Congenital pulmonary airway malformation (CPAM) (Fig. 1) affects between 1/8,000 and 1/35,000 livebirths1.
CPAM results from the disordered development of the lower respiratory tract. The current theory of pathogenesis favors a maturational arrest in bronchopulmonary development, resulting in dysplastic tissue distal to this segment1,2,3. CPAM appears to consist of two major subtypes: type I consists of lesions due to an arrest at the pseudoglandular stage of development (5-17 weeks post-conception); and type II lesions that occur during the saccular period (3rd trimester)5. The aberrant cysts in CPAM have intracystic communication4.
Figure 1. Left-sided CPAM (arrow); heart compressed against ribs on the right. Normal right lung behind the heart.

Figure 2. Left-sided CPAM (straight arrow) with hydrops (curved arrow).
Stocker5 initially subdivided congenital cystic adenomatoid malformation (CCAM) into three subtypes based upon postnatal cystic size and histology: type I, cysts between 30-100 mm; type II, cysts < 10 mm (Fig. 2); type III, microcystic. When Stocker added two additional types to his categories of congenital cystic adenomatoid malformation (type O, acinar dysplasia and type IV, peripheral), he suggested a change in the name of this lesion from CCAM to CPAM6. This nomenclature has, generally, carried over to the prenatal diagnosis of CPAM. The classification of Adzick et al7 based on the size of cysts > 5 mm and < 5 mm differs from this pathologic criteria.
CPAMs are evenly distributed between the left and right lung. Lesions are unilobar in 80% to 95% of cases5,8. CPAM may affect more than one lobe, but bilateral lesions are rare9,10.
Because of its abnormal development, the CPAM portion of the lung does not assist with gas exchange.
Microcystic lesions tend to have a rapid growth phase between 20 and 26 weeks’ gestation. The volume of the lesion then plateaus and may regress11,12,13. Microcystic lesions have an increased risk to develop hydrops. Microcystic CPAMs may become isoechoic with adjacent normal lung between 30-34 weeks’ gestation. Although they may not be detectable sonographically, these lesions remain visible on MRI.
Macrocystic lesions do not have a characteristic mid-pregnancy growth spurt and generally do not regress in volume as pregnancy advances.
Since the defect in CPAM is at the bronchiolar level, the vascular supply to the normal and abnormal lung segments is usually unaffected.
Large microcystic or macrocystic lesions have the potential to compress the esophagus and inferior vena cava. Polyhydramnios may result from impairment of fetal swallowing14. When intrathoracic pressure is sufficient to impair venous return to the heart, hydrops (Fig.3) may result15. In the presence of incipient or frank hydrops, fetal intervention may be considered16. Compression of the contralateral lung by a large CPAM can result in pulmonary hypoplasia4,13.
Figure 3. CPAM. Complete involvement right lung (straight arrow) with secondary hydrops (scalp edema [curved arrow] and ascites [short arrow]).

Figure 4. Right-sided CPAM (arrow); heart compressed against ribs on the left.
Increasing mediastinal shift as gestation advances is associated with decreased survival. A severe shift has been defined as complete displacement of the heart across the thorax with no visible contralateral lung (Fig. 4)11.
The degree of lung involvement is associated with survival – involvement of a single lobe suggests a better outcome than if the entire lung is involved14. Bilateral lung involvement has a poor prognosis17.
CPAM is characteristically diagnosed after 16 weeks’ gestation14. The ability to correctly diagnosis lung lesions by prenatal ultrasound is limited10. Achiron et al18 reported four cases of fetal hyperechogenic lung with subsequent resolution. In one case, retained mucoid lung secretions was confirmed as the etiology for the echogenic lobe.
In cases of segmental lung echogenicity without a mediastinal shift and signs of hydrops, there is a possibility of spontaneous resolution.
Differential Diagnosis
The differential diagnosis for CPAM is outlined in Table I.
Table I. Differential diagnosis of an echogenic chest mass55.
