Chapter 16: FETAL NECK AND CHEST ANOMALIES

Fetal neck masses can have a good or a poor prognosis. Ultrasound and/or magnetic resonance imaging (MRI) can help in the differential diagnosis of fetal neck masses, thus helping to counsel parents carrying fetuses with such an anomaly. In some cases prenatal treatment is available and can result in the resolution of the neck mass. In others, when the neck mass is large, it can cause airway obstruction with potential fetal demise after delivery. In such cases, it is important to define prenatally the relationship of the neck mass to airway structures indicating the need for pretherapeutic planning of an ex-utero intrapartum treatment (EXIT) procedure, which allows getting access to the fetal airways while fetomaternal circulation is preserved in order to optimize fetal outcome.

In the immediate perinatal period, the most important keys to survival are undoubtedly the pulmonary and cardiovascular systems. Lung development is a meticulous process that starts early in pregnancy, from approximately 26 days after fertilization, with the formation of 2 buds from the ventrolateral wall of the primitive foregut. Throughout pregnancy, the lung becomes a more and more complex and highly specialized organ that allows achieving blood oxygenation in the postnatal period. To be able to do so sufficiently, the lung has to fulfill 5 different stages of maturation.¹ It is only logical that different events during those various stages will end up as a heterogeneous group of lung pathologies. Impairment of the pseudoglandular stage will affect branching or ramification of both bronchial and arterial structures of the lung.² Interference of lung development during the canalicular phase is responsible for decreased complexity of the respiratory acinus and causes impaired lung maturation.

The aim of this chapter is first to describe fetal neck anomalies, and these are divided into the anterocervical or posterocervical pathologies. Second, we describe chest anomalies that primarily relate to fetal lung development and function. Pathologies of the diaphragm, such as congenital diaphragmatic hernia (CDH), lead to an intrathoracic herniation of abdominal organs and, due to a space-occupying effect, can also interfere with normal lung development. Since the emergence of fetal therapy for severe cases of CDH, much research work has been performed in this field and is summarized in this chapter.

Fetal neck anomalies are clinically important because of their potential influence on head position during labor, and also for the possibility of airway obstruction at the time of delivery. These abnormalities can be divided into anterocervical or posterocervical pathologies.

Anterocervical Pathologies

Enlargement of the Thyroid Gland or Goiter

Fetal goiter can be associated with hyperthyroidism, hypothyroidism, or a euthyroid state. Possible causes are intrauterine exposure to antithyroid drugs, congenital metabolic disorders of thyroid synthesis, and iodine excess or deficiency.

Prenatal sonographic studies demonstrate an anterior solid and symmetric mass, which may result in hyperextension of the fetal head. Because of mechanical obstruction of the esophagus, polyhydramnios can often occur. On MRI scan, normal fetal thyroid is easily detected using a T1-weighted image (WI) (Figure 16-1), allowing differential diagnosis from other neck masses.

The prognosis depends on the basic cause of the goiter. Most cases are in women with a history of thyroid disease. Fetal blood sampling can aid in determining fetal thyroid status, especially in women suffering from Grave disease where a transplacental transfer of drugs or thyroid-stimulating antibodies may result in fetal goiter.

Maternal therapy usually corrects fetal hyperthyroidism. Direct fetal therapy in cases of fetal hypothyroidism can be undertaken by amniocentesis or by cordocentesis, and this can result in resolution of the fetal goiter.

Cervical Teratoma

Teratomas are made up of a variety of parenchymal cell types representative of more than a single germ layer. Arising from totipotential cells, these tumors typically are midline or paraxial. They range from benign, well-differentiated cystic lesions to those that are solid and malignant. The most common location of teratomas is sacrococcygeal. In newborns, cervical teratomas represent only 5% of the total.³

Sonographic features include a unilateral and well-demarcated partly solid and cystic or multiloculated mass. Calcifications are present in about 50% of cases. Cervical teratomas may be intimately associated with the thyroid gland and, in addition to tracheal compression, they may cause esophageal compression and polyhydramnios.

MRI is most useful in cases of cervical teratoma for determining the best mode of delivery. This information can be used to guide clinical discussions for the potential need of an EXIT procedure (Figure 16-2).⁴

When fetal hydrops is associated, prognosis is poor with an intrauterine or neonatal mortality rate of about 75%.⁵ This is mainly due to airway compromise associated with the lesions. For those without fetal hydrops, survival after surgery is high, but since these tumors tend to be large, extensive neck dissection and multiple additional procedures are necessary to achieve complete resection of the tumor with acceptable functional and cosmetic results.

