SEARCH

SEARCH BY CITATION

Keywords:

  • Iloprost;
  • pregnancy;
  • pulmonary hypertension;
  • sildenafil;
  • survival;
  • therapy

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Disclosure of interests
  10. Contribution to authorship
  11. Details of ethics approval
  12. Funding
  13. Acknowledgements
  14. References

Please cite this paper as: Kiely D, Condliffe R, Webster V, Mills G, Wrench I, Gandhi S, Selby K, Armstrong I, Martin L, Howarth E, Bu’Lock F, Stewart P, Elliot C. Improved survival in pregnancy and pulmonary hypertension using a multiprofessional approach. BJOG 2010;117:565–574.

Objective  Pregnancy in women with pulmonary hypertension (PH) is reported to carry a maternal mortality rate of 30–56%. We report our experience of the management of pregnancies using a strategy of early introduction of targeted pulmonary vascular therapy and early planned delivery under regional anaesthesia.

Design  Retrospective observational study.

Setting  Specialist quaternary referral pulmonary vascular unit.

Population  Nine women with PH who chose to proceed with ten pregnancies.

Methods  A retrospective review of the management of all women who chose to continue with their pregnancy in our unit during 2002–2009.

Main outcome measures  Maternal and fetal survival.

Results  All women commenced nebulised targeted therapy at 8–34 weeks of gestation. Four women required additional treatment or conversion to intravenous prostanoid therapy. All women were delivered between 26 and 37 weeks of gestation. Delivery was by planned caesarean section in nine cases. All women received regional anaesthesia and were monitored during the peripartum period in a critical care setting. There was no maternal mortality during pregnancy and all infants were free from congenital abnormalities. One woman died 4 weeks after delivery following patient-initiated discontinuation of therapy. All remaining women and infants were alive after a median of 3.2 years (range, 0.8–6.5 years) of follow-up.

Conclusion  Although the risk of mortality in pregnant women with PH remains significant, we describe improved outcomes in fully counselled women who chose to continue with pregnancy and were managed with a tailored multiprofessional approach involving early introduction of targeted therapy, early planned delivery and regional anaesthetic techniques.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Disclosure of interests
  10. Contribution to authorship
  11. Details of ethics approval
  12. Funding
  13. Acknowledgements
  14. References

Pulmonary hypertension (PH) is a rare but important condition which can affect women of childbearing age. Recent classification based on shared pathophysiology describes five main classes of PH: pulmonary arterial hypertension (PAH), which can be idiopathic or associated with other conditions such as systemic sclerosis or congenital heart disease; PH in association with left heart disease; PH in association with lung disease; chronic thromboembolic pulmonary hypertension (CTEPH); and a miscellaneous group.1 In PAH, a vasculopathy resulting from the proliferation of the intima and media of the pulmonary arterial bed leads to an increase in pulmonary vascular resistance.2 Chronic thromboembolic disease resulting in obstruction to pulmonary blood flow (in addition to an associated distal vasculopathy) can lead to the development of CTEPH.3 In these conditions, the right ventricle is poorly adapted to an increased after-load, resulting in right heart failure, premature death and an inability to increase cardiac output appropriately or cope with large fluid shifts in situations such as pregnancy. Historical studies have suggested poor survival in women with PH who become pregnant, with mortality reported to be 30–56%.4–6 As a consequence, recommendations emphasise the importance of counselling women about the high risks of pregnancy in the setting of PH with clear contraceptive advice and, in the eventuality of pregnancy, consideration of termination.7

Unfortunately, there are no large prospective studies examining the different approaches to the management of PH in pregnancy. Although systematic reviews provide important information, there are inherent limitations that make it difficult to draw conclusions regarding cause and effect.5–8 Three classes of therapy with antiproliferative and vasodilator effects have entered clinical practice over recent years, resulting in improvements in both functional status and survival, and potentially curative surgery is increasingly available for women with CTEPH.9–11 Endothelin receptor antagonists (bosentan, sitaxsentan, ambrisentan) and phosphodiesterase-5 inhibitors (sildenafil, tadalafil) are administered orally, whereas prostacyclin (epoprostenol) and prostacyclin analogues (iloprost, treprostinil) are given via nebuliser or continuous subcutaneous or intravenous infusion. Endothelin receptor antagonists may be teratogenic.12 Several case reports have been published of successful pregnancies in women treated with targeted therapies.13–20 The majority of these have only reported successful outcomes, and reporting bias cannot be excluded. There is therefore no consensus on the use and choice of targeted therapy, nor on the use of anticoagulation or mode and timing of delivery in women who wish to continue with pregnancy.

