Mechanical heart valves and pregnancy

Abstract

Pregnancy is an exciting time, with great anticipation of the future for the woman and her baby. However, for women with mechanical heart valves (MHVs) pregnancy is a challenging undertaking fraught with risk to them and their baby. This article considers the issues for the woman, highlighting pre-surgical opportunities, pre-pregnancy counselling, the issues to consider with fertility treatment, the options during pregnancy for anti-coagulation management and the recognition and management of complications. This update stresses the complex challenges in the management of women with MHV during pregnancy and the need for specialist multi-disciplinary team involvement.

Introduction

Pregnancy is a normal physiological process that is generally associated with excitement and joy. As the demographics of the obstetric population change, with increasing age, increasing rates of diabetes and increasing obesity, the incidence of cardiac disease is increasing. Cardiac disease is recognised in up to 4% of pregnancies. ¹

Valvular heart disease typically affects older adults, however improved outcomes in patients with congenital cardiac conditions or acquired disease such as rheumatic fever means there are more women of childbearing age with significant valve disease. Some of these women will require close monitoring of their native valve function during the haemodynamic challenges of pregnancy. But some will already have had surgical intervention to their heart valves. Surgical intervention can include valve repair or valve replacement. Valve replacement surgery can include tissue replacement (e.g. a tissue prosthesis or more complex options such as the Ross procedure) or mechanical valve replacement. Approximately 250,000 prosthetic heart valves are being implanted worldwide each year, many of them in woman of childbearing age. ² Valve replacement may be required for acquired disease such as rheumatic heart disease or congenital heart disease (CHD) and there are increasing numbers (90%) of children with CHD surviving to adulthood. ³

The exact percentage of women of reproductive age having a mechanical valve is uncertain. ⁴ In the UK Obstetric Surveillance System (UKOSS), incidence of mechanical heart valve (MHV) in pregnancy was estimated at 3.7/100,000 maternities and was associated with a 9% maternal mortality and 41% maternal morbidity. ⁵ In the Registry of Pregnancy and Cardiac Disease (ROPAC), maternal mortality was reported as 1.4% in patients with mechanical valves, whilst valve thrombosis had an incidence of 5% during pregnancy and was associated with a 20% mortality. ⁴ The modified World Health Organisation (mWHO) maternal cardiovascular risk suggests an estimated rate of an event free pregnancy with a live birth of 58%. ⁶ These high-risk pregnancies are fraught with the balance of maintaining maternal haemostatic equilibrium whilst trying to ensure foetal wellbeing.

Options for valve intervention

Valve repair, when possible, would be preferable. ⁴ The Ozaki procedure involves fashioning of new valve leaflets from the patient's own pericardium or bovine pericardium, is a relatively new alternative that may be discussed at specialist centres. ⁷ The Ross procedure in which the patient's own pulmonary valve is used in the aortic position and a pulmonary homograft inserted provides an alternative surgical option in women with favourable anatomy and at centres with expertise in the procedure. ⁸ However, the Ross procedure can occasionally be associated with aortic dilatation which itself is a significant risk factor in pregnancy, but this may be addressed with novel surgical techniques. Whilst the Ross procedure offers excellent haemodynamics, it is technically difficult with ∼2% mortality and survival from re-operation of 76% at 10 years post-procedure. ² Mortality associated with re-operation was 20% in the first post-operative year. Therefore, it is a procedure that should be performed by a surgical team with significant expertise.

There are many occasions that valve repair is either not possible or not durable and valve replacement is the best option. ³ An important consideration in this cohort of young adults is consideration of future pregnancy therefore thorough counselling is required including the discussion of risks and benefits of a mechanical versus bioprosthetic prosthesis, including highlighting the challenges of anti-coagulation management during pregnancy.^3,9

Prosthetic heart valve replacement in young adults, whilst necessary, may be challenged by complications over their remaining lifespan including bleeding, thromboembolism and the need for re-intervention. ¹⁰ The most common options are valve replacement with a mechanical prosthesis which offers excellent haemodynamic performance and long-term durability but the necessity for anticoagulation increases maternal and fetal morbidity and mortality with a higher risk of major cardiac events during pregnancy in comparison to bioprosthetic valves (Figure 1).^1,11 Alternatively, a bioprosthetic valve may be considered which does not require anti-coagulation however has limited durability and in young women are associated with a high-risk of structural valve deterioration within 10 years of initial valve implant, resulting in the possibility of going through pregnancy with a dysfunctional valve depending on duration of initial implant, and eventually the requirement for re-operation.^1,7 The risk of structural valve deterioration of a bioprosthetic valve was seven-fold greater in the mitral position than in the aortic position. ² As bioprosthetic valves are associated with the likelihood of re-operation, the early mortality of re-operation of which reports range from 3.8% to 8.7% needs to be considered. ²

Figure 1.

Outcomes of patients with mechanical and bioprosthetic valves were assessed relative to patients with no prosthetic valve. Pregnancies with valve prostheses had increased odds of maternal complications, such as ante/post-partum haemorrhage and requiring blood transfusion, as well as foetal complications, such as preterm birth and stillbirth. HDP indicates hypertensive disease of pregnancy; and PPROM, preterm premature rupture of membranes. *P < 0.05 (see Figure 2 of Ng et al. ¹¹ ).