- Bronchopulmonary sequestration
- Bronchial atresia/stenosis
- Lobar emphysema
- Lobar mucus plug
- Neuroblastoma
- Mediastinal teratoma
- Congenital diaphragmatic hernia
Bronchopulmonary Sequestration
An echogenic lung mass, generally in the left lower lobe, with a systemic arterial supply is characteristic of a pulmonary sequestration. However, hybrid lesions with features of both congenital pulmonary airway malformation and pulmonary sequestration may occur17,19,20.
Bronchial Atresia
Bronchial atresia is a rare echogenic lung lesion with a prevalence of 1/100,00021. It is due to either a malformed dysplastic section of the main stem, lobar or segmental bronchus, or a bronchial atresia secondary to a traumatic fetal event22. As would be expected, there is a poor perinatal prognosis with main stem, in contrast to lobar or segmental bronchial atresia. Bronchial atresia and CPAM frequently co-exist23.
A diagnosis of bronchial atresia is rarely made antenatally.
The characteristic sonographic finding with bronchial atresia is a dilated tubular structure within an echogenic lung mass. Three-dimensional sonography is helpful in attempting to delineate a central hypoechoic structure within an echogenic lung mass as a bronchus24. The occlusion of the bronchus leads to accumulation of lung fluid distal to the obstruction. The subsequent enlargement of the lung results in an eversion of the diaphragm, a contralateral medialstinal shift and compression of the normal lung, resulting in pulmonary hypoplasia.
In the neonate with mild bronchial atresia, CT is the imaging modality of choice. Characteristic findings include: occlusion of a bronchus central to a mucocele and emphysematous changes in the peripheral lung fields25.
Congenital Lobar Emphysema
Lobar emphysema (Fig. 5) is generally a postnatal diagnosis due to an obstruction that permits inflation, but not deflation, of a lung segment. An intrinsic obstruction is due to defective bronchial cartilage or a mucus plug. An extrinsic mass (teratoma, CPAM) compressing a bronchus may produce the same effect. While seldom diagnosed in-utero, the neonatal incidence of congenital lobar emphysema is the same as for pulmonary sequestration and it is only slightly less frequent than CPAM26.
Figure 5. Congenital lobar emphysema (arrow).
The increased lung echogenicity with congenital lobar emphysema is due an excessive accumulation of lung fluid in the alveoli. A sufficient amount of pressure may build up behind the obstruction to permit the fluid to escape and thereby reduce the echogenicity of the lung27.
Frequently, a postnatal diagnosis, including histology, is required to clearly distinguish between the above diagnoses. It has, therefore, been suggested that the generic term – Congenital Lung Lesion – be used prenatally. Many, if not all, of the lung malformations may represent a continuum with airway obstruction as the primary etiology28. Specific descriptions of the lung mass would include its architecture (cystic or solid), location, size, presence or absence of a medialstinal shift, and signs of hydrops20,29.
Adrenal Neuroblastoma
Neuroblastomas are the most common solid tumor in children under one year, with an incidence of 1/17,000 infants30. A family history has been identified in 1-2% of cases31,32.
In-utero, neuroblastomas are characteristically diagnosed in the 3rd trimester. However, a neuroblastoma has been diagnosed as early as 20 weeks’ gestation32.
90% of neuroblastomas identified prenatally are adrenal in origin32.
When a large suprarenal mass is first detected, CPAM is generally the first diagnosis that is considered. Sonographically, a neuroblastoma may be cystic, solid (Fig. 6) or mixed cystic/solid in appearance32.
Figure 6. Neuroblastoma (arrow).
Serial ultrasound examinations every two weeks should be obtained in order to monitor the size of the mass and to assess the fetus for possible metastases, specifically to the placenta33.
Cystic neuroblastomas generally have a better prognosis34. When the diagnosis is in doubt fetal MRI is helpful35.
Mediastinal Teratoma
Congenital teratomas occur in approximately 1/30,000 births, with 40% located in the sacrococcygeal area; 7% of cases occur in the mediastinum36,37.
