Disease in Pregnancy


Fig. 12.1

Physiological changes in the cardiovascular system during pregnancy. From: Ashrafi R, Curtis SL. Heart Disease and Pregnancy. Cardiol Ther. 2017 Dec;6(2):157–173. Open Access, Springer Healthcare



Hemodynamic fluctuations are a part of normal labor and delivery. Pain, elevated heart rate, and catecholamine surge augment the cardiac output by 25–30% during the first stage of labor and up to 80% immediately postpartum [23, 24]. Maternal position, supine versus left lateral, also affects the venous return and thus the cardiac output. Volume shifts occur as uterine contractions autotransfuse 300–500 cm3 of blood from the uterine sinusoids into the systemic circulation [25]. Postpartum hemodynamics are affected by analgesic drugs, bleeding, and infection and peak within 24–72 h after delivery. Marked hemodynamic change s during pregnancy, labor, and the postpartum period increase the burden to the maternal circulation, particularly in women with preexisting cardiomyopathy, VHD, or CHD. The time course of clinical deterioration in susceptible women parallels the rise in hemodynamic burden by the late second trimester, during, or just after delivery.


Maternal Risk Assessment


The risk of pregnancy in women with CVD depends on the clinical status of the patient and the specific cardiac defect. Lesion-specific risk assessment , discussed below, may be limited by validation based on small or retrospective studies. Three prospective maternal risk scores have been developed in larger populations with diverse CVD and include general and lesion-specific risk when available. Risk assessment will guide pregnancy and labor management plans.


The modified World Health Organization (WHO) risk classification [26], Cardiac Disease in Pregnancy (CARPREG) [27], and the ZAHARA (Zwangerschap bij vrouwen met een Aangeboren HARtAfwijking-II, translated as pregnancy in women with congenital heart disease II) [28] are maternal risk scoring systems. The modified WHO risk classification is based on expert consensus and incorporates known maternal risk factors and gives contraindications to pregnancy; it was endorsed by the ESC 2011 guidelines [29] on the management of CVD during pregnancy.


The CARPREG risk score is the most widely used and validated maternal risk score. In an observational cohort of 252 pregnant women with acquired and congenital heart disease (CHD), 4 predictors of an adverse maternal cardiac event were identified:



  • Left heart obstruction (mitral valve area <2 cm2, aortic valve area [AVA] <1.5 cm2, or left ventricular outflow tract [LVOT] obstruction >30 mmHg)



  • History of a cardiac event (congestive heart failure, transient ischemic attack or stroke, arrhythmia)



  • Cyanosis or NYHA functional class III or IV



  • Systolic ventricular function <40%


A prospective validation study of 562 women and 617 pregnancies assessed the accuracy of the prediction model [8]. Maternal cardiac events, pulmonary edema, arrhythmia, stroke, or cardiac death occurred in 13% of pregnancies. Pregnancies with 0, 1, and >1 points had adverse cardiac event rates of 3, 39, and 66%. The agreement between the predicted and the observed event rates was excellent. Neonatal complications (death, intrauterine growth retardation [IUGR], preterm delivery, respiratory distress, or intraventricular hemorrhage) occurred in 17% of pregnancies and were associated with poor maternal functional class or cyanosis, left heart obstruction, anticoagulation, smoking, and multiple gestations. Adverse neonatal events occurred in a third of mothers (age <20 or >35 years) who smoked or received anticoagulants with a risk score >1 compared to the 11% rate in age-matched controls [27]. The CARPREG score may be associated with a higher rate of late cardiovascular events after pregnancy [30]. Late cardiac events (>6 months postpartum)—death, pulmonary edema, arrhythmia, and stroke—were evaluated in the CARPREG cohort of 318 women with 405 pregnancies (median follow-up, 2.6 years). Late cardiac events occurred in 12% of pregnancies and increased to 27% in women with cardiac complications during pregnancy. In women with 0, 1, >1 risk predictors, the 5-year late cardiac event rates were 7, 23, and 44%.


The ZAHARA score is a weighted score prediction model based on a retrospective observational cohort of 1302 pregnancies in 714 women with CHD exclusively (many complex). The ZAHARA score has not been validated. Cardiac complications, obstetric complications, and neonatal outcomes were assessed independently. Additive risk was assigned to women with the following factors:



  • Mechanical heart valve (4.25 points)



  • Severe left heart obstruction (mean pressure gradient >50 mmHg or AVA <1.0 cm2) (2.50 points)



  • History of arrhythmias (1.50 points)



  • History of cardiac medication use before pregnancy (1.50 points)



  • History of cyanotic heart disease (uncorrected or corrected) (1.00 points)



  • Moderate-to-severe pulmonary or systemic atrioventricular valve regurgitation (0.75)



  • Symptomatic heart failure before pregnancy (NYHA class ≥II) (0.75 points)


Scoring is based on five categories of cardiovascular risk (Table 12.1).


Table 12.1

ZAHARA scoring systems for estimating the risk of a cardiac complication during pregnancy

































Points


Number of pregnancies


CV risk (%)


0–0.5


828


2.9


0.51–1.50


280


7.5


1.51–2.50


126


17.5


2.51–3.50


58


43.1


>3.51


10


70.0



CV cardiovascular


Data from: Ashrafi R, Curtis SL. Heart Disease and Pregnancy. Cardiol Ther. 2017 Dec;6(2):157–173


The most prevalent obstetric complications in the ZAHARA cohort were hypertensive complications (12.2%). Arrhythmias (4.7%) and heart failure (1.6%) were the most common cardiac complications. The most frequently encountered neonatal outcomes, which complicated 25% of completed pregnancies, were premature birth (12%), small for gestational age (14%), and mortality (4%). Maternal cardiac complications and neonatal outcomes were highly correlated (r = 0.85, P = 0.002). Adverse neonatal outcome correlated with cyanotic heart disease (corrected/uncorrected) (P = 0.0003), mechanical valve replacement (P = 0.03), maternal smoking (P = 0.007), multiple gestation (P = 0.0014), and the use of cardiac medication (P = 0.0009). The ZAHARA study identified the following additional independent predictors of adverse cardiac complications during pregnancy in women with CVD: moderate or severe systemic or pulmonary AV valve regurgitation, mechanical valve prosthesis, and “at birth” cyanotic CHD.


The predictors and risk scores derived from CARPREG (which included women with acquired heart disease and CHD) and ZAHARA (which included only women with CHD) are population dependent. The CARPREG score reportedly overestimated risk in other CHD cohorts [3133], suggesting that women with acquired heart disease and arrhythmia may be at greater risk of cardiac events. Important maternal risk characteristics, including pulmonary hypertension and dilated aorta, were not included in CARPREG or ZAHARA because they were underrepresented in these studies. Right ventricular systolic dysfunction and/or severe pulmonary regurgitation have been reported as additional independent risk factors for adverse maternal and fetal events in women with CVD [34].


The modified WHO risk classification is the maternal risk assessment recommended by the European Society of Cardiology (ESC) guidelines on the management of CVD during pregnancy [29]. The modified WHO risk classification performed superiorly to both CARPREG and ZAHARA risk scores in a prospective evaluation of cardiovascular risk in 213 pregnancies in 203 women with CHD [26]. The WHO risk classification is based on expert consensus and incorporates all known specific maternal cardiac risk factors, congenital and acquired, and integrates them with other comorbidities [35, 36]. Risk is additive; therefore, for each individual, the risk of a pregnancy may move up a class if there are additional risk factors. The modified WHO classification divides women into four classes, ranging from low to high risk.



  • Modified WHO risk class I conditions are associated with no detectable increased risk of maternal mortality and no/mild increase in morbidity. This category includes uncomplicated, small or mild pulmonic stenosis, patent ductus arteriosus, mitral valve prolapse, and successfully repaired simple lesions. Cardiology follow-up during pregnancy may be limited to 1–2 visits.



  • Modified WHO risk class II conditions are of low to moderate risk and are associated with a small increased risk of maternal mortality or a moderate increase in morbidity. Conditions include arrhythmias, unrepaired atrial septal defect (ASD) or ventricular septal defect (VSD) and repaired tetralogy of Fallot (TOF). Cardiology follow-up is recommended every trimester.



  • Modified WHO risk class II to III conditions can fall either into class II or III depending on individual circumstances and, thus, require individualized assessment. These conditions include native or tissue valvular heart disease (VHD), repaired coarctation, an ascending aorta diameter <45 mm associated with bicuspid aortic valve (BAV), mild left ventricular dysfunction, and HCM. Cardiology and obstetric follow-up recommendations range from every trimester to monthly.



  • Modified WHO risk class III conditions are associated with significantly increased risk of maternal mortality or severe morbidity. Expert counseling is recommended to individualize maternal and fetal risk in pregnancy. Cyanosis with prepregnancy resting arterial oxygen saturation <85% is associated with only a 12% chance of a live birth. This class includes mechanical valve, systemic right ventricle, Fontan circulation, unrepaired cyanotic heart disease, complex CHD, BAV with ascending aortic diameter of 45–50 mm, and MFS with aortic dilatation 40–45 mm. The complication risk is high, and frequent (monthly or bimonthly) cardiac and obstetric monitoring is needed throughout pregnancy, childbirth, and the postpartum period.



  • Modified WHO risk class IV includes cardiac conditions in which pregnancy is contraindicated due to the extremely high risk of maternal mortality or severe morbidity. If pregnancy occurs, termination is advised and should be discussed. If the pregnancy is terminated, appropriate intervention to correct severe left heart obstruction (mitral or aortic stenosis or coarctation) or severe dilatation of the aorta should be performed before pregnancy is attempted again. If the pregnancy continues, monthly or bimonthly cardiology and obstetric follow-up is recommended, as for class III patients.


Fetal Risk Assessment


Limited data for fetal risk have been derived from small cohorts or registry studies of pregnancies in women with CHD, VHD, or both [27, 37]. Neonatal or fetal complications occur in 20–25% of pregnancies in women with CVD. Fetal death occurs in 1–4% of pregnancies in women with CVD [26, 28]. Miscarriage rates are higher in women with complex CHD [29]. Fetal and neonatal complications include death, preterm delivery, decreased birth weight or IUGR, and intraventricular hemorrhage. Maternal predictors of fetal complications include maternal NYHA functional class III or IV, left heart obstruction, mechanical heart valves, anticoagulation use, smoking, cyanosis, multiple gestations, and maternal age <20 years or >35 years [8, 29, 32, 34]. No fetal risk score has been established. Although maternal and fetal risks correlate, maternal risk scores do not adequately predict fetal risk [26].


Congenital Heart Disease and Pulmonary Hypertension in Pregnancy


Maternal High-Risk (III–IV) Conditions


General Recommendations


Women with CHD will usually have a diagnosis before pregnancy . Assessment of prepregnancy risk is imperative. Medical and surgical history, functional status, echocardiography, and oxygen saturation should be evaluated by an interdisciplinary expert team before pregnancy. Women with pulmonary hypertension, severe obstructive valvular heart disease or coarctation, depressed left ventricular or right ventricular function, aortic dilatation in MFS or bicuspid aortopathy, poor functional class, or cyanosis are at the greatest maternal and fetal risk during pregnancy. Risk in pregnancy depends on the specific cardiac defect but increases as the disease becomes more complex. Overall, 80% of pregnancies occurring in women with CHD will complete, whereas 15% will miscarry. Of the completed pregnancies, 12% of women will have cardiac complications, including congestive heart failure, arrhythmia, stroke, and death [30]. Neonatal mortality approaches 13% in mothers with pulmonary hypertension (mean pulmonary artery pressure >25 mmHg) [38] and 88% when maternal cyanosis (oxygen saturation <85%) [39] is present. Women in the highest risk group, WHO IV, should be informed of the extreme maternal risk in pregnancy and offered effective permanent birth control. In cases of severe pulmonary hypertension, pregnancy termination is recommended and should be offered. If the pregnancy is continued, supplemental oxygen, anticoagulation, and specific therapies targeting the etiology of pulmonary hypertension may be required.


Pulmonary Hypertension


Pulmonary hypertension encompasses a group of diverse diseases defined by a mean pulmonary artery pressure at rest >25 mmHg or 30 mmHg on exercise in the absence of a left-to-right shunt. Mild pulmonary arterial hypertension can also be defined as a pulmonary artery systolic pressure of ~36–50 mmHg. The WHO classifies patients into the following five groups based on etiology [40]: Group 1, idiopathic or inheritable pulmonary artery hypertension; Group 2, pulmonary hypertension secondary to congenital and left heart disease (elevated pulmonary capillary wedge pressure); Group 3, pulmonary hypertension due to chronic lung disease or hypoxemia; Group 4, chronic thromboembolic pulmonary hypertension; and Group 5, pulmonary hypertension of unclear or multiple etiologies. Pulmonary hypertension of any etiology , when severe, is poorly tolerated in pregnancy. Pulmonary hypertension in pregnant women is most commonly related to CHD, as advances in treatment of CHD have increased the number of women surviving into the childbearing age [6]. The risk of maternal death is increased, even in the presence of mild pulmonary hypertension. In the United Kingdom, maternal mortality data suggest that pregnancy can be associated with progression of pulmonary hypertension [7]. Maternal deaths are a result of pulmonary thrombosis, pulmonary hypertensive crisis and right heart failure usually in the last trimester or after delivery as pulmonary blood flow increases and the hormonal milieu of pregnancy activates the clotting cascade. General anesthesia, late hospitalization, and increasing severity of pulmonary hypertension are risk factors for maternal death in pulmonary hypertension [38].


