At the annual Roche Diagnostics-sponsored event pro-COAG event (2022), Prof Lai Heng Lee gave an insightful presentation on the risk management of venous thromboembolism (VTE) associated with hormone therapy and pregnancy. Below are some key takeaways.

Hormone-associated venous thromboembolism (VTE)

The incidence of VTE in young women is generally 1/10,000 per year. However, hormonal therapy such as the use of combined oral contraceptives (COCs) and hormone replacement therapy (HRT) is associated with increased risks of VTE. COC increases the risk of VTE by approximately 2 to 4 fold when compared to non-users. Nonetheless, the absolute risk is still low, estimated to be less than 0.1%. In HRT, the increase in VTE risk is more modest at around 1.5 to 2 times.

COCs cause changes in the hemostatic system leading to hypercoagulable states, mainly by inducing an increase in procoagulant proteins with increased levels of fibrinogen, prothrombin, factors VII, VIII and X.  It also causes a decrease in levels of natural anticoagulants such as antithrombin and protein S (PS). As PS is a cofactor for protein C (PC), the reduction in PS activity leads to acquired activated PC resistance (APC) state.

Many studies have shown that the VTE risks of COC vary with the composition of the pill. Higher estrogen doses (>30 μg of ethinyl estradiol) and non-levonorgestrel progestin containing COCs such as drospirenone, desogestrel, gestodene and cyproterone acetate are associated with a higher VTE risk. Similarly, the risks for VTE also vary with the composition of HRT products. Both esterified estrogen (EE) and conjugated equine estrogens (CEE) are associated with a significant risk for VTE, although the risk seems to be less for EE compared to CEE. The addition of progesterone to estrogen was thought to increase the VTE risk when compared to estrogen alone but a recent review and meta-analysis failed to document a significant additional risk when progesterone is added to estrogen.

Non-oral alternatives such as injectable progestin, patches and vaginal rings are also associated with increased VTE risks. Progestin-only subcutaneous implants with etonogestrel have 1.4 fold higher risks compared to non-users, but the absolute risks remain small with an incidence rate of 1.7 per 10,000 exposure years. Progesterone-only intrauterine devices, progesterone-only pills or transdermal estrogen-only HRT formulations do not appear to be associated with an increased VTE risk.

HRT was first introduced in the 1960s for postmenopausal women in order to prevent cardiovascular events and osteoporosis as well as to alleviate symptoms related to menopause. However, current data point out that the HRT-associated risk for VTE, breast cancer and lack of protection against cardiovascular diseases counterbalances the benefits from the reduction of osteoporosis. Moreover, HRT does not seem to improve the quality of life in postmenopausal women without clinical symptoms, so HRT should be a therapeutic choice only for postmenopausal women who need to control clinical symptoms.

It is important for women on hormone treatment to be aware of other VTE risk factors, which may be additional or synergistic in increasing the VTE risks. Women older than 50 years of age have an increased VTE risk of 6.3 fold if taking COCs and 4 fold if on HRT. Obesity and smoking also confer greater VTE risks. It is important to counsel the patient on healthy lifestyles to avoid obesity and smoking to reduce their risks of VTE while on COCs or HRT. If going for surgery, the patient should let their doctors know if they are on hormone treatment or if they have a family history of VTE, so that appropriate VTE prophylaxis can be applied.

After an initial VTE event for women on hormonal treatment, estimates for the cumulative risk for recurrent VTE straddled across a wide range from 0.4% to 5% per year, and 2.3% to 9% at 5 years. The differences in risk across the studies may be based on the year of study where different generations of COCs could have been used over time, and different inclusion criteria resulting in heterogeneous risks across different studies. In a more recent updated subgroup analysis of the original REVERSE cohort, women at low risk for HERDOO2 score, COC users had a recurrence rate of 0.4% per year vs 1.4 % per year for non COC users, while those with high risks, COC users had a recurrence rate of 3.5% per year vs 6.1% for non COC users. Hence other risk factors for VTE have to be considered in deciding the duration of anticoagulation for such women.

It is reassuring that continued hormonal therapy with COCs while on therapeutic anticoagulation is not associated with an increase in recurrence. When anticoagulation treatment is stopped, COC or HRT should be discontinued to minimise the risk of VTE recurrence. For prevention of pregnancy, other forms of contraception that do not carry an increased risk of VTE recurrence can be considered such as progestin-only IUDs or progestin-only subcutaneous implants.  Overall, these risk estimates of VTE recurrences are reassuringly low in women on hormonal therapy and long term anticoagulation therapy is not justified in the absence of other risk factors.

There had also been much interest in the thrombotic risks of women with inherited thrombophilia while on hormone therapy. Factor V Leiden (FVL) and Prothrombin Gene Mutation (PGM) are 2 milder forms of genetic thrombophilias common in the caucasian populations, increasing the risks of VTE 4-7 times. Antithrombin, PC and PS deficiencies are uncommon but much more potent in thrombogenic risks. Antithrombin deficiency has been reported to have increased VTE risks by 50 times, while PC and PS deficiencies are associated with increased VTE risks by 10-15 times. Therefore, the presence of a genetic thrombophilia greatly increases the thrombotic risks in women on the OCs.

