Low molecular weight heparin (dalteparin) for the treatment of venous thromboembolism in pregnancy


* Dr A. F. Jacobsen, Department of Obstetrics and Gynaecology, Ullevål University Hospital, N-0407 Oslo, Norway.


Objective To evaluate the effect and dose of dalteparin given to pregnant women with acute venous thromboembolism.

Design An observational study of pregnant women in Norway.

Setting Delivery and haematological departments in Norway.

Population Twenty women, aged 22–41 years, with acute venous thromboembolism verified by objective means.

Methods Patients were treated with dalteparin from diagnosis until delivery. Treatment was monitored with anti-activated factor Xa (anti-Xa) activity, and the dose was adjusted to achieve target 0.5–1.0 U/mL 2–3 hours post-injection.

Main outcome measure Anti-Xa activity and side effects.

Result None of the patients suffered recurrent venous thromboembolism or major bleeding complications. In 9 of 13 women starting with conventional dose of dalteparin (100 iu/kg bd), dose escalation was necessary to reach target anti-Xa activity. None of the six women who started with 105–118 iu/kg bd required dose escalation. One woman who started with 133 iu/kg bd required dose reduction. Bioaccumulation of dalteparin was not observed.

Conclusion Our study suggests that dalteparin may be used for the treatment of acute venous thromboembolism in pregnancy. Approximately 10–20% higher doses of dalteparin may be needed as compared with non-pregnant individuals.


Venous thromboembolism is one of the leading causes of maternal death and morbidity in pregnancy. The incidence of venous thromboembolism is approximately 1 in every 1000–2000 pregnancies1,2. Thrombosis of the deep veins of the lower extremities is most commonly observed, while pulmonary embolism is less frequently diagnosed.

Treatment of venous thromboembolism in pregnancy is challenging because the baseline risk of thrombosis increases progressively throughout pregnancy. This increased risk is related to activation of coagulation3–5 and inactivation of fibrinolysis3. In addition, the potential for harmful effects of treatment on the fetus must be considered. Warfarin has known teratogenic effects throughout pregnancy and should be used with caution in pregnancy6,7.

The established treatment of venous thromboembolism in pregnancy is unfractionated heparin initially given by continuous intravenous infusion followed by subcutaneous injections two or three times daily for the rest of pregnancy7. Treatment needs monitoring and dose adjustment. Many patients need high doses because of neutralisation of unfractionated heparin by acute phase proteins. High doses are associated with adverse effects (e.g. osteoporosis8 and thrombocytopenia)9.

During the last two decades, low molecular weight heparins (LMWH) have been introduced and proved to be at least as efficacious and safe as unfractionated heparin for the treatment of both arterial and venous thrombosis8. LMWHs may be given once or twice daily by the subcutaneous route and essentially without need for monitoring. Many patients may be treated on an ambulatory basis. Subcutaneous treatment with LMWH given once or twice daily may now be considered standard medical care in non-pregnant patients with acute venous thromboembolism.

Although many studies have reported use of LMWHs for the prophylaxis10–12 and treatment12,13 of venous thromboembolism in pregnancy, the efficacy and safety, optimal dose and need for monitoring of LMWH treatment in pregnancy remain poorly documented. Similar to unfractionated heparin, LMWHs do not pass the placenta barrier14,15 and are not associated with teratogenic effects.

In the present study, we have carefully monitored treatment with one LMWH, dalteparin (Fragmin, Pharmacia, New Jersey, USA), in pregnant women with acute venous thromboembolism. Since LMWHs were already introduced for the treatment of venous thromboembolism in pregnancy in many Norwegian hospitals, the present protocol was developed to standardise and document need for dose adjustment and monitoring.


All Norwegian hospitals were invited to participate in the study. The protocol was adopted as a national protocol for the treatment of venous thromboembolism in pregnancy and made public through a web site (http://www.legeforeningen.no/index.db2id1696) of the Norwegian Medical Association. The protocol was approved by the Regional Ethical Committee of Health Region I in Norway. The inclusion period was from February 2000 to February 2001. Pregnant women with acute venous thromboembolism were eligible for inclusion in the study. Venous thromboembolism was diagnosed according to standard medical care in each individual hospital using either compression ultrasonography and/or venography for the diagnosis of deep vein thrombosis or lung scan, helical computed tomography or pulmonary angiography for the diagnosis of pulmonary embolism. Women were excluded if the local physician thought that treatment with conventional unfractionated heparin was desired. They were also excluded if they had contraindications for dalteparin (e.g. known hypersensitivity or a history of heparin-induced thrombocytopenia).