In this article, we describe our experience of the management of ten consecutive pregnancies using a systematic multiprofessional approach based on close monitoring (where possible in the outpatient setting) with the early introduction and intensification of targeted pulmonary vascular therapies. This approach also included a planned delivery by caesarean section with the use of regional techniques, and peripartum monitoring in a critical care environment with highly experienced clinicians.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Disclosure of interests
  10. Contribution to authorship
  11. Details of ethics approval
  12. Funding
  13. Acknowledgements
  14. References

Data collection

Hospital records for all pregnant women with PH who had been managed between April 2002 and April 2009 (the period during which targeted therapies were used in pregnancy) at our unit were reviewed. Data on patient characteristics, management and subsequent clinical course were retrieved for all women who had chosen to continue with their pregnancy. The mortality and functional status of women on 1 May 2009 were also retrieved.

Setting

All women were managed from the pulmonary vascular perspective at the Sheffield Pulmonary Vascular Disease Unit, Royal Hallamshire Hospital, Sheffield, UK. This is a quaternary referral centre for the assessment of suspected pulmonary vascular disease with a referral population in excess of 15 million people and over 500 patients receiving targeted pulmonary vascular therapy. Patients with suspected PH undergo systematic evaluation, including echocardiography, lung function and serological testing, computed tomography (CT), magnetic resonance imaging (MRI) and right heart catheterisation. Depending on the clinical context, certain investigations may be deferred or not performed. Patients are carefully characterised in accordance with international classifications, and PAH is defined as a mean pulmonary arterial pressure of ≥25 mmHg in association with a normal pulmonary capillary wedge pressure (≤15 mmHg).21 Obstetric and anaesthetic care is provided by the Department of Obstetrics and Anaesthesia at our institution. This provides a large tertiary referral service for high-risk obstetric care with more than 7000 deliveries per year. For two women (patients 1 and 6), delivery was at the Glenfield Cardiothoracic Centre, Leicester, UK, with input from a cardiologist with a specialist interest in adult congenital heart disease (FAB) and an obstetrician with an interest in high-risk pregnancies (ESH).

Our regional ethics committee classed this review as service evaluation, and therefore formal ethical approval was not required. The management of three of the women has been reported previously.20

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Disclosure of interests
  10. Contribution to authorship
  11. Details of ethics approval
  12. Funding
  13. Acknowledgements
  14. References

During the study period, nine women (with ten pregnancies) who chose to continue with their pregnancy were treated with advanced therapy. The baseline characteristics are presented in Table 1. Four of these women were diagnosed with PH de novo during their pregnancy and presented between 24 and 33 weeks of gestation with excessive breathlessness. The five women who had been diagnosed previously with PH presented earlier, at between 6 and 24 weeks. Right heart catheterisation was performed during pregnancy in five women and following pregnancy in two women. Pulmonary haemodynamics are summarised in Table 2.

Table 1.   Baseline patient characteristics
CharacteristicPatient
123456788*9
  1. Amph, amphetamine; CHD, congenital heart disease-associated PAH; CTEPH, chronic thromboembolic pulmonary hypertension; FPAH, familial pulmonary arterial hypertension; IPAH, idiopathic pulmonary arterial hypertension; ISWT, incremental shuttle walking test; MCTD, mixed connective tissue disease; Mod, moderate; PAH, pulmonary arterial hypertension; PH, pulmonary hypertension; RV, right ventricular; sPAPECHO, echocardiographically estimated systolic pulmonary artery pressure at presentation in pregnancy; WHO class, World Health Organisation functional class.

