Pulmonary hypertension and pregnancy—a review of 12 pregnancies in nine women

Authors

  • RA Curry,

    1. Academic Department of Obstetrics and Gynaecology, Faculty of Medicine, Imperial College London, Chelsea and Westminster NHS Foundation Trust
    Search for more papers by this author
  • C Fletcher,

    1. Adult Congenital Heart Disease Unit, Royal Brompton and Harefield NHS Foundation Trust, and the National Heart and Lung Institute at Imperial College, London
    Search for more papers by this author
  • E Gelson,

    1. Academic Department of Obstetrics and Gynaecology, Faculty of Medicine, Imperial College London, Chelsea and Westminster NHS Foundation Trust
    Search for more papers by this author
  • MA Gatzoulis,

    1. Adult Congenital Heart Disease Unit, Royal Brompton and Harefield NHS Foundation Trust, and the National Heart and Lung Institute at Imperial College, London
    Search for more papers by this author
  • M Woolnough,

    1. Magill Department of Anaesthesia, Intensive Care and Pain Management, Chelsea & Westminster NHS Foundation Trust, London, UK
    Search for more papers by this author
  • N Richards,

    1. Magill Department of Anaesthesia, Intensive Care and Pain Management, Chelsea & Westminster NHS Foundation Trust, London, UK
    Search for more papers by this author
  • L Swan,

    1. Adult Congenital Heart Disease Unit, Royal Brompton and Harefield NHS Foundation Trust, and the National Heart and Lung Institute at Imperial College, London
    Search for more papers by this author
  • PJ Steer,

    1. Academic Department of Obstetrics and Gynaecology, Faculty of Medicine, Imperial College London, Chelsea and Westminster NHS Foundation Trust
    Search for more papers by this author
  • MR Johnson

    1. Academic Department of Obstetrics and Gynaecology, Faculty of Medicine, Imperial College London, Chelsea and Westminster NHS Foundation Trust
    Search for more papers by this author

Dr RA Curry, Academic Department of Obstetrics and Gynaecology, Faculty of Medicine, Imperial College London, Chelsea & Westminster NHS Foundation Trust, 369 Fulham Road, London SW10 9NH, UK. Email r.curry@imperial.ac.uk

Abstract

Please cite this paper as: Curry R, Fletcher C, Gelson E, Gatzoulis M, Woolnough M, Richards N, Swan L, Steer P, Johnson M. Pulmonary hypertension and pregnancy—a review of 12 pregnancies in nine women. BJOG 2012;119:752–761.

Objective  To report outcomes in a recent series of pregnancies in women with pulmonary hypertension (PH).

Design  Retrospective case note review.

Setting  Tertiary referral unit (Chelsea and Westminster and Royal Brompton Hospitals).

Sample  Twelve pregnancies in nine women with PH between 1995 and 2010.

Methods  Multidisciplinary review of case records.

Main outcome measures  Maternal and neonatal mortality and morbidity.

Results  There were two maternal deaths (1995 and 1998), one related to pre-eclampsia and one to arrhythmia. Maternal morbidity included postpartum haemorrhage (five cases), and one post-caesarean evacuation of a wound haematoma. There were no perinatal deaths, nine live births and three first-trimester miscarriages. Mean birthweight was 2197 g, mean gestational age was 34 weeks (range 26–39), and mean birthweight centile was 36 (range 5–60). Five babies required admission to the neonatal intensive care unit, but were all eventually discharged home. All women were delivered by caesarean section (seven elective and two emergency deliveries), under general anaesthetic except for one emergency and one elective caesarean performed under regional block.

Conclusions  Maternal and fetal outcomes for women with PH may be improving. However, the risk of maternal mortality remains significant, so that early and effective counselling about contraceptive options and pregnancy risks should continue to play a major role in the management of such women when they reach reproductive maturity.

Introduction

Pregnancies in women with pulmonary vascular disease are rare, with an estimated incidence of 1.1 per 100 000 maternities.1 The maternal mortality reported in the 1990s was 30–56%,2 probably because of the limited ability of women with pulmonary hypertension (PH) to adapt to the cardiovascular changes of pregnancy. The outcome for the fetus was also reported to be poor, with high rates of preterm delivery, fetal growth restriction and perinatal mortality.3 Published data on pregnancy with PH are limited to case reports or series; however, a recent review suggested that because of improved management, and in particular advanced therapies, maternal mortality may now be lower, in the region of 25%. Although current guidelines recommend discouraging pregnancy4,5 and the use of maximally effective contraception, with early termination should pregnancy occur, some women with PH still choose to become pregnant, or to continue with an unplanned pregnancy. Because data on pregnancy in these women are sparse, we report the maternal and fetal outcomes of 12 pregnancies in nine women with PH under our care.