Once again, knowing the absolute risks however will help put things in perspective when counselling patients with a known personal or family history of thrombophilias. The absolute risks of VTE in women on COCs with FVL or PGM range from 0.45 to 2 per 100 pill years, and these VTE risks may be more acceptable to some women when compared to the risk of getting pregnant. However, with the more potent thrombophilias where strong data is lacking, the absolute risks had been reported to between 4.3 to 4.5 per 100 pill years and more caution and alternative methods of contraception advised.

Even so, unjustified routine massive screening for thrombophilia prior to the use of COCs is not recommended. Such screening is not cost effective as a large number of women needs to be tested and advised to abstain from COCs use in order to prevent 1 single thrombotic event. Besides, the absolute risk of VTE is small for the milder thrombophilias and the majority of VTE events develop in women who do not have genetic thrombophilia identified. However, selective testing prior to administration of COCs might be useful in asymptomatic relatives of carriers of severe thrombophilias that are associated with much increased thrombotic risks, such as Antithrombin, PC and PS deficiencies and homozygosity for FV Leiden and FII G20210A.

Pregnancy-associated venous thromboembolism

Pregnancy, with its physiological changes, leads to increased levels of coagulation factors (VII, VIII, X and von Willebrand factor) with concomitant decrease in natural anticoagulants (PS). This hypercoagulable state together with increased susceptibility to venous stasis in the lower limbs and endothelial damage to pelvic vessels fulfil all of Virchow’s classic triad of prothrombotic factors underlying the pathogenesis of abnormal venous clots, predisposing to VTE.

The risks of VTE in pregnancy approximate 1 in 1000 pregnancies, with similar incidence of VTE similar in antepartum and postpartum periods. As the postpartum period is shorter than antepartum period, the daily VTE risk is higher and this increased risk persists until 12 weeks postpartum, with greatest risk in the first 6 weeks after delivery.

Pulmonary embolism (PE) is fatal in 25% of patients if left untreated. Hence VTE is a major cause for maternal morbidity and mortality which accounts for 17% of maternal death in western populations. It is very important to know and stratify the risk factors for developing VTE in pregnancy so that appropriate thrombo-prophylactic measures can be applied.

There are many guidelines on VTE risk stratification and management for pregnancy-associated VTE, which are mainly based on data extrapolated from non-pregnant patients and case series of pregnant patients in western populations.  It is important that we develop local guidelines based on local data. In Singapore, the Chapter of Physicians and College of Obstetrics and Gynaecologist worked together to review these guidelines and available local data to develop consensus recommendations for management on pregnancy-associated VTE. The stratification of VTE risk was adopted from the Royal College of Obstetricians and Gynecologists (UK) Greentop Guideline 37a with adaptations based on our local data and clinicians input.

Risk factors for developing VTE included in the guidelines were based on the history of previous VTE, inherited thrombophilia, patient’s existing risk factors, obstetric risk factors and other transient risk factors. Points given for each risk factor are determined according to the weightage of thrombotic risks for each risk factor. Recommendations for thrombo-prophylaxis are based on the number of points or risk scores. Hence a total score 4 and above antenatally will be considered for thromboprophylaxis once pregnancy is diagnosed while those with 3 points will be considered for thrombo-prophylaxis from 28 weeks gestation; postnatal thromboprophylaxis strategies are also based on the score matrix.

As the risks for VTE are dynamic and change with appearance of new risk factors or resolution of previous risk factors as the pregnancy progresses, it is important to repeat assessments for VTE risks at various time points throughout pregnancy and puerperium. It is also important to assess the bleeding risks. If a patient is not suitable for anticoagulant prophylaxis, mechanical prophylaxis with compression devices or stockings can be considered.

The adaptations made in the Singapore consensus recommendations are the weightage points assigned to genetic thrombophilias and increasing maternal age. The more severe forms of genetic thrombophilias such as deficiencies of Antithrombin (AT), PC and PS deficiencies are more prevalent in Asia compared to the western populations (AT deficiency being the most severe is associated with 50 fold increased risk of VTE while PC and PS def increases the risk by 10-15 fold). Our Workgroup came to the consensus that antithrombin deficiency should be given a risk score of 4 and this risk factor alone would justify initiation of thromboprophylaxis once pregnancy is diagnosed.

Patients with Protein S or Protein C deficiency and additional VTE risk factors such as family history of VTE should be considered for antenatal and postpartum prophylaxis (up to 6 weeks postpartum). More Asian data is required to determine the clinical significance and risk stratification of these thrombophilia.

