary
angiography or lung scintigraphy with
syringes and 3) similarity in ratios between the intermediate and low demonstration of DVT in the lower
dose for all LMWH types. We chose a ratio between the most common extremities by compression ultrasound
intermediate dose group (70–100 kg) and the fixed low dose group of or venography
2.5 to 3.0 (Table 3B). It should be emphasized that no direct evidence Fatal PE is PE based on objective diagnostic testing,
autopsy, or
is available to guide choosing a low dose or intermediate dose of
Death which cannot be attributed to a
LMWH in pregnancy, and that guideline recommendations are extrapo- documented cause and for which
lated from thromboprophylaxis studies in patients undergoing general PE/DVT cannot be ruled out
surgery or hip arthroplasty. In the TIPPS trial, in which pregnant (unexplained death)
women with thrombophilia at increased risk of VTE or with previous DVT: deep vein thrombosis; CUS: compression ultrasonography; PE: pulmonary embo-
placenta-mediated pregnancy complications were randomised to either lism; CT: computed tomography; VPLS: ventilation/perfusion lung scan.
antepartum thromboprophylaxis with dalteparin or to no dalteparin,
the dose of dalteparin was increased from 5000 IU to 10,000 IU at
20 weeks of gestation [18]. This decision was based on results from cut-off value of 100 kg in the ‘fixed low dose’ group, above which the
pharmacokinetic studies, suggesting that at this point, the dose require- dose is elevated approximately 1.5-fold (Table 3C).
ment increases [26].
No specific recommendations are given in the ACCP guideline re-
2.4. Prophylaxis following early termination of pregnancy
garding modification of prophylactic doses at extremes of body weight.
Several studies on thromboprophylaxis in surgery or cancer patients
Limited data are available on the VTE recurrence risk after an in-
have applied either dose elevations at cut-off values ranging from 70
duced abortion, miscarriage or stillbirth. In a study by Pabinger and col-
to 100 kg, or no dose elevation at all. In the FRUIT trial that evaluated
leagues, 2 of 83 patients (2.4%) with a terminated pregnancy had a
the addition of LMWH to aspirin at less than 12 weeks gestation in
recurrence, and this was the case in 1 of 53 patients (1.9%) after miscar-
women with inherited thrombophilia and prior delivery for hyperten-
riage and in 3 of 10 (30%) following stillbirth [9]. Based on these obser-
sive disorders and/or small-for-gestational age infants, the dalteparin
vations, patients with early termination of pregnancy in the Highlow
dose was increased in women weighing above 80 kg from 5000 IU to
study should continue LMWH injections until 6 weeks after termination
7500 IU [27]. The Royal College of Obstetricians and Gynaecologists sug-
of pregnancy.
gests a dose elevation above 90 kg, and the product monograph of
nadroparin recommends an increase of the prophylactic dose at a cut-
off value of 100 kg in patients undergoing hip arthroplasty [28]. Thus, 2.5. Debate on extended prophylaxis after 6 weeks postpartum
in the absence of consensus or evidence we pragmatically chose for a
There has been some debate on the need of prolonging prophylaxis
beyond the 6th week postpartum until the 12th week postpartum. The
Table 3C
current ACCP guideline recommends prophylaxis until the 6th week
Ratios of the low dosages for obese patients
postpartum, based on the fact that most pregnancy-related VTE epi-
Nadroparin Enoxaparin Dalteparin Tinzaparin sodes occur in the first 6 weeks postpartum [12,29,30]. A recent study
Low dose group demonstrated that the risk of a primary VTE is 11-fold higher within
b100 kg/220 lbs Ref Ref Ref Ref 6 weeks after delivery than in the same period 1 year later. During the
≥100 kg/≥220 lbs ×1.3 ×1.5 ×1.5 ×1.3 period of 7 to 12 weeks after delivery, the absolute VTE risk is low,
Ref = reference category. with a 2-fold higher incidence as compared with the same period
 S.M. Bleker et al. / Thrombosis Research 144 (2016) 62–68 67

Table 5 waiver regarding supply of study medication by the sponsor is a fre-

Definitions of major bleeding, clinically relevant non-major bleeding and postpartum quent problem as not all countries have such waiver by law. Further-
haemorrhage in the Highlow study [34].