The decision on mode of delivery was the responsibility of the practising physician. In case of spontaneous vaginal delivery, it was recommended to discontinue injections at the time of active labour. After delivery, dalteparin was restarted after 3–4 hours. In cases of prolonged first stage of labour, one extra prophylactic dose of dalteparin was recommended. In case of elective caesarean, the last dose of dalteparin was given 24 hours before surgery. Epidural or spinal anaesthesia was not offered to these patients. After delivery, commencement of oral warfarin therapy was recommended. Dalteparin was given until therapeutic international normalised ratio (2.0–3.0) had been achieved for two consecutive days.

The protocol recommended use of the LMWH brand commonly used for the treatment of venous thromboembolism in each individual hospital. The initial dose was given according to weight as recommended by the manufacturer (e.g. dalteparin 100 iu/kg bd and adjusted for weight change during pregnancy). At the time of execution of the study, use of dalteparin accounted for 80–90% of the LMWH market in Norway, and by chance, all women in the present study were given dalteparin.

Treatment was monitored by assay of anti-activated factor Xa (anti-Xa) activity in blood samples collected 3–4 hours after injection of the morning dose of dalteparin. Target anti-Xa activity was 0.5–1.0 U/mL throughout pregnancy, and the dose of dalteparin was adjusted accordingly. Anti-Xa activity and blood platelets were assayed after 3, 7, and 14 days and every 4 weeks thereafter. An additional blood sample was taken before dalteparin was initiated and four to six weeks after anticoagulant therapy was discontinued (usually three months postpartum).

Blood samples were collected at each visit and sent to a central laboratory (Haematological Research Laboratory, Ullevål University Hospital). Blood was collected in 5 mL Vacutainer tubes (Becton-Dickinson, Meylan-Cedex, France) containing 0.5 mL buffered citrate (0.129 mol/L) and in 5 mL Vacutainer tubes containing EDTA anticoagulant. Tubes with citrate as anticoagulant were centrifuged at 2000×g for 20 min within 1–2 hours, whereas blood collected in EDTA were used for DNA analyses. Thrombophilia screening was either performed before anticoagulant treatment was initiated (excluding protein S) or after termination of anticoagulation (usually three months after delivery).

Anti-Xa activity was assayed with a commercial chromogenic substrate method, Coatest heparin (Chromogenix, Mölndal, Sweden). The assay was calibrated with serial dilutions (0–0.8 iu/mL) of a commercial standard (LMWH anti-Xa standard, Chromogenix) in pooled normal plasma. The calibrator had been calibrated against the First International Standard for LMWH. The quality of the assay was ensured by external quality control (Labquality, Helsinki, Finland, and ECAT, Leiden, the Netherlands). Antithrombin and protein C activities, free protein S antigen, lupus anticoagulant and anticardiolipin antibodies were assayed as described earlier16. The factor V Leiden mutation and the prothrombin gene G20210A allele variation were assayed with commercial kits (Roche Diagnostics, Mannheim, Germany) run on a real-time PCR analyser (Light Cycler, Roche Diagnostics).

We registered pregnancy-related complications, especially ‘small for gestational age’ and pre-eclampsia. The patients were followed regularly, mostly once a week for the first two to three weeks, then every two till four weeks until delivery. Complications during and after delivery were also registered.


Twenty pregnant women with acute venous thromboembolism were included in the study (Table 1). This accounts for approximately one-third of the estimated total number of women with acute venous thromboembolism in pregnancy in Norway during the period. Median age was 31 years (range 22–41 years). Fifteen women were nulliparous and five expected their second child. One woman had known protein S deficiency and a history of venous thromboembolism in an earlier pregnancy. This was a twin pregnancy, and she developed recurrent venous thromboembolism despite subcutaneous prophylactic doses of dalteparin 5000 iu od. Laboratory evaluation later confirmed that she had protein S deficiency combined with both heterozygous factor V Leiden mutation and heterozygous prothrombin gene 20210GA allele variation. Another woman had previously suffered pulmonary embolism, but she was not given prophylactic antithrombotic treatment during pregnancy. A third woman had suffered venous thromboembolism five months before she got pregnant, but she decided to discontinue anticoagulant treatment when she was four weeks pregnant.

Table 1.  Characteristics of the 20 women with acute venous thromboembolism during pregnancy.
PatientAgeParityWeekType venous thromboembolismThrombophilia
  1. DVT = deep vein thrombosis; PE = pulmonary embolism; FVG1691A= heterozygous factor V Leiden mutation; FII20210GA= prothrombin gene 20210GA allele variation; PS = protein S.