  2. *Second pregnancy.

  3. **Right to left shunt of uncertain cause.

  4. ***Previous ventricular septal defect closure.

  5. ****Ostium secundum atrial septal defect.

Age (years)29232422293238212324
Height (m)1.671.601.561.651.611.531.551.631.631.6
Weight (kg)82564480645560505047
Parity1 + 01 + 02 + 11 + 12 + 01 + 07 + 20 + 01 + 00 + 0
Form of PHPAHPAHPAHCTEPHPAHPAHPAHPAHPAHPAH
Cause of PHFPAHIPAHCHD****ProximalMCTDCHD***AmphIPAH**IPAH**CHD***
Presentation (weeks’ gestation)106172431162430933
PH newly diagnosed?NoNoNoNoYesNoYesYesNoYes
1st presentation to our unit?NoNoYesNoYesYesYesYesNoYes
WHO classIIIIIIIIIIIIIIIVIVIIII/III
ISWT (m)45033014040013037000260200
sPAPECHO (mmHg)455080100526011511915064
RV dilatationn/aNoneSevereModMildn/aSevereModModMod
RV systolic impairmentn/aNoneModSevereNonen/aModModModMod
Table 2.   Right heart catheter details
 Patient
123456789
  1. CI, cardiac index; CO, cardiac output; mPAP, mean pulmonary artery pressure; mRAP, mean right atrial pressure; n Ilo, nebulised iloprost; PCWP, pulmonary capillary wedge pressure; PVR, pulmonary vascular resistance; S, sildenafil; TPR, total pulmonary resistance.

  2. Patients with idiopathic pulmonary arterial hypertension underwent vasodilator challenge with NO, but none of these studies was performed during pregnancy.

  3. *Performed at Leicester Royal Infirmary, Leicester, UK.

  4. **Patient 2, although a vasodilator responder to NO, did not benefit from high-dose calcium antagonist therapy prior to her pregnancy and was receiving bosentan when she presented pregnant.

  5. ***PCWP not obtained.

Performed during pregnancy?Yes*YesNoNoYesYes*NoNoYes
Gestation (weeks)1825  3311  32
TreatmentNiln Ilo  NilNil  Nil
mRAP (mmHg)718  912  11
mPAP (mmHg)4068  3835  36
PCWP (mmHg)14***  6***  15
CO (l/min/m2)n/a4.3  4.94.1  4.9
CI (l/min/m2)n/a2.7  2.92.7  3.2
PVR (dyn/s/cm5)n/an/a  519n/a  342
TPR (dyn/s/cm5)n/a1265  616680  588
Performed post-pregnancy?YesNoNoYesYesNoYesNoNo
Months postdelivery71  2035 9 7
Treatment at time of RHCNil  n IloS S n Ilo
mRAP (mmHg)5  156 12 9
mPAP (mmHg)40  5027 46 29
PCWP (mmHg)9  ***8 13 12
CO (l/min/m2)8.1  3.25.5 6.1 4.4
CI (l/min/m2)4.26  1.963.10 3.84 3.02
PVR (dyn/s/cm5)305  n/a275 435 309
TPR (dyn/s/cm5)393  1237390 607 527
Vasodilator response to NONoYes**n/an/an/an/aNon/an/a

A further five pregnancies in women with known PH were managed during this period; one woman had two miscarriages, and three women made informed decisions to terminate their pregnancy. There were no maternal mortalities among these women.