There are conflicting reports regarding the possibility of pregnancy accelerating structural valve deterioration of bioprosthetic valves. ² However, in recent prospective studies it was concluded that structural valve deterioration was not accelerated by pregnancy but was attributed to the natural course of the bioprosthesis. ¹² A recent report has suggested that the presence of a valve replacement was associated with maternal and foetal risk irrespective of the nature of the valve replacement. ¹¹

Homografts are thought to have superior haemodynamics, and lower rate of infection and the suggestion of better durability observed in some retrospective studies in comparison to bioprosthetic valves. ²

Current European Society of Cardiology (ESC) guidelines consider a desire for pregnancy as a Class IIa recommendation for selection of a biological valve. ¹ Tissue valves are associated with less morbidity and less foetal loss than mechanical valves and this should be discussed with patients prior to valve replacement. The decreased durability of a tissue valve should be explained as a potential issue later in life when it fails with the acknowledgement that this will require further surgery. ⁴

Pre-pregnancy counselling

A number of algorithms have been developed to risk stratify cardiac conditions in pregnancy. The most well-known are CARPREG I and II, ZAHARA and mWHO. ³ mWHO has been suggested to be the most accurate although it is imperative that each patient undergoes individual risk assessment by a specialty team in an experienced centre. ³ Presence of a mechanical valve has been classified as risk category III in the mWHO classification, within this category there is a 19% to 27% maternal cardiac event rate with significantly increased risk of maternal mortality or severe morbidity. ¹ Pre-pregnancy assessment should include assessment of symptoms and echocardiography.^1,8,9 They should receive careful counselling by a multi-disciplinary team with knowledge and experience of the risks of pregnancy in patients with mechanical heart valves.^1,8 In the UKOSS report 49% of women with MHV either did not receive pregnancy counselling or it was not recorded. ⁵

In women with CHD, pre-pregnancy counselling should begin during teenage years with information addressing miscarriage rates, fertility, long-term prognosis, recurrence risk and estimated maternal and foetal outcomes.^6,13 In women who are being considered for heart valve surgery, pre-surgical counselling should include discussion with a pre-pregnancy counselling team. This would seek to allow the woman to consider the balance of benefits and risks of each of the surgical strategies with consideration of their anticipated pregnancy plans.

Maternal obstetric, fetal and neonatal risks should be discussed. ¹² Risk stratification of thrombotic events based on prosthesis (type, position and function) and patient factors (previous thromboembolism, left ventricular systolic function, atrial fibrillation or issues with medication adherence) should be considered when deciding upon anti-coagulation regimen in pregnancy. ⁹ The choice of anticoagulation and route of administration of each anti-coagulation choice is important to discuss with patients as switching from vitamin K antagonist (VKA) to low molecular weight heparin (LMWH) involves switching form oral to subcutaneous preparations and patients may not be aware of this or willing to accept subcutaneous injections for prolonged periods.

In women with MHV, inappropriate anti-coagulation dosing and monitoring as well as medication non-adherence contributes to maternal morbidity and mortality therefore optimisation should be sought prior to pursuing pregnancy.^8,9

Patients should be aware of the importance of performing pregnancy tests early if they think they could be pregnant and contacting the specialist team at an early stage. ⁹

Fertility treatment

Subfertility affects approximately one in seven couples. An increasing number of older females are seeking assisted conception with almost 40% aged 38 or over. ¹⁴ Assisted reproduction using ovarian stimulation is an additional risk factor for thrombosis. ⁹ Involvement of the specialist team who can recommend on the safety/suitability of assisted conception should be consulted prior to proceeding. ⁹

It is imperative to recognise the potential increased risk of assisted conception in women with heart disease. In a multi-centre retrospective review of outcomes and complications in women with cardiac disease undergoing fertility treatment, 34 women were identified, with a total of 50 cycles undertaken, mean age 32. ¹⁵ Out of the 32 patients, 4 developed complications during fertility treatment including a patient on anti-coagulation who subsequently required a laparotomy 5 days post egg retrieval. Correlation between higher mWHO class and higher rates of complications was identified. Potential cardiac risks of assisted conception in patients with known cardiac disease should be considered and discussed with patients prior to commencing treatment.

The MBRRACE-UK report identified cardiac disease as the commonest indirect cause of death for pregnant woman in the United Kingdom. Along with other recommendations, it emphasises the necessity for guidance on maternal screening and medical assessment prior to assisted reproduction. ¹⁶

Best practice recommendations for medically assisted reproduction in patients with known cardiovascular disease were published in 2024. ¹⁴ Within these, it was recognised that anti-coagulation poses a significant risk of bleeding complications from the ovarian bed surrounding the time of egg collection and for the subsequent 7 to 10 days. It is recommended that there is discussion with the patient and involvement of the multi-disciplinary team including a coagulation expert and detailed protocol for management of anti-coagulation is put in place prior to egg collection. The British Society of Haematology (BSH) recommends that for invasive procedures such as egg collection in a patient who is anti-coagulated with VKA, that the VKA should be bridged with therapeutic LMWH for the minimum period prior to restarting the VKA, the last dose of LMWH should be >24 hours prior to the scheduled procedure and immediate therapeutic anti-coagulation should be avoided post-procedure. ⁹

Pregnancy

Women with a known cardiac history who become pregnant should be referred urgently to the specialist team and reviewed by the multidisciplinary team (MDT) ideally prior to 6 weeks gestation. ⁹ With pre-pregnancy counselling, a written plan should be shared with the patient, primary care team and anti-coagulant service describing the plan for anti-coagulation on confirmation of a positive pregnancy test. The UKOSS study found that 19% of women with an MHV were not referred to either a tertiary centre or to specialist cardio-obstetric service. ⁵ The time of highest risk of MHV is early in pregnancy at time of transition between anti-coagulant regimes if that is the chosen plan. Close monitoring at a tertiary centre with involvement of the multi-disciplinary team including cardiologists, maternal foetal medicine obstetricians, surgeons and anaesthetists with expertise in the management of high-risk cardiac conditions during pregnancy is recommended. ⁸