Other rare anterocervical masses can include brachial cleft cysts and vascular anomalies such as hemangioma.⁶

Posterocervical Pathology

Second Trimester Nuchal Fold Thickness

Nuchal fold thickness is the second trimester form of nuchal translucency measured at the 11- to 13⁺⁶-week scan. In the measurement of the nuchal skin fold thickness, critical landmarks should include the cavum septum pellucidum, the cerebral peduncles, and the cerebellar hemispheres. Nuchal fold thickness is measured by placing the calipers from the outer skull table to the outer skin surface (Figure 16-3). It is considered to be pathological when the measurement is more than 6 mm. For isolated increased nuchal thickness, the risk for trisomy 21 may be 10 times the background risk. It may be associated with chromosomal defects, cardiac anomalies, infection, or genetic syndromes.^{7,8,9,10, and 11}

Encephalocele

Encephalocele is part of the large group of neural tube defects (NTDs). Whereas the group of NTDs have an incidence of about 1 per 1000 births, encephalocele accounts for about 10% of cases in this group.

The precise etiology for more than 90% of these NTDs is unknown. Chromosomal abnormalities, various genetic syndromes, such as Meckel-Gruber, von Voss, Chemke, Roberts, and Knobloch syndromes, and maternal diabetes or ingestion of teratogens, such as warfarin, are implicated. Other associated brain abnormalities that may occur in isolation or as part of genetic or nongenetic syndromes include spina bifida, callosal dysgenesis, Arnold-Chiari II malformation, Dandy-Walker malformation, and brain migrational anomalies. The most common associated chromosomal anomaly is trisomy 18.

Encephaloceles are recognized on ultrasound as cranial defects with herniated fluid-filled or brain-filled cysts (Figure 16-4). They are most commonly found in an occipital location (75% of the cases), but alternative sites include the frontoethmoidal and parietal regions. Differential diagnosis should be made with a benign epidermal scalp cyst (see Figure 16-4) in which case there is no cranial defect.¹²

The prognosis of encephalocele is inversely related to the amount of herniated cerebral tissue. Overall, the neonatal mortality can vary between 10% and 80% depending on whether pediatric or fetal series are looked at, and more than 50% of survivors have mild to severe developmental delay.^12,13

When a parent or previous sibling has had any manifestation of an NTD, the risk of recurrence is about 5%. Periconceptual supplementation of the maternal diet with folate reduces by about half the risk of developing these defects. The dose of folic acid is daily 5 mg for high-risk groups and 0.5 mg for low-risk groups.^14,15

Cystic Hygroma or Lymphangioma

Cystic hygroma or lymphangioma is a cystic lymphatic lesion that can affect any anatomic site in the human body. It usually affects the head and neck with a left-sided predilection.

Cystic hygromas are thought to arise from a combination of the following: a failure of lymphatics to connect to the venous system, abnormal budding of lymphatic tissue, and sequestered lymphatic rests that retain their embryonic growth potential. These lymphatic rests can penetrate adjacent structures or dissect along fascial planes and eventually become canalized.^16,17

Prenatal diagnosis is based on classic sonographic findings that include a complex appearance with multiple fluid-filled cysts and septae (Figure 16-5). The prognosis is often poor, especially if the lesion is associated with hydrops, which is often associated with chromosomal abnormalities such as Turner syndrome, trisomy 21, or Klippel-Trenaunay syndrome.^16,17

Fetal chest anomalies involve pathologies of the fetal lungs and diaphragm as well as those of the fetal heart and mediastinum. In this chapter, only pathologies of the fetal lungs and diaphragm are discussed.

Prenatal lung growth can be interfered with by reduction of the intrathoracic space, oligo- or anhydramnios, and/or impaired fetal breathing movements.¹⁸ Depending on the origin of the malformation, this will result in pathology-specific prognosis.

In most circumstances, ultrasonography will be sufficient to diagnose the thoracic pathology. However, MRI has been proposed as an important complementary examination.^19,20 The potential benefit of MRI for these abnormalities can mainly be explained by its superior tissue contrast, absence of acoustic shadowing, and large field of view. These characteristics provide clinicians with a better idea regarding the extent of the pathology as well as the presence of other associated structural anomalies.²¹ In this chapter, we will discuss how MRI can yield additional information in the prediction of postnatal survival for fetuses with CDH.

