Pregnancy in women with PAH is associated with high maternal morbidity and mortality. In one of the oldest series of 44 well-documented cases with Eisenmenger’s syndrome and 70 pregnancies, Gleicher et al. found a maternal mortality of 52 % in 1979 [7]. Twenty years later perinatal mortality decreased to 36 % [8], but has since declined little as reported by Bédard et al. in 2009 who found that the maternal mortality was still 28 % [9]. Remarkably, in this study none of the patients died during the pregnancy, but all of them within the first months after the delivery [8, 9]. Causes of death were severe right ventricular failure, cardiac arrest, pulmonary embolism, endocarditis, and uncontrollable bleeding (Fig. 17.1).

Pregnancy induces systemic vasodilation and increases cardiac output. In patients with a persistent defect, the pulmonary-to-systemic shunt worsens and more hypoxia occurs. This leads to further vasoconstriction and higher pulmonary vascular resistance. Hemodynamic stress during labor and delivery induces more CO₂ retention and acidosis, which, in turn, acutely increases the pulmonary artery pressures and precipitates refractory heart failure [9, 10]. The effect of pregnancy on the cardiovascular system may persist for several months after delivery [11], and it fits with the hypothesis that most severe complications occur shortly of delivery. However, the prepregnancy severity of PH might be associated with outcome. A more recent study from Katsuragi et al. suggested that women with mild pulmonary arterial hypertension, who were more often NYHA class I or II in early pregnancy, had a less marked increase in pulmonary artery pressures during pregnancy. Only a few people with mild PAH deteriorated from NYHA class II to class III/IV. Women with severe pulmonary hypertension delivered earlier than women with mild pulmonary hypertension and no patients died in the mild form of pulmonary arterial hypertension [12].

The neonatal outcome in PAH-congenital heart disease pregnancies is also compromised. Bedard et al. found a neonatal or fetal death of 7 % and a rate of fetal growth restriction of 24 %. Premature delivery, defined as delivery before 37 weeks of gestation, occurred in 86 % [9]. Earlier series report an even higher perinatal mortality of 28.3 %, which was significantly associated with prematurity [7].

The best management of a PAH patient with congenital heart disease is to discourage pregnancy. Remarkably, in 27 % of the PAH cases in pregnant women with congenital heart disease, the diagnosis of PAH was first made during the pregnancy [9]. There are currently no guidelines or recommendations relating to the management of PAH in pregnant women. In the review from Bédard et al., the current status of management is discussed [9]. In general, patients are hospitalized earlier (median 32 weeks of gestation, range 17–40), probably related to complications occurring during pregnancy, the awareness of premature delivery, and the choice for elective delivery by Caesarean section. Caesarean section is electively planned in 31 % of the pregnant general heart disease population, but seems not to have any advantage over planned vaginal delivery [13]. However, one might suggest that Caesarean section is preferred in more complex heart diseases [9]. Indeed, 72 % of cases were delivered by CS in the report of Bédard et al. [9]. However, the hemodynamic impact of a vaginal delivery is held to be lower than that of a Caesarean section. Less blood loss causes smaller volume shifts, less clotting and bleeding disorders; moreover, a vaginal delivery is associated with a lower risk of infection [14]. During delivery, the Valsalva manoeuvre increases both heart rate and vascular resistance, both of which might increase the workload on the right heart [15].

The anesthetic management is closely related to the mode of the delivery. Bédard et al. found that in 31 % of cases, the Caesarean section was performed under general anesthesia in PAH patients with congenital heart disease. This was associated with a higher maternal mortality [9]. Although it is possible that more severe cases had a general anesthetic accounting for the greater maternal mortality, an alternative explanation would be that general anesthesia is not the best option since it might disturb the balance between the pulmonary and systemic circulations by differentially influencing the vascular resistance [16]. Pulmonary vascular resistance is increased by hypoxia, hypercarbia, high hematocrit levels, positive pressure ventilation, cold, metabolic acidosis, and alpha-adrenergic stimulation. The systemic vascular resistance decreases by the use of vasodilators, spinal and epidural anesthesia, deep general anesthesia, and hyperthermia. In case of a persistent shunt defect, the pulmonary-to-systemic shunt might increase and lead to more pronounced systemic desaturation. In the presence of PAH without a shunt defect, the increase of pulmonary vascular resistance will aggravate the pressure load on the subpulmonary ventricle and might induce acute heart failure. Moreover, all anesthetic agents depress myocardial contractility [17]. The choice between epidural and general anesthesia has to be left to an experienced anesthetist with sufficient expertise in the management of complex congenital heart disease to find the best compromise in this highly challenging situation [14].

Intensive monitoring is recommended for patients with PAH during labor, delivery, and in the postpartum period. An arterial line to measure blood pressure and oxygen saturation and a central venous line to control right atrial pressures seem to be useful in the peripartum period [18]. Acid–base status should be checked frequently, since undetected acidosis can exacerbate pulmonary hypertension [18]. Bédard et al. noted the use of radial artery and/or central venous pressure catheter in 38 % and the use of a pulmonary artery catheter in 31 % of the cases [9]. However, invasive pulmonary arterial pressure monitoring remains controversial and does not seem to influence outcome in a positive or negative way [9]. Nevertheless, more invasive monitoring is related to a higher risk for potential complications [19]. In-hospital monitoring is suggested for at least 2 weeks after delivery [18]. Indeed, the effect of pregnancy on the cardiovascular system may persist for several months after delivery [11].