Eisenmenger Syndrome


Eisenmenger syndrome —the triad of systemic-to-pulmonary shunt, pulmonary arterial hypertension, and cyanosis—is the most severe form of shunt-related pulmonary hypertension. Eisenmenger syndrome is caused by unrestrictive left-to-right heart shunting of volume and pressure increases the pulmonary blood volume and pulmonary pressure. Altered pulmonary volume/pressure, in turn, disrupts the pulmonary vascular endothelium and results in long-term fixed pulmonary arteriolar hypertension [41]. The increased pulmonary vascular resistance eventually reduces the left-to-right flow across the intracardiac shunt, with eventual right-to-left shunting and resultant cyanosis. In pregnancy, the natural reduction in maternal SVR increases the right-to-left shunt flow, decreases pulmonary blood flow, and increases cyanosis. Asymptomatic women with compensated cardiac defects and mild-to-moderate pulmonary hypertension may clinically deteriorate during the later stages of pregnancy or immediately postpartum. Maternal mortality is due to sudden arrhythmia-related death, progressive heart failure, or pulmonary hemorrhage, and ranges from 17 to 50% in patients with severe pulmonary hypertension and Eisenmenger syndrome [1, 38, 39, 42]. The size of the shunt and the ratio of the SVR to pulmonary vascular resistance determine the volume of the left-to-right shunt flow in VSD. The development of pulmonary hypertension is related to the volume and duration of pulmonary shunt flow in ASD, VSD, and patent ductus arteriosis (PDA), but transmitted systemic pressures contribute and augment the pulmonary arteriolar endothelial damage in VSD and PDA. Even with a large ASD, the pulmonary pressures do not increase until adulthood, whereas pulmonary hypertension develops early in large nonrestrictive VSD and PDA [43, 44]. Eisenmenger syndrome occurs in only 10% of unrepaired ASDs and in 50% of unrepaired VSDs, but it’s seen in nearly all patients with unrepaired truncus arteriosus [45]. In nonpregnant adults with Eisenmenger syndrome, life expectancy is reduced by 20 years in those with simple cardiac shunts but by 40 years in those with more complicated defects when compared to healthy adults [46]. Pregnancy is contraindicated in patients with Eisenmenger syndrome , and termination is recommended.


Cyanotic Heart Disease Without Pulmonary Hypertension


Most cases of cyanotic heart disease will be repaired or palliated in childhood, before childbearing age is reached. Possible cyanotic congenital lesions without pulmonary hypertension include unrepaired TOF, pulmonary atresia with aortopulmonary collaterals, some single ventricular lesions, tricuspid atresia, Ebstein’s anomaly with right-to-left shunts via an ASD, and congenitally corrected transposition of the great arteries (TGA) with VSD or ASD [39]. Right-to-left intracardiac or extracardiac shunts result in hypoxemia, erythrocytosis, and cyanosis. Cyanosis causes fetal loss , prematurity and fetal growth restriction [47, 48]. The degree of maternal hypoxemia is the most important predictor of fetal outcome. Only 12% of fetuses survive to live birth when maternal cyanosis or resting oxygen saturation is <85%. Fetal survival is >90% when maternal oxygen saturation is >90% [39]. Maternal complications occur in up to 30% of cyanotic pregnancies and include arrhythmias, heart failure, pulmonary or arterial thrombosis, and IE. Pregnancy is contraindicated if cyanosis (oxygen saturation <85% at rest) is present as fetal loss is likely and maternal risk is high. Termination is recommended if pregnancy occurs. For resting oxygen saturation >90%, the risk of fetal loss remains increased, and women should be counseled. The decrease in maternal SVR in pregnancy may increase the right-to-left shunt flow and increase maternal cyanosis. For women with mild cyanosis without pulmonary hypertension, completed pregnancy may be possible. The ESC guidelines on the management of CVD in pregnancy recommend exercise oxygen saturation testing for patients with resting saturation >85% but <90%. If saturation declines with exercise, pregnancy should be avoided as fetal prognosis is poor [29].


Severe Left Ventricular Outflow Tract Obstruction


Severe LVOT obstruction from any etiology in pregnancy poses high maternal and fetal risk. Pregnancy is contraindicated, and termination is recommended. Women with severe LVOT obstruction should undergo repair before pregnancy.


Maternal Low- and Moderate-Risk (I, II, III) Conditions


Women with repaired and unrepaired defects in the absence of cyanosis, pulmonary hypertension, or mechanical valve replacement may tolerate pregnancy well as long as ventricular function is preserved and the functional class is good. Prepregnancy assessment with echocardiography and careful follow-up during each trimester is advised.


Atrial Septal Defect


An ASD , a persistent interatrial communication, is the most common repaired or unrepaired lesion in pregnant women with CHD [47, 49]. The reported birth prevalence is approximately 2 per 1000 live births [47, 5052].


The most common ASD involves the secundum septum (fossa ovalis) and accounts for 70% of ASDs. The secundum ASD is twice as common in females as in males. A secundum ASD <8 mm in diameter usually closes spontaneously during childhood. The primum ASD (15–20% of ASDs) occurs because the septum primum fails to merge with the endocardial cushion during fetal development. Primum defects tend to be larger than secundum ASDs and are commonly associated with cleft mitral valve and VSD. The sinus venosus ASD, accounting for 5–10% of ASDs, usually involves the superior venosus septum and is almost always associated with anomalous drainage of the right superior pulmonary vein into the right atrium, which increases the volume of the left-to-right shunt [53]. The last type of ASD, the unroofed or coronary sinus ASD, is rare, accounting for <1% of ASDs. Primum, sinus venosus, and coronary sinus ASDs do not close spontaneously and cannot be closed percutaneously.


The clinical manifestations of an unrepaired ASD are related to defect location, ASD size, and the presence of other congenital defects. Atrial arrhythmias, exercise intolerance, fatigue, and late right heart failure may result from larger atrial shunts. Paradoxic embolization can occur, even in small ASDs, with a reported rate up to 5%. The presence of an ASD during the reproductive years is rarely associated with severe pulmonary hypertension and is generally well tolerated in pregnancy. Many pregnant women remain asymptomatic; those with significant shunts may develop a detectable systolic pulmonary flow murmur with pregnancy because of the pregnancy-related increase in intravascular volume. Closure of a symptomatic or large asymptomatic ASD is recommended before pregnancy to prevent right heart failure, atrial arrhythmia, and paradoxic embolization. Closure of asymptomatic small ASDs is not indicated prepregnancy to prevent paradoxic embolization [29]. Risk of thromboembolism in pregnant women with ASD is decreased by preventing venous stasis (via ambulation or compression stockings), restricting the use of long-term intravenous catheters, and using prophylactic anticoagulation in the immobilized. Percutaneous closure of secundum ASD during pregnancy is very rarely indicated. Pregnant women with repaired ASD are not at increased maternal or fetal risk [29]


Ventricular Septal Defects


VSD is the second most common form of CHD with a prevalence of 3–3.5 per 1000 live births, which represents 10% of CHD in adults [54]. VSDs are most commonly perimembranous (80%), muscular (5–20%), inlet (8%), or infundibular (6%). Muscular VSDs often close during childhood. The functional size of the defect, the presence of associated congenital conditions, and the ratio of systemic-to-pulmonary vascular resistance determines the severity of the left-to-right shunt, the resultant increase in right ventricular volume, and the degree of pulmonary overcirculation. In adults, most VSDs occur as an isolated defect but also occur with ASD (35%), PDA (22%), right aortic arch (13%), TGA, or TOF. Small VSDs, usually with an orifice dimension <25% of the aortic annulus diameter, are restrictive to both pressure and volume and are well tolerated in pregnancy. Moderate-size defects, measuring 25–75% of the aortic annulus diameter, allow a moderately sized left-to-right volume shunt but no or minimal evidence of pulmonary hypertension. These defects are also relatively well tolerated in pregnancy. Large defects, >75% of the aortic annulus diameter, lead to a large unrestricted shunt volume, left ventricular volume overload, and pulmonary hypertension. The maternal risk of heart failure and arrhythmias is high in women with a large VSD with pulmonary hypertension, a history of ventricular dysfunction, moderate or greater pulmonic stenosis, or arrhythmias.


Women with successfully repaired VSDs with normal ventricular function are not at increased maternal or fetal risk with pregnancy. Prepregnancy evaluation by echocardiography to assess residual VSD, ventricular function, and pulmonary pressures is recommended. Children of women with VSD have a greater risk of CHD (3–7%). The risk of endocarditis is 11% for unrepaired VSD, but the rates are halved after successful repair. Subacute bacterial endocarditis in unrepaired VSD is not related to defect size [55].


Atrioventricular Septal Defects


Atrioventricular (AV) septal or AV canal defects are a complex congenital heart defect involving the development of the endocardial cushion and are associated with defects involving the AV valves and the AV septum. Representing around 5% of CHD, AVSDs have a prevalence of 0.3–0.4 per 1000 live births [50, 54]. Down syndrome (trisomy 21) is strongly associated with AVSD and is seen in 40–50% of AVSD cases [56]. Maternal diabetes and obesity may be associated with nonsyndromic AVSDs [57]. AVSD involves both the atrial primum septum and the ventricular inlet septum in half the cases and is referred to as a complete AV canal defect. Otherwise, AVSD may be isolated to the atrial primum septum and is called an incomplete AV canal. Abnormalities of the AV valve are variable and include a common or cleft AV valve. AV valve regurgitation is common and contributes to symptoms. When the AV canal is complete, there is left-to-right shunting at both the atrial and ventricular levels, which produces a marked intracardiac shunt, leading to early heart failure and pulmonary hypertension in all cases. Surgical correction has enabled survival into the reproductive years. Residual shunt, AV valve regurgitation, and pulmonary pressures must be evaluated before pregnancy to assess maternal and fetal risk with pregnancy. Women with moderate or less residual AV valve regurgitation and normal left ventricular function after repair tolerate pregnancy well and are considered WHO category risk II. If ventricular function is abnormal (ejection fraction [EF] < 60% but >30%) and AV valve regurgitation is severe, surgical correction with mitral valve repair is recommended before pregnancy [58]. Worsening heart failure, arrhythmia, and perinatal mortality are consequences of pregnancy in AVSD with severe AV regurgitation and ventricular dysfunction [59]. For women with AVSD and pulmonary hypertension, pregnancy is contraindicated.


Coarctation of the Aorta


Coarctation of the aorta is defined as a significant narrowing of the proximal thoracic aorta at the insertion of the ductus arteriosum distal to the left subclavian artery. Aortic obstruction leads to systemic hypertension, early coronary artery disease, stroke, heart failure, aneurysm formation, and aortic dissection and rupture in unrepaired coarctation of the aorta [60, 61]. Genetic factors contribute to the pathogenesis of coarctation. Half of all cases of coarctation are associated with BAV and, almost one-fifth of patients with Turner syndrome have coarctation of the aorta [62]. Significant coarctation or recurrent coarctation after surgical or catheter repair with outflow obstruction (peak-to-peak gradient >20 mmHg or <20 mmHg with evidence of collateral flow) should be corrected before pregnancy. After successful repair of coarctation, pregnancy is well tolerated and is categorized as a WHO class II risk. Women with residual coarctation gradient, aortic aneurysm or residual hypertension are at increased risk of aortic rupture and rupture of cerebral aneurysm during pregnancy or delivery. During pregnancy, close BP monitoring and treatment of hypertension are recommended.


Pulmonary Valve Stenosis and Regurgitation


Pulmonary valve disease is a common congenital heart defect with a slight female prevalence and occurs in 7% of CHD cases [6365]. Pulmonary stenosis also occurs in association with other congenital defects including TOF, congenital rubella syndrome, and Noonan, Williams, Alagille, and LEOPARD syndromes. Pulmonary stenosis may occur at the valve, subvalvular, or supravalvular position. Valvular pulmonary stenosis is usually an isolated lesion with a benign clinical course, and patients are expected to survive to adulthood. Bicuspid pulmonary valves occur in less than 20% of cases [66]. Dysplastic pulmonary valves are common in Noonan syndrome [67]. If stenosis is severe, pulmonary blood flow may be limited with exertion resulting in exercise-induced fatigue, dyspnea, syncope, or chest pain, and eventual symptomatic right heart failure. Women with severe pulmonary stenosis (peak pulmonary valve gradient >64 mmHg) are at increased risk of right heart failure and possible fetal compromise and should undergo valvuloplasty or valve replacement before pregnancy [6870]. A normal pregnancy is expected following surgical or balloon repair of a congenital pulmonary valve stenosis with little residual obstruction or regurgitation. Pulmonary stenosis in the absence of right heart failure is well tolerated in pregnancy [71]. Percutaneous pulmonary valvotomy during pregnancy can reduce risk in symptomatic women with severe pulmonary stenosis [72]. Although no maternal complications were reported among 100 pregnant women with repaired and unrepaired pulmonary stenosis , fetal complications include fetal demise (0.8%), perinatal death (4.1%), premature delivery (14.5%), and recurrent CHD (2.8%) stenosis. Moderate or severe pulmonary regurgitation is usually a complication of repaired TOF or occurs after pulmonary valvotomy for childhood pulmonary stenosis. Severe pulmonary regurgitation may be associated with right ventricular dilatation or systolic dysfunction. Overall, pulmonary regurgitation, even when severe, is well tolerated in pregnancy. The risk of right heart failure during pregnancy is increased in women with severe pulmonary regurgitation and any one of the following: multiple gestations, right ventricular systolic dysfunction, right ventricular hypertrophy, or branch pulmonary stenosis [73]. In these circumstances, pulmonary valve replacement with a biologic prosthesis is recommended before pregnancy.