Pathologies of the Fetal Lungs

Congenital Cystic Adenomatoid Malformation

Congenital cystic adenomatoid malformation (CCAM) is an uncommon tumor; nonetheless, it represents the majority of all congenital lung lesions detected in utero. A commonly quoted incidence of these lesions is 1 in 25,000 to 1 in 35,000 live births, which likely underestimates their true incidence.²² The exact pathogenesis of this hamartomatous lung tumor is still unknown. The theory is that, due to a vascular insult around the sixth week of gestation, there is a cessation of distal bronchiolar structures, characterized by the lack of alveolar differentiation. The lesion is mostly limited to one single lobe (>95%) and communicates with the normal tracheobronchial tree.²³ The blood supply is maintained by branches of the normal pulmonary artery and veins. Stocker et al first classified the pathology in 3 types ranging from a single cyst to a homogeneous microcystic mass.²⁴

The prenatal sonographic appearance of CCAM will typically demonstrate a hyperechogenic pulmonary tumor, which can be either cystic (type 1), mixed (type 2), or solid—microcystic (type 3) (Figure 16-6). Microcystic disease results in uniform hyperechogenicity of the affected lung tissue. In macrocystic disease, single or multiple cystic spaces may be seen within the thorax. Both microcystic and macrocystic disease may be associated with deviation of the mediastinum, thus including the heart. In contrast, in bilateral disease, the heart may be severely compressed, although not deviated. When there is compression of the heart and major blood vessels in the thorax, fetal hydrops may develop. Polyhydramnios is a common feature, and this may be a consequence of decreased fetal swallowing of amniotic fluid due to esophageal compression, or increased fluid production by the abnormal lung tissue. It is important to search for associated anomalies, because in about 10% of the cases, CCAM is associated with other defects (eg, cardiac and renal anomalies).

Differential diagnosis is often made with CDH. Direct visualisation of the diaphragm and intraabdominal position of the stomach but also the liver can exclude the diagnosis of CDH. Moreover, the absence of peristaltic movements in the thorax is also helpful to differentiate a CCAM from a CDH. In case of doubt, differential diagnosis can be made easily using MRI.

CCAM has a broad spectrum of MRI findings, especially because this anomaly can present in 3 different ways. The presentation can range from a simple cyst to a homogeneous hyperintense solid mass adjacent to the normal lung parenchyma (Figure 16-7). Types I and III CCAM have image characteristics that can resemble a bronchogenic cyst or pulmonary sequestration, respectively. Hence, the differential diagnosis between these 2 lesions can pose a difficult challenge. The role of MRI in fetuses with CCAM lies mainly in the prognosis of the pathology, which will depend on the amount of lung hypoplasia that can be quantified using MRI. Another strength of MRI can be found in the good spatial resolution of this technique. It is known that a CCAM can diminish or even disappear during pregnancy with typical lower signal intensities on MRI.^21,25 Lower signal intensities can be a problem on sonography to localize the remaining CCAM. MRI has shown to be more accurate in detecting the residual lesion, and this information can be very important for postnatal management.²⁶

Prognosis of CCAM is variable and depends on the degree of impairment of lung development and fetal hemodynamics.²⁷ Isolated unilateral CCAM without hydrops is associated with a good prognosis. In about 60% of cases, the relative size of the fetal tumor remains stable; in 30% of cases there is antenatal shrinkage or resolution; and in 10% of cases there is progressive increase in mediastinal compression. Large intrathoracic cysts can lead to major mediastinal shift and associated hydrops. In affected patients, effective treatment can be carried out by the insertion of thoracoamniotic shunts. In the case of a large solid lesion causing hydrops, the prognosis may be improved by ultrasound-guided laser ablation of the feeding vessel.²⁸

In postnatal life, and in symptomatic neonates, thoracotomy and lobectomy are carried out and survival is about 90%. It is uncertain whether surgery is also needed for asymptomatic neonates.

Sequestration

A pulmonary sequestration represents only a minor part (6%) of congenital lung malformations detected in utero and is found in less than 1 per 50,000 births.^{29,30, and 31} It is a mass of nonfunctioning lung tissue that has no communication with the normal bronchial tree in contradiction to a CCAM. Its origin can be explained by an accessory lung bud that develops from the ventral aspect of the primitive foregut. Another important difference with a CCAM is the presence of a systemic vascular supply mainly deriving from the abdominal aorta. The lesion is mostly located in the lower lobes and can be divided according to its site. The sequestration can be extralobar (25% of cases) or intralobar (75% of cases). An extralobar sequestration has its own pleural wrapping and venous drainage as opposed to an intralobar sequestration.