Sonographically, mediastinal teratomas are large cystic and solid masses. When the tumor has achieved a sufficient size, cardiac compression, hydrops and polyhydramnios may result. If the teratoma has calcifications with acoustic shadowing, it is more easily distinguished from CPAM38. A predominantly cystic mediastinal teratoma with hydrops can be aspirated to reduce the intrathoracic pressure and successfully manage the hydrops39.
CPAM may occur in association with a congenital diaphragmatic hernia5,6.
Sonographic Assessment
An extended fetal anatomic survey is required when CPAM is suspected to rule out associated cardiac anomalies5, as well as to evaluate the fetus for possible sonographic stigmata of a karyotypic abnormality. In general, an isolated CPAM has not been associated with an increased risk of karyotypic malformations7.
In cases of possible microcystic CPAM and a mediastinal shift, weekly ultrasound examinations until the plateau of microcystic growth (26–28 weeks) can look for potential signs of hydrops11,12. After 26–28 weeks’ gestation, the fetus increases in size relative to the CPAM, allowing hydrops to resolve40. Ultrasonic surveillance can then be individualized based upon the size of the lesion. If CPAM has dominant cysts, serial ultrasound examinations until delivery are required to look for hydrops.
The CPAM volume is calculated using the formula for an elipse (height x width x length x 0.52). The CPAM volume is divided by the head circumference in centimeters to correct for gestational age. In one study a CPAM volume ratio(CVR) > 1.6 was associated with a hydrops rate of 80% versus 2% for a CVR < 1.640. Even with a CVR < 1.6, acute enlargement of a dominant cyst could result in hydrops.
The overall prognosis with a CPAM depends, not on its type, but on the size of the lesion. Approximately 6-11% of CPAM will undergo spontaneous regression7,41.
Fetal Therapy
CPAM with hydrops is almost invariably fatal42.
The current therapy of choice for microcystic CPAM with hydrops prior to 32 weeks’ gestation is one course, or if there is no response, a second course, of two doses of 12 mg of betamethasone. Steroid treatment is believed to arrest the growth of CPAM, resulting in an earlier plateau in its growth rate. Fetuses with non-hydropic and hydropic microcystic CPAM have survival rates of 85% and 49%, respectively, after a course of steroids43,44.
Type I or type II fetal CPAMs have significantly lower response rates to a course of maternal betamethasone, in contrast to microcystic CPAMs.
For hydrops that develops or recurs after 32 weeks’ gestation, an EXIT procedure followed by neonatal surgery should be considered45.
Macrocystic lesions that have resulted in hydrops can be treated with initial aspiration of the largest cysts (Fig. 7)4,46. If there are multiple large cysts, shunting of one cyst will decompress the other cysts due to their intercommuncation4. Reaccumulation of fluid may occur in a multicystic CPAM and, therefore, require a cyst-amniotic fluid catheter for permanent drainage and to reduce the mass effect of the cyst or cysts47. If there is a dominant cyst, even with a CVR < 1.6, acute enlargement of the cyst may result in hydrops. Twice weekly sonographic surveillance is, therefore, required to assess the fetus for the earliest signs of hydrops. Even after decompression and resolution of hydrops, surviving neonates often have respiratory insufficiency.
Figure 7. Right CPAM (19 weeks).
Contraindications to cyst drainage would include an abnormal fetal karyotype, a predominantly solid CPAM, or major fetal cardiac abnormality. The overall survival after a thoracoamniotic shunt is in the vicinity of 60%48.
The reported complications with catheter placement are the common complications associated with in utero procedures – structural and/or vascular fetal trauma, chorioamnionitis, preterm labor, and preterm premature rupture of the membranes4.
Neonatal Therapy
The postnatal history of CPAM is highly variable. The lesion may be completely asymptomatic or a child may have recurrent respiratory infections5. Occasionally, postnatal CT may not reveal any evidence for a previously detected echogenic fetal lung mass. These cases are likely secondary to segmental or lobar mucous plugs that subsequently resolved18.
80% of symptomatic postnatal patients with CPAM present at birth with cardiac or respiratory compromise5.