Ebstein’s Anomaly of the Tricuspid Valve


Ebstein’s anomaly is a congenital developmental defect involving the tricuspid valve and the right ventricle. It occurs in 1 in 20,000 births with equal frequency in males and females [7476]. The risk of Ebstein’s anomaly is increased in fetuses exposed to lithium early in gestation. Ebstein’s anomaly occurs in conjunction with a patent foramen ovale (PFO) or secundum ASD in about 80% of cases [77] and is also associated less frequently with VSD, PDA, BAV, and l-TGA. An accessory conduction pathway (e.g., Wolff-Parkinson-White) in 6–36% of patients with Ebstein’s anomaly predisposes to symptomatic tachycardia [78]. Apical displacement of the septal and posterior tricuspid valve leaflets into the right ventricle in Ebstein’s anomaly (septal leaflet displacement >2 cm or 0.8 cm/m2 >anterior mitral leaflet attachment) divides the right ventricle into two chambers: a superior “atrialized” thin, non-contracting right ventricle chamber above the tricuspid valve and a smaller distal right ventricular pumping chamber below the valve. Variable amounts of tricuspid regurgitation and right ventricle dysfunction are consequences of Ebstein’s anomaly . Bidirectional shunting across the PFO or ASD can cause cyanosis without pulmonary hypertension. In women with Ebstein’s anomaly without cyanosis or heart failure, pregnancy is well tolerated (WHO risk class II). Women with right ventricle failure and severe tricuspid regurgitation should undergo tricuspid valve repair before pregnancy. During pregnancy, the severity of the tricuspid regurgitation and the functional capacity of the right ventricle determine the hemodynamic burden and the outcome [79, 80]. Arrhythmias and right heart failure are associated with a worse prognosis [80] as premature delivery and fetal mortality are increased [79]. Women with Ebstein’s anomaly and interatrial shunting may develop right-to-left shunt and cyanosis during pregnancy. Paradoxic embolic risk is also increased in pregnancy [29]. Isolated severe tricuspid regurgitation can be managed with diuretics if needed during pregnancy.


Tetralogy of Fallot


TOF is a cyanotic congenital heart defect with 4 components: a malpositioned (rightward) aorta that overrides the ventricular septum, a large malaligned VSD, infundibular subpulmonary pulmonary stenosis, and right ventricular hypertrophy. Because of the malaligned aorta and VSD, the pulmonary artery may be underdeveloped, and the aortic root may be dilated, which may cause aortic insufficiency (AI). The prevalence of TOF is about 4–5 per 10,000 live births and accounts for 7–10 % of CHD [50, 81]. Children with TOF usually present with symptoms of agitation and cyanosis within the first year of life [82]. Cyanosis is dependent on the degree of right ventricular outflow tract (RVOT) obstruction (infundibular right ventricular hypertrophy and pulmonary stenosis). Patients with mild RVOT stenosis may remain “pink” and go undiagnosed until late adolescence or early adulthood when they present with evidence of pulmonary overcirculation secondary to the large unrestricted VSD. Prenatal diagnosis is common as widespread screening ultrasonography has improved [83]. Intracardiac surgical repair of TOF is typically performed before 6 months of age [84]. Residual defects, right ventricular systolic function, and pulmonary insufficiency affect late prognosis. Severe pulmonary insufficiency with moderate right ventricle dilatation or right ventricle systolic dysfunction should be repaired before pregnancy. Women with repaired TOF tolerate pregnancy well (WHO risk class II). Cardiac arrhythmias, heart failure, VTE, and endocarditis may occur in up to 12% of pregnancies in women with repaired TOF [85].


Transposition of the Great Arteries


Transposition of the great arteries (TGA) is a congenital heart defect in which the aorta arises from the right ventricle and the pulmonary artery arises from the left ventricle. The great vessels are transposed with the pulmonary artery positioned posterior to the aorta. Orientation of the ventricles determines the physiology and prognosis. In the most common form, dextro-TGA (d-TGA) , the left ventricle is aligned leftward producing two parallel circulations. Systemic venous blood recirculates via the right ventricle and the aorta to the peripheral tissues, whereas the oxygenated blood recirculates through the left ventricle and the pulmonary artery to the lungs. Cyanosis is present at birth with oxygenation dependent on intracardiac shunting (via an ASD, PFO, or VSD) or flow through a PDA. In levo-TGA (l-TGA) , the left and right ventricles are “inverted” with the left ventricle rightward and the venous and arterial circuits “physiologically corrected” and arranged in series, which avoids cyanosis unless other cardiac defects are present. However, the systemic ventricle in l-TGA is the less resilient morphologic right ventricle, and patients are at risk of progressive right ventricle dysfunction and right heart failure as adults.


d Transposition of the Great Arteries

d-TGA occurs in 2.3–4.7 per 10,000 live births; it accounts for less than 3% of all CHD but is seen in up to 20% of cyanotic CHD [50, 81]. VSD is present in 50% of patients with d-TGA. LVOT obstruction due to pulmonary stenosis or pulmonary atresia occurs in 25% of d-TGA cases. Antenatal diagnosis is difficult even with fetal echocardiography, and diagnosis is usually made by echocardiogram in a cyanotic newborn with respiratory distress. Without surgical correction, mortality is 90% in the first year of life [86]. Atrial switch procedures, either Senning or Mustard, were widely used from the mid-1960s to the 1980s to surgically repair d-TGA and allowed children to survive into adulthood. Redirecting the venous atrial inflow corrected the cyanosis and provided for circulation in series at the expense of allowing the morphologic right ventricle to remain the systemic ventricle. Long-term complications include eventual right ventricular dysfunction, atrial arrhythmias, and atrial baffle obstruction [8789]. The arterial switch operation (ASO) , in widespread use since 1990, is a surgical procedure for correcting the anatomy and involves reorienting the left ventricle as the systemic ventricle and restoring the cardiac circulation in series. Improved ASO techniques, in which the coronaries are reimplanted onto the root of the native pulmonary artery, have decreased the perioperative mortality in uncomplicated d-TGA to near 0% [90]. Long-term complications of the ASO involve coronary arterial insufficiency in as many as 12% at 15 years after correction [91], neo-aortic root dilatation (Z score –3) in up to 50% at 10 years, and moderate to severe neo-aortic insufficiency in 8–15% at 20 years [92].


The data on pregnancy after ASO are limited, but successful pregnancies have been reported [93, 94]. In a retrospective review of women who underwent ASO for d-TGA , aortic valve regurgitation worsened in 11 of 21 (52%) pregnant women and in 0 of 15 nonpregnant controls followed for 100 months [95]. Cardiac events in patients with d-TGA were common in both pregnant and nonpregnant women (62% vs. 53%, nonsignificant) with worsening ventricular function in both groups (29% and 27%). Premature birth (38%) and small for gestational age (38%) were adverse fetal outcomes reported in the offspring of women with d-TGA who underwent arterial switch procedures.


Women with d-TGA treated with ASO may be at greater long-term risk of systemic right ventricular deterioration during and after pregnancy than those who were surgically corrected with ASO in whom the systemic ventricle is the left ventricle [96]. An irreversible decline in right ventricular function has been described in 10% of pregnancies after atrial switch procedures. Women with moderate or greater right ventricle dysfunction or severe systemic AV valve regurgitation (TR) are at greatest risk of worsening right ventricle function and should be advised against pregnancy.


l-Transposition of the Great Arteries

l-TGA is also known as congenitally corrected transposition of the great vessels. l-TGA is rare and occurs in <1% of CHD with a prevalence of 0.02–0.07 per 1000 live births [97, 98]. Associated cardiac defects occur in 80–90% of patients with l-TGA : VSD, 70–80%; pulmonary stenosis, 30–60%; and Ebstein-like tricuspid valve anomaly, 20–53% [99, 100]. In patients without associated cardiac defects (20% of l-TGA patients), survival into adulthood without correction and often without symptoms is the rule. These women tolerate pregnancy well. The risk of AV block is increased in l-TGA patients, and careful use of AV nodal agents is advised. The risk in pregnancy depends on the severity of the associated defects, the systemic ventricular function, systemic tricuspid valve regurgitation, and the severity of the RVOT obstruction as it relates to the VSD size. In 2 studies of pregnant women with l-TGA, live births were seen in 27 of 45 (60%) [101] and 50 of 60 (83%) pregnancies [102]. Four women developed heart failure and one woman had a stroke. Prepregnancy evaluation and counseling are required. Patients with l-TGA and right ventricle systolic function <40%, severe tricuspid regurgitation, or NYHA functional class III or IV should be advised against pregnancy [29, 99, 100].


Fontan Circulation


The Fontan procedure is a palliative surgical procedure performed in patients with functional or anatomic single-ventricle: hypoplastic left heart syndrome, tricuspid atresia, pulmonary atresia with intact ventricular septum, and double-inlet left ventricle [69, 103]. An extracardiac conduit is created surgically to bypass circulation from the vena cava (cavopulmonary) or the right atrial appendage (atriopulmonary) directly to the pulmonary artery. The Fontan procedure separates venous from arterial cardiac circulation into series while eliminating a sub-pulmonic ventricle. Fontan physiology is characterized by reduced cardiac output and chronically increased systemic venous pressures [104]. Complication rates after the Fontan procedure are frequent, and 15-to 20-year survival rates range from 60 to 85% [105, 106]. Survival into reproductive age is possible with good functional capacity. Women with successful Fontan circulation and a well-performing systemic ventricle, preserved contractile reserve, and high functional class may have the cardiac reserve required to accommodate the hemodynamic burden of pregnancy. Any pregnancy in a patient with a Fontan circulation is high risk (WHO risk class III or IV). Data are limited on pregnancy after a Fontan procedure. Outcomes from 33 pregnancies in patients with Fontan from two US centers and 25 pregnancies from a literature review have been reported. Spontaneous abortion, preterm labor, IUGR, and fetal demise [29] suggest that the uteroplacental blood flow may be compromised in mothers with Fontan physiology. Maternal complications include postpartum hemorrhage (in up to 50%), atrial arrhythmias, and ventricular dysfunction. Pregnancy in patients with Fontan circulation must be carefully considered, as successful pregnancies are possible only in selected patients. A comprehensive cardiovascular evaluation with an adult CHD specialist is recommended to identify suboptimal Fontan physiology and risk. Patients with poor functional class (NYHA III or IV), systemic ventricular function <40%, moderate to severe AV valve regurgitation, cyanosis with room air saturation <90%, or a history of arrhythmia, venous thromboembolism, heart failure, or protein-losing enteropathy should be advised against pregnancy [32], and termination is recommended.


Aortopathy


Aortic disorders primarily affecting the thoracic ascending aorta predispose patients to aortic aneurysm formation, aortic dissection, and aortic rupture. The most common inheritable aortopathies are associated with BAV, MFS, and Loeys-Dietz, Turner, and Ehlers Danlos syndromes. The congenital defects, coarctation of the aorta and TOF, are also associated with aortic aneurysm formation . Pregnancy increases the risk of aortic dissection and rupture in patients with preexisting aortic pathology. Dissection, although rare in pregnancy, is an important cause of maternal mortality [4]. Dissection occurs most frequently in the last trimester of pregnancy (50% of cases) or early postpartum (33%) due to the hemodynamic and hormonal changes associated with the end of pregnancy [107109].


Because aortic pathology is silent until it becomes catastrophic, screening high-risk individuals—those with prior dissection, with Marfan, Loeys-Dietz, Ehlers Danlos, or Turner syndromes, and those with a family history of familial aortopathy—is the key to successfully preventing aortic dissection or rupture during or after pregnancy.


Marfan Syndrome


MFS is an autosomal dominant disorder affecting connective tissue with a reported incidence of 1 in 3000–5000 individuals [110, 111]. Of patients with MFS , 90% carry a fibrillin (FBN1) genetic mutation that is responsible for the clinical characteristics of MFS: aortic root dilatation/dissection, ectopia lentis, skeletal findings (kyphoscoliosis, pectus, arachnodactyly), mitral valve prolapse, dural ectasia, pneumothorax, and skin striae [112, 113]. Aortic root dilatation and lens ectopia are the cardinal features of MFS. Aortic aneurysmal dilatation, AI, aortic dissection, and aortic rupture are the primary causes of major morbidity and mortality in MFS. According to the revised Ghent criteria, patients with MFS must have aortic root dilatation, a family history of aortic root dilatation, or a FBN1 mutation [113, 114]. Aortic measurements in women of short stature should be indexed to body surface area. In MFS patients with a normal aortic root size (<20% of MFS patients), the risk of aortic dissection or other cardiac complications during pregnancy is 1% [115]. The risk of aortic dissection increases with increasing aortic diameter in MFS. The risk of major aortic complications during pregnancy appears to be low when the aortic root diameter is <4.0 cm [116]. Pregnant patients with MFS are at increased risk for aortic dissection if the aortic diameter exceeds 4 cm and if the diameter changes (>5 mm) during pregnancy [117119]. In pregnant women with MFS and an aortic diameter >4.0 cm, half will have an aortic rupture or life-threatening aneurysm growth or will require prophylactic aortic surgery during pregnancy. Women with MFS and a history of aortic dissection are at greater risk of recurrent dissection with pregnancy and should be discouraged from getting pregnant [120].


Elective repair of aortic root enlargement >4.0 cm in MFS patients before conception is recommended by the American College of Cardiology/AHA/American Association of Thoracic Surgeons guidelines [121]. After successful surgical correction of the ascending aorta, there is a residual risk of aortic dissection in the remaining aorta during subsequent pregnancies [117, 120, 122]. All women with MFS should undergo monthly or bimonthly cardiovascular and echocardiographic monitoring throughout pregnancy and for at least 4 weeks postpartum [29, 117, 121, 123].


Treatment with beta-blockers, labetolol, or metoprolol tartrate is recommended throughout pregnancy to reduce arterial shear stresses, control heart rate, and decrease the risk of aneurysmal dilatation and dissection [29, 82, 117, 121, 123]. Strict BP control in all pregnant women with MFS is advised. Angiotensin-converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs) are contraindicated in pregnancy. Calcium channel antagonists should be avoided as limited evidence suggests an increase in aortic complications with their use [124]. Delivery should occur in a center with emergency cardiovascular surgery services.


