sonographic appearance may be cystic, solid, or mixed, and may demonstrate irregular echogenic

patterns secondary to areas of tumor necrosis, cystic degeneration, internal hemorrhage, or

calcification. Ultrasound is important to assess abdominal or pelvic extension, evidence of bowel or

urinary tract obstruction, and the integrity of the fetal spine, and to document lower extremity function.

Fetal MRI is helpful to delineate the anatomy and may be useful to distinguish these lesions from MMC,

meconium pseudocyst, and obstructive uropathy. Fetal echocardiography and Doppler ultrasound

measurements are important in the diagnostic assessment and follow-up of fetuses with these

conditions. Increased aortic velocity, increased combined cardiac output, increased cardiac-to-thoracic

ratio, a dilated inferior vena cava, or reversed end-diastolic umbilical blood flow appear to be sensitive

early predictors of impending hydrops and fetal demise.8,54 For fetuses greater than 28 weeks’

gestation, these findings are an indication for emergency cesarean delivery and postnatal resection. For

fetuses less than 28 weeks, open fetal surgery and resection is the treatment of choice. Minimally

invasive approaches to interrupt the tumor’s blood supply with radiofrequency ablation have had

limited success

55 but may be alternatives for a fetus that is a poor candidate for open fetal resection.

Adzick et al.56 reported the first successful outcome for fetal surgery resection of a fetus with a large

SCT and hydrops in 1997. More recently, investigators from that center have reported 75% survival

(three of four) for fetal resection of SCT. A similar high-output heart failure physiology can occur with

teratomas in other locations. A fetus with a cervical teratoma and hydrops had successful fetal resection

of the lesion at 24 weeks’ gestation.57

Congenital Diaphragmatic Hernia

At present, the role of fetal intervention in the management of fetuses with congenital diaphragmatic

hernia (CDH) is not clear. A hole in the posterolateral diaphragm, left sided in 80% to 85% of cases,

permits abdominal viscera to herniate into the chest during fetal development, resulting in pulmonary

hypoplasia and pulmonary hypertension. In current practice, more than half of the infants with CDH are

diagnosed before birth. At times it can be difficult to distinguish CDH from a cystic lung mass or other

cystic lesions. Fetal MRI is useful to enhance diagnostic accuracy, confirm liver position, calculate lung

volumes, and further exclude associated anomalies.6,58

Outcome for patients with CDH varies widely, depending on the severity of disease when the disease

is diagnosed.59 The overall survival of CDH in the United States is 68%.60 Although the survival has

improved in selected centers,61–64 the morbidity of some survivors remains high.65 Fetuses with CDH

have lower survival rates than those for live-born infants or for infants presenting to a neonatal surgical

center. This paradigm has been termed the “hidden mortality” by Harrison et al.66 and has been

confirmed by multiple investigators.67–69 In 2000, Dillon et al.70 reviewed the outcomes from a large

series of fetuses with CDH registered in the Northern Region Congenital Abnormality Survey in the

United Kingdom. Between 1985 and 1997, 201 fetuses were evaluated for diaphragmatic problems, of

which 187 had congenital diaphragmatic hernia (CDH) (14 had diaphragm eventration). From this

cohort, 38 pregnancies were terminated, 26 of which had “multiple abnormalities,” and 14 pregnancies

(7%) were complicated by spontaneous miscarriage or stillbirth. The overall 1-year survival was 37%,

but for live-born fetuses the survival was 50%. Furthermore, for those live-born babies with isolated

CDH, the overall survival was 59%.

Ultrasound (and more recently MRI) can be used to predict fetal CDH severity. Herniation of the liver

into the fetal chest and marked lung hypoplasia as estimated by a lung-to-head ratio (LHR) less than 1.0

are strong predictors of perinatal death.71,72 Fetuses without liver herniation through the diaphragm

have a reported survival rate between 75% and 93%, whereas those with “liver up” have a survival rate

as low as 43%.73 In a report that included evaluation of 174 patients with CDH, an LHR less than 0.9

was associated with only 13% survival, whereas survival was 68% for an LHR 0.9 to 1.2 and 88% for an

LHR greater than 1.2.