Management and outcome

The management of the nine women who received targeted therapy during ten pregnancies is summarised in Table 3. All women were treated with four to seven doses of nebulised iloprost per day. Patients 1–3 received 10–20 μg per dose via an ultrasonic nebuliser (Schill, Stuttgart, Germany), and patients 4–9 received 5 μg via an iNeb nebuliser (Philips, Eindhoven, the Netherlands) following its introduction to the market. Two women (patients 2 and 8) required transition to intravenous iloprost because of clinical deterioration. Patient 2 experienced rapid deterioration culminating in cardiorespiratory arrest, requiring a brief period of cardiopulmonary resuscitation and the institution of intravenous therapy. An elective caesarean section was performed 5 days later. Patient 8 suffered two syncopal episodes on treatment with nebulised iloprost and sildenafil, and was transitioned to intravenous therapy in addition to sildenafil, with a caesarean section being performed 48 hours later. In addition, one woman (patient 7) was commenced on intravenous iloprost therapy at the onset of spontaneous labour, and one woman who was clinically stable but had severe disease (patient 8) was commenced on intravenous iloprost the day prior to delivery. Low-molecular-weight heparin (LMWH) was used in nine of ten pregnancies (five at full therapeutic dose and four at a prophylactic dose, including one woman with a suprarenal inferior vena cava filter inserted during pregnancy).

Table 3.   Management during pregnancy and current functional status
 Patient
123456788*9
  1. B, bosentan; C, central line; CS, caesarean section; E, epidural; F, full dose; ISWT, incremental shuttle walk test; IVC, inferior vena cava filter; L, LiDCO; LMWH, low-molecular-weight heparin; P, prophylactic dose; neb I, nebulised iloprost; Ref, refused; RIP, deceased; Rx, treatment; S, sildenafil; S/E, combined spinal and epidural; SG, Swan–Ganz catheter; SVD, spontaneous vaginal delivery; u/k; unknown.

  2. *Second pregnancy despite counselling.

  3. **At onset of spontaneous labour.

  4. ***48 hours predelivery.

  5. ****24 hours predelivery.

  6. *****Patient followed up locally.

  7. ******Bosentan stopped at presentation when pregnancy confirmed.

  8. Patients treated with sildenafil received 25 mg three times daily. New York Heart Association class and ISWT given are at the last visit prior to the censor date, where the patient received an assessment of functional status and exercise capacity. Current treatment is the therapy the patient was receiving at the time of the last clinical assessment when data on exercise testing and assessment of functional status were taken, prior to the censor date. For patient 8, the current treatment, functional status and ISWT recorded for the first pregnancy apply to the values recorded at the last visit prior to presentation at the second pregnancy.

Baseline targeted RxNilB******NilNilNilNilNilNilB******/SNil
Treatment commenced Nebulised iloprost (weeks’ gestation)1981831342826301434
Intravenous iloprost (weeks’ gestation)/dose (μg/hour) 25/8    34/2**32/4***34/2**** 
Sildenafil (weeks’ gestation)      33310 
LMWHFFNFFFP/IVCPPP
Delivery (weeks)36263534353434323437
ModeCSCSCSCSCSCSSVDCSCSCS
AnaesthesiaES/EEES/EEES/ES/ES/E
MonitoringSGSGC/Lu/kC/LC/LRefC/LC/LC/L
Current RxSBneb Ineb ISNilSB/SRIPneb I
NHYA classIIIIIIIIIIIIIIIIII II
ISWT (m)45044013042090*****470350 340
Follow-up (years to 1 May 2009)6.55.94.83.23.33.11.8  0.8

Delivery was by elective caesarean section in nine cases. One women, booked for an elective caesarean section at 34 weeks, declined to attend for surgery and subsequently delivered by spontaneous vaginal delivery following admission with worsening symptoms later that week. No woman underwent general anaesthetic; epidural or combined spinal and epidural anaesthesias were employed. Radial arterial and internal jugular central lines were used in the majority of cases. Cardiac output monitoring was performed using Swan–Ganz catheterisation in two women early in the series and minimally invasive lithium dilution measurement of cardiac output (LiDCO, Cambridge, UK) in women later in the series.

All ten pregnancies resulted in live births with no congenital abnormalities. Birth weights and Apgar scores are shown in Table 4. There was no maternal mortality in the antenatal or early postpartum period. Unfortunately, patient 8 died 4 weeks following delivery having decided to stop her therapy at home and having declined admission despite worsening of her condition. The remaining eight women were still alive on the censor date of 1 May 2009 with a median follow-up of 3.2 years (range, 0.8–6.5 years). Five women were transitioned onto oral therapies after the peripartum period. Patient 4 subsequently underwent successful pulmonary endarterectomy. Exercise capacity, functional status and treatment at the last visit prior to the censor date are summarised in Table 3.