Methods

Women with a diagnosis of PH (defined by a mean pulmonary artery pressure of >25 mmHg at rest) cared for by the Joint Cardiac and Obstetric Service of the Chelsea and Westminster and Royal Brompton Hospitals, London were identified from a database started in 1996. Data were collected from a detailed review of case notes supplemented by the Ciconia Maternity Information System (CMIS©). Information on patient characteristics, functional and echocardiographic assessment, cardiological, obstetric and anaesthetic management during pregnancy and labour, maternal and neonatal outcomes, and follow-up postnatal assessment was obtained.

Results

There were 12 pregnancies in nine women. Their baseline characteristics, together with maternal and neonatal outcomes, are recorded in Table 1. Their mean age at the initial clinic visit was 27 years (median 25 years, range 19–38). There were seven primigravidas. In four women (A, B, D and E) the diagnosis of PH was made using echocardiography during the index pregnancy. The remaining women were diagnosed before pregnancy at cardiac catheterisation, with the exception of patient I who was diagnosed by echocardiography, but not referred to our centre until she became pregnant. The PH aetiologies were varied—one each for idiopathic, HIV, left heart disease and primary lung disease and five congenital heart disease.

Table 1.   Baseline characteristics, maternal and neonatal outcomes
YearPatient (pregnancy no.)AgeParityGestation at diagnosisAetiologyWHO classPASP (mmHg)RV functionGestation at deliveryMode of deliveryEBL (ml)BW (g)BW centileMaternal complicationsMaternal outcomeNeonatal complicationsNeonatal outcome
  1. 6MWT, 6-minute walk test; ASD, atrial-septal defect; AVSD, atrioventricular septal defect; BW, birthweight; CHD, congenital heart disease; EBL, estimated blood loss; ERPC, evacuation of retained products of conception; FGR, fetal growth restriction; iPH, idiopathic pulmonary hypertension; LA, left atrium; LPA, left pulmonary artery; LSCS, lower segment caesarean section; MAPCAs, major aortopulmonary collaterals; misc., miscarriage; NICU, neonatal intensive care unit; PASP, pulmonary artery systolic pressure; PDA, patent ductus arteriosus; PPH, postpartum haemorrhage; PTD, pre-term delivery; RV, right ventricle; SaO2, oxygen saturations in air as measured by pulse oximetry; SVT, supraventricular tachycardia; ToF, tetralogy of Fallot; VSD, ventricular-septal defect; VT, ventricular tachycardia; WHO, World Health Organization.

1995A280Pre-pregnancyCHD (VSD, Eisenmenger)I104Moderately impaired26Emergency LSCS10006205Fulminating pre-eclampsiaDeath D15
PAH Crisis
PTD
SGA
NICU
Alive and well
1998B330Pre-pregnancyCHD (Secundum ASD closed with Dacron patch aged 19)I–II55–60Good36 + 4Emergency LSCS300257035Non-sustained VT—3rd trimester
Sustained SVT at LSCS
Intra-operative deathPTDAlive and well
2003C24024 weeksCHD (ToF variant with complex pulmonary atresia and MAPCAs)IIIAssumed to be systemicGood34Elective LSCS600208027PPHStill livingPTD
NICU
Alive and well
2007D(1)190Pre-pregnancyL heart disease (restrictive LA)II38GoodMisc. 6/40 ERPC    NilStill living  
2007E250 + 1Pre-pregnancyiPHII106Severely impaired32Elective LSCS400165037Worsening RV functionStill livingPTD
NICU
Alive and well
2007F38223 + 5 weeksThromboembolic/ primary lung disease (emphysema)III102Mildly impaired33 + 2Elective LSCS500209050Return to theatre to evacuate wound haematomaStill livingPTD
NICU
Alive and well
2008G(1)24136 weeksCHD (Repaired PDA, VSD; absent LPA and hypoplastic L lung)II65Good37Elective LSCS80023569PPHStill livingSGAAlive and well
2008G(2)252Pre-pregnancyCHD (Repaired PDA, VSD; absent LPA and hypoplastic L lung)   Misc. 12/40    NilStill living  
2008H28023 weeksCHD (AVSD with Eisenmenger physiology)I–II90Good34Elective LSCS1000218450PPHStill livingPTD
NICU
Alive and well
2009D(2)210 + 1Pre-pregnancyL heart disease (restrictive LAIII60GoodMisc. 7/40
ERPC
    Atrial tachycardia
Thyrotoxicosis
Still living  
2009I321Pre-pregnancyHIVI–II45Good39Elective LSCS800344060PPHStill livingNilAlive and well
2009G(3)262 + 1Pre-pregnancyCHD (Repaired PDA, VSD; absent LPA and hypoplastic L lung)II–III70Mild dysfunction37Elective LSCS1000278650PPH
Post-op. ileus
Still livingNilAlive and well