For selected infants with severe CDH, fetal surgery has been performed in an attempt to reverse the

high mortality rate. Initially, complete in utero repair of the diaphragmatic defect was attempted.74 This

approach, however, led to fetal bradycardia and immediate death after attempts to reduce the liver

from the fetal chest resulted in kinking of the ductus venosus.75 Subsequently, in utero tracheal

occlusion has been used to treat fetuses with severe CDH. It has been shown that temporary tracheal

occlusion, performed either as an open procedure or fetoscopically, can prevent the normal egress of

lung fluid and enhance growth of the fetal lungs.76 In a review of 15 patients treated by open fetal

tracheal occlusion at Children’s Hospital of Philadelphia (CHOP), survival was 100% for patients with

right-sided defects (two of two) and 23% for patients with left-sided defects (3 of 13), compared with

2885

0% survival for a matched group of seven infants with left-sided lesions managed by conventional

treatment alone. Endoscopic approaches to fetal tracheal occlusion, termed “fetoscopic” or “FETENDO,”

showed more promising outcomes.77 However, Harrison et al.17 from UCSF reported results from a

National Institutes of Health (NIH)-sponsored, prospective randomized trial that compared fetoscopic

tracheal occlusion with standard postnatal care for fetuses with left-sided CDH, “liver up,” and an LHR

less than 1.4. Twenty-four fetuses were randomized, but at this point the trial was stopped because of

an unexpected high survival in the standard treatment group and the expectation that fetal therapy in

this design would show no benefit. In this study of fetuses with liver-up CDH, the overall 90-day

survival rate was 75%, a number much higher than those reported previously. Whereas 8 of 11 (73%)

fetuses survived following fetoscopic tracheal occlusion, 10 of 13 (77%) survived with postnatal

treatment alone, thus showing no benefit to fetal therapy in this study. As expected, a striking

difference in the gestational ages was noted between the two groups. Whereas the standard treatment

group delivered at a mean gestation of 37 weeks, the tracheal occlusion group delivered at a mean of

30.8 (range of 28 to 34) weeks’ gestational age. Although fetoscopic tracheal occlusion likely had an

effect on lung growth, this response may have been limited by preterm labor, early delivery, and

associated lung immaturity. As has been demonstrated in the treatment of other conditions,

chorioamniotic membrane separation and preterm labor limit outcomes from fetal surgery.20,21 The lack

of benefit from fetal therapy may also result from the excellent outcomes that were achieved in the

control group. These results certainly suggest that a standardized protocol involving prenatal steroids

and expert pre- and postnatal care may improve outcome for fetuses with severe CDH.

Deprest reported results from the use of fetoscopic tracheal occlusion (FETO) in 210 patients. The

procedure was performed at a median gestational age of 27 weeks. Although spontaneous preterm

prelabor rupture of membranes occurred in 47% of patients, the median gestational age at delivery was

35.3 weeks. The authors estimated that in fetuses with left CDH treated with FETO, the survival rate

increased from 24% to 49%.24 The results from this trial are promising but still limited because of a lack

of a randomized design and lack of standardized, centralized postnatal care in the control group of

fetuses.

6 It is likely that fetal intervention may play a role in the management of a small cohort of fetuses

with the most severe form of CDH. Current work is directed at accurate identification of this cohort and

optimizing minimally invasive treatment approaches.

