Novel approach to prenatal predictors of outcomes for fetuses with severe Ebstein anomaly

Abstract Objective Ebstein anomaly (EA) is a cardiac malformation with highly variable presentation and severity with limited perinatal management options. We present incorporation of fetal lung measurements into a multidisciplinary evaluation for counseling and predicting postnatal outcomes in patients with severe EA. Methods Five fetuses with severe fetal EA were reviewed. Third trimester sonographic observed/expected total lung area (O/E TLA) and lung to head ratio (O/E LHR), fetal MRI total fetal lung volume ratio (O/E-TFLV), echocardiographic cardio-thoracic ratio (CT ratio), sonographic estimated fetal weight (EFW) by Hadlock formula and presence of hydrops, were used to guide perinatal management. Results Three of five had appropriate fetal growth, were delivered at term in a cardiac operative suite, and underwent immediate intervention with good neonatal outcomes. Two had severe fetal growth restriction (FGR), CT ratios > 0.8 and O/E LHR and TLA < 25%. One of which delivered prematurely with neonatal demise and one suffered in utero demise at 34 weeks. Conclusions FGR, hydrops, increased CT ratio and reduced O/E LHR and TFLV are potential prognosticators of poor outcomes in severe EA, and should be validated in larger cohorts that would allow for a statistical analysis of the predictive utility of these measurements. KEY POINTS Pulmonary hypoplasia is associated with severe morbidity There are limited prognosticating tools to risk stratify and guide management in cases of severe prenatal Ebstein anomalies Fetal MRI may improve prognostication for fetuses with EA


Introduction
Ebstein anomaly (EA) is a rare congenital cardiac disease characterized by malformation of the tricuspid valve (TV) and myopathy of the right ventricle (RV).The estimated incidence of EA in the general population is approximately 1 in 200,000 livebirths, and is unrelated to gender [1].Most cases of EA appear to be idiopathic.
Failure of delamination of TV leaflets from the underlying RV myocardium is the characteristic feature of EA.The TV is made up of three leaflets; the anterior, posterior/inferior and septal leaflets.In EA, the anterior leaflet is usually attached to the anatomic tricuspid valve annulus at the atrioventricular junction, with variable tethering to the underlying RV endocardium.The "annular" attachments of the septal and posterior/ inferior leaflets are displaced into the RV toward the apex and the RV outflow tract (RVOT) [2].
The morphology of EA, and consequently the clinical presentation is highly variable, ranging from severe forms with in utero hydrops and fetal demise to asymptomatic adults [3].Earlier diagnosis usually carries a worse prognosis [4].Early onset, severe disease can be associated with severe cardiomegaly, respiratory failure, cyanosis, pulmonary hypertension, and arrhythmias.The identification of fetuses with extremely high risk for perinatal mortality from pulmonary artery hypoplasia, pulmonary atresia, and/or massive cardiomegaly is critical to developing an appropriate postnatal management plan.Fetal echocardiogram is currently the mainstay for imaging modality to predict outcomes [5][6][7].However, studies have suggested pulmonary hypoplasia in severe EA and the use of prenatal MRI for assessment of fetal lung volumes may result in improved understanding of the anomaly and guide management for a successful postnatal surgical plan [8].
In our practice, prenatal cardio-thoracic ratio (CT ratio) measured from fetal echocardiogram;fetal lung observed/expected (O/E) areas measured from obstetric ultrasound; and fetal magnetic resonance imaging (MRI) O/E lung volumes have been used to assess for abnormal lung findings in cases of severe EA, which impacted immediate postnatal intervention plans.We based our hypothesis considering previous studies that estimate the O/E total fetal lung volume (TFLV) on fetal MRI, and O/E lung to head area ratio (LHR) on ultrasound, to successfully predict the degree of postnatal pulmonary hypoplasia in fetuses with congenital diaphragmatic hernia (CDH) [9][10][11][12].While echocardiography parameters were incorporated in counseling, they are outside the scope of this discussion.