Maternal mortality and morbidity

There were two maternal deaths, both before 1999. Patient A (1995) died on postnatal day 15 following a PH crisis, having been delivered at 26 weeks gestation for fulminating pre-eclampsia. Patient B (1998) died at caesarean section having developed intractable supraventricular tachycardia following manual delivery of the placenta and administration of repeated bolus doses of oxytocic drugs (policies which have since been changed). All women were delivered by caesarean section (seven elective and two emergency deliveries), all under general anaesthetic except for one elective and one emergency caesarean performed under regional block. Estimated blood loss at delivery ranged from 300 to 1000 ml (mean 711). In the five most recent cases the third stage was managed with a low-dose oxytocin infusion rather than a bolus and in four women uterine compression sutures were also used.6 In one case the third stage was managed with compression sutures alone, and no oxytocics were given.

Symptomatic and functional assessment, echocardiographic measurements at baseline, during pregnancy and at follow up, are summarised in Table 2. Patients were monitored throughout their pregnancies by echocardiography and clinical examination. Estimations of pulmonary artery pressure were assessed from the tricuspid and pulmonary artery Doppler. Detailed assessments of right ventricular size and function were the key drivers of management. Medical treatment is summarised in Table 3. Patient A received intravenous unfractionated heparin following delivery only. Subsequently, seven women were anticoagulated with subcutaneous low-molecular-weight heparin (LMWH). Their level of anticoagulation was determined by their thrombotic risk; patients E, F and H (who had been taking warfarin before pregnancy) received therapeutic doses of LMWH and were monitored by measurement of anti-Factor Xa levels. The other women were given prophylactic doses of LMWH. Patient A received intravenous iloprost postpartum. Six of the nine women (all from 2007 onwards) were treated with sildenafil throughout pregnancy; one of them (E) was given add-on therapy with nebulised iloprost during the pregnancy because of a significant deterioration in her symptoms. Table 3 also reports the anaesthetic management at delivery. Six cases had invasive monitoring with peripheral arterial lines and central venous pressure lines. In five cases, cardiac output was monitored intraoperatively using transoesophageal echocardiography.

Table 2.   Functional and echocardiographic assessment
YearPatient (Pregnancy no.)WHO ClassRVSPRV Function
BaselineWorst during pregnancy6–12 month follow upBaselineWorst during pregnancy6–12 month follow upBaselineWorst during pregnancy6–12 months follow up
  1. RVSP, right ventricular systolic pressure; WHO, World Health Organization.

  2. *Patient did not attend follow-up echocardiogram.

1995AII–IIN/A104104N/AModerately impairedModerately impairedN/A
1998BI–IIIIN/A55–6065N/AGoodGoodN/A
2003CIIIIIIIIAssumed systemic (120)Systemic*GoodGood*
2007D(1)IIIIII3838*GoodGood*
2007EIIII–IIIIII106120110Severely impairedSeverely impairedSeverely impaired
2007FIIIII–IVIII102102101Mildly impairedMildly impairedMildly impaired
2007G(1)IIIIIIII656570GoodGoodGood
2008G(2)Spontaneous miscarriage—no data available
2008HI–IIII–IIIII909087GoodGoodGood
2009D(2)IIIIIIIII607090GoodGoodGood
2009II–III–III–II454525GoodGoodGood
2009G(3)II–IIIIII–IVIII708051Mild dysfunctionMild dysfunctionGood
Table 3.   Medical, obstetric and anaesthetic management
YearPatient (Pregnancy No.)Medical therapyThird stageGA/RACO MonitoringPre-inductionLinesInductionMaintenanceVasoconstrictorsIV fluidsIntraoperative Complications
  1. CO, cardiac output; CVC, central venous catheter; ETT, endotracheal tube; GA, general anaesthetic; ICU, intensive care unit; IPPV, intermittent positive pressure ventilation; IV, intravenous; LMWH, low molecular weight heparin; PAC, pulmonary artery catheter; PCA, patient-controlled analgesia; RA, regional anaesthetic; RBCs, red blood cells; SVT, supraventricular tachycardia; TOE, trans-oesophageal echocardiography; UFH, unfractionated heparin.