The occurrence of thrombosis and/or pulmonary embolism raises the question about the usefulness of thrombosis prophylaxis in PAH patients. Although oral anticoagulation is the standard of care for patients with idiopathic PAH [1], thromboprophylaxis in idiopathic PAH was used in 52 % and 41 % during pregnancy and after delivery, respectively [9]. These numbers were even lower in PAH patients with congenital heart disease, 24 % received thromboprophylaxis during pregnancy and 31 % after delivery [9]. This is probably explained by the controversy that exists about the use of oral anticoagulants in PAH-congenital heart disease patients [20]. This is especially so in patients with the Eisenmenger’s syndrome, where the balance between thrombosis and bleeding is very delicate and explaining why severe bleeding is in the list of peripartum complications [21]. However, anti-thromboembolic stockings or compression pumps can be used to help prevent peripheral venous thrombosis with no adverse effects [18]. The ESC guidelines on pregnancy suggest that in PAH associated with congenital cardiac shunts and in the absence of significant hemoptysis, anticoagulant treatment should be considered in the presence of pulmonary artery thrombosis or signs of heart failure [22].

At least 50 % of pregnant women with PAH and congenital heart disease patients receive advanced therapy for PAH [9]. Prostacyclin analogues, phosphodiesterase inhibitors, and endothelin-receptor antagonists are proven to have a beneficial effect on exercise performance and functional capacity [23–25], whereas inhaled NO decreases pulmonary vascular resistance [26]. However, the number of case reports and the series, in which advanced therapy was used, is too small to draw conclusions on safety and efficiency for pregnant women and the fetus [27–32]. However, there tended to be a beneficial effect of advanced therapy on outcome, but there may be a publication bias toward positive results. The guanylate cyclase stimulator, riociguat [33], might cause fetal harm and was shown to have teratogenic effects. Calcium channel blockers are mainly used in idiopathic pulmonary arterial hypertension, but are relatively contraindicated in congenital heart disease patients with PAH and a persistent shunt, as they might increase pulmonary-to-systemic shunt by a decrease of systemic vascular resistance [20]. Although the use of oxygen is controversial in adult patients with the Eisenmenger syndrome, increased oxygen might lower pulmonary vascular resistance during labor and delivery [34].

If an unplanned pregnancy occurs in a PAH-congenital heart disease patient, termination should be discussed and performed in an experienced tertiary center [22]. However, even when the pregnancy is stopped at this stage, maternal mortality rates of up to 7 % are reported [7]. Dilatation and curettage in the first trimester is probably the procedure of choice, preferably with general anesthesia [18]. Women who opt to continue the pregnancy need to be referred to a tertiary center urgently for follow-up and delivery by a multidisciplinary team with experienced congenital cardiologists, experts in pulmonary hypertension, obstetricians, anesthetists, neonatologists, and experts in intensive care. Moreover, the psychological stress of such a high-risk pregnancy should not be underestimated and support by a psychological team beneficial. Frequent visits throughout pregnancy are recommended, and hospitalization during the third trimester may facilitate management and safer delivery [9].

During pregnancy the plasma volume increases and peaks at about 50 % above baseline early in the third trimester. The red blood cell mass increases by 20–30 % too [35]. The extra volume might compromise the function of the right ventricle and precipitate right heart failure since the higher pulmonary vascular resistance increases right ventricular pressure and work. In addition, systemic vascular resistance decreases in normal pregnancy, and in patients with a persistent shunt defect, this will result in a more pronounced pulmonary-to-systemic flow, which will exacerbate the preexisting hypoxia and result in greater pulmonary vasoconstriction. Once set in train, this sequence of events will lead to a self-perpetuating deterioration in the patient’s clinical state.

The physiological changes in labor may also challenge the circulation of those women with PAH and persisting shunt. Midwall et al. demonstrated that uterine contractions are associated with a substantial decrease in the ratio of pulmonary-to-systemic blood flow [34]. Normally, uterine contractions cause an increase in cardiac output and right ventricular pressures [36]; however, if this occurs when the pulmonary vascular resistance does not decrease or remains fixed, more pulmonary-to-systemic flow will occur. Also, when traction forceps were used during a uterine contraction, the pulmonary flow decreased further [34]. Moreover, in late pregnancy and during labor, the gravid uterus might compromise cardiac output by compression of the inferior vena cava in supine position. Some recommend delivery in the lateral position to avoid compression of the inferior vena cava and maintain sufficient systemic venous return [18]. Furthermore, at the time of labor and delivery, acidosis and hypercarbia may further increase pulmonary vascular resistance. Any hypovolemia resulting from blood loss or hypotension from a vasovagal response to pain may result in insufficient cardiac output and sudden death [18]. On the other hand, a temporary increase in venous return may occur immediately following delivery due to relief of inferior vena cava pressure which at times results in a substantial rise in ventricular filling pressures, which might unbalance the shunt flow [15]. Therefore, postpartum patients should continue to be monitored hemodynamically for 24–48 h to detect any deterioration due to the postpartum increase in venous return to the heart [15].