Anti-coagulation regimen during pregnancy

During pregnancy there is increased risk of thromboembolic complication (TEC) due to increased production of procoagulant factors, acquired protein C resistance, decreased levels of protein S and impaired fibrinolysis. ¹⁷ Patients with mechanical valves are at increased risk of TEC including valve thrombosis with resultant valve obstruction +/− regurgitation and the subsequent risk of systemic embolism including myocardial infarction and stroke in left-sided MHV and pulmonary embolism in right-sided MHV. ¹⁸ The risk of thrombosis is related to the type of valve and site, as well as the presence of atrial fibrillation or left ventricle systolic dysfunction. Conversely, in the ROPAC report there was no significant difference in the occurrence of mechanical valve thrombosis (MVT) in the mitral position (4.4%) compared with aortic position (2.6%; P = 1.00). ⁴ ROPAC also observed that MVT was observed in all stages of pregnancy and on every possible anti-coagulation regimen. In addition to increased risk of TEC, there is also a high risk of post-partum haemorrhage. ⁹

Many studies are observational and the potential for confounding cannot be excluded. In addition to this most studies are small therefore limiting conclusions. ¹⁸

There are three possible regimens of anti-coagulation during pregnancy for women with MHVs ¹⁹ ;

• VKA throughout the duration of pregnancy

• Heparin throughout pregnancy

• Sequential therapy with heparin in the first trimester and VKAs in the second and third trimesters.

A comparison of the different anti-coagulant regimens is summarised in Table 1.

Table 1.

Maternal and fetal benefits and risks of different anticoagulant regimens in pregnant individuals with MHV.⁹

Anti-coagulation regimen	Maternal benefits/risks	Foetal benefits/risks
Warfarin from a positive pregnancy test until 36 weeks: Strongly consider when high risk of maternal MVT Continue warfarin till 36 weeks Introduce LMWH from 36 weeks to delivery to minimise the risk of foetal and maternal haemorrhage	Lowest composite maternal risk (death, mechanical valve failure or thromboembolism) at 5% Lowest maternal mortality rate 0.9% (0.4–1.4) Lowest thromboembolic complications rate 2.7% (1.4%–4%) However, there is an increased risk of bleeding in some studies (both antepartum and post-partum)	Highest overall foetal loss (32.54%) Highest overall composite foetal risk (miscarriage, termination, foetal abnormality) 39% Live births 83.6% (75.8–91.4) with warfarin <5 mg/day and 43.9% (32.8–55) with warfarin >5 mg/day Foetal embryopathy/foetopathy 2.3% (0.7–4) with warfarin <5 mg/day and 12.4% (3.3–21.6) with warfarin >5 mg/day Foetal intracranial haemorrhage thought to be a risk with vaginal delivery if recent (<1 week) warfarin dosing
LMWH throughout pregnancy	Use of LMWH throughout has a higher composite maternal risk of 15.5% Higher mortality rate 2.9% (0.2–5.7) Higher thromboembolic complications 8.7% (3.9–13.4), but reported up to 53% Optimal dosing with the use of anti-Xa levels is not yet fully established so close monitoring is required	Lowest composite foetal risk at 13% Overall foetal loss rate (mostly miscarriage) 12.2% to 13.9% Highest livebirth rate 92% (86.1–98) No placental transfer to foetus during pregnancy
Combination (sequential LMWH/warfarin/LMWH) LMWH from positive pregnancy test to week 13 Warfarin weeks 13–36 LMWH week 36 until delivery (as above)	Combination (sequential LMWH/Combination/sequential LMWH/warfarin has a similar composite maternal risk to LMWH alone, of 15.9% Mortality rate 2% (0.8–3.1) which is lower than LMWH throughout, but higher than with warfarin alone MVT risk higher than warfarin 50% of MVT in pregnancy occur during first trimester LMWH Thromboembolic complications, 5.8% (3.8–7.7) Requires very close monitoring during transitions to limit subtherapeutic anti-coagulation Requires high maternal compliance with treatment/monitoring regimes	Composite foetal risk 23% Overall foetal loss rate 22.65% Foetopathy (foetal haemorrhage) 1.4% (0.3–2.5) Live birth 79.9% (74.3–85.6) Better livebirth rate than with >5 mg/day warfarin No placental transfer to foetus during first trimester
UFH only	Very rarely recommended due to MVT > 10% (11.2%) Low rate of live births however number of and risk of maternal osteoporosis and thrombocytopaenia	Low rate of live births however number of pregnancies analysed small
Combination (UFH/warfarin)	Composite maternal risk of 33.6% Slightly higher than LMWH/warfarin regimen More complex drug delivery/monitoring Osteoporosis/thrombocytopenia	Composite foetal risk of 34%
Other anti-coagulants: for example, dabigatran, rivaroxaban, apixaban, fondaparinux	Have not been shown to be as effective as warfarin in patients with MHVs, and cross the placenta so are not recommended Not recommended in individuals that are breast feeding	Limited evidence, currently contraindicated

LMWH: low molecular weight heparin; MHV: mechanical heart valve; MVT: mechanical valve thrombosis; UFH: unfractionated heparin.