1995AIV UFH (PN)
IV iloprost (PN)
Oxytocin bolus × 2GANoneRanitidine
Maxolon
NaCitrate
Diazepam
CVC
Arterial line
Fentanyl
Etomidate
Vecuronium
ETT
IPPV with O2/N2O/halothane
Diazepam
Phenylephrine bolusesCrystalloid 1000 ml
Colloid 1750 ml
RBCs 4 units
Sinus tachycardia
Treated with
Esmolol
1998BDiltiazemOxytocin bolus × 2RA (labour epidural topped up)None CVC
Arterial line
14 ml 0.5% bupivicaine (3-ml aliquots over 30 minutes)
100 μg fentanyl
N/APhenylephrine bolusesCrystalloid 1000 mlSVT 5 minutes post 2nd dose syntocinon during removal of placenta, hypotension and asystole
Intubated and CPR Initiated.
Resuscitation for >1 hour
2003CSC LMWH GAAnaesthetic notes not available
2007D(1)SC LMWH
Sildenafil
Miscarriage
2007ESC LMWH
Sildenafil
Nebulised iloprost
Furosemide
Compression sutureGATOE Wide bore peripheral access
Arterial line
Fentanyl
Etomidate
Suxamethonium
Propofol infusion
Atrcurium
Fentanyl
Morphine
Paracetamol
Ilio-inguinal blocks
  Bradycardia after induction
2007FSC LMWH
Sildenafil
Oxytocin infusion
Compression suture
GATOE
Trans-oesophageal doppler
 Wide bore peripheral access
Arterial line
PAC introducer
Fentanyl
Etomidate
Suxamethonium
ETT
IPPV with O2/isoflurane
Vecuronium
Morphine
Paracetamol
Ilioinguinal blocks
 Crystalloid 1000 mlBradycardia after induction
2007G(1)SC LMWH
Sildenafil
Oxytocin infusion
Compression suture
GATOERanitidine
NaCitrate
Large bore peripheral
Arterial line
CVC
PAC introducer
Etomidate
Fentanyl
Suxamethonium
Atropine
ETT
IPPV with O2/isoflurane
Morphine
Paracetamol
Ilioinguinal blocks
 Minimal crystalloid
Colloid 500 ml
RBCs 2 units
Persistant
Tachycardia
2008G(2)SC LMWH
Sildenafil
MISCARRIAGE         
2008HSC LMWH
Sildenafil
Oxytocin infusion
Compression suture
GATOERanitidine
NaCitrate
Large bore peripheral
Arterial line
CVC
PAC
introducer
Etomidate
Fentanyl
Suxamethonium
Atropine
ETT
IPPV with O2/isoflurane
Morphine
Paracetamol
Vecuronium
Ilioinguinal and rectus sheath blocks
 Crystalloid 500 ml
Colloid 500 ml
None
2009D(2)SC LMWH
Sildenafil
MISCARRIAGE         
2009ISC LMWH
Sildenafil
Antiretrovirals
Oxytocin infusion
Compression suture
RA (Spinal)NoneRanitidine16G peripheralSpinal of 2.5 ml 0.5%
heavy bupivicaine,
300 mcg diamorphine
N/APhenylephrine
300 mcg
Crystalloid 1000 ml
Colloid 500 ml
None
2009G(3)SC LMWH
Sildenafil
Furosemide
Oxytocin infusionGATOEAnaesthetic notes not available

Perinatal mortality and morbidity

Three pregnancies resulted in miscarriage (two in patient D, one attributed to the development of thyrotoxicosis) and nine pregnancies resulted in a surviving child. Five women were delivered iatrogenically preterm because of maternal decompensation, their babies requiring admission to the neonatal intensive care unit (NICU). Two of the babies were small for gestational age (5th and 9th centiles). None of the babies in this series had any congenital abnormalities.