Materials and methods
This is a case series of five fetuses with prenatally diagnosed EA who received care at our institution between March 2016 and March 2020.These five consecutive patients represent some of the most severe forms in the spectrum of prenatal EA.This study was reviewed and approved by the institutional review board (ID: 20-003937) and this study was deemed exempt due to no more than minimal risk to the study population.All patients had provided consent to participate in research.
All patients received multidisciplinary consultations, diagnostic imaging, and perinatal care at our institution.All five had thorough obstetric and cardiac ultrasound evaluations, and prenatal MRI results were available for calculation of lung volumes in 3 of 5 cases.Four out of the 5 cases in this cohort delivered and received postnatal care at our institution; 1 case delivered at another institution due to obstetric complications necessitating urgent preterm delivery.None of the patients had a history of lithium use in pregnancy.
EA was diagnosed on detailed obstetric (OB) anatomy ultrasound when failure of complete TV delamination and apical displacement of the proximal attachments of the TV leaflets from the atrioventricular valve ring were seen [13].All cases with suspected EA on OB anatomy ultrasound underwent a fetal echocardiogram for confirmation and further refinement of the diagnosis.A CT ratio was measured on transverse plane of the chest where ribs, thorax, and four-chamber view of heart was seen, with no stomach or abdominal contents visible in the image [13].In OB anatomy ultrasound, CT ratio is measured as cardiac "circumference" and expressed as a ratio of thoracic "circumference" (normal <0.5) -referred as OB CT ratio in this paper.In echocardiography imaging, CT ratio is measured as cardiac "area," expressed as a ratio of thoracic "area" (normal <0.35) [14]-referred to as CV CT ratio in this paper.In both instances, measurements were obtained during diastole and all were over-read and interpreted by a single provider for the analyses reported in this paper.
Sonographic fetal lung and head areas were measured from third trimester OB ultrasounds and classified as O/E total lung area (TLA) and O/E LHR.The observed TLAs were measured using the tracing technique on the same plane used for CT measurements (Figure 1).The expected TLAs were derived from reference ranges for 2-dimensional sonographic lung measurements in healthy fetuses [15].The O/E TLA was a ratio of the observed and expected values.The O/E LHR measurements were calculated using the TOTAL trial established lung volume calculator for CDH fetuses [16].
The 3 fetal MRI scans were performed within the third trimester (>28 weeks gestation) of pregnancy to obtain measurements of O/E-TFLV.Single-shot fast spin echo in the axial and coronal planes was performed and repeated until motion free sequences are obtained, as this is critical to determine accurate lung volumes.Manual region of interest tracings were performed of the fetal lung tissue, with care to exclude the heart, thymus, and mediastinal structures including the central pulmonary vessels (Figure 2).The areas of each slice were added and multiplied by slice thickness to obtain the observed total fetal lung volume (TFLV).The expected TFLV was calculated using the Rypens formula, which takes into account gestational age [17].These numbers then provided the O/E TFLV as a ratio or percentage.
Standard counseling with Maternal Fetal Medicine and Pediatric Cardiology was provided at the time of diagnosis or the time of the initial referral to our facility, usually in the mid-second trimester.Once the fetus reached the third trimester, and short term natural history was established and diagnostic imaging workup were complete, patients underwent further multidisciplinary evaluation and consultation with the following services: Maternal Fetal Medicine, Genetics, Pediatric Cardiology, Cardiothoracic Surgery, Neonatology, Obstetric Anesthesiology and Social Work.The patients were counseled on management options and expected outcomes based on the clinical picture, diagnostic imaging results and patient desires.The mode and frequency of antepartum surveillance, the timing of delivery, location of delivery, peripartum care, immediate postnatal care, ability to offer resuscitation, and the need for postnatal surgical intervention were discussed at these meetings.Pediatric Cardiology and Cardiothoracic Surgery utilized the estimated cardiac size and lung volumes to predict degree of respiratory difficulties and alter postnatal surgical planning accordingly.
Clinical information that was used to guide these decisions and prognosticate outcomes included: (1) Gestational age at diagnosis; (2) Estimated fetal weight (EFW) within 4 weeks of delivery; (3) Presence or absence of fetal growth restriction (FGR), defined as EFW <10 th percentile for gestational age (4) Doppler studies of umbilical artery (UA) in those with FGR, classified as antegrade, absent or reversed diastolic flow, (5) CT circumference ratio measured on obstetric ultrasound (OB CT ratio) (6) CT area ratio measured on fetal echocardiography (CV CT ratio) (7) O/E TFLV using Rypen's formula, measured on fetal MRI after 28 weeks gestation; (8) O/E TLA measured on OB ultrasound after 28 weeks gestation (9) O/E LHR measured on OB ultrasound after 28 weeks gestation (10) Presence of fetal hydrops on obstetric ultrasound (defined as accumulation of fluid in two or more fetal compartments, including ascites, pleural effusion, pericardial effusion, or skin edema); (11) Presence of anatomic or functional pulmonary atresia; (12) Presence and amount of pulmonary regurgitation; indicating risk for the circular shunt and (13) Presence/absence and significance of fetal cardiac arrhythmias.
In a publication by Freud,et al. [3], evaluating the outcomes and predictors of perinatal mortality in fetuses with Ebstein anomaly or tricuspid valve dysplasia, several other echocardiographic parameters were included.Our list of parameters was selected using this reference, and condensed to those that would provide objective information, while facilitating feasible interpretation and application of the data by both the perinatologist as well as the pediatric cardiologist.