more, the academic sponsor (Academic Medical Center, Amsterdam,
Major bleeding Is defined as overt bleeding and the Netherlands) is only able to provide patient research insurance for
Associated with a fall in haemoglobin of 2 g/dL or Dutch patients, which cannot be extended to foreign patients. Conse-
more, or
quently, local patient research insurance often needs to be arranged
Leading to a transfusion of 2 or more units of packed
red blood cells or whole blood, or which may be costly. The institutional ethics committee unfortunately
Occurring in a critical site: intracranial, intraspinal, did not waive the need for patient research insurance, even though
intraocular, pericardial, intra-articular, intramuscular we believe that the risk for patients participating in this study is negligi-
with compartment syndrome, retro-peritoneal, or
ble, as there is a large body of evidence on the safety LMWH in pregnan-
Contributing to death
Clinically relevant Is defined as overt bleeding not meeting the criteria for
cy. LMWH is widely used in pregnant patients outside of a trial setting
non-major bleeding major bleeding but associated with medical for the very indication and in the doses that are investigated in the
intervention, unscheduled contact (visit or telephone Highlow study [12,13]. A modest inclusion fee of €250 is available for
call) with a physician, (temporary) cessation of study every randomised patient with a completed case record form (CRF), to
treatment, discomfort such as pain or impairment of
partly overcome the abovementioned expenses.
activities of daily life
Haematuria if it is macroscopic, and either
spontaneous or lasts for more than 24 h after
instrumentation (e.g. catheter placement or surgery) of 3. Conclusion
the urogenital tract, or
Macroscopic gastro-intestinal haemorrhage: at least In conclusion, this study is likely to greatly impact patient care and
one episode of melena/hematemesis, if clinically
apparent, or
modify guideline recommendations. Although we have met several
Rectal blood loss, if more than a few spots, or challenges, especially in setting up the study internationally, we are
Vaginal blood loss, if more than a few spots, or convinced that our goal of including approximately a 1000 patients
Haemoptysis, if more than a few speckles in the will be reached within the foreseeable future.
sputum, or
Intramuscular hematoma, or
Subcutaneous hematoma if the size is larger than 25
cm2, or larger than 100 cm2 if provoked, or
Trial status
Multiple source bleeding
Postpartum Blood loss of more than 500 mL within 24 h of deliverya The Highlow study is registered on ClinicalTrials.gov
haemorrhage (NCT01828697).
a
According to the criteria of the World Health Organization. Between 24 April 2013 and 1 June 2016, 210 patients have been
randomised in 40 centres in 5 countries. Currently, 32 centres in the
Netherlands, 21 centres in France, 1 centre in Ireland, 1 centre in
1 year later [7]. In the Highlow study patients are followed up until Belgium and 1 centre in Norway are actively recruiting. At present we
12 weeks postpartum, which will allow us to assess the risk of recurrent are in the process of getting the study approved and started in other
VTE in the 6 weeks after cessation of LMWH. countries including Germany, the United States of America and Canada.
The estimated last inclusion date is April 2018. An updated list of the
2.6. Cutaneous reactions to LMWH number of included patients, participating centres and countries can
be found at www.highlowstudy.org
Hypersensitivity skin reactions following LMWH injections are fre-
quently seen, but are probably underreported in most observational
studies. It is estimated that at least half of all pregnant women experi- Sponsorship and financial support
ence these side effects and switch to at least one alternative LMWH
type [31–33]. One unique feature of our study is the close monitoring An unrestricted grant is provided by Aspen Pharma to the Academic
of skin reactions such as redness, swelling, pain and itchiness, and the Medical Center in Amsterdam (the Netherlands), who is the sponsor of
necessity to switch to other LMWH types. In addition, the occurrence the Highlow study (Principal Investigator: Prof. Dr. S. Middeldorp).
of easy bruising is carefully recorded. In France, where CHU de Saint Etienne is the sponsor, a grant was ac-
quired from the French Ministry of Health (PHRC national 2014).
The CIAC is based in Saint Etienne, France.
2.7. Therapeutic doses of LMWH
The DMC is based in the Academic Medical Center, Amsterdam, the
Netherlands.