Pulmonary sequestrations are sonographically detected as a hyperechogenic mass, most frequently in the posterior part of the left lung. The use of color Doppler ultrasonography to detect the feeding systemic vessel from the thoracic or abdominal aorta permits a more definitive diagnosis (Figure 16-8). Large lung sequestration may form an arteriovenous fistula that causes high-output heart failure and hydrops. Intralobar sequestrations are usually isolated, whereas more than 50% of extralobar sequestrations are associated with other abnormalities, mainly CDH and cardiac defects.^32,33

With T2-weighted MR imaging, a sequestration is specified by a solid, well-defined hyperintense mass, in the same way as a CCAM. Fortunately, visualization of a feeding systemic vessel originating from the thoracic or abdominal aorta can confirm the diagnosis of a sequestration (Figure 16-9). The role of MRI is again in the search for remaining lesions since a sequestration involutes in utero even more frequently than a CCAM.²⁵

The prognosis of a sequestration is, like a CCAM, variable and depends on the degree of impairment of lung development and fetal hemodynamics. It can also give rise to prenatal complications, ie, torsion of the sequestration with secondary pleural effusion, or postnatal complications, ie, infection or rarely an esophago-bronchopulmonary malformation.³⁴ Postnatal outcome also depends on the presence of associated abnormalities. In general, intralobar sequestration has an excellent prognosis, whereas extralobar sequestration has a poor prognosis because of the high incidence of other defects and hydrops.^31,35 In the case of a large lesion with evidence of a hyperdynamic fetal circulation, the prognosis may be improved by ultrasound-guided laser ablation of the feeding vessel.³⁶

Bronchogenic Cyst

Bronchogenic cysts, arising from abnormal budding from the ventral foregut, are seldom diagnosed in utero. These cysts are mostly isolated along the tracheobronchial tree with a predilection for the region of the carina.

A bronchogenic cyst sonographically appears as an anechogenic structure in a normal lung parenchyma.

On the MRI study, it can be easily differentiated thanks to its markedly higher signal than the surrounding lung tissue on T2-weighted imaging (Figure 16-10). However, the value of MRI in these fetuses is limited except in cases when the cyst directly obstructs pulmonary structures.

In utero bronchogenic cysts rarely give rise to symptoms. These cysts can rarely communicate with the tracheobronchial tree and may result in air trapping and/or infection of the infant.³⁷

Congenital High Airway Obstruction

Congenital high airway obstruction (CHAOS) is characterized by an obstruction of the fetal airway involving the higher airways, ie, trachea or larynx. CHAOS is an extremely rare condition, found in about 1 per 50,000 births.^38,39 The obstruction can be due to a variety of causes including cyst(s) in the larynx, malformations that close off the trachea or larynx (atresia), or a narrowing of the glottis. This causes the lungs to enlarge, the tracheobronchial tree to dilate, and eventually may cause congestive heart failure resulting in fetal hydrops and polyhydramnios.

Ultrasound images reveal a hyperechogenic, hyperplastic bilateral lung structure with possible inversion of the diaphragm convexity and, often, fetal hydrops.³⁹

On MRI scan, CHAOS is characterized by enlarged lungs with increased signal intensity on T2-weighted imaging of the gross lung parenchyma (Figure 16-11). The diaphragm will become flattened due to pulmonary expansion. Because of airway obstruction, the trachea and/or bronchi will be filled with fluid (except in the case of a tracheoesophageal fistula) and will allow good contrast with the surrounding isointense mediastinal structures. One must diligently search for associated anomalies that are present in more than 50% of these fetuses.²³

A large type III CCAM can present in a similar way as CHAOS, especially when it only affects a single bronchus. Because MRI has excellent spatial resolution it accurately characterizes this malformation with respect to its anatomical boundaries.

CHAOS may require intervention even before the baby takes a first breath. This abnormality is associated with poor neonatal outcome and increased incidence of stillbirth.^{39,40, and 41} CHAOS should be differentiated from bronchial atresia affecting only one lung (Figure 16-12).

Pleural Effusion

Pleural effusion, a result of the impairment of the equilibrium across the pleural membranes, may be unilateral or bilateral, primary (1 in 12,000 live births) or, more frequently, secondary (1 in 1500 live births).^42,43 Pleural effusion is found in about 1 per 3000 pregnancies. Unfortunately, the differentiation between primary and secondary etiologies of fetal hydrothorax is not always clear. This determination can be important since it determines the severity and prognosis of fetal hydrothorax. A thorough examination of the chest and abdomen should also exclude lesions that may be associated with pleural effusion such as pulmonary sequestration. These associated pathologies can be seen in up to 40% of fetuses with pleural effusion.⁴⁴

Fetal pleural effusion is easily diagnosed on the sonographic examination (Figure 16-13). Possible causes such as anemia, cardiac defects, aneuploidy, viral infection, and lung lesions can be ruled out by a detailed ultrasound examination as well as an amniocentesis. Ultrasound-guided pleural aspiration allows temporary drainage but also allows the aspirate to be sent for hematological, cytogenetic, and microbiological analysis. When pleural effusion reaccumulates, thoracoamniotic shunting could be indicated, allowing the pleural fluid to decompress to the level of the amniotic fluid and thus reducing the risk of pulmonary hypoplasia.⁴⁵

A pleural effusion has the appearance of fluid on MRI, being hypointense on T1-weighted images and hyperintense on T2-weighted images. The clinical value of MRI in fetuses diagnosed with hydrothorax is limited because this can be easily diagnosed using conventional sonography (Figure 16-14). However, MRI can be used to search for associated structural anomalies and for the assessment of potential lung hypoplasia.

