Background: Venous Thromboembolism (VTE) is a leading cause of maternal death in the developed world. 1 Although a rare event, occurring in approximately 1-2 per 1,000 pregnancies, 2 every pregnant woman is potentially at risk, and every event is potentially fatal. The death of a mother is a devastating event, with far-reaching impact on an individual, societal and health systems level. Moreover, non-fatal consequences may be associated with significant short- and longterm morbidity. 3
Prevention, diagnosis and treatment of pregnancy-associated VTE (PA-VTE) should therefore be a priority for health services globally.
Risk factors for PA-VTE
The risk of VTE is estimated to be 5-10 times higher in pregnant and postpartum women compared with non-pregnant women. 4-6 Risk factors for PA-VTE differ in the antenatal and postnatal periods. Maternal VTE risk factors include advanced age and parity, Body Mass Index (BMI), smoking status and medical co-morbidities. 7 VTE risk is dynamic and may change during pregnancy and delivery. For example, the risk of VTE following an emergency caesarean delivery is reported to be twice that of a planned caesarean and four times that of a vaginal delivery. 8,9 Other VTE risk factors may only arise, or may only be identifiable, in the postpartum period, including stillbirth, systemic infection, postpartum haemorrhage or receipt of a blood transfusion. 10
VTE risk factors are common among pregnant and postpartum populations. In a recent analysis of risk factors for postpartum VTE among 21,000 consecutively sampled women, we found that more than 75% of women had at least one risk factor for VTE, while more than 40% carried at least two risk factors. 11 This high prevalence suggests a considerable burden of VTE risk at a population level, even in the absence of well defined ‘high-risk’ characteristics. However, little is known on the interaction between individual risk factors and the cumulative risk of VTE when two or more risk factors co-occur.
Together, the high and increasing prevalence of VTE risk factors emphasise the need for interventions towards effective management of VTE risk to avoid associated morbidity and mortality.
In some cases, PA-VTE may be preventable. 12 Low Molecular Weight Heparin (LMWH) is the preferred agent for pharmacological thromboprophylaxis. 10,13-15
LMWH is administered as a subcutaneous injection, and it is safe to use in pregnancy and breastfeeding. 16 For women with prior VTE whose risk warrants antenatal pharmacological thromboprophylaxis, the optimal dose is unknown, and guidelines have previously advocated strategies including a low prophylactic or an intermediate (half-therapeutic) dose. 15,17 The recently completed Highlow study (NCT 01828697; clinicaltrials.gov), has addressed this question in an investigator-initiated, multicentre, multinational RCT comparing a fixed low dose of LMWH with an intermediate weight-adjusted dose in the prevention of pregnancy-related VTE recurrence in women ≥18 years with a history of VTE and an indication for ante- and postpartum thromboprophylaxis who are recruited ≤14 weeks gestation.
Direct oral anticoagulant agents (DOACs) are not routinely recommended in the context of PA-VTE due to insufficient evidence to evaluate safety for the mother, the fetus or the breastfeeding infant.
Previous studies have also demonstrated the effectiveness of low-dose aspirin (ASA) in the prevention of VTE in other populations, 18-21 and its potential role in PA-VTE is also being considered. The pilot PARTUM trial (Postpartum Aspirin to Reduce Thromboembolism Undue Morbidity) will assess the feasibility of a randomised, multicentre, placebo-controlled trial using low-dose aspirin in the prevention of postpartum VTE in women at moderate risk of VTE (NCT: NCT04153760, Clinicaltrials.gov).
Several prominent international organisations have published clinical recommendations to guide VTE prevention in clinical practice. 10,15,17,22-24 The most widely cited are from the Royal College of Obstetrics and Gynaecology (RCOG) in the UK10, and the American College of Chest Physicians (ACCP) 15 and the American College of Obstetrics and Gynecology (ACOG) 17 in North America. There is a lack of high-quality evidence to guide optimal prevention of PA-VTE in women with intermediate- and low-risk characteristics. 22,24,25 In the absence of robust evidence, clinical recommendations have been based on expert opinion and consensus. 26 This has resulted in striking variation in recommendations on the prevention of PA-VTE. 11,27 We have recently reported a five-fold difference in the number of women who would theoretically receive a recommendation for postpartum thromboprophylaxis by various international guidelines, which ranged from 7% under ACOG to 37% under RCOG guidelines. 11
This highlights the need for large-scale, prospective, high-quality research to determine best practices for effective prevention of PA-VTE and to guide clinical decision making regarding the use of pharmacological thromboprophylaxis.