Discussion

Maternal mortality

There were two maternal deaths in 12 pregnancies, indicating the high maternal mortality associated with PH, but both deaths occurred more than 10 years ago. Patient A died on day 15 postoperatively as a result of a PH crisis, emphasising the fatal combination of PH with pre-eclampsia.7–9 She required emergency delivery at 26 weeks for fulminating pre-eclampsia associated with a coagulopathy. Patient B developed sustained supraventricular tachycardia shortly after manual delivery of the placenta, resulting in haemodynamic instability and unrecoverable cardiac arrest. As a result of this death, we modified our management significantly, with the avoidance of bolus doses of syntocinon, and the use of controlled cord traction for delivery of the placenta instead of manual removal. The use of sildenafil since 2007 appears to have improved maternal tolerance of the plasma volume expansion of pregnancy. The highest risk of death is likely to be in the first 2 weeks following delivery, but it is unlikely that the risk returns to a non-pregnant level for several months. Both deaths in our series occurred during the early postnatal period, supporting earlier observations that this is the time of highest risk.2 Previous series have reported refractory right heart failure as the commonest cause of death,3 but we did not observe this in our series.

A recent systematic review found no significant differences in mortality in the different subgroups of PH.3 However, both deaths in our series occurred in women with PH secondary to congenital heart disease. Patients with Eisenmenger syndrome have significantly increased risks of bleeding, thrombotic complications and arrhythmia, with multi-organ involvement being the norm.10

Fetal and neonatal outcomes

Maternal PH is associated with an increased risk of fetal and neonatal complications, including stillbirth and neonatal death (up to 13%), fetal growth restriction (up to 33%) and preterm delivery (in some series as high as 100%).2,3,11 The risk is highest in women with poor functional class pre-pregnancy (New York Heart Association functional class III or IV), cyanosis or left heart obstruction12 and may be a result of direct effects of the underlying cardiac disease compounded by complications of iatrogenic preterm delivery. Two of the babies in our series were small for gestational age. There were no stillbirths or neonatal deaths but five out of nine women were delivered preterm, all for maternal reasons. Of these five, only one (patient A) was severely preterm (26 weeks), being delivered because of fulminating pre-eclampsia. This relatively high rate of preterm delivery reflects reported trends.3 It seems likely that improvements in the neonatal care and survival of preterm infants over the past decade have led to an increased willingness to undertake iatrogenic preterm delivery for maternal interest.

Predictors of outcome

Our study failed to identify any particular predictors of maternal outcome. This may be because of the relatively small number of women and their ability to adapt to their PH (particularly when associated with congenital heart disease), which gives a misleading New York Heart Association class status. Right ventricular function remained relatively stable, with only one woman (E) having objective evidence of right ventricular decompensation before delivery. However, this issue was not specifically addressed in our retrospective study, and legitimate concerns regarding incomplete right ventricular recovery after successful pregnancy remain and need to be addressed when discussions take place with women during counselling.13

Two of the women (B and D) experienced sustained arrhythmia and of these one died. Data from non-pregnant women with PH suggest that sustained atrial arrhythmias occur in 11.7% of women.14 The additional volume load and haemodynamic stress imposed by pregnancy on the already pressure-overloaded right heart may increase the tendency for both atrial and ventricular arrhythmia. This has previously been reported in other congenital lesions, where right ventricular dilatation has been shown to be associated with a higher incidence of arrhythmia.15,16 Although our patient group is clearly too small to make any statement about the relation of arrhythmias to prognosis, maternal arrhythmia has been associated with increased cardiac events during pregnancy17 and adverse fetal outcomes.18

Management

Anticoagulation

Seven of the nine women in our series were anticoagulated with subcutaneous LMWH; (the degree of anticoagulation was based on their estimated thrombotic risk). Patients who had been taking warfarin before pregnancy were fully anticoagulated with subcutaneous LMWH, monitored by measuring anti-factor Xa levels (target 0.6–1). They were restarted on warfarin as soon as it was established that there were no ongoing bleeding problems following surgery. Those women on prophylactic LMWH during pregnancy continued this after delivery for 6–12 weeks. There were no thrombotic complications during pregnancy or the puerperium. Anticoagulation during the puerperium is strongly advocated, regardless of PH aetiology, because of the substantial increase in risk (11-fold) of venous thromboembolism at this time.19 There remains, however, uncertainty regarding the most effective and safest regimen for thromboprophylaxis, particularly for women with Eisenmenger physiology in whom a balance must be struck between their bleeding diathesis and thrombotic risk.