135023proximal DVTnone
222026proximal DVTnone
327018distal DVTnone
441015proximal DVTFVG1691A
637012distal DVT+PEnone
737126proximal DVTFVG1691A+ FII20210GA+ PS deficiency
829032proximal DVTFVG1691A
937013proximal DVTFII20210GA
1033132proximal DVTnone
1129130proximal DVTFVG1691A+ FII20210GA
1230035proximal DVTFVG1691A
1334129proximal DVTnone
1437036proximal DVTFVG1691A
152908distal DVTnone
1632035proximal DVTnone
1729032proximal DVTnone
1824014proximal DVTnone
1927039proximal DVTFVG1691A
2039011jugular vein DVTnone

Thrombophilic defects were found in 8 (40%) of the 20 patients (Table 1). Seven women had heterozygous factor V Leiden mutation, three women had heterozygous prothrombin gene 20210GA allele variation and one woman had protein S deficiency. Three of the women had combined defects: one had protein S deficiency combined with heterozygous factor V Leiden mutation and heterozygous prothrombin gene 20210GA allele variation and two combined heterozygous factor V Leiden mutation and heterozygous prothrombin gene mutation.

In 11 out of 20 women, venous thromboembolism developed in the third trimester (Table 2). All these events were proximal deep vein thrombosis. Altogether, 15 women suffered proximal deep vein thrombosis, two distal deep vein thrombosis, one pulmonary embolism and one woman had combined deep vein thrombosis and pulmonary embolism. One woman suffered severe jugular vein thrombosis.

Table 2.  Time and type of venous thromboembolism during pregnancy.
Type of thrombosisFirst trimesterSecond trimesterThird trimester
  1. *One patient with combined distal venous thromboembolism and lung embolism included.

Proximal DVT411
Distal DVT2*1
Pulmonary embolism1*1
Jugular vein thrombosis1

All women were given dalteparin. Individual anti-Xa activity levels are shown in Fig. 1. Thirteen women were initially treated with 100 U/kg bd. Dose escalation proved necessary in 9 of these 13 women to reach target peak anti-Xa activity of 0.5–1.0 U/mL. Stable target levels were achieved after dose adjustments to median 113 U/kg bd (range 105–125 U/kg bd). Six women initially started with higher doses of dalteparin, median 112 U/kg bd (range 105–125 U/kg bd) and target anti-Xa level was achieved in all women. One woman, who started with a dose of 133 U/kg bd, achieved peak anti-Xa activity of 1.49 U/mL. After dose reduction, she finally reached stable target levels at a dose of 100 U/kg bd. Bioaccumulation of dalteparin was not observed in any of the women.

Figure 1.

Individual anti-Xa activities in the 20 women treated with therapeutic doses of LMWH (dalteparin) during pregnancy. Upper panel: initial dose 100 iu/kg, no dose adjustments were performed (n= 4); middle panel: initial dose 100 iu/kg, dose adjustments were performed in all women (n= 9); lower panel: initial dose 105–125 iu/kg, except for one woman initially given 133 iu/kg (n= 7). Dose reduction was performed only in the one woman initially given 133 iu/kg (arrow).

Nineteen women delivered healthy full-borne babies, but one pregnancy ended with intrauterine death after 37 weeks of gestation. This was a high-risk pregnancy complicated by smoking, obesity (BMI 36 kg/m2) and gestational diabetes. One ‘small for gestational age’ child was delivered at 37 weeks. The mother suffered deep vein thrombosis at 35 weeks, and she had only used dalteparin for two weeks at the time of delivery. None of the women developed recurrent thrombosis or pre-eclampsia. No serious bleeding episodes (>1000 mL) were observed, but one woman had an atonic postpartum bleeding, one a wound haematoma, one a minor perineal haematoma and one woman had curettage after delivery due to intrauterine residual placental tissue. Ten women delivered by caesarean section and 10 by the vaginal route. None of the patients developed thrombocytopenia.


Different LMWHs are now widely used for venous thromboprophylaxis in pregnancy, but there is still limited data on their efficacy and safety and their optimal dose and need for dose adjustment in the treatment of acute venous thromboembolism in pregnancy17. In the present study, women with acute venous thromboembolism in pregnancy were given dalteparin twice daily and the dose was adjusted according to weight throughout pregnancy. We have found that therapeutic, weight-adjusted doses of dalteparin during pregnancy yielded predictable plasma concentrations and were not associated with any bioaccumulation of dalteparin. Dose escalation proved necessary in the majority of women initially given the recommended dose for non-pregnant women, whereas one woman initially given a 33% excess dose required dose reduction. None of the women initially given a 5–25% higher dose than recommended in non-pregnant women required dose adjustment. Our data therefore suggest that women suffering acute venous thromboembolism in pregnancy should be given a 10–20% higher weight-adjusted dose of dalteparin, as compared with non-pregnant women. Although our data also provide some evidence that monitoring of dalteparin concentration may not be necessary, more data are needed to conclude that dalteparin can be given in weight-adjusted doses without monitoring of plasma concentration.