Promoting a systematic approach to VTE prevention
Irrespective of the approach taken to prevent VTE in individual hospitals or countries, VTE prevention is based on the accurate identification of women who are at risk of VTE and subsequent implementation of risk-appropriate preventative interventions. Despite the potentially devastating consequences associated with PA- VTE, and the availability of nationallevel recommendations to guide its prevention, an implementation gap regarding utilisation of riskappropriate thromboprophylaxis remains suboptimal in many developed countries. 28-33
The introduction of VTE risk scores and harnessing electronic health record (EHR) functionality are examples of interventions intended to reduce variation in care and improve the reliability of action in the prevention of PA-VTE.
A number of VTE risk scores intended to guide PA-VTE prevention strategies have been reported in the literature and recommended in clinical guidelines, reflecting trends towards efficiency and algorithmic approaches seen elsewhere in healthcare. 34 There is considerable variation among these risk scores, in their development, target population, risk factors included and the weight of risk assigned to each risk factor. Critically, few have been validated, and the ability of each risk score to identify women at risk of VTE remains unknown.
The previously mentioned implementation gap’ has led to various other systemslevel interventions aimed at improving thromboprophylaxis. 35 These interventions are primarily implemented in nonobstetric settings and include the introduction of automatic computerised or human alerts, or multifaceted interventions. 36 Many interventions have also been described which are embedded in electronic health records (EHR) or employ Computerised Clinical Decision Support Systems (CCDSS) to stratify the patient according to VTE risk and make suggestions for thromboprophylaxis. Examples of advanced CCDSS functionality include the auto-population of clinical information into risk assessment checklists and translation of thromboprophylaxis recommendations into an order. 37 In a systematic review of clinical trials and observational studies, the use of CCDSS was associated with a two-fold increase in the rate of ordering prophylaxis for VTE when compared with controls (OR 2.35, 95% CI: 1.78-3.10, p<0.001). As yet, there are no large published studies reporting the use of CCDSS in obstetric settings.
Diagnosis of pulmonary embolism (PE) in pregnancy
Evaluating a pregnant woman with suspected PE in pregnancy can be challenging because symptoms that can suggest PE can be common during normal pregnancy, particularly as the pregnancy advances. 38,39
Both maternal and fetal radiation exposure is low using modern imaging techniques. 38 Perfusion scintigraphy and computed tomography pulmonary angiography (CTPA) confer extremely low fetal radiation doses, well below the threshold associated with fetal radiation complications (50-100mSv). 40,41 Moreover, advances in CT technology have greatly reduced radiation exposure to the mother using a number of new methods that do not simultaneously reduce image quality. These include: reduced kilovoltage, reduced anatomic coverage of the scan (reported to reduce radiation dose by up to ~70% 42 ), using iterative reconstructive techniques and, of particular relevance to breast radiation dose, reducing the CTPA contrast-monitoring component. 42,43 It is now possible to achieve such low radiation dose that the breast tissue may only be exposed to doses as low as 3-4 mGy. 38
A normal perfusion scan and a negative CTPA appear equally safe for ruling out PE in pregnancy. 44-46 Inconclusive or indeterminate results are common 47-49 and the chance of an inconclusive or suboptimal CTPA scan increases with advancing gestation. 49 Decreased arterial enhancement in pregnancy is a contributing factor to the high proportion of inconclusive CTPA scans. Nevertheless, sensitivity and negative predictive value of lung scintigraphy and CTPA are reported to be high.
In light of these data, recently published 2019 guidelines of the European Society of Cardiology recommend that “Perfusion scintigraphy or CTPA (with a low-radiation dose protocol) should be considered to rule out suspected PE during pregnancy; CTPA should be considered as the first-line option if the chest X-ray is abnormal”. 50 In contrast, 2018 ASH guidelines suggest “V/Q lung scanning over CT pulmonary angiography (conditional recommendation, low certainty in evidence about effects)”. 13 Either imaging approach seems reasonable as a first-line diagnostic test, and is, in practice, driven by local availability.