Pregnant women with MFS who develop chest pain should undergo a thorough evaluation that includes aortic imaging for suspected aortic complications. Magnetic resonance imaging without gadolinium is recommended over computed tomographic imaging to avoid exposing the fetus to ionizing radiation. Aortic dissection in pregnancy poses a grave risk to both mother and fetus. Surgical treatment recommendations are the same as in nonpregnant MFS patients. For ascending aortic dissections during the first or second trimester, urgent surgical repair with aggressive fetal monitoring is recommended. Fetal loss is a common complication of hypothermia and cardiopulmonary bypass [125]. In the third trimester, when ascending aortic dissection occurs, urgent cesarean section with concomitant aortic repair appears to offer the best chance for survival for the unborn child and the mother. Medical therapy or stent grafting if anatomy is amendable is preferred in MFS patients who develop descending aortic dissection during pregnancy [121].The long-term rate of aortic dilatation may be increased in MFS patients after pregnancy compared to MFS women who have never been pregnant (0.36 vs. 0.14 mm per year), among those with baseline aortic diameter ≥40 mm [120].


Bicuspid Aortic Valve


Aortopathy, dilatation of any or all segments of the proximal aorta from the aortic root to the aortic arch, is present in approximately half of patients with BAV [126]. The two major complications of bicuspid aortopathy are aortic aneurysm formation and aortic dissection. The risk of aortic dissection risk is relatively low compared to the risk in MFS, with an incidence of 0.1% per patient-year of follow-up in a Toronto study involving 642 BAV patients [119, 127]. In another study, no dissections occurred in patients with an aortic diameter <4.5 cm and only two dissections occurred overall among 416 patients with BAV who were followed for 16 ± 7 years, yielding an incidence of 3.1 cases per 10,000 person-years [128].


Maximal aortic dilatation usually involves the distal ascending aorta, which is not well visualized with standard transthoracic echocardiography. Assessment of ascending aortic size by computed tomography or magnetic resonance imaging is recommended before pregnancy in patients with BAV. Baseline aortic diameter predicts aneurysm expansion in BAV patients [128, 129].


The risks of pregnancy in BAV with aortopathy have not been evaluated. In women with BAV, the hemodynamic and hormonal changes in pregnancy pose a risk of aortic dilatation and dissection. Women with BAV and aortic dilatation are at risk of spontaneous aortic dissection, usually in the third trimester or after delivery, especially if there is an associated aortic coarctation [130]. Guidelines for preconception prophylactic aortic repair for women with BAV have not been established. In nonpregnant patients with ascending aortic aneurysm, aortic repair is indicated when the aortic root or ascending aorta diameter is >5 cm or if the rate of increase is ≥5 mm per year [121]. The threshold for recommending ascending aortic repair in patients with BAV aortopathy before pregnancy varies. The 2010 ESC guidelines for managing adult CHD [68] and the ESC guidelines for managing CVD in pregnancy [29] recommend prepregnancy prophylactic ascending aorta repair in patients with BAV and aortic size >5 cm. However, the 2010 ACC/AHA thoracic aortic guidelines recommend an aortic diameter threshold >4.5 cm for aortic repair if aortic dilatation is progressive and/or there is progression of aortic regurgitation [121]. The consensus is that pregnancy should be avoided in women with BAV and aortic dilatation >5 cm. During pregnancy in patients with BAV and aortopathy, strict BP control with a beta-blocker is warranted.


Ehlers Danlos Syndrome


Type IV Ehlers Danlos syndrome is an autosomal dominant defect in type III collagen (COL3A1 gene) characterized by tissue fragility, which predisposes women to arterial, gastrointestinal, and uterine rupture [121]. According to the 2011 ESC guidelines on managing CVD in pregnancy [29], pregnancy is an absolute contraindication in women with type IV Ehlers Danlos syndrome. Celiprolol is recommended in pregnant and nonpregnant patients with type IV Ehlers Danlos syndrome to reduce the risk of high-risk dissections [121].


Turner Syndrome


Turner syndrome is characterized by short stature, skeletal abnormalities, primary ovarian failure, and BAV with and without coarctation in females caused by a loss of at least part of the X chromosome. The true prevalence is not known as mild cases may go undetected [131]. The reported incidence is approximately 1 in 2000 to 1 in 2500 live female births [132, 133] Abnormalities of the aortic valve and/or the aorta associated with Turner syndrome are responsible for the morbidity and mortality; BAV is present in up to 30% and coarctation in 18% [134136]. The risk of aortic dissection is 100x greater in women with Turner syndrome than in normal females [137]. Aortic root dilatation, defined as an aortic size index (ASI) >2.0 cm/m2 (which is >95th percentile), is a predictor of aortic dissection in patients with Turner syndrome . Aortic dissection or rupture leads to increased cardiovascular mortality usually during the third and fourth decade of life [138, 139]. The ESC guidelines for treating CVD in pregnancy recommend prophylactic aortic repair for those at the greatest risk of rupture/dissection: ASI >2.7 cm/m2 [29]. Risk of aortic dissection or rupture is increased with coarctation, BAV, and maternal hypertension [137]. The prevalence of hypertension is 30–50% in patients with Turner syndrome, which also increases the risk of stroke. Infertility is the rule in Turner syndrome, but spontaneous pregnancies occur occasionally. In vitro fertilization with oocyte donation has increased the rate of pregnancy in Turner syndrome. Pregnancy increases the risk of aortic complications, and the maternal death risk is reported to be as high as 2–11% [29, 140]. Preeclampsia risk is also increased during pregnancy, and treatment of hypertension with beta receptor antagonists is recommended.


Valvular Heart Disease in Pregnancy


Worldwide, the most common cause of VHD in pregnancy is RHD [141, 142]. In the United States and Canada, RHD now causes less than a quarter of heart disease in pregnant women [8, 143]. Declining rates of RHD in developed countries have made congenital causes of VHD more common. Overall, stenotic VHD carries a greater risk in pregnancy than regurgitant VHD according to the WHO [35, 144] scoring systems of maternal risk in pregnancy. Left-sided valve lesions have a greater rate of adverse events than do right-sided ones. Severe LVOT obstruction almost exclusively due to mitral or aortic stenosis is one of four predictors of a maternal adverse event in pregnancy [29]: VHD with left heart obstruction, mechanical heart valves, anticoagulation in pregnancy, and poor NYHA functional class contribute significantly to neonatal complications including fetal death, preterm delivery, IUGR, reduced birth weight, and respiratory distress syndrome [29]. During pregnancy, increases in stroke volume, heart rate, and cardiac output by the second trimester can cause symptomatic decompensation in women with known and unknown VHD. The maternal hemodynamic adaptations to pregnancy may unmask a previously unrecognized valvular heart condition. Complication rates vary by the type and severity of VHD.


Women with known VHD benefit from preconception cardiac evaluation . Women with moderate or high-risk VHD should be followed at a center with a multidisciplinary team of cardiologists, maternal-fetal medicine specialists, obstetric anesthesiologists, and cardiac surgeons. Preconception medical and surgical history, assessment of functional status, an electrocardiogram, and a detailed transthoracic echocardiogram are recommended [145].​ Echocardiography should determine the specific valve abnormality (stenotic, regurgitant, or mixed), the number and location of affected valves, and the severity of the valvular abnormalities. Evaluation of left and right ventricular systolic and diastolic function, estimation of the pulmonary artery pressure, and identification of other associated cardiac defects make the preconception echocardiogram vital in planning a future pregnancy [145]. Cardiac magnetic resonance imaging may also be useful in the prepregnancy evaluation of women with VHD and suspected aortopathy or right ventricle dysfunction for assessing aortic pathology as well as right ventricle volumes and function.


For women with advanced VHD who require valve replacement , a detailed discussion of the risks and benefits of surgical options with a specialized cardiovascular team before conception is recommended. All prosthetic valve types are associated with increased maternal and fetal risks during pregnancy. Mechanical valves require lifelong anticoagulation to prevent valve thrombosis and thromboembolic events and significantly increase maternal and fetal complications in pregnancy. Biologic valves generally do not require anticoagulation but have limited durability. Prosthetic valves placed in women of childbearing age will require a repeat valve intervention, which has a mortality of 0–5% depending on the valve position and degree of emergency [29]. The trade-off between the potential for reintervention for bioprosthetic deterioration and the risk associated with pregnancy and long-term anticoagulation should guide the discussion with the patient before pregnancy [146, 147]. The choice of specific prosthesis in a woman who desires pregnancy should be made only after extensive discussion and evaluation of specific patient risk. The desire for pregnancy is a class IIb indication for a biologic valve in the 2007 ESC guidelines on the management of VHD and a class Ic indication in the 2014 ACC/AHA Guidelines [58, 145].


Cardiac Surgery During Pregnancy


Cardiac surgery during pregnancy is high risk and should be reserved for women with severe intractable heart failure symptoms unresponsive to medical therapy [145]. The maternal mortality rate (3–6%) with cardiac surgery is similar to that in nonpregnant women, but the extreme emergency of cardiovascular surgery during pregnancy combined with the added risk of emergency delivery in many cases increases poor maternal outcomes [1, 125]. Maternal mortality occurs in 9% of surgical valve procedures but in 22% of aortic or arterial dissections and pulmonary embolectomies [1]. Fetal-neonatal risks of maternal surgery during pregnancy are high and unpredictable. The risk of fetal death during cardiac surgery is 20–30% [1]. The duration of pregnancy at the time of surgery does not appear to influence the fetal-neonatal outcomes [125].


Fetal outcome in cardiac surgery during pregnancy is related to reduced uteroplacental flow, which is compounded by uterine contractions, fetal bradycardia, and fetal lactic acidosis related to the fetal stress response [1, 148]. Techniques to improve fetal outcomes include increasing cardiopulmonary bypass flow rates above 2.5 L/min per m2 and maintaining mean arterial pressures >70 mmHg. Continuous fetal monitoring is imperative as prolonged fetal bradycardia (<80 beats per minute) that is unresponsive to increasing flow rates during cardiac surgery is an indication for cesarean delivery if the fetus is viable. Hypothermia during cardiac surgery does not appear to increase fetal risk [1], but rewarming may induce preterm labor [125, 149].Timing of cardiac surgery during pregnancy is difficult to predict. Optimizing fetal and maternal clinical outcomes is the goal. The safest period for cardiac surgery is likely during weeks 20–28 of pregnancy, which limits the risk of fetopathy during early pregnancy and premature delivery and the increased maternal risk during the later stages of pregnancy [150]. Delaying cardiac surgery to 26–28 weeks gestation allows for fetal maturation, increased fetal viability, and better fetal neurologic outcomes. Performing a cesarean delivery before cardiac surgery after 26 weeks gestation has been recommended [29, 125], and successful cesarean delivery at the time of cardiac surgery has been reported [151].


Mitral Stenosis


Mitral stenosis is the most common cause of VHD in pregnancy worldwide and is almost always a distant consequence of acute rheumatic fever, although most patients do not recall the acute rheumatic reaction [152, 153]. A streptococcal infection [154] in childhood can trigger an exaggerated immune reaction that can lead acutely to a clinical syndrome of arthritis (in 35–66% of cases) and pancarditis (in 30–80% of cases) with pericarditis, myocarditis, and valvulitis [155]. The spectrum of valve inflammation without active infection varies geographically and temporally. Pure mitral regurgitation is the valvular abnormality commonly seen in the 2 decades after acute rheumatic fever. These patients have pliable non-scarred leaflets, elongated chordae, and mitral leaflet prolapse. Many (47%) of these valves have pathologic evidence of active valvular inflammation [156]. Over time, pure mitral stenosis or mixed mitral stenosis and mitral regurgitation develop as the commissures fuse, the subvalvular structures fibrose and retract, and the leaflets become calcified and immobilized predominantly at the tips with relative preservation of the motion at the base of the leaflet. Inflammatory reactions and leaflet prolapse are no longer seen. As this progresses, diastolic mitral leaflet motion becomes restricted causing mitral stenosis with and without mitral regurgitation. Rapid progression from mitral regurgitation to stenosis in endemic areas may be related to the lack of antibiotic use, recurrent streptococcal infection, or a more virulent strain of streptococci [156]. Long-term permanent valve damage associated with acute rheumatic fever is called RHD . The mitral valve is involved in almost all cases of RHD, and the aortic valve is affected in 20–30% of cases [157]. Females account for two-thirds of all RHD cases [158]. RHD is endemic in the poor and underdeveloped nations of Oceania, South Asia, and central sub-Saharan Africa. Globally, 33.4 million estimated cases of RHD were diagnosed in 2015, with more than 319,000 deaths [159]. Since 1990, mortality secondary to RHD has declined worldwide by an estimated 48%. [158]. RHD accounts for 55–88% of the cardiac disease in pregnant women in developing nations [141, 142] but less than 25% of the cardiac disease in US and Canadian pregnancies [8, 28, 143]. Mitral stenosis with and without mitral regurgitation accounted for 42% of VHD and only 11% of all heart disease cases in the European Registry on Pregnancy and Heart Disease from 2007 to 2011 [28]. Mixed mitral valve disease occurs at a similar frequency in pregnant women as does mitral stenosis and shares a similar maternal and fetal complication pattern [28, 160]. Maternal and fetal complications correlate with the severity of mitral stenosis in pregnancy [37, 160] (Table 12.2). Mild mitral stenosis (mitral valve area >1.5 cm2) is well tolerated in pregnancy, whereas moderate or severe mitral stenosis (mitral valve area <1.5 cm2) is poorly tolerated. Mitral stenosis obstructs left ventricular filling and creates a gradient across the mitral valve. As the diastolic gradient across the valve increases, left atrial and pulmonary venous pressures increase. In mid-to-late pregnancy and especially during labor, the rise in stroke volume and heart rate augment the elevation of the left atrial pressure and increase the risk of congestive heart failure and atrial fibrillation in women with known or unknown mitral stenosis. The hypercoagulability associated with pregnancy further increases the risk of stroke in patients with mitral stenosis and atrial fibrillation.