It is recommended that VKA is continued when trying to conceive. If a change in anti-coagulant regime is planned, then this should be before 6 weeks gestation as the early effects of VKA on the foetus start from 6 weeks gestation. ⁶ It is advised that woman perform pregnancy testing early if there is a possibility they may be pregnant and have a low threshold for performing a pregnancy test, especially when menstruation is irregular.

Importantly, trials have demonstrated that anti-Xa direct oral anti-coagulants are not safe in patients with mechanical heart valves and therefore are not an option in pregnancy. ⁸

VKAs-warfarin

VKAs are the anti-coagulant of choice for mechanical valves in non-pregnant patients and within pregnancy VKAs are superior to LMWH for preventing MVT and are associated with the lowest risk of maternal adverse outcomes.^4,9,17 However, VKAs crosses the placenta, are teratogenic and can adversely affect the foetus. ¹⁸ Embryopathy is a recognised complication of warfarin exposure between 6 and 12 weeks’ gestation. Embryopathy entails nasal bone hypoplasia, stippled epiphysis and choanal atresia. Foetal exposure to warfarin later in pregnancy is associated central nervous system abnormalities and intracranial haemorrhage (foetopathy). ⁶ There is an increased risk of miscarriage and stillbirth that can occur at any gestational age with an estimated foetal death rate of 32.5% (29.6–35.5). ⁶ The risk of foetal loss is thought to be dose-related, with lower foetal loss 13.4% to 19.2% with low-dose VKA. ¹ This dosing difference has been considered in both the ESC and American Heart Association/American College of Cardiology (AHA/ACC) guidelines. Conversely, the data that has led to the cut off of 5 mg daily dose of warfarin is controversial as it is based upon studies with small numbers and there have been conflicting studies to this.^9,20 So, whilst foetal adverse event rates increase above warfarin doses >5 mg/day, foetal loss and adverse events can still occur in the context of daily doses <5 mg/day, therefore appropriate counselling is required. ²⁰

In a meta-analysis, the use of VKAs with standard INR targets throughout pregnancy (2.5–3.5) was associated with the lowest pooled proportions of maternal morbidity and TECs, as well as the lowest event rates of major bleeding, cardiac events or adverse drug events ¹⁹ : therefore, it was the safest option for maternal outcome. However, anti-coagulation-related embryopathy and foetopathy was observed in ∼2% of the foetuses exposed to VKA throughout pregnancy. In contrast there were no cases of anti-coagulant induced embryopathy/foetopathy in the foetus exposed to LMWH only and lower foetal and neonatal complications when heparin was used throughout pregnancy or with sequential treatment.

There is a high risk of traumatic foetal haemorrhage, foetal death and maternal haemorrhage during labour for patients on warfarin. ⁶ ESC recommends switching warfarin to LMWH/unfractionated heparin (UFH) at 36 weeks’ gestation ¹ whereas AHA/ACC recommend switching to LMWH 1 week prior to delivery then UFH 36 hours prior to delivery which is then stopped 4 hours pre-delivery. ⁸

Low molecular weight heparins

The use of LMWHs in the management of mechanical valves in pregnancy remains an off-label indication. The optimal dosage of LMWH in the management of mechanical valves in pregnancy is undetermined therefore expert supervision is required in clinical practice involving its use. It is however accepted clinical practice with expert supervision.

To minimise the higher foetal loss and foetal anomaly rates associated with VKAs, regimens substituting LMWH for warfarin either for the entire duration of pregnancy or as an alternative to warfarin during the first trimester/period of embryogenesis as well as the peri-partum period have been examined. ¹⁹ LMWH does not cross the placenta and its use throughout pregnancy is associated with the highest number of livebirths (followed by sequential treatment with LMWH then VKA). Non-compliance and sub-therapeutic monitoring levels are associated with higher rates of TEC therefore close monitoring is required. ¹⁹ However, dose-adjusted LMWH is still associated with TEC in 4% to 17% of pregnancies and the optimal anti-Xa level, evaluation of peak or trough levels and frequency of monitoring are uncertain. ⁶ Despite monitoring of peak anti-Xa levels and dose adjustments, there remains a valve thrombosis risk of 4.4% to 8.7%. ¹ Peak anti-Xa levels within range were associated with sub-therapeutic trough levels <0.5U/ml therefore the Italian Federation of Centers for Diagnosis and Surveillance of the Antithrombotic Therapies (FCSA) recommend monitoring both peak and trough levels. ⁶

The recommendation for monitoring of anti-Xa is based on expert recommendations (from many societies) rather than on clinical evidence as, although there are studies that have incorporated peak and/or trough anti-Xa levels in patients with mechanical valves during pregnancy, no study has directly observed associations between anti-Xa level and clinical outcomes.