The ongoing OPTICA Study (Optimised Computed Tomography Pulmonary Angiography (CTPA) in Pregnancy, Quality and Safety; NCT04179487, Clinicaltrials.gov) is a prospective multicentre study aiming to validate the clinical utility and safety of an optimised lowdose CTPA protocol for suspected pulmonary embolism in pregnancy.
Pregnant women undergoing the specified low-dose CTPA protocol for suspected PE at study sites with equivalent CT capabilities will be recruited. The primary outcome is the incidence of VTE at 3 months in women in whom PE was excluded at baseline CTPA. 51
Until recently, clinical prediction rules and D-dimer levels had not been clinically validated to exclude PE during pregnancy. Therefore, the overall prevalence of PE during pregnancy was low, approximately 2-5%. 47,48,52 A combination of data from clinical prediction rules and D-dimers may be used to rule out PE in non-pregnant patients without the need for radiological imaging. However, D-dimers continuously increase during pregnancy 53,54 and levels are above the threshold for VTE “rule-out” of 500 µg/L in almost one-quarter of pregnant women in the third trimester. 54 A negative D-dimer has not been clinically validated as a stand-alone rule-out test for PE during pregnancy.
The results of two recently published multicentre, multinational prospective diagnostic management outcome studies have, for the first time, provided evidence supporting the incorporation of D-dimers and clinical prediction rules into diagnostic algorithms for women with suspected PE during pregnancy. The first was published by Righini et al in 2019. 55 In this study, PE was deemed excluded without CT pulmonary angiography (CTPA) in pregnant women who had a low or intermediate pretest clinical probability assessment (using the revised Geneva score) and a negative D-dimer (defined as <500 µg/L). Women fulfilling these criteria did not undergo diagnostic imaging, were not treated and were followed up at 3 months. PE was diagnosed in 7.1% of women and the rate of symptomatic VTE events at 3 months was 0.0% (95% CI: 0.0-1.0%) among untreated women. Among the 392 women who did not have a high pretest probability, 11.7% had a negative D-dimer result and did not require diagnostic imaging according to this algorithm. Bilateral CUS was mandated in women with a high pretest probability or low/ intermediate pretest probability with a positive D-dimer because CTPA could be avoided if the CUS revealed DVT. Of these 397 women undergoing CUS, only 7 had a positive result (an overall diagnostic yield of only 2%).
A second multicentre, international prospective management study with a design that also included withholding of anticoagulation in algorithm-defined lower-risk women and clinical follow-up at 3 months evaluated a pregnancy adapted YEARS algorithm 56 with parallel assessment of the three YEARS criteria (clinical signs of DVT, haemoptysis and pulmonary embolism as the most likely diagnosis) and D-dimer levels followed by selective CTPA in 498 women with suspected PE during pregnancy. 57 In women with no YEARS items and a D-dimer <1000ng/mL, or in patients with ≥1 YEARS item and D-dimer <500ng/ mL, PE was deemed excluded. All other patients underwent diagnostic imaging, consisting of CTPA (or CUS if women had clinical signs of DVT, without further imaging if the CUS revealed DVT). The primary outcome was the cumulative incidence of symptomatic VTE, with confirmation by objective tests, during a 3 month follow up period in women in whom anticoagulation had been withheld on the basis of a negative algorithm result. In 510 consecutive patients screened, 4.0% were diagnosed with PE.
At 3-month follow up, only one woman developed a popliteal DVT (0·21%; 95% CI: 0·04-1·2%) and no women developed PE, demonstrating the safety of this algorithm for ruling out PE during pregnancy. Exposure to diagnostic imaging could be avoided in 39% (95% CI: 35-44%) of patients. D-dimers increased throughout pregnancy. Consequently, the proportion of women in whom PE could be excluded without diagnostic imaging was lower in the first trimester (32%) and higher in the third trimester (65%). The diagnostic yield of targeted CUS was higher than that reported in the Righini et al study, at 7%.
Collectively, these prospective data derived from high-quality clinical management studies prompted a change in the recommendation given in the 2019 ESC Guidelines on Diagnosis and Management of Acute PE38: that D-dimer measurement in conjunction with clinical prediction rules “should be considered” during investigation of suspected PE in pregnancy. Further prospective clinical trials and studies may consider further evaluating the optimal d-dimer threshold and clinical prediction rule in the evaluation of a pregnant woman with a suspected PE.