Irrespective of the underlying cause, newborn infants with large pleural effusions can present in the neonatal period with severe, and often fatal respiratory insufficiency. This is either a direct result of pulmonary compression caused by the effusions, or due to pulmonary hypoplasia secondary to chronic intrathoracic compression. The overall mortality of neonates with pleural effusions is 25%, with a range from 0% to 27% in infants with isolated pleural effusions to about 60% to 95% in those with gross hydrops.⁴³

Isolated pleural effusions in the fetus may either resolve spontaneously or be treated effectively after birth. Nevertheless, in some cases, severe and chronic compression of the fetal lungs can result in pulmonary hypoplasia and neonatal death. In others, mediastinal compression leads to the development of hydrops and polyhydramnios, which are associated with a high risk of premature delivery and perinatal death. Attempts at prenatal therapy by repeated thoracocenteses for drainage of pleural effusions have been generally unsuccessful in reversing the hydropic state, because the fluid reaccumulates within 24 to 48 hours of drainage. A better approach is chronic drainage by the insertion of a thoracoamniotic shunt.^43,46 This is useful both for diagnosis and treatment. First, the diagnosis of an underlying cardiac abnormality or other intrathoracic lesion may become apparent only after effective decompression and return of the mediastinum to its normal position. Second, it can reverse fetal hydrops, resolve polyhydramnios, and thereby reduce the risk of preterm delivery, and may prevent pulmonary hypoplasia. Third, it may be useful in the prenatal diagnosis of pulmonary hypoplasia because, in such cases, the lungs often fail to expand after shunting. Furthermore, it may help to distinguish between hydrops due to primary accumulation of pleural effusions. Under these circumstances, the ascites and skin edema may resolve after shunting. Other causes of hydrops, such as infection, may occur in which drainage of the effusion does not prevent worsening of the hydrops. Survival after thoracoamniotic shunting is more than 90% in fetuses with isolated pleural effusions and about 50% in those with hydrops.⁴³

Congenital Diaphragmatic Hernia

Overview

Development of the diaphragm is usually completed by the ninth week of menstrual age. In the presence of a defective diaphragm, there is herniation of the abdominal viscera into the thorax at about 10 to 12 weeks, when the intestines return to the abdominal cavity from the umbilical cord. CDH is found in about 1 per 4000 births and can be divided into 4 classes: agenesis of the diaphragm, Bochdalek hernia, Morgagni hernia, or eventrations. The left-sided Bochdalek hernia occurs in approximately 90% of cases. In cases of a diaphragmatic abnormality, herniation of both small and large bowel as well as intraabdominal solid organs into the thoracic cavity can occur, reducing the intrathoracic space.

CDH is usually a sporadic abnormality. However, in about 50% of affected fetuses there are associated anomalies, structural as well as chromosomal. The most frequently associated anomalies are cardiac problems, but renal, central nervous system, and gastrointestinal anomalies are also frequently found.⁴⁷

In the case of an isolated CDH, the main problem is the variable degree of pulmonary hypoplasia that can present at birth. The lungs are characterized by a reduction of bronchial branches and less gas exchange, which is further decreased by surfactant dysfunction. In addition to parenchymal disease, increased muscularization of the intra-acinar pulmonary arteries occurs resulting in pulmonary hypertension.

Diagnosis

With the evolution of high-resolution ultrasonography, CDH can usually be diagnosed prior to birth. Absence of the diaphragm is usually indirectly suggested by the intrathoracic presence of abdominal viscera. A sagittal view of the fetal body may reveal a defect in the posterior aspect of the diaphragm, at least for the most common posterolateral (Bochdalek) type of hernia. For left-sided lesions, mediastinal shift and rightwards displacement of the heart can be seen, and in most cases a fluid-filled stomach and/or bowel are present within the thoracic cavity (Figure 16-15). An important feature to look for is the presence of a portion of the liver in the thorax. Doppler interrogation of the umbilical vein and hepatic vessels may be helpful in this respect (Figure 16-16). With right-sided lesions the right lobe of the liver usually herniates into the chest, combined with mediastinal shift to the left. In cases of equivocal sonographic findings, MRI can be very helpful in detecting the intrathoracic position of the liver (Figures 16-17 and 16-18). It is also helpful in the differential diagnosis of diaphragmatic eventration (Figure 16-19).