The half-life of subcutaneous LMWH is 3 to 6 hours, therefore it is recommended by the BSH ⁹ to use twice daily administration with the expectation of maintaining a more predictable anti-coagulant level over 24 hours than once daily administration. ⁶

The manufacturer of the LMWH enoxaparin recommends the measurement of peak and trough anti-Xa levels. ²⁰ In the UKOSS 83% of women using LMWH required a dose in excess of the dose recommended by the British National Formulary (BNF) at 10 weeks’ gestation, and at 20 weeks’ gestation 89% required a higher dose. ⁵ The BSH recommend that the transition from VKA to LMWH in pregnancy should consider the starting dose of LMWH higher than the standard therapeutic dose in recognition of the higher rates of TEC in the first trimester during transition between anti-coagulant strategies. ⁹ Some studies have suggested that even with LMWH titrated to achieve guideline-recommended peak anti-Xa levels this can still be associated with sub-therapeutic trough levels in over two-third of patients and is probably related to the enhanced renal clearance of LMWH. ²⁰ This questions whether monitoring of trough anti-Xa levels in addition to peak level is necessary to minimise both TEC and bleeding risks; however, the additional burden this will have on the mother needs to be considered as trough levels are to be taken 4 hours pre-dose (aiming anti-Xa levels >0.6 U/ml) and peak levels 4 to 6 hours post-dose. The BSH recommend peak anti-Xa to be taken 3 to 4 hours post-dose. ⁹ At present data regarding peak and trough levels is limited; however, the ESC does have trough level monitoring as a IIB recommendation ¹ and it is also suggested within the AHA/ACC guidelines to aid maintenance of therapeutic anti-coagulation. However, measuring peak and trough levels would be impractical and time consuming for the patient, and the BSH conclude that monitoring of peak levels is ‘a reasonable compromise’. ⁹ Furthermore, the UKOSS study found similar incidence of maternal complications in woman who had post-dose levels checked as those with pre and post levels, there was also variation in frequency of monitoring anti-Xa levels ranging from weekly to not at all, with the median number of measurements during pregnancy equating to 10. ⁵ Thrombotic complications were associated with a greater need for dose alterations following monitoring of anti-Xa levels.

Unfractionated heparin

UFH throughout the second and third trimester of pregnancy has been associated with unacceptable high maternal morbidity^8,18 as well as foetal complications including intraventricular haemorrhage. ¹⁹ It is also associated with heparin induced thrombocytopenia, osteopenia and line infection therefore should be reserved for situations where LMWH is not feasible.^8,20 UFH requires continuous infusion and close monitoring of the Activated Partial Thromboplastin Time (APTT) ratio.

Sequential treatment

Embryopathy occurs in 0.6% to 10% of cases when VKA is used in the first trimester. ¹ Although sequential treatment eliminates the risk of anti-coagulant-related embryopathy, there remains a 0% to 3% risk of foetopathy incorporating neurological and ocular abnormalities and intracranial haemorrhage.^1,8,19 The risk of foetopathy has also been observed with administration of UFH but not LMWH. ¹ When LMWH is used in the first trimester only, valve thrombosis occurs in 5.8% to 7.4%, similar to using LMWH throughout pregnancy. ¹

The rate of MVT appears higher during the first trimester during transition between anti-coagulation regimens. ⁹ Considering this, the BSH recommend that the starting dose of LMWH should be higher than the standard therapeutic dose (therefore total 2.5 mg/kg/day of enoxaparin divided into 2 doses). It was also observed that most pregnant woman required dose escalation between 10 and 20 weeks’ gestation.

Addition of aspirin

The AHA/ACC recommend that low dose aspirin is safe during pregnancy and may be added to the anti-coagulation regimen if required for other indications such as prevention of preeclampsia. ⁸ The BSH advise the addition of low dose aspirin (75–100 mg daily) to anti-coagulation with LMWH in women with MHVs in the absence of contraindication, especially in individuals with higher risk of MVT. ⁹ This may be stopped 3 days prior to scheduled delivery to decrease the risk of post-partum haemorrhage.

Current recommendations

Both the ESC and AHA/ACC recommend low-dose warfarin (<5 mg/day) in the first trimester (class IIa recommendation) in preference for LMWH/UFH (IIb).^1,8 In women with daily dose of warfarin > 5 mg/day then LMWH with monitoring of levels is preferred in the first trimester. When using LMWH in women with MHV and pregnancy, the ESC has recommended the use of dose-adjusted LMWH to achieve peak anti-Xa levels between 1.0 and 1.2 U/ml (mitral and right-sided valves) or 0.8 and 1.2 U/ml (aortic valves) and the AHA/ACC recommend levels between 0.7 and 1.2 U/ml, respectively.^1,8 Weekly monitoring of levels is recommended. ¹

The AHA/ACC and ESC recommend VKA as class I recommendation for the second and third trimester due to the lower risk of TEC.^1,8 In contrast, the FCSA recommend subcutaneous LMWH with dose adjustment according to weekly monitoring (monitoring to include peak and trough levels) ⁶ ; however, in patients with very high risk of TEC such as those with first-generation valves or other co-existing high-risk prothrombotic conditions then VKA may be considered. The FCSA suggest use of LMWH throughout pregnancy in preference to VKA due to the better safety profile for the foetus as the risk of foetal loss with VKA is present throughout pregnancy. They also suggest that LMWH when appropriately monitored, has a better efficacy profile for maternal thrombotic risk in comparison to VKA.

Pre-pregnancy counselling is essential to discuss the above options with the patient and should address the advantages and disadvantages of the various anti-coagulation regimes. The use of VKAs has better outcomes with regards maternal morbidity and mortality due to the decreased incidence of valve thrombosis, and it should be highlighted that risks to the mother inherently lead to risks with the baby. Despite these advantages of VKA, the increased incidence of embryopathy, foetopathy, foetal loss and foetal haemorrhage associated with the use of VKAs need to be acknowledged and the alternative of LMWH (which is associated with higher risk of valve thrombosis) may be considered and preferential for some mothers. ¹ An individualised anti-coagulation plan should be described and shared with the patient, primary care team, and the anti-coagulation team.