Patient 1
A 34 year-old healthy Gravida 2, Para 1-0-0-2 at 25 1/ 7 weeks of gestation was referred for prenatally diagnosed severe EA.Surgical history was significant for one prior Cesarean Delivery (CD) for twin breech pregnancy.
A detailed anatomy ultrasound at 25 1/7 weeks gestation revealed a single live intrauterine pregnancy with cardiomegaly, OB CT ratio 0.66, intermittent arrhythmia, apical displacement of the TV with large regurgitant jet and dilated right atrium consistent with EA.OB US showed EFW was 900 gms (54 th percentile), normal amniotic fluid volume (AFV) and no other structural abnormalities.Amniocentesis performed for genetic studies was normal.
Fetal echocardiogram confirmed diagnosis of EA with severe TV regurgitation, CV CT ratio 0.55, no pericardial effusion, normal pulmonary valve annulus diameter with antegrade flow and no regurgitation (annulus Z score ¼ −2).There was severe RA and moderate RV enlargement, LV size and biventricular functions were normal.Fetal lung area measurements by OB sonography at 31 0/7 weeks gestation revealed O/ E TLA of 0.41, and O/E Total LHR of 1.05.Fetal MRI was offered and declined.
The fetus developed atrial flutter and retrograde ductal perfusion of the pulmonary circulation in the third trimester.Fetal atrial flutter was treated with digoxin and sotalol.The patient was admitted to the hospital at 34 1/7 weeks gestation due to concerns for evolving hydrops with new findings of ascites and polyhydramnios; EFW remained appropriate for gestational age.OB CT ratio was 0.64 and CV CT ratio was 0.52; stable.The fetus remained in atrial flutter, although heart rate control was eventually achieved, with rates decreasing from 170 to 100 bpm.
A repeat CD was performed at 37 2/7 weeks gestation in the cardiac suite due to anticipated need for cardiac intervention due to refractory arrythmia.Delivery occurred of a live born female weighing 3360 gms, apgars were 9 and 9 at 1 and 5 min respectively.The baby was hemodynamically stable at birth, with findings of severe EA, severe TR, and persistent atrial flutter.Her pulmonary valve was severely hypoplastic (Z score −3.4) and anatomically atretric at the time of delivery.No forward or regurgitant flow was ever demonstrated post-natally.She underwent immediate cardioversion and initiation of IV inotropic and antiarrhythmia support.Three days after initial stabilization, an intracardiac repair (tricuspid valve repair, subtotal ASD closure and RV-PA homograft conduit reconstruction) was performed.She required short-term postoperative ECMO support but otherwise had an uneventful recovery and was discharged home.She has had a subsequent conduit replacement and continues to clinically do well at 4 years of age.