Fetal Surgery at the End of Pregnancy

Ex Utero Intrapartum Treatment Procedure

7 As an outgrowth of the fetal intervention efforts, the EXIT procedure was devised to treat fetal airway

obstruction caused by large neck masses or intrinsic airway problems.78 This approach involves a

planned hysterotomy and delivery with preservation of the maternal–fetal placental circulation for

oxygenation of the fetus. While on “placental bypass,” up to 2 hours is available for direct

laryngoscopy, bronchoscopy, endotracheal intubation, tumor resection, or tracheostomy to secure the

airway (Fig. 100-5). The umbilical cord is then divided, and the fetus delivered. Indications for an EXIT

procedure include the presence of a large cervical mass with evidence of airway deviation or

compression. Polyhydramnios, caused by esophagus compression, is one marker for fetal tracheal

compression. Recently, the tracheoesophageal displacement index (TEDI) has been described to assess

the degree of tracheal deviation on fetal MRI.79 Multiple centers have used the EXIT procedure to

stabilize the airway in fetuses with large cervical tumors (e.g., teratomas or lymphatic malformations)

with excellent fetal outcomes and very little additional maternal risk (Fig. 100-5).80–82

2886

Figure 100-5. A 36-week fetus with a giant right neck teratoma at ex utero intrapartum treatment (EXIT) procedure. Due to

significant tracheal deviation, a tracheostomy was performed on placental support.

An EXIT procedure is the treatment of choice for fetuses with congenital high airway obstruction

syndrome (CHAOS). This syndrome is usually caused by laryngeal or tracheal atresia, tracheal stenosis,

or a mucosal web. A fetus with this condition develops overdistended, echogenic lungs, which may

compress the mediastinum, flatten or evert the diaphragm, and cause hydrops because of compromise of

venous return. Fetuses with CHAOS should be delivered with an EXIT procedure to permit tracheostomy

and airway control while maintaining the maternal–fetal placental circulation. For fetuses with CHAOS

who develop hydrops, early delivery or prenatal tracheostomy are treatment options, depending on

gestational age.83–85

EX UTERO INTRAPARTUM TREATMENT–EXTRACORPOREAL

MEMBRANE OXYGENATION

The EXIT–ECMO procedure has been developed to treat anticipated respiratory failure at birth. The

rationale for this procedure is to transition from a stable intrauterine environment to ECMO while

avoiding hypoxemia, barotrauma, hemodynamic instability, and acidosis. This strategy has been applied

to severe CDH. Kunisaki et al. reported the largest series of 14 patients who had EXIT–ECMO for

treatment of severe CDH.86 Inclusion criteria included liver herniation, an LHR less than 1.4, percentage

of predicted lung volume by fetal MRI less than 15, and/or congenital heart disease. Overall survival

was 64% for this severe subset of CDH patients. Although this therapy is promising, at this time its

efficacy is unproven. Variations of this procedure include EXIT-resection for lung lesions,87 EXIT–ECMO

for fetal thoracic masses,88 and EXIT-resection–ECMO for giant chest masses (Fig. 100-6).89

Myelomeningocele

MMC was the first nonlethal anomaly to be treated by fetal surgery.90–92 MMC, or spina bifida, is a

midline defect of the spine and spinal cord that leads to exposure of the contents of the neural canal. It

is relatively common, and occurs in 1 of every 2,000 live births. The natural history of fetuses with this

condition is variable, but generally mortality is low. Instead, affected newborns suffer lifelong

disabilities, which include a combination of paraplegia, abnormalities in bowel or bladder control,

sexual dysfunction, skeletal deformations, hydrocephalus, and mental impairment. More than 80% of

fetuses with MMC can be identified before birth by maternal serum α-fetoprotein screening. Fetal

ultrasound may detect the characteristic spine abnormalities as early as 16 weeks’ gestation.

2887

Figure 100-6. A 37-week fetus with severe congenital diaphragmatic hernia (CDH) delivered using the ex utero intrapartum

treatment (EXIT)–extracorporeal membrane oxygenation (ECMO) procedure. The airway was secured and ECMO cannulas were

placed on placental support.