Prediction of Postnatal Outcome by Two- and Three-Dimensional Ultrasonography

Once the diagnosis is made, the fetal medicine specialist has a tremendous responsibility to predict the outcome of the individual case to assist parents in choosing between available options: expectant management, fetal surgery, or termination of pregnancy.⁴⁸

It has been shown that lethal pulmonary hypoplasia may be predicted on the basis of indirect assessment of lung development. Several methods to quantify this have been investigated, including two- and three-dimensional ultrasonography and fetal MRI, as well as direct or indirect assessment of flow resistance within the pulmonary circulation.

The most widely accepted method for antenatal prediction of postnatal outcome is assessment of lung size by the measurement of the so-called lung area to head circumference ratio (LHR). In this method, which was first proposed in 1996,⁴⁹ the contralateral lung area is measured by two-dimensional ultrasound in a transverse section through the fetal thorax at the 4-chamber view and expressed as a ratio over the head circumference (Figure 16-20). The rationale for expressing lung area in relation to head circumference was to correct for the effect of gestation on lung size. A series of studies have examined the effectiveness of LHR in predicting outcome and provided contradictory results, possibly because of varied methodologies in the estimation of lung area, wide range in gestational age, and small number of cases examined in each study (Table 16-1).^{49,50,51,52,53,54, and 55}

In our group, we have undertaken a series of studies that allowed validating the prediction of postnatal outcome from prenatal assessment of lung size. In a first study, 10 tertiary centers provided their data on a total of 184 consecutive fetuses with isolated, left-sided CDH that were examined at 22 to 28 weeks of gestation and subsequently managed expectantly.⁵⁶ Assessment of the LHR had been undertaken according to the method described by Metkus et al⁴⁹ in which the lung area is derived from the multiplication of the longest diameters of the contralateral lung. We used logistic regression analysis to determine the effect on survival of a multitude of variables including LHR and intrathoracic herniation of the liver, but also gestation at delivery, year of management, and the place at which the patient was managed. The only significant independent predictors of survival were provided by the LHR and presence of intrathoracic herniation of the liver.

We subsequently demonstrated that LHR is dependent on gestational age at measurement and consequently introduced a gestational-age-independent method of evaluating lung size. In fact, at between 12 and 32 weeks of gestation there is a 16-fold increase in lung area and only a 4-fold increase in head circumference.⁵⁷ Therefore, the assumption that the LHR provides a gestational-age-independent assessment of lung size is not true. In order to produce a gestational-age-independent method for the assessment of fetuses with CDH we examined 650 normal fetuses at 12 to 32 weeks of gestation and 354 fetuses with isolated CDH at 18 to 38 weeks.⁵⁸ In the normal group we expressed each measured LHR as a ratio of the appropriate (left or right) normal mean for gestation and established a normal range of observed to expected or o/e LHR, which did not change with gestation. The mean o/e LHR in the left lung was 100% (95% CI, 61% to 139%) and in the right lung was 100% (CI 67% to 133%). In fetuses with CDH the LHR increased while o/e LHR was independent of gestation. The mean o/e LHR was 39% (range 7% to 79%) and regression analysis demonstrated that significant predictors of survival were the o/e LHR, side of CDH, and gestation at delivery. We have therefore introduced a method for evaluating lung size that is validated during the whole second part of the pregnancy and no longer limited to a narrow window of 22 to 28 weeks.

Finally, we examined the value of LHR and liver herniation in the prediction of postnatal short-term morbidity. It is well known that survivors from corrected CDH may suffer from serious long-term morbidity, including chronic respiratory, feeding, hearing, and neurodevelopmental problems.^{59,60,61,62,63, and 64}

A multicenter study investigated the relationship between o/e LHR and liver position and morbidity indicators during the neonatal period.⁶⁶ We examined the data of 100 cases of isolated CDH in which the diagnosis was made prenatally, the antenatal management was expectant, and the babies were liveborn and discharged from the hospital alive. Independent significant predictors of the need for prosthetic patch repair were the o/e LHR and intrathoracic position of the liver. The incidence of gastroesophageal reflux was also related to the need for prosthetic patch repair. The o/e LHR predicted the need for postnatal assisted ventilation, supplemental O₂ at 28 days—which may increase the risk for longer-term respiratory morbidity—and the postnatal age at full enteral feeding.

In summary, in fetuses with isolated CDH, o/e LHR provides a gestation-independent method of assessment of lung size and is validated for the prediction of postnatal survival as well as the severity of the diaphragmatic defect, the functional consequences of impaired lung development, and incidence of feeding problems.