Patient 2
A 26-year-old healthy Gravida 1 Para 0 female at 30 weeks gestation was referred for prenatally diagnosed severe EA.A detailed anatomy ultrasound at 30 weeks revealed cardiomegaly with OB CT ratio of 0.9, abnormal four-chamber heart; apically displaced TV and severely enlarged right atrium.EFW was 1154 gms, (<3 rd percentile) consistent with severe FGR; UA Doppler studies showed antegrade diastolic flow.There was normal AFV, FHR was 140 bpm with sinus rhythm and all other fetal anatomic findings were normal.A fetal echocardiogram performed on the same day confirmed the diagnosis of EA.The details of the echocardiogram are included in Table 1, but importantly included massive right heart enlargement, CV CT ratio − 0.8, severe TR with a severely tethered septal leaflet and anatomic pulmonary atresia.Routine genetic testing was offered and declined.However, given the aforementioned constellation of findings, an underlying genetic cause cannot be ruled out.Sonographic lung measurements revealed O/E TLA of 0.07 and O/E Total LHR of 0.16.Fetal MRI at 32 weeks revealed TFLV of only 8.4 ml, and O/E TFLV of 12%, consistent with extreme pulmonary hypoplasia.
The patient was offered comfort care measures for the neonate due to projected dismal prognosis secondary to severe EA with extreme pulmonary hypoplasia and severe FGR.The neonate would not meet eligibility criteria for ECMO due to an EFW less than 2000 gms and expected inability to wean off from the mechanical circulatory support due to pulmonary hypoplasia.
The patient was admitted at 34 5/7 weeks gestation with hydrops and intrauterine fetal demise (IUFD).She underwent induction of labor and delivered a stillborn male infant weighing 1471 gms.The rest of her postpartum course was otherwise uncomplicated.She was offered bereavement support and resources through this very difficult pregnancy and was discharged home in a stable condition.

Patient 3
A 29 year-old Gravid 6 Para 3-0-2-3 female at 31 6/ 7 weeks gestation was referred for prenatally   diagnosed severe EA.Her past medical history was significant for mild intermittent asthma and hypothyroidism, treated with levothyroxine.She had no significant surgical history.
A detailed anatomy ultrasound revealed a singleton intrauterine pregnancy.The fetus had cardiomegaly with OB CT ratio of 0.72 and cardiac anatomy consistent with EA.The EFW was 1422 gms, (<3 rd percentile) consistent with severe FGR; UA Doppler studies revealed antegrade diastolic flow with elevated systolic/diastolic ratio.There was oligohydramnios with a maximum vertical pocket (MVP) of 1.2 cm which contributed to suboptimal imaging of the rest of fetal anatomy.Cell-free DNA testing was low risk for Trisomy 13, 18 and 21.Fetal echocardiogram performed on the same day revealed cardiomegaly with CV CT ratio of 0.62, severe right heart enlargement, severe TR, functional pulmonary atresia with mild pulmonary regurgitation.Details of the echocardiographic findings can be found in Table 1.FHR was 140 bpm with normal sinus rhythm and no evidence of fetal hydrops.Fetal sonographic lung area measurements revealed an O/E TLA of 0.26 and O/E Total LHR of 0.74.
The patient was offered a primary CD with a readily available cardiac operating room to enable rapid resuscitation, ventilation, provision of intravenous prostaglandins, and potentially a right atrial reduction procedure within the first hours of life.The need for surgical intervention would depend on lung compliance, ventilation characteristics and cardiovascular physiology once the neonate transitioned to postnatal circulation.The team anticipated that ductal patency would be required to maintain pulmonary blood flow in the setting of functional pulmonary atresia, at least early after delivery.The decision between either a single ventricle approach, or a biventricular repair would be based on post-natal findings and clinical course.
At 33 0/7 weeks gestation, the patient presented to her "local" university medical center for interim OB follow up.The EFW was 1113 gms (<3 rd percentile), with new findings of absent diastolic flow on UA Doppler studies, a poor biophysical profile (BPP) 4/8, new pericardial effusion; and an increased OB CT ratio to 0.8.She was admitted to the hospital there for betamethasone administration for fetal lung maturation and was offered nonintervention on behalf of the fetus due to persistent category II fetal heart tracing and extremely poor prognosis.However, she desired full resuscitation and elected to proceed with a CD to avoid further fetal distress and to optimize resuscitative efforts.Delivery occurred at her local hospital at 33 1/7 weeks gestation with a liveborn male infant weighing 1315 gms.Apgars were 4 and 6 at 1 and 5 min respectively.Of note, fetal MRI was never performed due to the delivery occurring prior to the scheduled MRI date.We do not have postnatal records available, but the mother communicated with us that her neonate died at DOL 14.