Table 4.   Neonatal characteristics
 Patient
123456788*9
  1. n/a, not available.

  2. *Second pregnancy.

  3. **Positive pressure ventilation via bag and mask.

Birth weight (kg)2.80.652.162.082.53n/a2.031.571.582.44
Birth weight (centiles)50–752–99–2525–5050–75n/a259–252–99–25
Apgar 1 minute98989n/a9798
Apgar 5 minutes991099n/a9**98

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Disclosure of interests
  10. Contribution to authorship
  11. Details of ethics approval
  12. Funding
  13. Acknowledgements
  14. References

Over a 7-year period, we have managed ten consecutive pregnancies with a novel regimen of the early introduction of targeted therapy, frequent monitoring in the outpatient setting, escalation of targeted therapy where indicated and elective delivery by caesarean section under regional anaesthesia at around 34 weeks of gestation. Despite promising outcomes, it must be emphasised that pregnancy is still associated with a significant risk of maternal mortality and, indeed, in this series two women deteriorated rapidly, requiring escalation of therapy and semi-urgent delivery, and one woman died 4 weeks following delivery, albeit following patient-initiated discontinuation of therapy, self-discharge from hospital and declined re-admission to hospital.

Early reports of the outcomes of pregnancy in PH suggested that the survival of women with PH in pregnancy was poor. A single French centre reported a mortality rate of 36% in 15 consecutive pregnancies treated with optimal conventional therapy between 1995 and 2005.6 A systematic review of relevant data published between 1997 and 2007 has recently suggested that outcomes may have improved, with a mortality of 25%.8 As a consequence of these data, the recommendation of termination to all women with significant PH is advocated.7 This approach, however, is not acceptable to some women, and a significant proportion will present with PH for the first time at an advanced stage in pregnancy.

Our management technique, entailing a multiprofessional approach involving close co-operation between pulmonary vascular physicians and nurses, obstetricians, anaesthetists and haematologists is summarised in Figure 1 and discussed below. This approach is also compared with the relevant literature, some of which has been published during the duration of our series.

image

Figure 1.  Current approach to management of pregnancy in pulmonary hypertension (PH).

Download figure to PowerPoint

Preconception counselling

All women in our service, of childbearing age, receive counselling regarding the high risk of pregnancy, together with access to contraceptive advice in accordance with national guidelines.22 Women also have access to specialist family planning clinicians and, in the eventuality of pregnancy, consideration of the early interruption of pregnancy.

Management and investigation of women during pregnancy

Invasive measurement of pulmonary haemodynamics during pregnancy is performed where indicated. Women previously diagnosed with PH were not subjected to repeat right heart catheterisation during pregnancy unless there was diagnostic doubt with regard to the severity of disease (patients 1, 6 and 9), or the woman was critically unstable (patient 2) and measurements were used to help guide therapy and the timing of delivery. In view of the small but finite risk of the procedure, right heart catheterisation is deferred in women with clear evidence of severe disease on clinical evaluation, echocardiography and cross-sectional imaging. Where right heart catheterisation is performed during pregnancy, facilities for emergency caesarean section in the eventuality of complications are available.

Selection and timing of targeted therapy

The need for targeted therapy in women early in our cohort was based on the assessment of progress during previous pregnancies (patients 1 and 3, where the women had been unwell during previous pregnancies) and the need for alternative treatment in a woman who conceived whilst taking bosentan (patient 2). Subsequently, all women were offered targeted therapy when a clear diagnosis of significant precapillary PH had been established, following counselling regarding the unlicensed nature of these products during pregnancy, with the purpose of optimising pulmonary vascular status prior to delivery. It was felt advantageous to commence therapy whilst the patient was stable to allow the effects of pharmacological intervention to be assessed. Patients in World Health Organisation (WHO) functional class III, who were not already on treatment, were offered immediate targeted therapy. Patients in WHO functional class II, who were treatment naïve, were commenced on targeted therapy after the first trimester. Our patients were all offered nebulised iloprost as initial therapy. Treatment changes, with escalation to more potent therapy with parenteral prostanoid and, more recently, the addition of oral sildenafil, were made if there was clinical deterioration or inadequate clinical response. The choice of nebulised iloprost as initial therapy was based on the limited choice of available therapies at the time of the initial patients in our series, namely nebulised or systemic prostanoid. The nebulised route was chosen as we were keen to minimise complications (such as line infections and systemic hypotension) in otherwise stable patients. Nebulised iloprost has been shown to be an effective therapy in patients with PAH.23 Although intravenous iloprost is not licensed for the treatment of patients with PH, it is often used in the UK as its stability is superior to that of epoprostenol. Its long-term use in Germany has been reported recently.24 It is possible that the early introduction of oral sildenafil as initial therapy in our cohort may have been associated with similar outcomes.