Table 12.2

Pregnancy outcomes in women with mitral stenosis





































Severity of MS


Number of pregnancies


Congestive heart failure


Arrhythmia


Preterm delivery


Small for gestational age or IUGR


Fetal demise


Mild MS


MVA >1.5 cm2


61


20%


8%


11%


8%


2%


Moderate or severe MS


MVA <1.5 cm2


65


51%


20%


31%


18%


5%



IUGR intrauterine growth restriction, MS mitral stenosis, MVA mitral valve area


Data from [37, 160]


In developed countries, maternal mortality in mitral stenosis is low, ranging from 0 to 3% [37, 160]. In a report from sub-Saharan Africa, the maternal mortality among 46 pregnant women with rheumatic mixed mitral valve disease was high at 32%; this finding may reflect the lack of antenatal diagnosis and reduced surgical resources [142]. Congestive heart failure and arrhythmia are the most common maternal complications of mitral stenosis. Overall, congestive heart failure occurs in 31–36% of pregnancies among women with variable degrees of mitral stenosis [37, 160, 161]. Atrial fibrillation provokes acute deterioration with congestive heart failure in up to 20% of mitral stenosis cases in pregnancy as the elevated heart rate associated with atrial fibrillation reduces the diastolic filling time and increases the transmitral gradients [160]. Women with mild mitral stenosis in pregnancy have significant rates of congestive heart failure (20%) and arrhythmia (8%) as the stroke volume and heart rate increases in pregnancy may unmask previously asymptomatic mitral stenosis. Symptoms in women with mild mitral stenosis are usually not severe and can be easily managed. Women with moderate and severe mitral stenosis have the greatest risk of congestive heart failure and arrhythmia in pregnancy (Table 12.2). Heart failure in women with moderate to severe mitral stenosis is progressive and increases maternal mortality [29]. All women with moderate or severe mitral stenosis regardless of symptoms should avoid pregnancy, and valve intervention should be performed before conception. Percutaneous mitral commissurotomy is preferred for those with favorable valve morphology [29, 145].


Fetal complications in women with moderate or severe mitral stenosis include fetal demise (1–3%), preterm labor (20–30%), and fetal growth restriction (5–20%) [29].


The management of mitral stenosis in pregnancy depends on its severity. Clinical and echocardiographic follow-up is recommended every trimester and before delivery in women with mild mitral stenosis. For women with moderate-severe mitral stenosis, monthly or bimonthly clinical and echocardiographic follow-up is recommended. Transmitral gradients and the pulmonary artery systolic pressures will increase during pregnancy because of the elevated stroke volume and heart rate. Mitral valve areas estimated by planimetry or by the pressure half time are less load-dependent and can be followed throughout pregnancy. The 2011 ESC guidelines for the management of CVD in pregnancy recommend activity restriction and β1 selective antagonists in symptomatic women and in those with pulmonary pressures greater than 50 mmHg [29]. In pregnant women with moderate or severe mitral stenosis, the normal augmentation of cardiac output may be blunted due to the obstruction to left ventricle filling. Reduction in heart rate with activity restriction and beta-antagonists may improve the diastolic filling time, reduce the left atrial pressure, and increase the cardiac output. Careful titration of the beta-blocker to symptoms and heart rate is important in patients with severe mitral stenosis who effectively have a fixed obstruction to inflow and in whom the heart rate augments the necessary increase in cardiac output. Diuretics can be used in women with persistent symptoms. Anticoagulant use is indicated in women with mitral stenosis with atrial fibrillation, left atrial thrombus, or previous thromboembolic event [145, 162]. Using low-molecular-weight heparin (LMWH) or intravenous unfractionated heparin (UFH) avoids the teratogenic risk and fetopathy associated with warfarin.


Aortic Stenosis


Aortic stenosis in women of childbearing age is predominantly caused by a congenital defect in valve development, usually a BAV . Women with unicuspid aortic valves usually progress to severe valvular stenosis and require valvular repair before puberty. BAV is the most common congenital abnormality and occurs in 1–2% of the general population. Although BAV is more common in men by an estimated 2–4:1 margin, it is seen in approximately 5 in 1000 females [163]. A normal aortic valve has three semilunar valve cusps, whereas a BAV typically comprises two leaflet cusps of unequal size [164]. A genetic cause for BAV is supported by the high rate (9%) of BAV in first-degree relatives [165], familial clustering (36% of patients with BAV have multiple first-degree relatives with BAV), and the association of BAV with known genetic abnormalities such as Turner (XO), Shone, and DiGeorge syndromes. First-degree relatives of patients with a BAV should be screened for the presence of this abnormality [121, 166]. Although it may go undetected for decades, BAV can cause serious complications in more than one-third of patients [127, 128].


Aortic stenosis is the most common complication of BAV. Stenosis is caused by premature fibrosis and leaflet calcification, which is increased in cusps with asymmetry or in those in an anteroposterior position [167]. Tobacco use and abnormal lipid profiles have been associated with progression of aortic stenosis in BAV patients [168].


Before pregnancy, aortic stenosis may be asymptomatic, and the diagnosis may be unknown. The augmented cardiac output associated with pregnancy will increase both the transaortic gradient [169] and the audible systolic ejection murmur. Echocardiography is important to discriminate aortic stenosis from the flow murmur associated with pregnancy. Echocardiography will indicate the valvular pathology and associated lesions (coarctation or PDA) and provide the resting transaortic stroke volume, the peak and mean aortic gradients, and an estimated AVA by the continuity equation. The AVA and the stroke volume should be indexed to body surface area to correct for different body sizes.


Severe aortic stenosis in pregnancy is rare [170]. Patients may be asymptomatic despite having severe aortic stenosis [58]. The 2011ESC guidelines on the management of CVD in pregnancy [29] recommend exercise testing before conception to confirm exercise tolerance, provoke symptoms or arrhythmia, and assess the BP response. All patients with severe symptomatic aortic stenosis should avoid pregnancy and undergo valve replacement or valvuloplasty before conception according to both the 2011 ESC guidelines and the 2014 AHA/ACC valve disease guidelines. In asymptomatic women with severe aortic stenosis who have normal ventricular function and good exercise tolerance, the 2011 ESC guidelines do not advise against pregnancy, whereas the 2014 AHA/ACC valve disease guidelines do advise against pregnancy. Aortic surgery is recommended if the aortic root is >5 cm. Mild and moderate aortic stenosis is generally well tolerated in pregnancy [29, 171]. Severe aortic stenosis (AVA <1.0 cm2, mean aortic gradient >40 mmHg, and a peak gradient >4 m/s or 64 mmHg) increases the maternal risk of arrhythmia (3–25%) and heart failure (10%) [172]. Mortality is rare [37, 58, 171, 172].


Fetal complications of maternal moderate or severe aortic stenosis include preterm birth (28–44%), IUGR (27–33%), and low birth weight (25%) [37, 171].


Aortic Insufficiency


AI during the childbearing years is usually a consequence of BAV , endocarditis, or rheumatic valve disease. AI secondary to BAV may be associated with cusp prolapse, fibrotic retraction of the leaflets, or dilatation of the aortic root. Isolated severe AI complicating BAV [127, 173] leads to aortic valve replacement in approximately 2–6% of BAV patients during long-term follow-up. AI may be more common than aortic stenosis in younger patients with BAV who also have a greater risk of endocarditis and aortopathy [163, 174]. AI associated with rheumatic valve disease is characterized by significant mitral valve disease, and the mitral pathology dominates the clinical sequelae and risk [175].


Acute AI resulting from IE, aortic dissection, or trauma causes heart failure and low forward cardiac output. Acute AI is very poorly tolerated during and before pregnancy and requires urgent diagnosis and emergent aortic valve surgery.


Patients with chronic AI often have a long, asymptomatic phase as left ventricle compensation to the diastolic pressure and volume load leads to progressive increases in left ventricular end-diastolic volume with eccentric and concentric hypertrophy. The reduced SVR and the increased heart rate during pregnancy decrease the regurgitant volume of AI. Thus, AI from any cause, without associated enlargement of the ascending aorta (>4.5 cm), is generally well tolerated in pregnancy if left ventricular function and contractile reserve are normal [8, 32, 162]. Symptoms (even mild or transient), left ventricle size (at end systole indexed to BSA), and left ventricular systolic function are the most important predictors of complications in patients with severe AI [145, 176]. Asymptomatic women with moderate or severe AI and normal left ventricular systolic function who otherwise do not meet the criteria for valve surgery should not be referred for prophylactic aortic valve surgery before pregnancy [145, 162]. Left ventricular dysfunction increases the risk of heart failure (25%) and mortality in nonpregnant women with severe AI [176]. Mortality for symptomatic nonpregnant patients with severe AI exceeds 10% overall and is approximately 25% for patients with NYHA class III or IV symptoms [152]. Therefore, women with severe AI with symptoms or left ventricular dysfunction should be referred preferably for aortic valve repair over valve replacement before pregnancy [29]. Pregnancy in women with severe AI and LVEF <30% is not advised due to the very high risk of maternal complications [29]. Women with moderate or severe AI who become pregnant may be treated medically with diuretics and calcium channel blocking vasodilators [145]. ACE inhibitors and ARB vasodilators are contraindicated during pregnancy. Treatment of hypertension is effective in reducing the regurgitant AI volume. Cardiac surgery during pregnancy should be reserved only for women with refractory (NYHA III or IV) heart failure because of the significant fetal risk associated with surgery [145]. If the fetus is viable, cesarean delivery before aortic valve surgery is recommended [29, 177].


Mitral Regurgitation


Mitral regurgitation in women of childbearing age is usually related to myxomatous mitral valve prolapse, rheumatic valve disease, or rarely CHD. The reduction in SVR associated with pregnancy decreases the volume of mitral regurgitation during pregnancy. Asymptomatic women with variable degrees of mitral regurgitation tolerate pregnancy well if the left ventricular systolic function and pulmonary systolic pressures are normal [8, 32]. There is no evidence that severe mitral regurgitation accelerates left ventricle dysfunction in women during pregnancy [145]. Women with symptoms and severe mitral regurgitation should undergo corrective mitral surgery before pregnancy [178, 179]. Mitral repair is preferred over mitral valve replacement if the mitral anatomy is suitable to a durable repair. Pregnancy is contraindicated in women with severe mitral regurgitation and LVEF <30% or significant pulmonary hypertension [29]. During pregnancy, treatment with diuretics and calcium channel blocking vasodilators may control symptoms. Cardiac surgery is rarely required during pregnancy to treat cardiogenic shock and low cardiac forward output associated with severe mitral regurgitation.


Prosthetic Valve Replacement


Pregnancy can be successful in women with VHD after prosthetic valve replacement . Pregnancy risk is related to the type (mechanical or biologic) and position (mitral, aortic, tricuspid, or pulmonic) of the prosthetic valve, left ventricular systolic function, pulmonary arterial pressures, maternal functional class, and the presence of other associated cardiac defects. The hemodynamic and coagulation adaptations to pregnancy can lead to heart failure and valve thrombosis in susceptible patients with prosthetic valves during pregnancy. A preconception clinical evaluation with echocardiogram and electrocardiogram is strongly recommended. An assessment of baseline valve function, let ventricular systolic function, and pulmonary artery pressure allows for comparison during pregnancy [145]. Gradients are expected to increase with the increase in stroke volume and heart rate during pregnancy. Women with a prosthetic valve and LVEF <30%, significant pulmonary hypertension, or symptoms of NYHA class III or IV are classified as modified WHO risk class IV. Pregnancy is contraindicated in these patients, and pregnancy termination may be warranted.


Bioprosthetic Valves


Bioprosthetic valves offer excellent hemodynamic performance and do not require anticoagulants (AC) other than low-dose aspirin unless other thromboembolic risks (e.g., atrial fibrillation, deep vein thrombosis, valve thrombosis) are present. However, a bioprosthetic valve is less durable than a mechanical valve [145]. Bioprosthetic structural deterioration occurs more frequently in the mitral than the aortic or tricuspid position [145]. Structural deterioration also occurs earlier and more frequently in younger patients [180]. The 15-year rate of reoperation due to structural deterioration is 22% for patients 50 years old, 30% for patients 40 years old, and 50% for patients 20 years old at the time of bioprosthetic implantation [180]. Women who undergo prosthetic valve replacement in the childbearing years are expected to require a valvular reintervention (repeat valve replacement or valvotomy) during their lifetime, with an expected mortality of 0–5% depending on the valve position and timing of the procedure (emergent vs. nonemergent) [29]. Pregnancy has been reported to accelerate bioprosthetic structural deterioration [162, 181, 182], but more contemporary, larger studies have not confirmed this finding [183, 184]. Having bioprosthetic valves is considered to be a modified maternal risk class II, suggesting a small increased risk of maternal mortality or a moderate increase in morbidity [35]. Pregnancy in women with bioprosthetic valves is generally well tolerated. The data addressing the maternal and fetal risks in patients with bioprosthetic valves are limited. In a contemporary meta-analysis of 11 trials including 59 pregnancies in women with a bioprosthetic valve implanted from 1997 to 2012, no maternal deaths or thromboembolic events were reported; there were two perinatal deaths in 47 births and 14 pregnancy losses among 59 pregnancies [185]. Among 134 pregnancies in women with bioprosthetic valves who were prospectively followed in the ROPAC study, maternal mortality was 1.5%, and freedom from a serious adverse event during pregnancy did not differ from the group without a valve replacement (79% vs. 78%) [186]. Maternal risk is related to prosthetic valve function and left ventricular systolic function [187], and left ventricular dysfunction increases the risk of heart failure and arrhythmia during pregnancy.