Patient 4
A 28 year-old healthy Gravida 3 Para 2-0-0-2 female at 24 2/7 weeks of gestation was referred for prenatally diagnosed severe EA.She had no significant medical or surgical history.A detailed anatomy ultrasound at 24 2/7 weeks gestation revealed a singleton pregnancy with no additional abnormalities.Cardiac findings included cardiomegaly with OB CT ratio of 0.78, displaced TV, atrialization of RV, and intermittent arrhythmias.Amniotic fluid volume was normal and there was no evidence of other structural abnormalities or hydrops.Genetic testing was offered and declined.Fetal echocardiogram revealed cardiomegaly with a CV CT ratio of 0.6, severe right heart enlargement, severe TR, and pulmonary valve atresia (likely anatomic atresia, Z score ¼ −5.0).Further details regarding the echocardiogram can be found in Table 1.FHR was 148 bpm with predominantly sinus rhythm (occasional non-conducted premature atrial complexes).
Fetal lung area measurements at 35 1/7 weeks gestation revealed O/E TLA of 0.32, and O/E Total LHR of 0.85.At 38 weeks gestation, the patient returned for immediate prenatal reevaluation.Appropriate interval fetal growth was noted, with EFW 3193 gms, (47 th percentile), OB CT ratio was stable at 0.76, CV CT ratio was 0.65 and there was no evidence of hydrops.Fetal MRI was performed with a measured lung volume of 84.6 ml and a total O/E TFLV of 78%.
The patient was offered full resuscitation with a planned CD in a cardiac operative suite at 39 weeks gestation due to planned atrial reduction immediately in postnatal period.At 39 2/7 weeks gestation; she delivered a live-born male infant weighing 3780 gms with Apgars 8 and 8 at 1 and 5 min, respectively.Postnatal transthoracic echocardiography (TTE) confirmed severe EA/TR with severe RA enlargement and pulmonary atresia.The newborn was started on prostaglandins, intubated and temporarily stabilized.However, he develop supraventricular tachycardia and became hemodynamically unstable.He then underwent cardioversion, right atrial reduction surgery and initiation of ECMO support via median sternotomy.In anticipation of separating from ECMO, he then underwent ductus arteriosus stent placement on DOL2.The patient subsequently underwent fenestrated patch closure of the tricuspid valve at the time of separation from ECMO on DOL4.