Intuitively, it seemed more logical to measure lung volume rather than assess its surface in just 1 plane. This can be done either by three-dimensional ultrasonography (Figure 16-21)^67,68 or by MRI (Figure 16-22).⁶⁹

One study compared three-dimensional ultrasonography with MRI for volumetry of both lungs in 78 fetuses with CDH.⁷⁰ These results suggest that it was not possible to accurately visualize the ipsilateral lung in about 45% of cases using three-dimensional ultrasonography. Moreover, the reproducibility of lung volume measurement was comparable between these 2 approaches for the contralateral lung volume but less for the ipsilateral lung volume using three-dimensional ultrasonography.

In a subsequent study and in 47 fetuses with CDH evaluated between 21 and 36 weeks of gestation, we found that the prediction obtained using the o/e LHR was better than the o/e contralateral lung volume measured by three-dimensional ultrasonography.⁷¹ Consequently, in fetuses with expectantly managed and isolated CDH, prenatal prediction of survival by measurement of lung volume using three-dimensional ultrasonography is typically not performed for our patient population.

Although at present the o/e LHR is the best available method of assessing the severity of the diaphragmatic hernia, the sensitivity of predicting postnatal survival from prenatal measurement of o/e LHR is only about 50% for a false-positive rate of 10%. This prediction has not improved by measurement of lung volume by three-dimensional ultrasongraphy as discussed earlier.⁷¹

Preliminary data suggest that improved prediction may be provided by Doppler assessment of the fetal pulmonary vasculature. A study of 21 fetuses with diaphragmatic hernia at 23 to 33 weeks of gestation used three-dimensional power Doppler to measure the pulmonary vascular index and reported that the index was significantly lower in those who subsequently died than in survivors with an apparent sensitivity of 100% and false-positive rate of 0%.⁷² In another study the fetal pulmonary arterial resistance before and 10 minutes after maternal hyperoxygenation was assessed in 10 fetuses with diaphragmatic hernia at 30 to 36 weeks of gestation. Hyperoxygenation resulted in a more than 20% decrease in resistance in all 6 babies that survived, whereas a decrease of resistance in less than 20% was observed in 4 cases, 2 of which survived and the other 2 died in the neonatal period due to pulmonary hypoplasia.⁷³ Although very promising, the results from the assessment of pulmonary vascularity need to be confirmed in further studies. In the meantime, the o/e LHR remains the standard method used at the ultrasound examination for predicting postnatal outcome in fetuses with isolated CDH.

Prediction of Postnatal Outcome by MRI

Since the fetal lung is primarily composed of water, it has high signal intensity on MRI, allowing for reliable morphological and volumetric evaluation. Most of the work has been dedicated to measurement of lung volume, as a relationship of what is measured in the fetus of interest (observed lung volume) in proportion to what is expected in a comparable yet normal fetus (expected lung volume), expressed in the o/e total fetal lung volume (TFLV) ratio. A number of studies have examined the potential value of TFLV as measured by MRI, in the prediction of outcome (Table 16-2)^{69,74,75,76,77, and 78}; however, so far, the number of patients examined was relatively small to draw definite conclusions.

We could show in a large multicenter series involving 148 cases of isolated CDH that o/e TFLV and intrathoracic liver position provide independent prediction of postnatal outcome.⁷⁹ In the group with intrathoracic herniation of the liver, the survival rate increased from 12% for those with o/e TFLV of 25% or less, to about 40% for o/e TFLV of 26% to 35%, 60% for o/e TFLV of 36% to 45%, and more than 70% for o/e TFLV of 46% or more. These findings therefore provide the background for future interventional studies such as fetal endoscopic tracheal occlusion (FETO).⁸⁰ Furthermore, o/e TFLV showed a trend towards a better prediction of survival as compared to the most validated two-dimensional ultrasound measurement being o/e LHR.

In some studies normative fetuses were selected based on their gestational age, but much better is to select them based on biometric variables.^75,81 Intuitively, biometric indices should be more reliable since they discount for problems such as incorrect dating or altered fetal growth. Biometric variables include abdominal circumference, liver volume, or as we recently proposed, fetal body volume (FBV) (Figure 16-23).⁸² We could show that FBV correlated the best with the predicted TFLV.⁸² The use of FBV in an algorithm may help to improve prediction of survival in fetuses with isolated CDH.⁸³

Finally, we have introduced a quantitative method for liver herniation in fetuses with CDH (Figure 16-24).⁸⁴ In fetuses with left-sided isolated CDH, liver to thoracic ratio or LiTR provides a reliable method for a volumetric quantification of intrathoracic liver using MRI. For expectantly managed fetuses, o/e TFLV and LiTR provided independent useful prediction of subsequent survival.