Advanced therapies

In this series five out of nine women (all cases since 2007) were treated with sildenafil during pregnancy, covering a wide range of diagnostic groups: one pulmonary hypertension associated with HIV, one idiopathic pulmonary arterial hypertension, one pulmonary hypertension associated with lung disease and two pulmonary hypertension associated with congenital heart disease. Two were treated with prostanoids. Both of the women who died did so before treatment with sildenafil was introduced. Development of advanced therapies such as sildenafil has revolutionised the management of PH, with improvements in symptoms, functional capacity and survival being demonstrated.20 However, there is no clear consensus on the use of advanced therapies for PH in pregnancy, with the evidence limited to case series and a few small reviews; however it is tempting to attribute the recent reduction in maternal mortality to the increased availability of these therapies.

Mode of delivery

All of our cases were delivered by caesarean section, which is in keeping with the recommendations in recent reports.2,3 However, the ideal mode of delivery continues to be contentious.8,21 Although vaginal delivery is associated with less risk of haemorrhage, infection and venous thromboembolism, emergency caesarean section may need to be performed for a variety of reasons in women with PH. When that happens it is associated with an increased rate of complications compared with either elective caesarean or vaginal delivery.22 As many women will be delivered early, when the cervix is still unfavourable, induction of labour is likely to be difficult. Furthermore, elective caesarean section offers the advantage of timing the delivery to ensure optimal conditions for such high-risk women (for example, attendance of all relevant senior staff), and also allows the use of transoesophageal echocardiography.

Management of the third stage and blood loss

Mean blood loss at delivery in our series was 711 ml (range 300–1000 ml), with a blood loss of greater than 500 ml in six out of nine cases, and 1000 ml in three of those. In normal pregnancy, uterotonics have been shown to significantly reduce the risk of postpartum haemorrhage by up to 50%.23 However, they have adverse haemodynamic effects which make them unsuitable for women with PH. Ergometrine causes vasoconstriction and hypertension, whereas oxytocin causes an increase in pulmonary vascular resistance, a reduction in systemic vascular resistance and tachycardia.24 Additionally two recent randomised clinical trials have reported electrocardiographic changes suggestive of myocardial ischaemia associated with oxytocin administration in healthy women undergoing elective caesarean section with regional anaesthesia.25,26 Current policy in our unit is to use a low-dose oxytocin infusion at a rate of 8–12 mU/minutes over 4 hours, with a low threshold for inserting uterine compression sutures. Both women who died had received bolus intravenous doses of oxytocin. We speculate that in patient B this, in combination with a sudden autotransfusion consequent on delivery of the placenta, may have contributed to her demise. However, in patient A it was of no obvious consequence as death occurred 15 days after delivery. General anaesthesia, specifically the use of volatile agents such as halothane, isoflurane and nitrous oxide, is known to increase uterine atony. In this series, in the six cases with an estimated blood loss of greater than or equal to 500 ml at delivery in association with a general anaesthetic, five (71%) women had anaesthesia maintained with either halothane or isoflurane.

Anaesthetic considerations

All forms of regional anaesthesia have been used successfully for the delivery of women with PH.19,27–29 Due consideration is needed for anticoagulation, the woman’s ability to tolerate lying flat and potentially profound sympathetic block resulting in decreased venous return. The avoidance of vasodilatation and reduced venous return is best achieved by using a slow incremental infusion of local anaesthetic. However, the majority of the women received general anaesthesia for their caesarean sections, because it facilitated the use of transoesophageal echocardiography (TOE) to monitor cardiac output, and diminished the work of breathing, while allowing the anaesthetist to manipulate metabolic parameters.30–32 Further, direct pulmonary vasodilators, such as nebulised prostacyclins and nitric oxide, can be given more easily and in the event of major complications the woman can be put onto cardiac bypass quickly.

In most cases in this series, induction of general anaesthesia was achieved with high-dose fentanyl, etomidate and suxamethonium with cricoid pressure. In the literature, use of suxamethonium and a large dose of opiate to blunt the intubation response has been reported;33 etomidate has been used by others30 as has propofol,31,32 but the actual effect of propofol on the pulmonary circulation is controversial because it may increase or decrease pulmonary vascular resistance.33 Maintenance of anaesthesia was generally with high fractional inspired oxygen concentration and a volatile agent, usually isoflurane, which has been reported to suppress the pulmonary vasoconstrictor response to hypotension.33 Vecuronium and atracurium have been used for muscle relaxation, the former being more cardiostable and the latter being better metabolised by the pregnant woman. Nitrous oxide was used in the first case (its suitability has been recommended by others30,31), but as it has since been shown to be cardiodepressant as well as potentially increasing pulmonary vascular resistance in women with PH,34 we have subsequently avoided its use.