Patient 5
A 27 year-old healthy Gravida 2, Para 1-0-0-1 at 32 0/ 7 weeks of gestation with was referred for prenatally diagnosed severe EA.A detailed anatomy ultrasound at 32 0/7 weeks gestation revealed a single live intrauterine pregnancy with cardiomegaly -OB CT ratio of 0.71 and cardiac anatomy consistent with EA and subjectively adequate lung tissue.The EFW was 1905 gms (44 th percentile), AFV was normal and no other structural abnormalities were identified.Routine prenatal genetic testing was offered and the couple elected to collect cord blood for chromosomal microarray at the time of delivery, which was normal.
Fetal echo performed on the same day revealed CV CT ratio of 0.6, marked abnormalities of the tricuspid valve (TV) with only a single long, thin, redundant anterior leaflet and no other mobile TV leaflets seen, severely dysfunctional RV myocardium, and pulmonary atresia (likely to be anatomic atresia, Z score ¼ −5.4); all consistent with very severe EA.Further details are included in Table 1.FHR was 148bpm with normal sinus rhythm, no evidence of pericardial effusion or hydrops.Fetal lung area measurements revealed O/E TLA of 0.27, and O/E Total LHR of 0.60.Follow-up imaging at 36 3/7 weeks gestation showed adequate fetal growth with EFW 2836 gms, (43 rd percentile), OB CT ratio of 0.79 and CV CT ratio of 0.66.Fetal MRI 37 0/7 weeks gestation revealed lung volume of 63 ml and O/E TFLV of 63%.
The patient was offered a CD in a cardiac suite at 39 weeks gestation.The lung compliance, ventilation characteristics, and cardiovascular physiology would be assessed as the newborn transitioned to postnatal circulation; prostaglandin would be administered to maintain the ductus arteriosus as a source of pulmonary blood flow.A right atrial reduction was planned on DOL0 in the setting of adequate lung ventilation; and a palliative procedure was to be offered if there was poor lung compliance.
However, secondary to premature labor she underwent a primary CD at 37 4/7 in the obstetrics unit.The patient delivered a live-born female infant weighing 2920 gms, with Apgars 8 and 9 at 1 and 5 min, respectively.
Postnatal echocardiography demonstrated EA with an atrialized RV and no discrete TV; there was anatomic pulmonary atresia and a large PDA.The newborn was started on prostaglandin infusion and underwent right atrial reduction via median sternotomy on DOL0 as planned.The baby was stabilized with central ECMO and underwent PDA stent placement on DOL4, followed by Starnes procedure on DOL5 and separation from ECMO on DOL7.

Overall results
Table 1 summarizes the information from these 5 cases and Table 2 summarizes prognosticators of poor outcomes that were identified through this case series.
The fetus in case 2 had the worst outcome with IUFD at 34 5/7 weeks secondary to severe cardiomegaly (OB CT ratio 0.9, CV CT ratio 0.8), severe pulmonary hypoplasia (O/E TFLV 12%, O/E TLA of 0.07 and O/E Total LHR of 0.16), severe FGR with EFW <3 rd percentile and presence of hydrops.
The fetus in case 3 had slightly less cardiomegaly and slightly larger lungs, with OB and CV CT ratios of 0.72 and 0.62 respectively.The O/E TLA was 0.26 and O/E Total LHR was 0.74.However, the cardiomegaly was progressive (the OB CT ratio increased to 0.8 over just 2 weeks).The fetus developed severe FGR and hydrops leading to an obstetrically indicated preterm delivery at 34 0/7 weeks at an outside facility before an MRI was performed.The neonate did not survive, passing away at DOL 14.
The surviving fetuses, cases 1, 4 and 5, all had severe EA, but less cardiac enlargement than cases 2 and 3. Third trimester OB CT ratios were between 0.64 to 0.79 and CV CT ratios between 0.52 to 0.65.OB and CV CT ratios remained less than 0.8 and 0.7 respectively throughout their gestations.All 3 showed appropriate fetal growth throughout pregnancy.They did not develop hydrops, in the absence of sustained tachycardia (atrial flutter in case 1).The O/E TLA and O/E total LHR were all > 0.25.The two cases with available MRI data had O/E TFLV between 63-78%.They delivered at term (>37 0/7 weeks gestation) in a cardiac suite and had immediate cardiac support/interventions available.Cases 4 and 5 underwent RA reduction and required ECMO support.All 3 fetuses have subsequently had additional, more definitive cardiac surgical procedures and are alive and doing well.