The rationale for fetal intervention for MMC is based on compelling evidence that associated

neurologic deficits do not simply result from incomplete neurulation but rather from chronic mechanical

and chemical trauma caused by exposure of the neural tissue to the amniotic environment. Support for

this conclusion includes the observation that fetal leg movements decrease with advancing gestational

age. Furthermore, in animal models of MMC, in utero repair leads to improved neurologic function and

reversal of hindbrain herniation.93

Human fetal interventions for MMC were initially designed to reduce intrauterine exposure of neural

elements. Both fetoscopic and open techniques were attempted; however, open methods are considered

the current standard. The first successful outcome of fetal surgery repair of MMC was reported by

Adzick et al. in 1998.90 Since that time multiple centers have reported positive outcomes that suggest

that surgical repair of MMC before 25 weeks’ gestation can preserve neurologic function, reverse the

hindbrain herniation, decrease the incidence of clubfoot, and obviate the need for postnatal placement

of a ventriculoperitoneal shunt compared with historical controls.91,92,94 No consistent improvement was

seen in neurologic or bladder function, however, and these interventions did come at the cost of fetal

prematurity (preterm labor) and increased maternal risk. Furthermore, lack of accurate prenatal

indicators of neurologic function and absence of matched controls and long-term follow-up hamper

evaluation of this approach.

Urinary Obstruction

Obstructive uropathy, one of the most common fetal structural anomalies, occurs in about 1 of 1,000

live births. Of all fetuses with urinary tract dilatation, as many as 90% do not require fetal intervention,

such as those with low-pressure dilation, continued good urine output, adequate amniotic fluid volume,

unilateral urinary obstruction, or advanced irreversible renal dysplasia.95,96 Fetuses with urethral

obstruction, severe bilateral hydronephrosis, and preserved renal function may be candidates for fetal

intervention. Urethral obstruction, usually caused by posterior urethral valves in a male fetus, leads to

decreased fetal urine output, oligohydramnios, and eventually pulmonary hypoplasia, which is often the

most important factor affecting postnatal outcome. Fetuses with oligohydramnios present in the early

second trimester have mortality rates in excess of 90%.97,98

Prenatal ultrasound is very accurate in the detection of fetal hydronephrosis and in determining the

level of urinary obstruction. Because most of the amniotic fluid in middle and late pregnancy is the

product of fetal urination, the presence of a normal amniotic fluid index implies the excretion of urine

from at least one kidney. Decreasing amniotic fluid volume on serial ultrasound usually indicates

deteriorating renal function. Furthermore, renal function can be assessed by the sonographic appearance

of the renal parenchyma and by the laboratory analysis of fetal urine via percutaneous bladder

aspiration. The presence of cortical cysts or increased echogenicity is highly predictive of renal

dysplasia, but the absence of these findings does not exclude it.99 Normal fetal urine chemistry includes

a urinary sodium less than 100 mEq/dL, chloride less than 90 mEq/dL, osmolarity less than 200

mOsm/L, and β2

-microglobulin less than 4 mg/dL. Values greater than these suggest that the fetal

kidney is unable- to reabsorb these molecules and predict poor postnatal renal function.

The greatest challenge is selecting fetuses that have severe urinary obstruction yet reversible or

2888

salvageable renal function. Presently, selection criteria for fetal intervention are a male fetus less than

30 weeks’ gestation with evolving oligohydramnios, normal renal function, and no associated

anomalies. Fetuses older than this are best treated by postnatal approaches. If there is evidence of lung

maturation, then early delivery is recommended. The goal of fetal therapy is to adequately drain the

urinary obstruction and to restore normal amniotic fluid volume. Methods of urinary tract

decompression include percutaneous vesicoamniotic shunt placement, fetoscopic or open vesicostomy,

and fetoscopic fulguration of posterior urethral valves.100,101 Open fetal surgery, including vesicostomy

or ureterostomy, has resulted in 50% survival, with two of eight children in one series having normal

renal function.102 Today, the most widely used and accepted means of treating bladder outlet

obstruction is percutaneous insertion of a double-J vesicoamniotic shunt. Placement of a vesicoamniotic

shunt is associated with 50% survival, but shunt complication rates as high as 45% may result. Direct

fetoscopic ablation of posterior urethral valves holds promise for the future, though currently limited

data are available to assess selection criteria and outcomes for fetuses treated by this approach.