The AHA/ACC guidelines for patients with valvular heart disease recommend low-dose aspirin (75–100 mg daily) during the second and third trimesters for pregnant patients with any type of prosthetic valve to reduce the rate of thromboembolic events [145].


Mechanical Valves


Mechanical valves offer excellent hemodynamic performance and long-term durability but require strict anticoagulation plus low-dose aspirin to prevent valve thrombosis and thromboembolic events [145]. The relative hypercoagulable state of pregnancy is associated with an increased thromboembolic risk in patients with mechanical valves [188], and the ideal anticoagulation regimen has not been determined in pregnancy [162, 170]. The risk of mechanical valve thrombosis and systemic embolization is related to valve type (ball and cage has a greater risk than tilting disc), valve size (<21 mm), valve position (mitral has a greater risk than aortic), the number of prosthetic valves (multiple has a greater risk than single), the anticoagulation regimen, atrial fibrillation, heart failure symptoms, and a history of thromboembolic events. Mechanical valve replacement is considered a modified WHO risk class III, indicating significantly increased risk of maternal mortality or severe morbidity [36]. Monthly or bimonthly cardiology and obstetric clinical follow-up with individualized and frequent anticoagulant monitoring is recommended during pregnancy [29] in patients with mechanical heart valves. Maternal complications associated with mechanical valve replacement include maternal death, valve thrombosis with valvular obstruction and systemic thromboembolism, heart failure, arrhythmia (including atrial fibrillation), hemolysis, endocarditis, and bleeding secondary to anticoagulation. Fetal risks dominate the sequelae with anticoagulation and include perinatal fetal loss and miscarriage, warfarin embryopathy, fetal hemorrhage, and small gestational weight. All anticoagulants are associated with increased fetal loss and miscarriage.


Anticoagulation During Pregnancy


Anticoagulation with continuous vitamin K antagonist (VKA) (warfarin) is the most reliable and effective approach for preventing maternal thromboembolic complications in pregnancy, with a 2–4% rate of pregnancy-related valve thromboembolic complications [189, 190].


The risk of valve thrombosis was 3.6% with oral VKA used throughout pregnancy, 9.2% with a sequential strategy of UFH during the first trimester followed by VKA in the second and third trimester, and 33% with UFH used throughout pregnancy. Maternal mortality due mostly to valve thrombosis was 2%, 4%, and 15% in these three groups, respectively [191]. UFH and LMWH are associated with greater risk of prosthetic valve thrombosis and systemic embolization than warfarin in pregnancy, but they do not cross the placenta and thus are not associated with embryopathy, significant fetal loss, or fetal hemorrhage. LMWH is preferred over UFH because of better bioavailability, more predictable anticoagulation levels, lower rates of valve thrombosis, and less bone loss, bleeding, and thrombocytopenia [192194]. The dosage of LMWH required to keep the anti-Xa levels in the therapeutic range in pregnancy is markedly elevated because of increased renal clearance and a larger volume of distribution [195]. New oral direct thrombin inhibitors or anti-Xa anticoagulants are not approved in patients with VHD and are contraindicated in patients with mechanical prosthetic valves because of the increased risk of valve thrombosis compared to warfarin in nonpregnant patients [196198].


Warfarin is associated with severe fetal complications especially when administered after 5 weeks gestation [162] and through the first trimester. Fetal embryopathy with characteristic fetal bone and cartilage anomalies (chondromalacia punctata with stippled epiphyses and nasal and limb hypoplasia) occur in 5–10% of fetuses exposed primarily during the first trimester [29, 162, 166, 189191, 193, 199].


Miscarriage due to fetal loss before 20 weeks gestation occurs in approximately 30% of cases [162, 186], whereas late fetal loss after 20 weeks gestation may occur in another 10%. Fetal hemorrhage is another devastating complication of warfarin exposure during pregnancy [186, 191, 193]. Fetal embryopathy and fetal loss secondary to warfarin appear to be dose dependent; the incidence at low doses (<5 mg daily) is 2.6 and 8% at higher doses (>5 mg daily) [200]. International normalized ratio (INR) ranges are recommended to target effective anticoagulation levels to specific patient risk. The INR target varies according to the valve site, valve type, and other risk factors for thromboembolic events (e.g., atrial fibrillation, multiple prosthetic valves, prior thromboembolism, or valve thrombosis).


In patients with mechanical bileaflet or current-generation single-tilting disc mechanical valves in the aortic position who have no additional risk factors for thromboembolism, the INR target is >2.5. However the On-X mechanical valve is approved for a target INR of 1.5 in non-pregnant patients. Whether that target pertains safely in pregnancy is unknown but it may allow for lower warfarin dosing and thus less risk of embryopathy. If a patient has additional risk factors for thromboembolism or an older generation ball and cage type AVR is present, the target INR is >3.0. The target INR for any mechanical valve in the mitral position is >3.0. Careful and frequent monitoring is important as fluctuations in the INR are associated with an increased risk of thromboembolic complications (low INR) and bleeding in patients (high INR) with mechanical heart valves. Pregnancy increases the variability in the INR because of changes in VKA drug availability, the volume of drug distribution, liver function, and food intake.


Low-dose daily aspirin is recommended in addition to VKA during the second and third trimesters to prevent valve thrombosis and thromboembolism in patients with mechanical valve replacement [145].


Treatment strategies in pregnancy are designed to minimize fetal and maternal complications of anticoagulation, although adequate randomized studies comparing different regimens are not available. In the available studies, live birth rates were the highest (92%) in an anticoagulation strategy using only LMWH throughout pregnancy, intermediate (80%) with a sequential strategy of LMWH during the first trimester followed by VKA in trimesters two and three, and lowest (65%) when VKA was used throughout pregnancy [189]. Unfortunately, maternal risk of thromboembolic complications with LMWH in pregnancy has been estimated as high as 12%, and many of these events are associated with poor dosing compliance or inadequate monitoring of anti-Xa activity [189, 192, 201].


In pregnant patients with no additional risk for thromboembolic events, a strategy that minimizes exposure to warfarin in the first trimester is recommended [145]. In women with a stable therapeutic INR on a dose of warfarin <5 mg daily, warfarin may be continued until 36 weeks gestation. If the stable warfarin dose is >5 mg daily or if the patient wants to limit fetal exposure to warfarin during the first trimester, dose-adjusted subcutaneous LMWH administered twice daily from 5 to 12 weeks is a more expensive but acceptable alternative. Monitoring of anti-Xa activity with LMWH is recommended to a target level of 1.0–1.2 units/mL for mitral valve prosthesis and 0.8–1.0 units/mL for aortic valve prosthesis at 4–6 h postdose. When LMWH is not available, another anticoagulant option is dose-adjusted, continuously administered, intravenous UFH from 5 to 12 weeks gestation. Subcutaneous administration of UFH may be offered if LMWH or inhospital continuous intravenous UFH is not available, but not all experts recommend this strategy. Monitoring of the activated partial thromboplastin time (aPTT) to a target of 2× control levels is recommended 6 h after subcutaneous UFH administration or randomly drawn if UFH is given continuously [145]. Conversion to VKA during the second trimester and up to 36 weeks gestation with careful INR monitoring is recommended. For patients at increased maternal risk of thrombosis or systemic embolization, a strategy of continuous warfarin until 36 weeks of gestation is recommended.


Peripartum Management of Anticoagulation for Mechanical Valves


For women receiving VKA up to 36 weeks of gestation, it is recommended to transition to a pre-planned delivery strategy of a shorter-acting and potentially reversible anticoagulant regimen to minimize the risks of maternal and fetal hemorrhage. The risk of valve thrombosis and systemic embolization must be balanced by the risk of obstetric hemorrhage and regional anesthesia in consultation with the care team of obstetricians, anesthesiologist, and cardiologist after discussion with the patient. VKA can be switched to subcutaneous LMWH twice daily until 12–24 h before planned delivery. Intravenous UFH can then be administered until hours before delivery. Reversal of anticoagulation from VKA with fresh-frozen plasma or intravenous prothrombin complex is an option in an obstetrical emergency. Administration of low-dose (1–2 mg) oral vitamin K may be beneficial as the effect of fresh-frozen plasma or prothrombin complex has a shorter half-life than the effects of VKA therapy [145]. Anticoagulation can be resumed postpartum after the bleeding risk has diminished.


Management of Valve Thrombosis During Pregnancy


Mechanical left-sided valve obstruction may present as a life-threatening emergency with high mortality and requires urgent treatment with either fibrinolytic therapy or surgical intervention [202207]. In patients with symptoms of new or worsening dyspnea or a systemic embolic event, transthoracic echocardiography or flouoroscopy should be followed by transesophageal echocardiogram to better visualize the mechanical valve and assess for possible IE or acute thrombosis. Fibrinolysis is the therapy of choice for right-sided prosthetic valve thrombosis [58]. High-risk features of left-sided mechanical valve thrombosis are severe symptoms (NYHA functional class III or IV) and a mobile thrombus >0.3 cm in diameter or any thrombus with an area ≥1.0 cm2. Unstable patients require fibrinolytic therapy or surgery. Surgery is usually indicated in nonpregnant patients with left-sided mechanical valve thrombosis because compared with thrombolysis, surgery has increased efficacy, less bleeding, and a reduction in embolic complications associated with a large thrombus burden. The overall 30-day mortality rate with surgery is 10–15% but <5% in patients with less severe NYHA class (I/II) symptoms [204, 207, 208]. Before 2013, the results of traditional fibrinolytic therapy showed an overall 30-day mortality rate of 7% and a hemodynamic success rate of 75%, but the thromboembolism rate was 13% and the major bleeding rate was 6% (intracerebral hemorrhage, 3%) [202207].


Fetal loss with cardiovascular surgery is high, and surgery should be reserved for patients in whom fibrinolysis failed or is contraindicated. Most fibrinolytic agents do not cross the placenta; therefore, fetal hemorrhage risk is not increased, but there is a risk of placental hemorrhage. A new strategy is echocardiogram-guided, low-dose, slow-infusion fibrinolytic treatment with 25 mg tissue-type plasminogen activator (t-PA) infused over 6 hours without a bolus; the data are promising but limited (success rates >90%, embolic event rates <2%, and major bleeding rates <2%). After the t-PA infusion is completed, UFH is administered via a 70 IU/kg bolus followed by an infusion of 16 IU/h (up to 1000 IU/h) with a target aPTT of 1.5–2.0 times the mean reference range [209]. Repeat fibrinolytic protocol (once every 24 h up to six times to a maximum total dose of 150 mg) is guided by the following indicators of fibrinolytic success: resolution of clinical symptoms and echocardiographic resolution of the increased transvalvular gradient or a reduction by >50% in the thrombus area or length [209]. The 2017 AHA/ACC focused update of the 2014 AHA/ACC guideline for the management of patients with VHD [145] suggested a role for low-dose, slow-infusion fibrinolysis for patients with mechanical valve thrombosis. Intravenous UFH is recommended in stable patients with mechanical valve thrombosis associated with subtherapeutic anticoagulation.


Cardiomyopathy in Pregnancy


Preexisting Cardiomyopathy


Heart failure during pregnancy is rare and most frequently occurs in women with preexisting cardiomyopathy (idiopathic, infectious, valvular, and cardiotoxic drug-associated [adriamycin, herceptin, and cocaine]) with decompensation due to the physiologic hemodynamic burden of increased cardiac output during the later stages of pregnancy. If cardiomyopathy was not diagnosed before pregnancy, the timing of presentation during pregnancy helps to predict the cause of the cardiomyopathy. Dilated cardiomyopathy and secondary cardiomyopathy usually present within the second trimester. Later presentation is more characteristic of peripartum cardiomyopathy (PPCM) . Women with primary or dilated cardiomyopathy present with symptoms of congestive heart failure and evidence of left ventricular systolic dysfunction without evidence of abnormal hypertrophy or VHD. Women with NYHA class II–IV symptoms and left ventricular systolic function <45% are at the greatest risk for decompensation and are thus advised to avoid pregnancy. Women with LVEF <20% are at the highest risk of maternal mortality, and pregnancy termination should be recommended.


Hypertrophic Cardiomyopathy


HCM is an autosomal dominant genetic cardiomyopathy caused by mutations in one of several sarcomere genes. HCM is characterized by left ventricular hypertrophy occurring in the absence of left ventricular hypertension (e.g., systemic hypertension, aortic stenosis pressure, aortic insufficiency, or VSD). In most patients with HCM, left ventricular hypertrophy develops during the adolescent period [210], and hypertrophy measurements do not change once early adulthood is reached. Abnormalities of diastolic function precede the development of hypertrophy and serve as phenotypic markers of HCM in at-risk family members. Symptoms of HCM are related to the pattern of hypertrophy, the presence and severity of LVOT obstruction, the severity of diastolic dysfunction, the presence and ventricular rate of arrhythmia-atrial fibrillation, and late systolic dysfunction.


Symptoms and Risk Stratification


Women with HCM may present with new-onset symptoms of congestive heart failure, arrhythmia, or an asymptomatic murmur of LVOT obstruction (augmented by Valsalva maneuver) in pregnancy. Echocardiographic evidence of otherwise unexplained hypertrophy of any pattern (diffuse, asymmetric, or apical) is diagnostic of HCM. A family history of HCM may be present but is not mandatory as genetic mutations may be sporadic. Although rare, patients with HCM have an increased risk of death from sudden cardiac death (SCD), heart failure, and stroke. SCD risk stratification is recommended in all patients with HCM. High-risk features of SCD in HCM include a history of sudden cardiac arrest or ventricular arrhythmia, severe hypertrophy (>3 cm), a family history of SCD, unexplained syncope [211], and non-sustained ventricular tachycardia (most commonly defined as ≥3 beats at 120 beats per minute), especially in patients younger than 30 years old [210, 212214]. The degree of resting LVOT obstruction as well as the age at presentation (<30 years) are also relative markers of SCD risk [215].