As data pertaining to age as a risk factor for thrombosis in the current literature are conflicting, we look at our recent local study over a 12-year period from 2004 to 2016 at 2 tertiary maternity hospitals (KKH and SGH), looking at 89 women with pregnancy-associated VTE and 926 controls. It found that age more than or equal to 35 years of age was not a statistically significant risk factor for VTE. A Korean study similarly found that increased age was not associated with VTE in pregnancy. However, a large UK population-based cohort study found that outside of pregnancy, women in the oldest age band (35–44 years) had a 50% higher rate of VTE than women aged 25–34 years. This study also found that although the rate of VTE did not increase with age in the antepartum period, women aged 35 and over had a 70% increase in risk compared to 25–34-year-olds in the postpartum period. Based on the available evidence, the committee thus decided to use age as a risk factor for thrombosis only in the postnatal risk assessment for thrombotic risk. Advanced maternal age of more than 35 years as a risk factor for thrombosis has been included only as part of the postnatal assessment for thromboprophylaxis.

On the management of a confirmed new VTE, anticoagulation is the cornerstone of treatment. If left untreated, 25% of deep vein thrombosis (DVT) will develop clinically relevant PE and there is 10-15x increase in recurrent VTE in inadequately treated VTE. One cannot overstress the importance of adequate therapeutic doses and appropriate duration of anticoagulation treatment. In the choice of the anticoagulant drug, it is important to consider the maternal and foetal complications as well as mothers’ preference. In the mother, there are thrombotic complications of under treatment and bleeding complications of over-anticoagulation. If the anticoagulant drug crosses the placenta, it may cause teratogenicity and bleeding complications in the foetus.

Low molecular weight heparin (LMWH) is the anticoagulant of choice in the treatment of VTE in pregnancy. This is due to the ease of administration, lower risk of osteoporosis when compared to unfractionated heparin (UFH), reduced rates of bleeding and lower risk of heparin induced thrombocytopenia. Both UFH and LMWH do not cross the placenta and hence are not associated with risk of teratogenicity or foetal bleeding. Vitamin K antagonists should be avoided as they cross the placenta and cause warfarin embryopathy (midfacial and limb hypoplasia, stippled bone epiphyses).  Warfarin has also been associated with loss of pregnancy and foetal anticoagulation and bleeding. Hence, unless under exceptional circumstances, we recommend that warfarin be avoided. Direct oral anticoagulants (DOACs) such as dabigatran, apixaban, edoxaban, and rivaroxaban are not recommended as they cross the placenta. Increased miscarriages and foetal anomalies have been reported and there is a lack of safety data in pregnancy.

Anticoagulation should be maintained at therapeutic doses of anticoagulants for the duration of pregnancy up to at least 6 weeks postpartum, and for a total of 3 months minimum. If a patient is still pregnant after adequate duration of therapeutic anticoagulation, the doses may be decreased to prophylactic doses. Patients should undergo another VTE risk scoring prior to cessation of anticoagulation.

Epidural anaesthesia for delivery can be delivered with particular attention given to the timing of administration. Decisions can be made on an individual basis after discussion with the patient. It is safe if PTT is normal and no administration of standard heparin for 4-6 hours has been done prior to epidural catheter insertion. For patients receiving LMWH, epidural anaesthesia can be administered at least 12 hours after the last prophylactic dose and 24 hours after the last therapeutic dose of LMWH.

Anticoagulation and delivery plans should be made in conjunction with the anaesthetist, besides the obstetrician and haematologist or physician. Caesarean Section Indicated for Obstetric Reasons only and anticoagulation plans should be made to support the mode of delivery.

For patients on therapeutic dose of anticoagulation, a scheduled delivery with prior discontinuation of anticoagulation for patients on therapeutic anticoagulation may facilitate epidural anaesthesia and reduce maternal bleeding risks. For patients on prophylactic-dose LMWH, spontaneous labour may be feasible, with cessation of anticoagulation at onset of labour as a 12-hour interval between last dose of standard prophylactic LMWH and epidural catheter would allow most women to receive epidural anaesthesia regardless of scheduled or spontaneous delivery.

After delivery, the epidural catheter should not be removed within 24 hours of the most recent therapeutic LMWH injection. LMWH should not be given till at least 4 hours after the use of spinal anaesthesia or after epidural catheter removal. If the patient is not at high risk of bleeding, with hemostasis secured, has normal renal function and no coagulopathy as well as complete neurological recovery after epidural anaesthesia, prophylactic dose of LMWH can be restarted 4-6 hours after delivery and therapeutic LMWH restarted 24 hours after delivery. Breastfeeding is safe for babies when the mother is on LMWH or warfarin. DOACs are not recommended in a nursing mother as it is being secreted into the breast milk and there is a lack of safety data.

References:

[1] Oral contraceptives and hormone replacement therapy: How strong a risk factor for venous thromboembolism?  Leslie Skeith a,, Gr´egoire Le Gal , Marc A. Rodger. Thrombosis Research 202 (2021) 134–138

[2] Reducing the Risk of Venous Thromboembolism during Pregnancy and the Puerperium RCOG Green-top Guideline No. 37a, 2015

[3] CONSENSUS STATEMENT on VENOUS THROMBOEMBOLISM IN PREGNANCY – RECOMMENDATIONS FOR PREVENTION, TREATMENT AND INVESTIGATION. College of Obstetricians and Gynaecologists (COGS), Chapter of Haematologists, College of Physicians, Singapore.