Ongoing care

Most women with pulmonary hypertension have ongoing limitations of function such that further pregnancies are strongly contraindicated. Therefore maximally effective contraception is necessary. Patient G was sterilised at delivery by salpingectomy (Pomeroy sterilisation carries a relatively high-risk of the fallopian tubes re-canalising). Patients who did not wish to be sterilised were recommended to have maximally effective contraception, and we have used both Implanon (now replaced by Nexplanon) and Mirena (intrauterine contraceptive system with progesterone) as these have extremely low failure rates, comparable with sterilisation.

In keeping with the relatively poor prognosis for women with PH, patients A and B died during pregnancy. Accordingly, there is no echocardiographic follow-up in Table 2 for these women.

Conclusions

Although PH is rare, ours is one of the largest consecutive series to be reported, so it adds significantly to the reported experience of managing this challenging condition. Nonetheless, small numbers limit the conclusions that can be drawn. Prospective, multicentre registry data are required to shed further light on both outcome and prognosis for this high-risk condition.

Maternal and fetal outcomes for women with pulmonary hypertension seem to be improving,3,35 but maternal risk remains extremely high, so timely and effective counselling on contraception and the hazards of pregnancy remain vital. PH pregnancies should be managed in a tertiary setting with a multidisciplinary team.

Disclosure of Interests

During the last 5 years Philip Steer has received funding and honoraria to support speaking at meetings from: Ferring Pharmaceuticals; Alliance Medical; Pfizer; Milupa; the Obstetric Anaesthetist’s Association; The University of Copenhagen; the North Midlands Urogynaecological Society; Postgraduate Education and Academic Affairs Department, National Guard Health Affairs, Riyadh, Saudi Arabia; the University of Copenhagen; and BabyLifeline. He has received funding to defray the expenses of attendance at meetings from FIGO (speakers honorarium), Bristol Hospitals Healthcare Trust, the Icelandic Medical Association, Capetown University, Rome University, the 3rd Military Hospital Chonqing, AsiaPacific Congress of Fetal/Maternal Medicine, the British Maternal and Fetal Medicine Society, the British Association of Perinatal Medicine, the European Congress of Perinatal Medicine, the Swedish Society of Obstetricians and Gynaecologists, the National Perinatal Society of Jordan, the National Perinatal Society of Turkey, Zurich University, Leuven University, the RCOG, the University of Nottingham and the International Association for Adult Congenital Heart Disease. He has received reimbursement for travel expenses and honoraria for examining from the Universities of Hong Kong, Lund and Stockholm. He has been treated to dinner while attending a meeting with representatives of Hewlett Packard Ltd. to discuss fetal monitoring. He has received fees for medicolegal opinions from the Welsh Health Legal Authority, Northwick Park Hospitals NHS Trust, Her Majesty’s Police, and various solicitors. He receives publication royalties from Elsevier, and ALCS. He chairs the medical advisory panel of ‘Group B Strep Support’ (unpaid). His research is funded by public bodies in the UK (including the HTA). The other authors did not report any potential conflicts of interest.

Contribution to authorship

RAC, CF, EG, MW and NR were responsible for the acquisition of the data. RAC and CF wrote the paper which was subsequently revised and edited by RAC, MAG, LS, PJS and MRJ.

Details of ethics approval

This study received approval from the Brompton, Harefield and NHLI Research Ethics Committee, REC reference 09/H0706/14.

Funding

No external funding was obtained for this review.

Acknowledgements

We would like to thank Dr Steve Yentis (Consultant Anaesthetist, Chelsea & Westminster NHS Foundation Trust, Dr David Alexander (Consultant Anaesthetist, Royal Brompton Hospital), Dr John Wort (Consultant in Pulmonary Hypertension, Royal Brompton and Harefield NHS Foundation Trust), Mr Martin Lupton, Miss Zoe Penn and Miss Gubby Ayida (Consultant Obstetricians Chelsea and Westminster Hospital) for their valued input into the care of the women in this report.

Ancillary