One may wonder why we chose not to compare an intermediate
dose to a therapeutic dose of LMWH. In preparation of the Highlow
study design, we found that most experts did not support the use of
Disclosure of conflicts of interest
therapeutic doses of LMWH in the prevention of recurrent VTE, mostly
based on concern of bleeding. Although in one study therapeutic
There are no relevant conflicts of interest to disclose for any of the
doses of LMWH in pregnancy were found not to be associated with an
authors.
increase in postpartum haemorrhage [20], a study with a therapeutic
dose arm seems premature. If we find an unacceptably high incidence
of recurrent VTE in the intermediate dose group in the Highlow study, Acknowledgements
a next step could be a randomised trial comparing an intermediate to
a therapeutic dose of LMWH. We would like to acknowledge all Highlow investigators
(Supplementary Appendix A).
2.8. Challenges in setting up the study internationally This study is supported by the Dutch Consortium for Healthcare
Evaluation and Research in Obstetrics and Gynaecology
Setting up the study in different countries is challenging. Legislation (NVOG Consortium 2.0), by two French gynaecological networks (GO-
has hampered approval of the study protocol in several countries; the CIC and CROG) and by the French network on thrombosis (INNOVTE).
68 S.M. Bleker et al. / Thrombosis Research 144 (2016) 62–68

Appendix A. Supplementary data heparin during pregnancy and postpartum: is it effective? J. Thromb. Haemost. 9
(2011) 473–480.
[17] L. Tengborn, D. Bergqvist, T. Mätzsch, A. Bergqvist, U. Hedner, Recurrent thrombo-
Supplementary data to this article can be found online at http://dx. embolism in pregnancy and puerperium. Is there a need for thromboprophylaxis?
doi.org/10.1016/j.thromres.2016.06.001. Am. J. Obstet. Gynecol. 160 (1989) 90–94 http://www.ncbi.nlm.nih.gov/pubmed/
2912109 (accessed March 5, 2013).
[18] M.A. Rodger, W.M. Hague, J. Kingdom, S.R. Kahn, A. Karovitch, M. Sermer, et al.,
References Antepartum dalteparin versus no antepartum dalteparin for the prevention of preg-
nancy complications in pregnant women with thrombophilia (TIPPS): a multina-
[1] J.A. Heit, C.E. Kobbervig, A.H. James, T.M. Petterson, K.R. Bailey, L.J. Melton, Trends in tional open-label randomised trial, Lancet 384 (2014) 1673–1683, http://dx.doi.
the incidence of venous thromboembolism during pregnancy or postpartum: a 30- org/10.1016/s0140-6736(14)60793-5.
year population-based study, Ann. Intern. Med. 143 (2005) 697–706. [19] P.G. Lindqvist, K. Bremme, M. Hellgren, Efficacy of obstetric thromboprophylaxis
[2] A.H. James, M.G. Jamison, L.R. Brancazio, E.R. Myers, Venous thromboembolism dur- and long-term risk of recurrence of venous thromboembolism, Acta Obstet.
ing pregnancy and the postpartum period: incidence, risk factors, and mortality, Gynecol. Scand. 90 (2011) 648–653, http://dx.doi.org/10.1111/j.1600-0412.2011.
Am. J. Obstet. Gynecol. 194 (2006) 1311–1315, http://dx.doi.org/10.1016/j.ajog. 01098.x.
2005.11.008. [20] S. Roshani, D.M. Cohn, A.C. Stehouwer, H. Wolf, P.W. Kamphuisen, S. Middeldorp,
[3] R. Cantwell, T. Clutton-Brock, G. Cooper, A. Dawson, J. Drife, D. Garrod, et al., Saving et al., Incidence of postpartum haemorrhage in women receiving therapeutic
Mothers' Lives: Reviewing Maternal Deaths to Make Motherhood Safer: 2006–2008, doses of low-molecular-weight heparin: results of a retrospective cohort study,
The Eighth Report of the Confidential Enquiries into Maternal Deaths in the United BMJ Open. 1 (2011), e000257 http://dx.doi.org/10.1136/bmjopen-2011-000257.
Kingdom., BJOG2011 1–203, http://dx.doi.org/10.1111/j.1471-0528.2010.02847.x [21] H.M. Knol, L. Schultinge, N.J.G.M. Veeger, H.C. Kluin-Nelemans, J.J.H.M. Erwich, K.