Mechanical valve thrombosis during pregnancy

MVT is rare but potentially life threatening in patients with a prothrombotic state such as pregnancy. ³ Valve thrombosis should be suspected if the clicks from the valve are no longer audible, ⁹ an embolic event has occurred, or clinical deterioration or symptoms of heart failure present or if there is pronounced increase in the gradient across the valve or new valvular regurgitation during pregnancy. ⁸ Management of valve thrombosis during pregnancy warrants emergent multi-disciplinary heart team assessment and discussion. ³

If a pregnant patient with an MHV presents with new severe chest pain +/− breathlessness or symptoms of a stroke, then emergency assessment should proceed to assess for possible causes including pulmonary embolism (as a pregnancy associated risk), ischaemic chest pain (potentially due to embolic event from MHV), MVT and aortic dissection (potentially increased risk if history of congenital aortic valve disease including bicuspid aortic valve). Assessment of valve thrombosis requires imaging of the valve leaflets and the function of the valve. A transthoracic echo allows assessment of valve gradient, presence of new regurgitation, and ventricular function. Imaging of valve leaflet motion is likely to require trans-oesophageal echo, fluoroscopy or gated computed tomography.

Management of MVT depends on how critically ill the patient is, in the non-critically ill patient it consists of optimisation of anti-coagulation with intravenous (IV) UFH, if this fails then surgery may be considered for obstructive MVT. ³ Urgent management of left-sided MVT is required, high-risk features include NYHA functional class II/III, mobile thrombus >0.3 cm diameter or area >1cm². Surgery is preferred when the patient is critically ill and is a surgical candidate. ³ However, as foetal loss is high with surgery (30%) fibrinolysis may be considered as an alternative in non-critically ill patients. ¹ Fibrinolysis should also be considered in critically ill patients when surgery is not immediately available. Alteplase has the highest molecular weight and does not cross the placenta. With fibrinolysis there is a 10% risk of embolisation and there is a concern for sub-placental bleeding. In right-sided prosthetic valve thrombosis fibrinolysis is the therapy of choice. Obstetric management will be based upon the foetal gestation and condition in conjunction with the maternal condition. ⁹

Most fibrinolytic items are unable to cross the placenta due to a molecular weight >1000 Da, however small amounts of streptokinase and fragments of urokinase may pass into the foetal circulation. ¹ Alteplase (a recombinant tissue plasminogen activator) has the highest molecular weight and does not cross the placenta. Intravenous alteplase is eliminated primarily by the liver with an initial half-life of <5 minutes and a terminal half-life of 72 minutes. ⁷ The duration of effect will depend on duration of infusion and protocol used.

Management of miscarriage or termination of pregnancy

The BSH guidelines include specific guidance about the management of both miscarriage and termination of pregnancy in women with MHV. ⁹ In the first trimester, miscarriage and termination should be actively surgically managed to minimise blood loss and unplanned surgical intervention. For later gestations, and the gestation will be dependent on the local expertise, medical management will usually be required.

During the management of miscarriage or termination, interruption of anti-coagulation should be minimised, but equally not resumed prematurely with excessive risk of bleeding, which would result in even longer interruption of anti-coagulation with the attendant risks to MHV function.

Planned anti-coagulation for delivery

Anti-coagulation strategies require advance planning and close monitoring. ⁷ For patients using VKA in their third trimester, there are various recommendations when this should be stopped and substituted by LMWH (or UFH) depending on the guidelines; AHA recommends stopping at least 1 week prior to planned birth and switching to either twice daily subcutaneous LMWH or IV UFH. In comparison, the ESC recommends switching VKA at 36 weeks. ²¹ The FCSA suggest switching VKA to LMWH at least 2 weeks prior to planned delivery. ⁶

If already on LMWH, it should be continued until 24 hours before planned delivery. As exact duration of labour is variable and difficult to estimate (in the case of vaginal delivery) and prolonged interruption of LMWH is a risk for valve thrombosis consideration of further doses (prophylactic/intermediate) as determined by the MDT or bridging with IV UFH may be considered with the aim of stopping UFH 4 to 6 hours prior to delivery. This should be planned with the coagulation expert, obstetric and anaesthetic teams.

For women planning a caesarean section who are at high risk of peri-partum thrombotic/haemorrhagic complications, LMWH should be discontinued no later than 24 hours prior to planned delivery and avoidance of neuroaxial analgesia in preference for general anaesthetic considered.

In the situation of spontaneous labour whilst on VKA treatment, caesarean section should be performed if the attending Obstetrician confirms that it is appropriate considering, the stage of delivery and the velocity of the labour. ¹

Anaesthetic considerations

Delivery is a particularly high-risk time for pregnant patients with MHV due to the need to balance the haemorrhagic risk in addition to the potential for MVT due to a prolonged period of reduced anti-coagulation. ⁸ A delivery plan should be created by the MDT prior to delivery that details anti-coagulation regimen, analgesic options, anaesthetic options for possible operative intervention, haemodynamic monitoring and appropriate uterotonic agents. It should also incorporate the recommended anti-coagulation regimen for the post-partum period. ⁹ Around the time of delivery, the therapeutic anti-coagulation regimen requires modification to decrease the risk of post-partum haemorrhage and foetal bleeding in addition to facilitating safe neuroaxial analgesia/anaesthesia. ²¹ All pregnant patients with MHV should have an anaesthetist review early in their pregnancy (by 28 weeks) to discuss analgesia and anaesthetic during labour which is also dependent on dose and timing of anti-coagulation discontinuation pre-delivery.