In the absence of hydrops, neonatal survival approaches 100%. However, many of the neonates with large lesions require significant ventilatory support7. With a type I CPAM there is a risk of air trapping in the CPAM, resulting in respiratory compromise. Pneumothorax may also occur49.
The presence of CPAM (Fig. 8) is a lifelong risk for both infection and malignant transformation. The malignant transformation of CPAM has been estimated between 1%50 and 4%51. The bronchioloalveolar carcinomas associated with CPAM generally appear between 6 years of age and young adulthood52. As a result, some groups recommend the removal of CPAMs even in asymptomatic neonates. An additional benefit of resection is the compensatory growth of the normal lung that occurs after surgery53.
Figure 8. CPAM (arrow) neonatal CT.
Complications from pediatric surgery range between 6-9%; mortality is rare. Most infants have normal pulmonary function on follow–up examination50,51.
There is lack of consensus with respect to the appropriate course of action in asymptomatic children with a presumptive diagnosis of CPAM. In one study 33% of pediatric surgeons recommended observation with a highly variable frequency of follow-up. The cost of expectant observation must factor in, not only office visits and periodic chest CTs, but also the intangible cost of radiation exposure2,54. The use of MRI for surveillance in asymptomatic neonatal CPAM has not been evaluated. However, MRI is generally less effective than CT in the evaluation of cystic lesions.
In general, surgical intervention in the asymptomatic patient is associated with shorter hospital stays, fewer complications and decreased medical costs when compared to surgical intervention only with the advent of symptoms2.
References
- Shanti CM, Klein MD. Cystic lung disease. Semin Pediatr Surg 2008;17:2-8.
- Marshall KW, Blane CF, Teitelbaum DH, VanLeeuwen K. Congenital cystic adenomatoid malformation: Impact of prenatal diagnosis and changing strategies in the treatment of the asymptomatic patient. AJR 2000;175:1551-1554.
- Cangiarella J, Greco MA, Askin F, Perlman E, Goswami S, Jagirdar J. Congenital cystic adenomatoid malformation of the lung: insights into the pathogenesis utilizing quantitative analysis of vascular marker CD34 (QBEND-10) and cell proliferation maker MIB-1. Med Pathol 1995;8:913-918.
- Schrey S, Kelly EN, Langer JC, Davies GA, Windrim R, Seaward PGR, Ryan G. Fetal thoracoamniotic shunting for large macrocystic congenital cystic adenomatoid malformations of the lung. Ultrasound Obstet Gynecol 2012;39:515-520.
- Stocker JT, Madewell JER, Drake RM. Congenital cystic adenomatoid malformation of the lung: classification and morphologic spectrum. Hum Pathol 1977;8:155-171.
- Stocker JT. Congenital pulmonary airway malformation: a new name and an extended classification of congenital cystic adenomatoid malformation of the lung. Histopathology 2002;41:424-431.
- Adzick NS, Harrison MR, Crombleholme TM, Flake AW, Howell LJ. Fetal lung lesions: management and outcome. Am J Obstet Gynecol 1998;179:884-889.
- Shimohira M, Hara M, Kitase M, Takeuchi M, Shibamoto Y, Kurano K, Shimizu S. Congenital pulmonary airway malformation: CT-pathologic correlation. J Thorac Imaging 2007;22:149-153.
- De Santis M, Masini L, Noia G, Cavaliere AF, Olivia N, Caruso A. Congenital cystic malformation of the lung: antenatal ultrasound findings and fetal-neonatal outcome. Fifteen years experience. Fetal DiagnTher 2001;15:246-248.
- Rempen A, Feige A, Wunsch P. Prenatal diagnosis of bilateral cystic adenomatoid malformation of the lung. J Clin Ultrasound 1987;15:3-8.
- Kunisaki SM, Barnewalt CE, Estroff JA, Ward VL, Nemes LP, Fauza DO, Jennings RW. Large fetal congenital cystic adenomatoid malformations: growth trends and patient survival. J Pediatr Surg 2007;42:404-410.
- Saltzman DH, Adzick NS, Benacerraf BR. Fetal cystic adenomatoid malformation of the lung: apparent improvement in utero. Obstet Gynecol 1988;71:1000-1002.