(118 Suppl). Meijer, The risk of postpartum hemorrhage in women using high dose of low-mo-
[4] A.A. Sultan, J. West, L.J. Tata, K.M. Fleming, C. Nelson-Piercy, M.J. Grainge, Risk of first lecular-weight heparins during pregnancy, Thromb. Res. 130 (2012) 334–338,
venous thromboembolism in and around pregnancy: a population-based cohort http://dx.doi.org/10.1016/j.thromres.2012.03.007.
study, Br. J. Haematol. 156 (2012) 366–373, http://dx.doi.org/10.1111/j.1365- [22] S. Schulman, C. Kearon, Definition of major bleeding in clinical investigations of
2141.2011.08956.x. antihemostatic medicinal products in non-surgical patients, J. Thromb. Haemost. 3
[5] A.F. Jacobsen, F.E. Skjeldestad, P.M. Sandset, Incidence and risk patterns of venous (2005) 692–694, http://dx.doi.org/10.1111/j.1538-7836.2005.01204.x.
thromboembolism in pregnancy and puerperium–a register-based case-control [23] R.D. Hull, G.F. Pineo, P.D. Stein, A.F. Mah, S.M. MacIsaac, O.E. Dahl, et al., Extended
study, Am. J. Obstet. Gynecol. 198 (233) (2008) e1–e7, http://dx.doi.org/10.1016/ out-of-hospital low-molecular-weight heparin prophylaxis against deep venous
j.ajog.2007.08.041. thrombosis in patients after elective hip arthroplasty: a systematic review, Ann.
[6] E.R. Pomp, A.M. Lenselink, F.R. Rosendaal, C.J.M. Doggen, Pregnancy, the postpartum Intern. Med. 135 (2001) 858–869 http://www.ncbi.nlm.nih.gov/pubmed/
period and prothrombotic defects: risk of venous thrombosis in the MEGA study, J. 11712876 (accessed May 8, 2013).
Thromb. Haemost. 6 (2008) 632–637, http://dx.doi.org/10.1111/j.1538-7836.2008. [24] S. Gates, P. Brocklehurst, S. Ayers, U. Bowler, Thromboprophylaxis and pregnancy:
02921.x. two randomized controlled pilot trials that used low-molecular-weight heparin,
[7] H. Kamel, B.B. Navi, N. Sriram, D.A. Hovsepian, R.B. Devereux, M.S.V. Elkind, Risk of a Am. J. Obstet. Gynecol. 191 (2004) 1296–1303, http://dx.doi.org/10.1016/j.ajog.
thrombotic event after the 6-week postpartum period, N. Engl. J. Med. 370 (2014) 2004.03.039.
1307–1315, http://dx.doi.org/10.1056/NEJMoa1311485. [25] R. Howell, J. Fidler, E. Letsky, M. de Swiet, The risks of antenatal subcutaneous hep-
[8] P. Brill-Edwards, J.S. Ginsberg, M. Gent, J. Hirsh, R. Burrows, C. Kearon, et al., Safety of arin prophylaxis: a controlled trial, Br. J. Obstet. Gynaecol. 90 (1983) 1124–1128
withholding heparin in pregnant women with a history of venous thromboembo- http://www.ncbi.nlm.nih.gov/pubmed/6360198 (accessed March 5, 2013).
lism. Recurrence of Clot in This Pregnancy Study Group, N. Engl. J. Med. 343 [26] B.J. Hunt, H.A. Doughty, G. Majumdar, A. Copplestone, S. Kerslake, N. Buchanan,
(2000) 1439–1444, http://dx.doi.org/10.1056/NEJM200011163432002. et al., Thromboprophylaxis with low molecular weight heparin (Fragmin) in high
[9] I. Pabinger, H. Grafenhofer, A. Kaider, P.A. Kyrle, P. Quehenberger, C. Mannhalter, risk pregnancies, Thromb. Haemost. 77 (1997) 39–43.
et al., Risk of pregnancy-associated recurrent venous thromboembolism in women [27] J.I.P. de Vries, M.G. van Pampus, W.M. Hague, P.D. Bezemer, J.H. Joosten, Low-molec-
with a history of venous thrombosis, J. Thromb. Haemost. 3 (2005) 949–954, ular-weight heparin added to aspirin in the prevention of recurrent early-onset pre-
http://dx.doi.org/10.1111/j.1538-7836.2005.01307.x. eclampsia in women with inheritable thrombophilia: the FRUIT-RCT, J. Thromb.