Several other groups have published case reports or very small case series of successful pregnancies in patients with PAH who have received calcium channel blockers,13,14 intravenous prostacyclin,13,15,16 nebulised iloprost18,19 and sildenafil.17 The endothelin-1 receptor antagonist, bosentan, has been shown to have teratogenic potential in animal studies.25 Two women conceived whilst receiving bosentan, despite receiving appropriate family planning advice. Both women were transitioned to inhaled iloprost with one of these women also receiving sildenafil at the time of conception. In both of these women, the infants were free from congenital abnormality. The targeted therapies in current use are listed in Table 5.

Table 5.   Pulmonary hypertension therapies used in the UK
ClassDrug nameAdvantagesDisadvantages
  1. CCB, calcium channel blocker; ERA, endothelin-1 receptor antagonist; IV, intravenous; neb, nebulised; PAH, pulmonary arterial hypertension; PDE5-I, phosphodiesterase-5 inhibitor; SC, subcutaneous.

  2. *No randomised controlled trial data in nonpregnant women, although data from retrospective studies suggest benefit. For all other therapies, efficacy has been demonstrated in randomised controlled trials.

CCBDiltiazem, etc.*Oral, costOnly beneficial in ≈10% of idiopathic PAH patients (predicted by positive acute vasodilatory challenge at right heart catheter)
PDE5-ISildenafilOral 
ERABosentan, sitaxsentan, ambrisentanOralERAs have teratogenic potential – contraindicated in pregnancy
ProstanoidEpoprostenol (IV), iloprost (IV)* Iloprost (neb) Treprostinil (SC)Intravenous route potentially most efficacious Nebulised route avoids need for indwelling access Subcutaneous route avoids need for indwelling accessRequires continuous infusion via Hickman line with risk of sepsis. Rebound pulmonary hypertension if abruptly discontinued. Body image issues Requires frequent nebulisation (≈7/day) Subcutaneous infusion can cause significant site pain

Monitoring during pregnancy and the peripartum period

Unless otherwise indicated, women were seen approximately every 4 weeks until 28 weeks of gestation, then every 2 weeks until 32 weeks, and then weekly until planned delivery. Outpatient assessment included standard history and examination, routine biochemistry and haematology, electrocardiogram, field exercise testing and echocardiography where indicated. Fetal assessment during pregnancy was performed by ultrasound biometry at 28 and 32 weeks. If fetal growth restriction was suspected, further assessment was performed using the amniotic fluid index and umbilical artery Doppler to help guide subsequent management. Women noting any deterioration were admitted for further evaluation and, if required, remained as inpatients until delivery. Otherwise, women were admitted 24 hours prior to elective delivery to a critical care environment. All women underwent continuous heart rate and oxygen saturation monitoring. Unless declined, all women had central venous access and arterial lines inserted. Initially, Swan–Ganz catheterisation was used at the time of delivery. As a result of concerns regarding potential complications,26 subsequent patients were managed using less invasive methods of cardiac output measurement. Women were managed with appropriate fluid resuscitation following delivery with the aim to maintain a systemic blood pressure and central venous pressure around immediate predelivery levels for the first 24 hours. A rising central venous pressure and/or fluid overload was managed with diuretic therapy. Women were managed for 1 week in a critical care environment. Women were at risk of deterioration in the days following delivery as it can take several weeks for cardiopulmonary parameters to return to baseline level and significant fluid shifts can occur.27,28 Following discharge, women were seen at frequent intervals until 6 months.