Due to the lower risk of haemorrhage, infection and thrombosis, a vaginal birth is the preferred mode of delivery barring obstetric indications for caesarean delivery. ²² This may be more complex in individuals with MHV due to anti-coagulation. In the UKOSS study 45% of women with MHV underwent vaginal delivery, 53% underwent a caesarean section for various indications including foetal compromise, maternal compromise, previous caesarean section, maternal request, failed induction and failure to progress in labour (2% unknown mode of delivery). ⁵

The BSH suggest there is insufficient evidence to recommend a specific mode of delivery in these individuals and instead an individualised plan for timing and mode of delivery with input from the MDT should be agreed and documented in advanced and this should be accessible if the patient were to attend at any maternity unit unplanned. ⁹ If a pregnant woman presents in labour whilst using a VKA (or VKA within the last 2 weeks) then a caesarean delivery using a general anaesthetic is recommended to avoid bleeding risks to the anti-coagulated foetus associated with a vaginal delivery including foetal intracranial haemorrhage. Reversal of anti-coagulation with prothrombin complex concentrate (PCC) may be considered but needs to be weighed up with the risks and will not alter the mode of delivery in this scenario. ²¹ Protamine can be used as a reversal agent for UFH (and may have a limited effect for LMWH) in the event of life-threatening haemorrhage. ⁹

Epidural anaesthesia is preferred in patients with mechanical heart valves when possible as it provides effective analgesia which results in greater cardiopulmonary and hemodynamic stability during labour. ²¹ Pain induced by labour provokes catecholamine release resulting in tachycardia, hypertension, hyperventilation, increased cardiac output and increased cardiac oxygen demand. These changes can be detrimental for patients who are at risk for cardiac ischaemia, arrhythmias or rupture of an aneurysm therefore epidural anaesthesia is commenced early. ²² In addition, epidural avoids the requirement of general anaesthesia if an emergency caesarean delivery becomes necessary. Either epidural-only or combined spinal-epidural (CSE) techniques can be performed however onset of neuraxial anaesthesia results in a decrease in venous tone and a decrease in systemic vascular resistance often with resultant hypotension, epidural would provide a more gradual onset of the hemodynamic changes than spinal. UFH should be stopped >4 hours and APTT and platelets within the normal range prior to proceeding with an epidural. If receiving LMWH, the previous dose should be  > 24 hours prior to the epidural. The use of epidural anaesthesia is preferred but can be challenging to time around the anti-coagulation for MHV, therefore early involvement (before labour) of an obstetric anaesthetist to assess and plan for care at time of delivery is recommended.

Prolonged interruption of anti-coagulation should be avoided, commencement of an intermediate dose of LMWH or IV UFH during induction of labour may be considered following MDT discussion with haematology and anaesthetic teams. ⁹ Figure 2 summarises the anaesthetic considerations for delivery planning in women with MHV. ²¹

Figure 2.

Delivery planning for anti-coagulation in mechanical heart valves (modified from Figure 1 of Bhatia et al. ²¹ ).

Post-partum care

The decision regarding timing to restart anti-coagulation should be individualised and involve the multi-disciplinary team, the risk of haemorrhage must be balanced against the risk of mechanical valve thrombosis. ²¹ The rate of PPH is high (see Table 2), therefore some units may delay recommencing UFH for 6 to 12 hours, LMWH for 24 hours post-delivery or introduce earlier at initial prophylactic dose, especially following caesarean delivery. ²¹ Therapeutic anti-coagulation with heparin in the post-operative period has been shown to be associated with a five-fold increase in bleeding without reducing the thrombosis risk. ⁹ The substitution of UFH/LMWH to warfarin may be 1 week after delivery. ²¹ BSH recommend commencing prophylactic/intermediate doses of LMWH for the first 24 to 48 hours post-partum, then consider restarting VKA at 5 to 7 days to balance the risk of thrombosis against major bleeding. ⁹ If using UFH, a gradual uptitration is recommended for the first few days. FCSA recommend a reduced dose of LMWH in the first 12 to 24 hours following delivery if no bleeding is detected, then gradual increase in LMWH with full therapeutic LMWH around 72 hours after delivery and consideration of bridging VKA from 72 hours after delivery. ⁶ As LMWH is better managed in clinical practice than UFH, they recommend UFH only in selected cases for women with higher risk of thrombotic complication when full dose anti-coagulation is required to be recommenced as soon as possible or for women at high risk of post-partum haemorrhage where a short acting drug would be preferable.

Table 2.

Observational cohort data on the risk of peripartum and postpartum haemorrhage in individuals with MHVs. ⁹