Women with HCM usually tolerate pregnancy well. The increased stroke volume associated with pregnancy acts to decrease the risk of LVOT obstruction by increasing left ventricular volumes. Women with symptoms before pregnancy and those with high resting LVOT gradients are at greater risk of symptomatic deterioration in pregnancy and need specialized cardiac care.


Management


Beta-blockers are useful in women with HCM to slow the maternal heart rate, blunt the inotropic response to catecholamines, increase left ventricular volumes, and decrease the risk for atrial fibrillation. Beta-blockers reduce the risk of resting and exercise-related LVOT obstruction and mitral regurgitation related to systolic anterior motion of the mitral valve. Verapamil can be substituted when beta-blockers are not well tolerated. If atrial fibrillation occurs in pregnancy and is poorly tolerated, electrical cardioversion is the preferred treatment. There is little risk to the fetus during electrical cardioversion as the amniotic fluid acts as an insulator. Maternal hemodynamic instability poses the greatest risk to the fetus. Women with high-risk ventricular arrhythmia require specialized electrophysiologic management and possible automatic implantable cardiac defibrillation placement.


Mode of Delivery


Low-risk women with HCM should be allowed to have a spontaneous labor and a normal vaginal delivery . Symptomatic or high-risk women with HCM should have a planned, controlled delivery. Pain, volume losses, and epidural anesthesia may increase the risk of LVOT obstruction.


Acquired-Peripartum Cardiomyopathy


Definition and Diagnosis


In 1971, Demakis proposed the original diagnostic criteria for PPCM, which included symptoms of heart failure within the last month or within 5 months of delivery in the absence of demonstrable heart disease or other cause for heart failure [216]. Since then, the definition of PPCM has evolved. Advances in cardiac imaging techniques have helped to demonstrate the primary cardiomyopathic etiology in PPCM, and newer definitions of PPCM require a depression of left ventricular systolic function to below 45% [217].


The timing of symptom presentation has been debated as a diagnostic criterion in PPCM. Symptoms of heart failure arise from the severity of left ventricular dysfunction and the rapidity of its decline. Because young, healthy pregnant women can accommodate a decline in cardiac function without significant symptoms, early less severe forms of PPCM may have been overlooked and thus undertreated by applying traditional criteria. More importantly, women with onset of symptoms earlier than the last month of pregnancy would be excluded from the traditional PPCM diagnosis. A comparison of risk and outcomes between women who present with pregnancy-induced cardiomyopathy early (n = 23; mean, 32 weeks gestation) versus late or the traditional definition (n = 100; mean, 38 weeks gestation) suggests that women with early cardiomyopathy may share a common etiology, risk, and prognosis when compared with women who meet the traditional definition of PPCM [218].


A broad definition of PPCM is recommended although no distinct definition is used globally. In the 2011 guidelines for management of CVD in pregnancy, the ESC recommended a definition for PPCM: symptomatic heart failure with depression of left ventricular systolic function that develops within the last months of, and up to 6 months after, pregnancy in women without known CVD [29, 219].


The diagnosis of PPCM is one of exclusion. Initial definitions of PPCM relied exclusively on clinical findings of congestive heart failure in pregnant or postpartum women who presented within the window of assumed risk. In the past, congenital cardiac defects, pericardial disease, occult valvular abnormalities, hypertensive diastolic dysfunction, and the subjectivity of symptoms led to false classifications of PPCM. In 1997, a National Heart, Lung, and Blood Institute working group added strict criteria to the diagnosis of PPCM including left ventricular dysfunction and now a LVEF <45% [220]. Noninvasive imaging with echocardiography, cardiac computed tomography, or cardiac magnetic resonance is necessary to exclude occult causes of heart failure in women in whom the hemodynamic demands of pregnancy have been superimposed on chronic cardiac disorders. These techniques can also be used to define the hemodynamics during pregnancy, including cardiac output, preload (right or left atrial pressures), and right ventricular afterload (right ventricular or pulmonary arterial systolic pressure).


Biomarkers in Peripartum Cardiomyopathy


Currently, no specific biomarkers for PPCM are available. The diagnosis of PPCM is often delayed and complicated by the overlap of symptoms of heart failure—fatigue, edema, and shortness of breath—with normal pregnancy-associated symptoms. Specific biomarkers could help distinguish PPCM patients early and expedite a quick diagnosis and initiation of treatment. N-terminal (NT)-proBNP and troponin T are nonspecific markers of structural heart disease and elevated filling pressures, and their levels are increased in hypertension/left ventricular hypertrophy and in heart failure/cardiomyopathy. NT-proBNP levels were significantly higher in 38 PPCM patients than in healthy postpartum controls [221]. However, it has not been established whether NT-proBNP levels distinguish congestive heart failure secondary to PPCM from symptoms secondary to volume overload or structural heart disease. Increased cardiac troponin T levels (>0.4 ng/mL) within 2 weeks of PPCM onset have also been shown to predict persistent left ventricular dysfunction and lower LVEF at 6-month follow-up (P < 0.001) [222]. Because increased troponin T levels are not specific for PPCM, the clinical diagnostic utility of troponin T in PPCM is not expected. Two biomarkers are potentially specific for PPCM: micro-RNA-146a and sFlt (see etiology below). Their levels may reflect mechanistic alterations in prolactin processing that have been demonstrated in women with PPCM [223] (Fig. 12.2).

../images/301959_1_En_12_Chapter/301959_1_En_12_Fig2_HTML.png

Fig. 12.2

Comparison of timing during and after pregnancy of hemodynamic changes , exemplified as cardiac output (CO; in black), elevations in prolactin and soluble Fms-like tyrosine kinase 1 (sFlt1) hormones (red), and incidence of peripartum cardiomyopathy (PPCM; blue bars). ∗Prl levels stay elevated in women who nurse. From: Arany Z, Elkayam. Peripartum cardiomyopathy. Circulation 2016;133:1397-1409. Free Access ©2016 American Heart Association, Inc.


Incidence and Associated Conditions


The incidence of PPCM in populations around the world varies greatly according to geography and socioeconomic class. Overall, the incidence of PPCM is about 1 in 3000 pregnancies [224, 225]. Ascertainment bias with reporting based on clinical symptoms alone overestimates the risk. Shortness of breath, peripheral edema, and palpitations are common nonspecific symptoms in pregnancy; therefore, milder expressions of PPCM are likely underdiagnosed. Baseline characteristics of women with PPCM, however, are remarkably similar. Women who develop PPCM are frequently older, of African ancestry, and have preeclampsia, hypertension, and multiple gestations. [218, 226228].


Age

Increasing age is strongly associated with PPCM; half of all PPCM cases occur in women over 30 years old [224, 229, 230]. Moreover, age >40 years is associated with a 10-fold increased risk of PPCM compared to age under 20 years [229, 230].


Geography and Race

The reported incidence of PPCM varies geographically [219, 231]. The incidence of PPCM in the United States varies from 1 in 968 to 1 in 4000 live births [229]. The incidence has increased from 8.5 cases per 10,000 live births in 2004 to 11.8 cases per 10,000 live births in 2011. This increase is attributed to increased awareness, access to diagnostic imaging, advanced maternal age, and multiple gestation pregnancies [229]. Japan has the lowest reported rate of PPCM at 1 in 20,000 live births, whereas Nigeria and Haiti have the highest rate at 1 in 100 live births [228, 232, 233]. Cultural rituals may contribute to the high risk of PPCM in Nigeria, where it is customary in the postpartum period to consume large amounts of salt, which promotes fluid overload and hypertension [234]. Genetic predisposition has not been well studied, but race appears to affect the risk of PPCM. The prevalence of PPCM is higher in women in Africa and women with African ancestry in Haiti and in the United States. In the United States, African American (AA) women have a higher prevalence of PPCM , a greater burden of gestational hypertension, more severe disease, and greater morbidity than white women [235, 236].


Preeclampsia and Hypertension

Hypertension, pregnancy-induced hypertension, and preeclampsia are increased in many cohorts of PPCM patients [218, 226, 237]. Historically, women with preeclampsia or eclampsia were purposefully excluded in many PPCM studies to avoid misclassification of preeclampsia-associated pulmonary edema as PPCM. Preeclampsia, but not gestational hypertension, induces subclinical abnormalities of diastolic function as measured by echocardiographic indices of myocardial strain and myocardial performance index, and these abnormalities persist after normalization of BP [238, 239]. Volume overload during pregnancy or at delivery may trigger overt congestive heart failure without systolic dysfunction. PPCM rarely occurs in women with preeclampsia or hypertension disorders of pregnancy (<10%). However, preeclampsia and hypertension in pregnancy frequently coexist in women who develop PPCM. Demakis and Rahimtoola [240], in their classic 1971 description of PPCM, reported that “toxemia,” an older term for preeclampsia, was detected in 22% of affected women. Preeclampsia has often been cited as an independent risk factor for the development of PPCM [241, 242], but not all clinical studies support this conclusion [243]. In a global meta-analysis of 22 studies involving 979 women with PPCM, the prevalence of preeclampsia was 22%, which is 4–5 times the average expected rate in the general population. The rate of hypertension during pregnancy in this large blended cohort of PPCM was 37% [226]. No geographic or racial differences were detected [244]. The findings in a US review of 535 women with PPCM in 6 states supported the association between PPCM and preeclampsia, with a 29% prevalence of preeclampsia and a 47% prevalence of hypertension [245]. In a single-center study of 75 PPCM cases, one-third were associated with preeclampsia, which is markedly more than the population rate of 3–5% [246]. These studies suggest that preeclampsia is associated with a predisposition to PPCM through a shared pathophysiologic mechanism that is independent of race and geography.


Parity

Multiparity has traditionally been considered a risk factor for PPCM [247]. However, in most US studies, PPCM developed in conjunction with the first or second pregnancy in 50% of patients [218, 248]. Therefore, these data do not support a strong association between multiparity and PPCM in the United States.


Multiple Gestations

Pregnancies associated with multiple gestations , twins or less commonly triplets, are sharply associated with a greater risk of PPCM. In the global meta-analysis of PPCM described above, the rate of twin gestation was 9%, which is three times the expected rate (3%) in women without PPCM [226]. Twin pregnancy has also been associated with a greater risk of preeclampsia, and this association may hint at an underlying placental etiology (see below) [249]. Overall, in cohort studies of women with PPCM, the frequency of multiple gestations ranges between 4% and 13% [218, 226].


Etiology


Hemodynamics

PPCM has a unique time course of symptom onset, with the peak incidence occurring within 1 month of delivery in more than 80% of cases. Normal pregnancy is associated with up to a 50% increase in cardiac output as a result of a 15% increase in heart rate, a 30% reduction in SVR, and a 15–25% increase in stroke volume. These hemodynamic accommodations to pregnancy plateau by the end of the second trimester. Patients with preexisting cardiac disease develop signs and symptoms of heart failure as the hemodynamic demands increase, usually within the second trimester [161]. Pregnancies in mothers with a preexisting cardiac structure abnormality are at even greater risk of clinical deterioration when multiple fetuses are present, given the even larger hemodynamic burden to the maternal heart. The symptoms of PPCM do not develop along this timeline, and thus the increase in hemodynamic cardiac output is unlikely to be the primary precipitant of PPCM.


Genetics

Genetics is unlikely to be the primary mechanistic cause of PPCM because most women with PPCM have no family history of PPCM or dilated cardiomyopathy. Furthermore, women with PPCM in whom left ventricular function recovers after delivery rarely develop a recurrence of clinical heart failure or PPCM with subsequent pregnancies. The increased prevalence of PPCM in women in Africa and in black women of African descent in the United States and Haiti provide clues that a genetic susceptibility may predispose a subset of women to develop PPCM. In the United States, 40% of PPCM cases occur in black women, and, in some series, the prevalence in black women is 3- to 14-fold greater than in white women [250]. Black women also have a worse prognosis, lower recovery rates, and delays in left ventricular recovery [235, 251]. A single genetic polymorphism involving the guanine nucleotide-binding proteinsβ-3 subunit (GNB3/TT) has a prevalence of 50% in black women and 10% in white women and is associated with increased rates of hypertension, decreased plasma renin activity, and abnormalities of cardiac remodeling [252255]. The Investigations of Pregnancy-Associated Cardiomyopathy (IPAC) investigators compared left ventricular recovery at 6 and 12 months in black and white women with and without the GNB3 TT polymorphism. Black women with the GNB3 variant had less left ventricular recovery than did white women with and without the variant [256].


Evidence that genetics are involved in PPCM also comes from familial clusters of PPCM and dilated cardiomyopathy and [250, 257260] female genetic carriers of the X-linked cardiomyopathies, Becker, Duchenne, and Danon, who demonstrate an increased risk of PPCM [261263]. Additionally, whole genome sequencing of 41 patients with PPCM identified a single-gene polymorphism near the PTHLH gene that may link the genomics with abnormalities of vascular homeostasis [264, 265].


Abnormalities of the genes coding for myofibril proteins, specifically the sarcomere protein, titin (TTN), have been described in two rare pedigree cohorts of patients with PPCM and familial dilated cardiomyopathy [250, 257]. Truncating variants involving the TTN gene were also found in 10% of 172 women with PPCM who were screened for high-impact nonsense, frameshift, and splicing variants of 43 genes associated with familial dilated cardiomyopathy. PPCM participants from the IPAC study [266] with the TTN variant had a lower EF at 6 and 12 months than those without the TTN variant. TTN variants were noted in both black and white women. Of note, the TTN variant was tenfold greater in nonhypertensive patients than in those with hypertension. These findings suggest that PPCM in the absence of hypertension may derive from a separate, more genetic pathophysiologic mechanism than that observed in the presence of hypertension [263]. Overall, 15% of women with PPCM and 17% of sporadic dilated cardiomyopathy patients from another cohort exhibited important similar genetic variants, further contributing to the role of genetic susceptibility in PPCM [263].