A more laborious step in evaluating the fetal lung is to assess the pulmonary system at a microstructural level. A leading study was done by Osada et al who evaluated the lung/spinal fluid signal intensity ratio.⁸⁵ Whereas the signal intensity of the cerebrospinal fluid remains equal during the pregnancy, it has been reported that the signal intensity of the fetal lungs increases during pregnancy. Combining the lung/spinal fluid signal intensity ratio with lung volumes was predictive in their study. Unfortunately, this study only covered 12 fetuses with CDH examined late in gestation. The lung/liver signal intensity ratio is another approach to the evaluation of pulmonary hypoplasia. Brewerton et al evaluated this ratio on 74 normal fetuses and concluded that the ratio provides a normal scale with a 95% prediction interval.⁸⁶ Unfortunately, there is some controversy in the use of this ratio since it is known that the liver intensity is not constant, in contrast to the spinal fluid, but changes with advancing pregnancy.⁸⁷

Another approach to fetal lung imaging is used in diffusion-weighted images (DWI). Diffusion refers to the movement of water molecules in the intercellular space, called brownian motion. In biologic tissue, this diffusion process is not random because of membranes, vessels, and cell structures hampering this free water motion. Therefore, the calculated net-diffusion value of water in tissue is actually an apparent diffusion, from which the term apparent diffusion coefficient (ADC) is derived. Studies have indicated an increase in ADC as the pregnancy evolves.^88,89 This could be explained by the increase of pulmonary vascularization.⁸⁸ Unfortunately, to date no reports mention evaluation of this variable in fetuses at risk for pulmonary hypoplasia such as CDH.

Spectroscopy of the lungs to detect the amount of lecithin is an even more exotic way of evaluating lung maturation. Unfortunately, this technique is hampered by fetal motion artifacts that limit its clinical use at the moment.

Fetal Surgery for CDH

Despite advances in neonatal care, fetuses with isolated CDH still die in the neonatal period mainly due to pulmonary hypoplasia and/or hypertension. Large surveys rate mortality in antenatal diagnosed isolated and liveborn cases of around 30%.^90,91 Apparent improvement in survival rate might be biased by increased termination rates in more severe cases.

Even the use of extracorporeal membrane oxygenation (ECMO) has unfortunately not proven its benefits. It has its inherent complications and is not widely available. Theoretically, ECMO might be promising when postnatal attempts are undertaken to enhance lung growth, eg, by partially filling the lung with perfluorocarbon. This might induce increased lung tissue stretch, which is currently used to trigger lung growth in the prenatal period.⁹² Large randomized studies are still awaited to prove the benefit on postnatal survival in CDH of such a technique.

Current postnatal therapy does not salvage the underlying problem in CDH, and therefore there is a rationale for prenatal interventions directed to improve lung development. Extensive animal studies have suggested that pulmonary hypoplasia and hypertension due to intrathoracic compression are reversible by in utero surgical repair.^{93,94,95,96, and 97} In a few cases of diaphragmatic hernia in the human, hysterotomy and fetal surgery have been carried out,⁹⁸ but this intervention has now been abandoned in favor of minimally invasive surgery.⁸⁰

Endoscopic obstruction of the trachea, by the placement of a balloon below the vocal cords at around 26 weeks, results in expansion of the fetal lungs by retained pulmonary secretions (Figures 16-25,16-26, and 16-27). Practically, FETO is done using a percutaneous approach under locoregional anesthesia and prophylactic tocolysis.⁹⁹ After fetal immobilization, a 1.2-mm fetoscope within a 3.0-mm sheath is inserted under ultrasound guidance directed towards the fetal mouth. Following landmarks, the fetoscope is inserted until the level of the trachea and a detachable balloon occlusion system in inflated (Figure 16-28). Patients are typically admitted for 48 hours, and prenatal removal of the balloon is scheduled at 34 weeks, done either by fetal tracheoscopy or by puncturing the balloon using an ultrasound-guided 20-gauge needle.

Preliminary data suggest that this treatment is associated with substantial improvement in survival, and randomized controlled trials to validate this technique are at the moment under way.

Investigators are exploring the use of an algorithm for the prenatal prediction of postnatal survival in fetuses affected by congenital diaphragmatic hernia. Such an algorithm will include ultrasound markers such as Doppler assessment of the pulmonary vessels as well as markers on magnetic resonance imaging such as lung volume corrected for the fetal body volume and quantification of the intrathoracic liver, but also assessment of the lung at the microstructural level using diffusion-weighted images and T2 intensity.

Highlighted References