Treatment of VTE in pregnancy
The management of VTE during pregnancy should be delivered by an experienced multidisciplinary team 38 and ideally, written care pathways should be agreed and discussed with the woman who has experienced the VTE event. LMWH is recommended by international guidelines for the treatment of VTE during pregnancy, with a more favourable risk profile and more predictable pharmacokinetics compared with unfractionated heparin. 13,15,38 Unlike vitamin K antagonists (VKAs) and direct oral anticoagulants (DOACs), LMWH does not cross the placenta and there is no associated risk of fetal haemorrhage or teratogenicity.
DOACs are contraindicated in pregnant patients. 58 For most women receiving LMWH treatment for VTE during pregnancy (except in exceptional circumstances including renal impairment and at extremes of body weight 59,60 ), the measurement of plasma anti- FXa activity to guide dosing is not recommended in view of the predictable pharmacokinetic profile of LMWH, lack of data on optimal anti-FXa levels and limitations of the assay 61. Unfractionated Heparin (UFH) use in previous decades was associated with heparin-induced thrombocytopenia and bone loss resulting in a reported 2-9% risk of osteoporotic fracture following prolonged UFH use during pregnancy (summarized in 62 ). However it remains uncertain whether the risk is increased with LMWH use: in a recent observational cohort study, lumbar spine bone mineral density was similar in LWMH-treated women to controls following adjusting for potential confounders. No osteoporosis or osteoporotic fractures were reported. 62
Management of LMWH in the peripartum period for women diagnosed with VTE during pregnancy is not supported by high-quality data. The incidence of spinal haematoma after regional analgesia in pregnant women is uncertain. Moreover, pregnant women are in a procoagulant state at the time of delivery. International guideline recommendations vary on timing of peripartum regional analgesia 63,64 and it is very important to consider the competing risk of performing a caesarean section under general anaesthesia. In general, regional analgesia should be avoided unless LMWH has been discontinued at least 24 hours before delivery (assuming normal renal function and including risk assessment at extremes of body weight). Current guidelines remind us that data are limited on the optimal timing of postpartum LMWH re-initiation and should be guided by a personalized risk assessment, taking into account mode of deliver and thrombotic versus bleeding risks. 64,65 Recent European Society of Cardiology guidelines recommend that “LMWH should not be given for at least 4 hours after removal of the epidural catheter; the decision on timing and dose should consider whether the epidural insertion was traumatic and take into account the risk profile of the woman. For example, an interim dose of a prophylactic LMWH dose may be considered postoperatively (after caesarean section), once at least 4 hours have elapsed since epidural catheter removal and allowing for an interval of at least 8–12 hours between the prophylactic and the next therapeutic dose” 66,67 _ ENREF_119
Close collaboration between the obstetrician, the anaesthesiologist, and the attending physician is important. After delivery, a planned transition from LMWH to VKA therapy can be considered by the multidisciplinary team. Anticoagulant treatment should be administered for at least 6 weeks after delivery and with a minimum overall treatment duration of 3 months. LMWH and VKAs can be given to breast-feeding mothers.
Excellent prevention, diagnosis and treatment strategies should be a priority for health service providers globally. Preventing PA-VTE is reported to be the most readily implementable means of systematically reducing the rate of maternal death at a population level. 68,69 However, effective prevention is limited by a paucity of evidence to support effective interventions and challenges with implementing of VTE prevention strategies in clinical practice. A systematic approach to VTE prevention should be adopted to reliably identify women at increased risk and ensure consistency of VTE prevention where indicated.
References available on request
Written by: Osas Edebiri 1, Fergal O’Shaughnessy 2,3, Brian Cleary 2,3 and Fionnuala Ní Áinle 4,5,6
1. Dept of Internal Medicine, Darlington Memorial Hospital, United Kingdom
2. Pharmacy Department, Rotunda Hospital, Dublin 1, Ireland
3. Division of Population Health Sciences, Royal College of Surgeons in Ireland, Dublin 2, Ireland
4. School of Medicine, University College Dublin, Dublin 4, Ireland
5. Department of Haematology, Rotunda Hospital, Dublin 1, Ireland
6. Department of Haematology, Mater University Hospital, Dublin 7, Ireland
Read the full magazine: HPN October Edition
Read our Features