Reference		N (MHV)	Results	Comments
Vause et al. 5	UKOSS prospective	58	Primary PPH 2 (2%) Secondary PPH 6 (10%) Wound haematoma 6 (10%) Intra-abdominal bleed 4 (7%) Vaginal haematoma 1 (2%)	17 (29%) bleeding complication which needed return to theatre or transfusion (10 had delivered by CS and 7 delivered vaginally). Anti-coagulation around delivery likely variable and not reported
van Hagen et al. 4 (ROPAC)	International, registry, anti-coagulation around delivery likely variable and not reported	212	49 (23%) haemorrhagic complication 15.1% major haemorrhagic event, 10.4% PPH	Post-partum anti-coagulation regimen not specified
Khader et al. 23	Prospective, Egypt	40	12.5% PPH	UFH commenced 4–6 hours after delivery
Abildgaard et al. 24	Retrospective	12	2/12 after CS	LMWH commenced with 24hours
Quinn et al. 25	Retrospective UK	12	3 (25%) PPH Wound haematoma 1 (8.3%) needing re-exploration	LMWH commenced with 24 hours following delivery
Irani et al. 26	Single unit, retrospective USA	14 pregnancies	6 (42%) deliveries – primary PPH 3 (21%) secondary PPH 2 (14%) wound haematoma and 2 (14%) intra-abdominal bleed	Anti-coagulation restarted median 6 hours, predominantly UFH IV) 2 intra-abdominal bleeds and one wound haematoma in those who restarted <6 hours, no intra-abdominal bleeds in  < 6 hours group and 1 wound haematoma
Kariv et al. 27	Single centre prospective cohort, South Africa	29 pregnancies	5 (17%) returns to theatre (including 2 wound haematoma) 6 (21%) major haemorrhage 1 (3.4%) primary PPH 5 (17.2%) secondary PPH	UFH initiated at 6 hours if stable, warfarin on day after delivery
McLintock et al. 29	Retrospective audit, New Zealand	47 pregnancies	15 (32%) PPH 6 (12.8%) primary PPH (3 abruptions) 9 (19%) secondary PPH	Restart UFH at 6 hours post-VD, 6–12 CS, warfarin on day 1 (VD) or days 2–3 (CS)
Cousins et al. 28	Retrospective, single centre, France	18 cases	2 (11%) primary PPH, 5 (30%) secondary PPH needing re-operation	Median time to restart anti-coagulation 7 hours post-delivery
Dos Santos et al. 30	Two centre, UK & the Netherlands	44 pregnancies	17% primary PPH 4 (31%) secondary PPH 4 (31%) wound haematoma 2 (10%) intra-abdominal	All with PPH/haematoma received prophylactic dose LMWH 4–6 hours after delivery

Abbreviations: CS, caesarean section; LMWH, low molecular weight heparin; MHVs mechanical heart valves; PPH, post-partum haemorrhage; ROPAC: Registry of Pregnancy and Cardiac Disease; UFH IV, unfractionated heparin intravenous; UKOSS: UK Obstetric Surveillance System; VD, vaginal delivery.

To prevent further interruptions in therapeutic anti-coagulation timely recognition and management of wound and intra-abdominal haematoma is essential. ⁹

It is imperative to remember that women are at high risk of thrombosis in the 6 weeks following delivery, and this risk does not return to pre-pregnancy levels until approximately 12 weeks post-partum therefore compliance with anti-coagulation dosing and monitoring at this stage still needs to be highlighted. ⁷

Breastfeeding

Breastfeeding can safely proceed when using therapeutic doses of LMWH and VKAs. ²¹

Contraception

Contraception should be discussed and ideally facilitated prior to discharge from hospital. ⁹ Oestrogen containing contraceptive should be avoided in patients with MHV due to its increased prothrombotic nature. ⁷ Long-acting reversible contraception (LARC) that are progesterone only such as intrauterine coil or subcutaneous implant are extremely effective contraception methods and are safe and well tolerated in cardiac patients.

Conclusion

In young woman who may desire pregnancy in the future, the choice of an appropriate prosthetic valve is difficult, and no ideal option is currently available. ¹² Thorough counselling is of the utmost importance with discussion incorporating the risks and benefits of each option and the implications they may have upon future pregnancy. Pre-pregnancy counselling should involve a multi-disciplinary team with expertise in the management of pregnant women with cardiac disease to optimise outcomes.

Patient preferences regarding anti-coagulation based upon maternal and foetal outcomes in pregnancy should be considered when determining the optimal regimen in pregnant patients with MHV. ¹⁹ It is a difficult decision and depending on the woman's priorities she may choose an anti-coagulation regimen that minimises maternal risk, one that minimises foetal risk or one that tries to achieve a balance between both risks. ⁸ The woman's choice should not be assumed, instead detailed counselling of risks and benefits of all strategies should facilitate shared decision making. Regardless of which regimen is chosen, this is a dangerous time that demands frequent monitoring and for anti-coagulant control be as close to perfect as possible. ⁴

As more women with congenital heart disease, rheumatic heart disease, and prosthetic valves are reaching reproductive age due to advances in cardiovascular care, it is expected that there will be increased numbers considering and proceeding with complex pregnancies. ⁷

Summary

Cardiac disease is recognised in up to 4% of pregnancies. Some cardiac conditions may require valve replacement with a MHV. A recent survey in the United Kingdom described an incidence of MHV in pregnancy estimated at 3.7/100,000 maternities. MHV in pregnancy was associated with a 9% maternal mortality and 41% maternal morbidity. These high-risk pregnancies are fraught with the balance of maintaining maternal haemostatic equilibrium whilst trying to ensure foetal wellbeing.

In young woman who may desire pregnancy in the future, the choice of an appropriate prosthetic valve is difficult. No ideal option is currently available. Thorough counselling is important incorporating the risks and benefits of each option and the implications they may have upon future pregnancy. Pre-pregnancy counselling should involve a multi-disciplinary team with expertise in the management of pregnant women with cardiac disease to optimise outcomes.

Patient preferences regarding anti-coagulation based upon maternal and foetal outcomes in pregnancy should be considered. It is a difficult decision and depending on the woman's priorities she may choose an anti-coagulation regimen that minimises maternal risk, one that minimises foetal risk or one that tries to achieve a balance between both risks. Their choice should not be assumed, instead detailed counselling of all strategies should facilitate shared decision making.

As more women with MHV are reaching reproductive age due to advances in cardiovascular care, it is expected that there will be increased numbers considering and proceeding with complex pregnancies. This review considers the aspects to help support their care.