Principal findings
This series of 5 patients includes only individuals with severe forms of prenatally diagnosed fetal EA, and yet their cases still led to varying outcomes.We used a combination of clinical and radiologic features THE JOURNAL OF MATERNAL-FETAL & NEONATAL MEDICINE including OB and CV CT ratio, sonographic O/E TLA and O/E Total LHR, and when available; O/E TFLV measured on third trimester fetal MRI, to help predict the severity of cardiomegaly and pulmonary hypoplasia, to better predict perinatal outcomes, guide the management plan and counsel the families.
While we were unable to determine MRI O/E TFLV cut offs for suspected severe pulmonary based on our small case series of 5 patients, it provides guidance to some information that can be obtained to guide prenatal counseling.Table 2 summarizes these variables in the 5 cases.
The presence of FGR with hydrops, the increase in OB CT ratio over a short time, low O/E TLA and O/E Total LHR, and preterm delivery could be significant contributing factors to the poor outcome.Delivery at term in a tertiary care center with access to immediate postnatal interventions and cardiac support can contribute to a significantly improved outcome.Some infants may also require ECMO at some point in the postnatal period; suggesting that this may need to be considered early in postoperative management for fetuses with severe EA that require neonatal surgical intervention.

Literature review and comparison
EA accounts for 1% of all congenital heart disease [18].Those that present with progressive cardiomegaly, right-sided enlargement and associated TR in the prenatal period are considered severe and within the subset of these severe cases; the presentation and outcomes can still be highly variable.All fetuses with EA carry the risk of fetal hydrops from fetal heart failure and severe pulmonary hypoplasia from pulmonary compression and poor lung development [19].
To date, most studies evaluating independent predictors of mortality in prenatally diagnosed severe EA focus on cardiac parameters including CT ratio, TV and RVOT diameters, and biventricular function.Amongst these, the largest multicenter study by Freud, et al. including 243 fetuses from 23 institutions from 2005 to 2011 reported pulmonary regurgitation with circular shunt physiology to be a high risk for worse outcomes; along with earlier gestational age at diagnosis < 32 weeks, presence of pericardial effusion and large TV annulus [3].The presence of circular shunt physiology or pulmonary atresia increases morbidity, as these fetuses will require immediate postnatal prostaglandin infusion to support pulmonary circulation; and are more likely to require immediate surgical intervention.Though this study included 10 cases of fetuses that underwent prenatal fetal MRI for fetal lung volumes; the measurements and analysis from the MRI scans and use of that data to predict outcomes were not reported.Holst, et al. performed a multi-institutional analysis using The Society of Thoracic Surgeons Congenital Heart Surgery Database (STS-CHSD) looking specifically at the surgical management and outcomes of EA in neonates and infants from 2010 to 2016 [20].This large study included 255 neonates and found the major preoperative factor resulting in a need for surgery was preoperative ventilation, accounting for 61.6% of patients; the most common operation performed was Ebstein repair (39.6%), followed by a systemic-to-pulmonary shunt (20.4%); and the overall neonatal mortality was reported at 27.4% [20].This data suggest that both cardiac and pulmonary parameters should be objectively evaluated to better predict outcomes in cases of severe EA.However, as previously mentioned most studies have evaluated cardiac parameters with very limited information on pulmonary parameters.
MRI has been extensively studied in the measurement of fetal lung volumes in fetuses with congenital diaphragmatic hernia (CDH) [9].MRI offers superior soft tissue contrast and spatial resolution, providing excellent delineation of the lungs and mediastinal structures compared to ultrasound.MRI also offers the advantage of multiplanar imaging which is less affected by maternal body habitus and fetal lie, compared to ultrasound.MRI of the fetus is highly sensitive to fetal movement and motion-free imaging is critical to determining accurate observed total fetal lung volumes using MRI.There are two methods of determining the expected TFLV; calculation by Rypen's formula using gestation age, and tables determined from large data sets of normal fetuses throughout gestation.Studies have validated the accuracy, reproducibility and reliability of MRI measured fetal lung volumes in fetuses with CDH.It has been suggested that fetuses with CDH and O/E TLV <35% were associated with a worse outcome due to severe pulmonary hypoplasia, at all gestational ages [11,12,[21][22][23][24][25].More evaluation will be needed to determine if the CT ratios are concordant with lung volume assessments or if CT ratio may fail to identify certain fetuses at higher risk of pulmonary compromise that may be better predicted with volume assessment.