Mode and timing of delivery and choice of anaesthesia

The optimal mode and timing of delivery in women with PH remain uncertain and involve an assessment of the maternal risk of continuing pregnancy and the neonatal risk of preterm delivery. Interestingly, a recent paper has identified a trend towards earlier delivery in women with PH.8 The absolute mortality risk of mild preterm birth is low, although it is acknowledged that the relative risk of neonatal and infant mortality is increased slightly compared with term birth, and a proportion of children may have subsequent developmental delay.29,30 Near-term infants have a higher rate of neonatal medical problems, and so the infant is carefully monitored in an appropriate neonatal setting.31 Blood volume increases in pregnancy by almost 50% by 32–36 weeks’ of gestation32, and cardiac output increases during the second trimester, reaching a peak at around 28 weeks’ of gestation.33 The development of increasing breathlessness and falling exercise capacity with or without advanced therapy during this period is a grave sign and reflects an inability to cope with the cardiovascular demands of pregnancy. In this setting, the outlook is very poor without immediate delivery.6 Our practice is therefore to plan delivery at around 34 weeks’ of gestation in women with significant PH who are clinically stable, with a plan for early delivery prior to this time if there is any evidence of symptomatic decline.

In a patient group which has a limited ability to acutely increase cardiac output, an approach which minimises further demands on the cardiovascular system is important. Normal vaginal delivery is associated with a 34% increase in cardiac output at full cervical dilation.34 Pushing during the second stage of labour may also have deleterious haemodynamic effects, especially in the presence of impaired venous return. Caesarean section is therefore our preferred method of delivery. It can be performed during daytime hours in the presence of experienced obstetricians, anaesthetists and pulmonary vascular physicians following clear premeditated delivery plans. The minimisation of blood loss at caesarean section is especially important in this group of women. Bimanual compression and suture compression of the uterus are helpful nonpharmacological methods, although syntocinon was safely used in the majority of our patients. Syntocinon may cause a reflex tachycardia and hypotension, and so must be carefully titrated.

General anaesthesia in this group of patients has been described.18 Adverse effects of intubation on pulmonary artery pressure and of positive pressure ventilation on venous return are well documented.18,35 Therefore, we used either epidural anaesthesia or combined spinal and epidural anaesthesia. The latter approach achieves a superior degree of anaesthesia than an epidural alone, but without the potentially deleterious hypotensive effect of a standard-dose spinal anaesthetic.6 This technique also provides good postoperative analgesia and is sufficiently flexible to allow return to theatre in the event of postoperative bleeding without the need for general anaesthesia. The option to proceed to general anaesthesia, however, needs to be considered as it may be necessary in some cases.

Use of anticoagulation

The risks and benefits of anticoagulation are discussed with patients on an individual basis. The approach in patients with idiopathic PAH and other forms of associated PAH was initially full-dose anticoagulation with LMWH. During later pregnancies, prophylactic LMWH was used to allow regional anaesthetic procedures to be performed safely. Patients were given a prophylactic dose of LMWH the evening prior to delivery and the evening following delivery if there were no bleeding complications. In patients with PAH associated with congenital heart disease, caution is exercised if there is any history of haemoptysis.

Management plans

All women delivered in our unit have clear individualised written management plans covering a variety of eventualities, including premature onset of labour, emergency caesarean section for mother/baby and planned elective caesarean section. These plans cover the mode of delivery, anaesthetic agents, regional anaesthetic techniques and syntocinon use, together with protocols for the treatment of hypotension and bradycardia and escalation of pulmonary vascular therapy.