Hormonal Vascular Theory

In two seminal papers, investigators have expanded the proposed mechanism of PPCM to include a link between late gestational placental and maternal hormone secretion and vascular injury in susceptible hosts [267, 268].


Hormones secreted at the end of pregnancy—prolactin by the pituitary and a soluble variant of the vascular endothelial growth factor (VEGF) receptor 1 (soluble fms-like tyrosine kinase—sflt1) by the placenta—have potent antiangiogenic properties that can lead to endothelial cell apoptosis and a decline in myocardial vascularity in susceptible hosts. Evidence in humans that angiogenesis inhibition may induce cardiomyopathy is suggested by the cardiac dysfunction reported with the use of VEGF neutralizing antibodies in treating human cancers [269]. Two murine models of PPCM have been induced by knockout of specific myocyte transcriptional factors, STAT3 [267] and cardiac-specific deletion of proliferator-activated receptor-gamma coactivator-1α (PGC-1α) [268]. Additionally, STAT3 has been shown to be reduced in patients with end-stage dilated cardiomyopathy [270]. Through similar pathways, deletion of these nuclear transcription factors allows the overexpression of reactive oxygen species and a subsequent increase in cathepsin D [267, 268, 271].


Cathepsin D cleaves the hormone prolactin to an antiangiogenic 17 kDa prolactin fragment that has been shown to enhance the secretion of a miR146a, which, in turn, leads to endothelial apoptosis, altered energy metabolism, reduced myocardial vascularity, myocardial dysfunction, and eventual cardiomyopathy [223, 272]. Cardiomyopathy could not be provoked in nonpregnant female or male mice, suggesting a mechanistic pathway specific to the development of PPCM. However, cardiomyopathy could be provoked in the nulliparous PGC1α-deficient mice administered sFlt; this evidence suggests that late gestational antiangiogenic placental hormones can directly trigger cardiomyopathy. In the second model, depletion of PGC-1α also leads to vascular injury through the loss of a proangiogenic VEGF-mediated pathway. Evidence that a toxic late gestational hormonal milieu causes maternal cardiomyopathy is supported by the observation that hormonal blockade reverses the cardiomyopathy. Reversal of PPCM in murine models has been achieved by inhibiting prolactin secretion via bromocriptine alone in the STAT3 model [267] and with bromocriptine combined with VEGF in the PGC-1α model [268]. Partial reversal of murine PPCM (improved contractile function and partial rescue of capillary density) has been demonstrated with administration of antisense oligonucleotides to silence miRNA 146a without the suppression of lactation, which occurs as a consequence of bromocriptine treatment. Of note, circulating levels of mi R146a have been shown to be increased in women with PPCM, and levels decline in women treated with bromocriptine. SFlt1 levels, which are usually increased in women with PPCM, correlated with congestive heart failure symptoms and outcome in women with PPCM who were enrolled in the IPAC study [249, 268].


Excess placental sFlt1 secretion in women with preeclampsia [273] and multiple gestations [249] highlights the epidemiologic association between preeclampsia, twin or multiple gestations, and PPCM [274]. Removal of vasculotoxic placental hormones after delivery may also explain the rapid reversal of cardiomyopathy seen in the majority of patients with PPCM compared to other forms of cardiomyopathy. In late pregnancy, an angiogenic balance is necessary to allow the safe separation of the uteroplacental circulation while protecting maternal myocytes from vascular injury. Abnormalities in this balance in susceptible women can lead to the development of PPCM and influence its severity through the proposed hormonal vascular injury hypothesis described above.


Prognosis and Complications


PPCM confers risk to the mother and the neonate. Maternal risks include death, cardiovascular arrest, the need for heart transplantation or mechanical circulatory support, fulminant heart failure, and thromboembolic events [275277]. In a retrospective review of 535 women diagnosed with PPCM from 2003 to 2007 in the United States, 36% experienced a major maternal adverse event [229]. In the recent prospective IPAC study, which enrolled 100 US women from multiple centers and followed their clinical and echocardiographic course for 12 months, the prognosis of women with PPCM was better. Only 13% of IPAC subjects had a major event or persistent cardiomyopathy with EF <35% [251]. Overall, more than 50% of women recover completely, with relief of symptoms and recovery of LV systolic function within six months [218, 248].


Mortality

PPCM is now the leading cause of maternal death in California (causing 23% of maternal deaths) [278]. Maternal mortality estimates in PPCM vary according to race and length of follow-up; the estimates range from 3 to 28% [218, 248, 279281]. US maternal inhospital mortality rates secondary to PPCM are low at 1.3% [229], 4% at 1 year in the IPAC study [251], and 11–16% over 7–8 years of follow-up [282]. Adverse prognostic factors for maternal death include higher NYHA class [283], EF <25–30% [251], black race [281, 284], and age >30–35 years [245, 282].


Obstetric and Neonatal Outcomes

Cesarean delivery was performed for obstetrical indications in 40% of 123 PPCM patients [218]. Stillbirths are more common in mothers with PPCM, occurring in 3.8% of 535 pregnancies [229]. Of the 100 women with PPCM prospectively followed in the IPAC study, there were two stillbirths, one neonatal death, and four newborns with congenital anomalies. Mean birth weight, intrauterine growth, and Apgar scores [241] are lower in neonates born to women with PPCM and may be a consequence of the 25% rate of preterm birth (<37 weeks) seen in PPCM. Delivery decisions should involve a team of maternal fetal specialists, pediatricians, and cardiologists. Early delivery should be restricted to cases of impending maternal or fetal loss since it has not been shown to improve maternal or fetal outcomes.


Thromboembolic Events

Thromboembolic events are more common in women with PPCM than in those with idiopathic or virally mediated cardiomyopathy [228, 257, 285]. The hypercoagulable state of pregnancy may potentiate the risk of left ventricular thrombus formation in women with severe cardiomyopathy and reduced cardiac output. In one study, thromboembolic complications were the most common adverse event associated with PPCM, occurring in 6.6% of patients [229].


Left Ventricular Recovery

Myocardial recovery is greater in women with PPCM than in nonperipartum women with cardiomyopathy [228, 285]. Nevertheless, recovery of left ventricular function (EF >50%) is heterogeneous in women with PPCM [275]. Partial or complete recovery in PPCM occurs in 40–72% of affected women [223, 228, 285, 286]. In the IPAC study, only 13% of women with PPCM had severe persistent cardiomyopathy or major adverse events (death, transplant, left ventricular assist device [LVAD]) at 1 year; 15% experienced a partial recovery; and 72% recovered completely [286]. Although most who recover left ventricular function do so within 6 months, up to one-third of women may have delayed recovery [228]. Of the women who recover left ventricular function, three-quarters have an EF >45% by 2 months from presentation, suggesting that recovery occurs early in most women [223]. In a US retrospective single-center study of data from January 1986-December 2016, African American women with PPCM were younger than non-African American women with PPCM, had more advanced left ventricular dysfunction with a lower EF at presentation (39.5% vs. 56.5%, respectively), and were twice as likely to fail to recover (43.0% vs. 24.2%) [211, 236].


In two retrospective studies of predominantly African-American women [280], recovery of left ventricular function was low with only 23–30% of women achieving an EF >50% at 6 months. In a larger more contemporary study, full recovery of EF was noted in 59% of black women and 77% of white women by 1 year [286]. Recovery may be poorer in black women because of later disease presentation (>6 weeks postpartum in 50% of black women vs. 22% in white women) and a higher prevalence of hypertension (70% in black women vs. 34% in white women) [286].


Predictors of Recovery

Overall EF at presentation is the best predictor of left ventricular recovery. Determinants of poor recovery include LVEF <30% at diagnosis [286], left ventricular internal diameter end diastole (LVIDD) >5.6–6.0 cm [223, 286], late diagnosis [257, 286], presence of left ventricular thrombus [223], and black race [223, 257, 286]. In the IPAC study, no woman achieved a full recovery at 1 year when the LVIDD was >6 cm and the EF was <0.3 at presentation, whereas full recovery occurred in 91% of women in whom the LVIDD was <6 cm and the EF was >0.3 at presentation [286]. In this cohort, echocardiography was performed at baseline, 2, 6, and 12 months after presentation, and recovery was predicted in 86% of women with a baseline LVEF ≥0.30 compared with only 37% of those with an LVEF <0.30 (p < 0.001) [286]. In 187 women with PPCM, the EF at presentation predicted failure to recover EF by echocardiogram at 6 months. For those with an EF of 10–19%, 63% failed to recover (EF >50%) compared to 32% of women with an EF of 20–29% and 21% of those with an EF at presentation of >30% [227]. Failure to achieve a LVEF ≥30% was seen in 30% of patients with an EF of 10–19% and 13% of patients with an EF of 20–29% at presentation [227].


Recurrence in Subsequent Pregnancies

Since PPCM usually occurs in a first or second pregnancy, the risk of PPCM in a subsequent pregnancy is important. Relapse is greatest in women who do not recover left ventricular systolic function. In a review of 191 recurrent pregnancies in women with PPCM, the risk of relapse (decline in left ventricular function) was almost twice as great for women who failed to recover left ventricular function than in those who had full recovery (48% vs. 27%). The mortality rate in the group who failed to recover was 16%, whereas no deaths were reported in women with full left ventricular recovery [287]. In a smaller study of 28 women with subsequent pregnancy after complete recovery of left ventricular systolic function, no maternal deaths occurred. In 16 women without complete recovery who had a subsequent pregnancy, the mortality rate was 19%. Congestive heart failure developed in 44% and a decline in EF >20% was measured in 25 [288]. Reduction in contractile reserve measured by dobutamine stress echocardiography in recovered patients suggests persistent subclinical myocardial abnormalities [263]. The best predictor of left ventricular deterioration and death with recurrent pregnancy is prepregnancy LVEF. Normalization of EF, however, does not predict a risk-free subsequent pregnancy. Women with persistent left ventricular dysfunction who want a subsequent pregnancy should be advised of the grave risk and counseled to avoid pregnancy, continue standard heart failure medications, and wait for normalization of left ventricular function before becoming pregnant [287]. Women with complete recovery after PPCM in whom normal left ventricular function persists after medication weaning can be counseled that risk of maternal death is likely zero but that deterioration of LVEF may occur. Long-term outcomes in women with relapse are not known. Careful monitoring of symptoms and left ventricular function by echocardiography are strongly recommended during and after pregnancy.


Treatment


Pharmacologic Therapies

The treatment recommendation for PPCM follows the guidelines for the treatment of other causes of cardiomyopathy because evidence-based clinical data are lacking for treating heart failure during or after pregnancy. Treatment escalation is tailored individually to the severity of symptoms at presentation. The recommended treatment for symptomatic heart failure in PPCM is to optimize oxygenation via supplemental oxygen, mechanical ventilation, and very rarely venoarterial extracorporeal membrane oxygenation or even extracorporeal mechanical oxygenation. Interventions to reduce preload and optimize cardiac contractility are emphasized as afterload reduction is a natural consequence of the low resistance placental circuit in pregnancy. Standard cardiomyopathy drug treatment includes the potential use of diuretics, intravenous or oral vasodilators, intravenous inotropes, ACE inhibitors, ARBs, beta-blockers, inhibitors of mineralocorticoid activity, and digoxin. Drug treatment in patients with PPCM requires knowledge of the drug’s unique risks during pregnancy and lactation, when detrimental effects are known.


Diuretics are recommended to reduce preload-pulmonary capillary wedge pressures to relieve symptoms and to maximize oxygenation. The use of diuretics before delivery may impair placental perfusion and potentially harm the fetus; therefore, the lowest dosage possible is recommended. Improvement in stroke volume and cardiac output may require inotropic drugs, vasopressors, and rarely mechanical circulatory support. Digoxin use is safe in pregnancy, but its usefulness in the treatment of systolic heart failure is uncertain. Its use is acceptable in persistently symptomatic women during pregnancy and lactation, but levels should be checked as pregnancy increases digoxin clearance [289].


Pharmacologic antagonism of the neurohormonal axis is recommended for longer-term improvement in left ventricular contractility in PPCM patients, according to the guidelines for use in cardiomyopathy for other causes. Inhibiting the adrenergic, angiotensin, and mineralocorticoid pathways in nonpregnant women with symptomatic heart failure has been shown to reduce symptoms, improve hemodynamics and left ventricular contractility, stimulate left ventricular remodeling, and, most importantly, prolong survival.


  1. 1.

    Beta-blockers


    Controlled studies on the use of beta-blockers in PPCM have not been performed. However, beta-blockade is recommended in all women with symptomatic PPCM, but dosages should be carefully titrated to avoid short-term worsening of cardiac output. In the United States, 3 beta-blockers have been approved for treating congestive heart failure: carvedilol, bisoprolol, and metoprolol tartrate. Data on their efficacy in pregnancy and their risk to the fetus are limited.


    Antepartum use of the cardioselective β1-antagonists (metoprolol tartrate or bisoprolol) are preferred over the nonselective beta-blockers (propralolol) as they interfere less with the β2-mediated effects on uterine tone [290] which can potentially lead to uterine contractions and preterm labor. The mixed alpha and nonselective beta-antagonist, labetolol, has been used extensively antepartum in the treatment of hypertension, and carvedilol shares a similar receptor affinity. The antepartum use of atenolol has been associated with IUGR [291] and should be avoided.

    Only gold members can continue reading. Log In or Register to continue

    Stay updated, free articles. Join our Telegram channel

Apr 23, 2020 | Posted by in CARDIOLOGY | Comments Off on Disease in Pregnancy

Full access? Get Clinical Tree

Get Clinical Tree app for offline access