Clinical implications
Outcomes in cases of severe EA are reliant on good prenatal and postnatal planning and institution experience in the care of these fetuses.Our case series highlights how the integration of cardiac and pulmonary assessments, with the use of OB and CV CT ratios, O/E TLA, O/E LHR and O/E TFLV measurements, may be valuable in predicting prognosis for and guiding the management of fetuses with severe EA.
Normal OB CT ratio is <0.50 and CV CT ratio is < 0.35 [14]; experience from our case series suggests that an increased OB CT ratio or CV CT ratio, especially when associated with FGR, carries a poor prognosis.
In regards to pulmonary assessment, we are intrigued by the additive value of fetal MRI, but our small sample size limits our ability to outline quantitative cutoffs, and target gestational age for performing fetal MRI evaluations.In CDH patients, MRI measured fetal lung volumes to prognosticate outcomes have been associated with worse outcomes if O/E TFLV <35% at all gestational ages [26].However, unlike CDH; severe EA is a progressive disease and perhaps the use of MRI in the third trimester (>28 weeks gestation) may be a more reliable predictor of outcomes.We cannot define a risk "cutpoint" based on these three cases.Certainly, the two cases with O/E TFLV > 50% had no difficulties with initial pulmonary function and the one fetus with the largest heart and O/E TFLV of 12% had the worst outcome (intrauterine demise before 35 weeks).Based on the physiology of disease progression, we hypothesize that the degree of bronchopulmonary disruption in EA may not be as severe as in CHD and occurs more gradually, thus the preliminary cutoff point for poor prognosis is likely to be higher.A larger sample size will be required to more reliably establish such cutoffs.

Research implications
Though severe EA represents a rare congenital cardiac anomaly, it carries high morbidity and mortality for the fetus.Larger multicenter trials are needed to validate the efficacy of our proposed assessments and to establish evidence-based cut off values that will translate into clinical practice.

Limitations
This report is limited by its small sample size, lack of standardization of timing of fetal MRI; and delivery occurring at varying gestational ages which could result in confounding bias from prematurity and birth weight.

Conclusion
We propose the integrated use of OB, cardiac and prenatal MRI scans to assess fetal lung areas/volumes and fetal cardiac sizes, and multidisciplinary assessments to risk stratify and manage fetuses with severe EA.Such a coordinated evaluation should allow us to better prognosticate outcomes and create individualized perinatal management plans for these complex fetuses.These diagnostic tests are already available in the tertiary centers likely to be providing care to these fetuses; making it feasible to quickly adopt these strategies into clinical practice and allow further collaborative investigations required to refine the approach.

Figure 1 .
Figure 1.Two-dimensional ultrasound axial view of the fetal chest at four-chambers level showing the area tracing measurements of the fetal right lung (RL) and fetal left lung (LL).

Figure 2 .
Figure 2. Fetal magnetic resonance imaging (MRI) showing an axial view of the fetal chest at the four-chambers view of the fetal heart with the measurements of the fetal right lung (R) and fetal left lung (L).

Table 2 .
Summary of prognosticators of poor outcomes.