Twin-to-Twin Transfusion Syndrome

TTTS, present in 5% to 35% of monochorionic twin pregnancies, occurs when there is unequal sharing

of the monochorionic placenta. Generally, this finding can be made on prenatal ultrasound, and often

the communicating placental vessels can be characterized by Doppler. Arteriovenous anastomoses lead

to a net shunting of blood from the donor to recipient. The donor twin often suffers intrauterine growth

retardation, cardiac failure, and oligohydramnios, whereas the recipient twin may suffer

polyhydramnios, cardiomyopathy, and fetal hydrops. The smaller twin’s placental cord insertion is often

marginal or velamentous, whereas the larger twin’s cord inserts into the placenta centrally.103 Quintero

et al.104 have described a staging system for this condition. In stage 1, polyhydramnios and

oligohydramnios occur, but the donor bladder is still visible. In stage 2, the bladder is no longer visible,

and in stage 3 abnormal Doppler signals are detected in the respective umbilical arteries or veins. Stage

4 is characterized by findings of hydrops, and stage 5 is fetal death. In addition to the Quintero staging,

recent studies suggest that comprehensive cardiac assessment by echocardiography may improve patient

risk stratification.9 Overall, TTTS before 26 weeks’ gestation is associated with a high rate of fetal loss

and perinatal death and a high incidence of brain damage in survivors.105,106 In severe forms of TTTS,

perinatal mortality rates of up to 90% have been reported.107

Treatment options include aminoreduction of the recipient sac, amniotic membrane septostomy, and

fetoscopic ablation of communicating vessels.108–110 Recently, endoscopic laser coagulation has become

the treatment of choice for severe TTTS diagnosed before 26 weeks’ gestation. In a multicenter

European trial, twins with severe TTTS of all Quintero stages were randomized to undergo fetoscopic

laser ablation or standard amnioreduction therapy. The trial was stopped after 142 women were

randomized, because of the clear benefit of the laser approach. Endoscopic laser ablation led to

improved outcomes in all variables examined, including increased fetal survival and decreased incidence

of neurologic complications.105

Congenital Heart Defects

Fetal intervention may play a role in the management of some fetuses with rare congenital heart

defects, including those with severe aortic or pulmonary stenosis, and hypoplastic left heart syndrome

with intact or highly restrictive atrial septum.111–113 There is evidence that fetuses with aortic or mitral

valve stenosis may progress to hypoplastic left heart syndrome, presumably as a result of decreased

ventricular filling. Furthermore, critical pulmonary stenosis or atresia with intact ventricular septum

may progress to the so-called “hypoplastic right heart syndrome.” Staged palliative surgery is the only

therapeutic option for newborns with these conditions, and mortality remains significant. The goals of

fetal intervention for these anomalies are to relieve the obstruction, restore normal cardiac physiology,

and reverse the progression toward ventricular hypoplasia. In an initial report from investigators at

Children’s Hospital, Boston, in which 20 fetuses underwent balloon dilation of a stenotic aortic valve at

21 to 29 weeks’ gestation, five fetuses died following the procedure.113 Of 14 who were thought to have

a technically successful procedure, 3 had evidence of a two-ventricle heart postnatally, thus supporting

the experimental rationale. More recently, Tworetzky reported the predictors of technical success and

postnatal biventricular outcome after in utero aortic valvuloplasty in 70 fetuses. The technical success

rate was 74%, and 17 had biventricular circulation postnatally. The authors developed a multivariable

threshold scoring system allowing highly sensitive and moderately specific identification of fetuses able

to survive postnatally with a biventricular circulation.114 Tworetzky also reported the outcomes of in

2889