Comparison with previous studies

As discussed earlier, historical data from case series and systematic reviews suggested a maternal mortality rate of between 30% and 56%,4–6 but a more recent systematic review found an improved mortality of 25%.8 Bedard et al.,8 the authors of this later review, observed that there were fewer vaginal deliveries compared with data from the earlier era. They also observed that mortality was higher in women undergoing general rather than regional anaesthesia.8 Our policy of caesarean section under regional anaesthesia may therefore partly explain our outcomes. In the case series by Bonnin et al.,6 where maternal mortality was 36%, only one woman received targeted pulmonary vascular therapy during pregnancy, although several women received prostacyclin or inhaled nitric oxide during the peripartum period. In their systematic review, Bedard et al.8 observed that targeted therapies tend to be used late in pregnancy, and suggested earlier use before cardiac decompensation occurred, a suggestion in keeping with our approach.

Limitations

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Disclosure of interests
  10. Contribution to authorship
  11. Details of ethics approval
  12. Funding
  13. Acknowledgements
  14. References

The present study involves only ten pregnancies managed in a single centre. Previous systematic reviews have been performed which have involved larger numbers of patients.5,8 Such studies are, however, open to significant reporting bias, with negative outcomes being less likely to be reported in the literature. This study reports an approach to the management of women with PH who are pregnant, and cannot attribute an improvement in survival compared with historical studies to any single intervention, be it the timing and mode of delivery or the use of targeted pulmonary vascular therapy.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Disclosure of interests
  10. Contribution to authorship
  11. Details of ethics approval
  12. Funding
  13. Acknowledgements
  14. References

In this article, we have reported improved outcomes, compared with other single centre series and systematic reviews, in ten consecutive pregnancies in women with PH. This has involved a multiprofessional approach with the early institution of targeted therapy (with escalation from nebulised to intravenous prostanoid or additional targeted therapy where appropriate), a planned caesarean section under regional anaesthesia at around 34 weeks and close postpartum observation. Although pregnancy in PH continues to pose a significant risk of maternal mortality, and termination should continue to be offered, our data suggest that, in fully counselled women who choose to continue with their pregnancy, such use of a multiprofessional approach may result in satisfactory outcomes. Issues such as the optimal time for delivery and the initial targeted therapy of choice require further investigation; given the rarity of pregnancies within individual pulmonary vascular centres, data from multicentre studies are needed.

Disclosure of interests

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Disclosure of interests
  10. Contribution to authorship
  11. Details of ethics approval
  12. Funding
  13. Acknowledgements
  14. References

DGK has acted as an expert witness at maternal death inquests in the UK. He has received fees for speaking at educational meetings by Actelion, Bayer Schering and Pfizer (the manufacturers of bosentan, iloprost and sildenafil), and has received consultancy fees from Actelion and Pfizer. Funding to international meetings has been provided by Actelion and Bayer Schering. The department has received an educational grant from Actelion Pharmaceuticals and funding for MRI support from Bayer Schering. RC has received fees for speaking at educational meetings by Actelion. Funding to attend international meetings has been provided by Actelion and Pfizer Pharmaceuticals. LM has received assistance to attend symposia from Actelion Pharmaceuticals. IJA has been reimbursed by Actelion and Pfizer for attending several international conferences. Actelion has paid honorarium for a number of educational meetings. CAE has received honoraria for speaking at meetings from Actelion Pharmaceuticals. He has also received funding to attend international conferences from Actelion. No other authors have financial disclosures.

Contribution to authorship

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Disclosure of interests
  10. Contribution to authorship
  11. Details of ethics approval
  12. Funding
  13. Acknowledgements
  14. References

DGK, RC and CAE wrote the initial draft. All other authors reviewed and contributed to the manuscript. All authors were involved in the clinical care of the subjects.

Details of ethics approval

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Disclosure of interests
  10. Contribution to authorship
  11. Details of ethics approval
  12. Funding
  13. Acknowledgements
  14. References

Our regional ethics committee classed this review as service evaluation, and therefore formal ethical approval was not required.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Disclosure of interests
  10. Contribution to authorship
  11. Details of ethics approval
  12. Funding
  13. Acknowledgements
  14. References

DGK, RC and CAE are investigators in the NIHR Cardiovascular Biomedical Research Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Sheffield, UK.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Disclosure of interests
  10. Contribution to authorship
  11. Details of ethics approval
  12. Funding
  13. Acknowledgements
  14. References