[10] V. De Stefano, I. Martinelli, E. Rossi, T. Battaglioli, T. Za, P. Mannuccio Mannucci, et al., Haemost. 10 (2012) 64–72, http://dx.doi.org/10.1111/j.1538-7836.2011.04553.x.
The risk of recurrent venous thromboembolism in pregnancy and puerperium with- [28] Royal College of Obstetricians and Gynaecologists, Reducing the Risk of Venous
out antithrombotic prophylaxis, Br. J. Haematol. 135 (2006) 386–391, http://dx.doi. Thromboembolism during Pregnancy and the Puerperium. Green-top Guideline
org/10.1111/j.1365-2141.2006.06317.x. No.37a, 2015.
[11] R.H. White, W.-S. Chan, H. Zhou, J.S. Ginsberg, Recurrent venous thromboembolism [29] R.B. Gherman, T.M. Goodwin, B. Leung, J.D. Byrne, R. Hethumumi, M. Montoro,
after pregnancy-associated versus unprovoked thromboembolism, Thromb. Incidence, clinical characteristics, and timing of objectively diagnosed venous
Haemost. 100 (2008) 246–252. thromboembolism during pregnancy, Obstet. Gynecol. 94 (1999) 730–734.
[12] S.M. Bates, I.A. Greer, S. Middeldorp, D.L. Veenstra, A.-M. Prabulos, P.O. Vandvik, VTE, [30] E.L. Simpson, R.A. Lawrenson, A.L. Nightingale, R.D. Farmer, Venous
thrombophilia, antithrombotic therapy, and pregnancy: antithrombotic therapy and Thromboembolism in Pregnancy and the Puerperium: Incidence and Additional
prevention of thrombosis, 9th ed: American College of Chest Physicians Evidence- Risk Factors from a London Perinatal Database., BJOG, 1082001 56–60.
Based Clinical Practice Guidelines, Chest 141 (2012) e691S–e736S, http://dx.doi. [31] I. Bank, E.J. Libourel, S. Middeldorp, J. Van Der Meer, H.R. Büller, High rate of skin
org/10.1378/chest.11-2300. complications due to low-molecular-weight heparins in pregnant women, J.
[13] I.A. Greer, C. Nelson-Piercy, Low-molecular-weight heparins for Thromb. Haemost. 1 (2003) 859–861.
thromboprophylaxis and treatment of venous thromboembolism in pregnancy: a [32] L. Schultinge, H.M. Knol, H.C. Kluin-Nelemans, J.J.H.M. Erwich, K. Meijer, Incidence of
systematic review of safety and efficacy, Blood 106 (2005) 401–407, http://dx. hypersensitivity skin reactions in patients on full-dose low-molecular-weight hepa-
doi.org/10.1182/blood-2005-02-0626. rins during pregnancy, Neth. J. Med. 71 (2013) 518–522.
[14] J.G. Ray, W.S. Chan, Deep vein thrombosis during pregnancy and the puerperium: a [33] M. Schindewolf, H. Kroll, H. Ackermann, J. Garbaraviciene, R. Kaufmann, W.-H.
meta-analysis of the period of risk and the leg of presentation, Obstet. Gynecol. Surv. Boehncke, et al., Heparin-induced non-necrotizing skin lesions: rarely associated
54 (1999) 265–271 http://www.ncbi.nlm.nih.gov/pubmed/10198931 (accessed with heparin-induced thrombocytopenia, J. Thromb. Haemost. 8 (2010)
May 7, 2013). 1486–1491, http://dx.doi.org/10.1111/j.1538-7836.2010.03795.x.
[15] C. Rozanski, A. Lazo-Langner, M. Kovacs, Prevention of venous thromboembolism [34] S. Schulman, C. Kearon, Definition of major bleeding in clinical investigations of
(VTE) associated with pregnancy in women with a past history of VTE, Blood 114 antihemostatic medicinal products in non-surgical patients, J. Thromb. Haemost. 3
(2009) 1217–1218. (2005) 692–694, http://dx.doi.org/10.1111/j.1538-7836.2005.01204.x.
[16] J.E. Roeters van Lennep, E. Meijer, F.J.C.M. Klumper, J.M. Middeldorp, K.W.M.
Bloemenkamp, S. Middeldorp, Prophylaxis with low-dose low-molecular-weight