Pregnancy is associated with several important physiological changes (e.g. increases in body mass, plasma volume and glomerular filtration rate) that may influence the optimal dosing of dalteparin. Several investigators have investigated the need for dose adjustments of LMWH during pregnancy18,19. Hunt et al.18 found that anti-Xa levels decreased during pregnancy after prophylactic fixed doses of dalteparin, and that dose adjustments were needed in each trimester. Crowther et al.19 found that the plasma concentration of prophylactic fixed doses of reviparin varied inversely with weight, but that reviparin retained its predictable pharmacokinetics during pregnancy. In the present study, the therapeutic dose of dalteparin was adjusted for weight gain, which resulted in stable anti-Xa levels in the target range.

Bioaccumulation is of potential importance for risk of bleeding complications, especially during labour. Bremme et al.20 recently reported that prophylactic fixed-dose dalteparin was associated with significant accumulation in four cases, but Crowther et al.19 did not observe any bioaccumulation in their patients given a similar dose of reviparin. It is therefore of interest that we did not find any evidence of accumulation of dalteparin in women given much higher comparative therapeutic doses of dalteparin. Moreover, major bleeding complications were not observed although treatment with high-dose dalteparin was continued until start of labour or the evening before an elective caesarean.

Our study is too small to provide definite evidence on the efficacy and safety of dalteparin treatment of acute venous thromboembolism in pregnancy. However, all the 20 women were successfully treated with dalteparin without any clinical recurrence and without major maternal bleeding or other complications. One intrauterine death was observed in a high-risk pregnancy (habitual smoking, obesity, and gestational diabetes), but it is unlikely that this complication was related to dalteparin treatment. Our study is corroborated by several other reports, mostly case reports, on the effect of LMWH treatment for acute venous thromboembolism in pregnancy13,21,22. Lepercq et al.12 reported retrospective data on the treatment with enoxaparin in 49 women with acute venous thromboembolism in pregnancy, but this study does not provide any details on efficacy and safety and dose adjustments and achieved heparin concentrations.

Conventional treatment with unfractionated heparin is initially given intravenously for 5–7 days followed by subcutaneous injections twice or thrice daily to maintain activated partial thromboplastin time in the therapeutic range7. In our study, dalteparin was administered twice daily to keep peak anti-Xa activity in the putative therapeutic range of 0.5–1.0 U/mL. The rationale for the maintained high dose is that these women are at rapidly escalating risk of venous thromboembolism during the rest of pregnancy, especially during the last trimester and at delivery. However, it is still open whether the dose of LMWH may be reduced after one to four weeks, which could potentially reduce the risk of complications (i.e. bleeding and osteoporosis)23. In our study, none of the women developed symptomatic osteoporosis. Since bone densitometry was not performed, we cannot exclude the possibility that continued high-dose dalteparin was associated with excess bone loss.

Forty percent (8/20) of our women had inherited thrombophilia, but no cases with antiphospholipid antibodies were detected. Of particular interest is the very high frequency of the factor V Leiden phenotype (35%), the heterozygous prothrombin gene 20210GA allele variation (15%) and combination of the two mutations (10%). These two mutations occur in 8% and 1% of the Norwegian population, respectively (PM Sandset, unpublished), and our data underline the importance of these two mutations as risk factors for pregnancy associated venous thromboembolism in the general population. Such combined mutations were recently reported to be associated with a very high risk of venous thromboembolism in pregnancy24.

We conclude that dalteparin may be used for the treatment of acute venous thromboembolism in pregnancy, but that slightly higher doses of dalteparin is probably needed as compared with non-pregnant women. Treatment should be adjusted according to body weight during pregnancy. Further studies, including other brands of LMWHs, are needed to substantiate our results. Finally, it should be emphasised that treatment with dalteparin was associated with an obvious improvement in quality of life for the affected women with minimal stay in hospital, simplified follow up and reduced number of injections and injection volumes.


The authors thank the staff of Haematological Research Laboratory, Department of Haematology, Ullevål University Hospital, for their expert technical assistance and all physicians at participating hospitals who followed the patients (see Appendix 1).


Appendix 1List of participating hospitals (city)—physicians (number of patients)

Ullevål University Hospital (Oslo)—AF Jacobsen (6), Aker University Hospital (Oslo)—L Ballo and U Abilgaard (2), Rikshospitalet (Oslo)—T Henriksen (1), Bærum sykehus (Bærum)—A Tveit (1), Sentralsjukehuset i Møre og Romsdal (Ålesund)—S Hjelle and S Wørner (2), Ringerike sykehus (Hønefoss)—V Stenberg and J Tuveng (1), Vest-Agder Sentralsykehus (Kristiansand)—AG Askvold (2), Sentralsykehuset i Buskerud (Drammen)—L Ellingsen and S Hjoertur (2), Sentralsykehuset i Hedmark (Elverum)—E Markova (1), Sentralsykehuset i Telemark (Skien)—K Heldal (1), St Olafs sykehus (Trondheim)—Mattson (1).