- Fine C, Adzick NS, Doubilet PM. Decreasing size of a congenital cystic adenomatoid malformation in utero. J Ultrasound Med 1988;7:405-408.
- Bromley B, Parad R, Estroff JA, Benacerraf BR. Fetal lung masses: prenatal course and outcome. J Ultrasound Med 1995;14:927-936.
- Mahle WT, Rychik J, Tian ZY, Cohen MS, Howell LJ, Crombleholme TM, Flake AW, Adzick NS. Echocardiographic evaluation of the fetus with congenital cystic adenomatoid malformation. Ultrasound Obstet Gynecol 2000;16:620-624.
- Adzick NS. Management of fetal lung lesions. Clin Perinatol 2009;36:363-376.
- Achiron R, Hagesh J, Yagel S. Fetal lung lesions: a spectrum of disease. New classification based on pathogenesis, two-dimensional and color Doppler ultrasound. Ultrasound Obstet Gynecol 2004;24:107-114.
- Achiron R, Strauss S, Seidman DS, Lipitz S, Mashiach S, Goldman B. Fetal lung hyperechogenicity: prenatal ultrasonographic diagnosis, natural history and neonatal outcome. Ultrasound Obstet Gynecol 1995;6:40-42.
- Biyyam DR, Chapman T, Ferguson MR, Deutsch G, Digha MK. Congenital lung abnormalities: embryologic correlation. Radiographics 2010;30:1721-1738.
- Alamo L, Gudinchet F, Reinberg O, Vial Y, Franani K, Osterhold MC, Meuli R. Prenatal diagnosis of congenital lung malformations. Pediatr Radiol 2012;42:273-283.
- Bonnefay C, Blanc P, Coste K, Delabaere A, Dechelotte PJ, Leurichesse-Delmas H, Labbe A, Jacquetin B, Lemery D, Sapin V, Gallot D. Prenatal diagnosis of lobar bronchial atresia. Ultrasound Obstet Gynecol 2011;37:110-112.
- Schuster SR, Harris GB, Williams A, Kirkpatrick J, Reid L. Bronchial atresia: a recognizable entity in the pediatric age group. J Pediatr Surg 1978;13:682-689.
- Riedlinger WF, Vargos SI, Jennings RW, Estroff JA, Barrewolt CE, Lillehei CW, Wilson JM, Colin AA, Redi LM, Kozakewich HP. Bronchial atresia is common to extralobar sequestration, intralobar sequestration, congenital cystic adenomatoid malformation, and lobar emphysema. Pediatr Dev Pathol 2006;9:361-374.
- Abitayeh G, Ruano R, Martinovic J, Barthe B, Aubry M-C, Benechi A. Prenatal diagnosis of main stem bronchial atresia using 3-dimensional ultrasonographic technologies. J Ultrasound Med 2010;29:633-638.
- Gipson MG, Cummings KW, Hurth KM. Bronchial atresia. Radiographics 2009;29:11531-1535.
- Bailey PV, Tracey T, Connors RH, deMello D, Lewis JE, Weber R. Congenital bronchopulmonary malformations. J Thorac Cardiovasc Surg 1990;99:597-603.
- Quinton AE, Smoleniec JS. Congenital lobar emphysema – the disappearing chest mass: antenatal ultrasound appearance. Ultrasound Obstet Gynecol 2001;l17:169-171.
- Demos N, Teresi A. Congenital lung malformations: a unified concept and a case report. J Thorac Cardiovasc Surg 1975;70:260-264.
- Puligandia PS, Laberge J-M. Congenital lung lesions. Clin Perinatol 2012;39:331-347.
- Gurney JG, Ross JA, Wall DA, Bleyer WA, Severson RK, Robison LL. Infant cancer in the U.S.: histology-specific incidence and trends, 1973 to 1992. J Pediatr Hematol/Oncol 1997;19:428-432.
- Maris JM, Kyemba SM, Rebbeck TR White, PS, Sulman EP, Jensen SJ, Allen C, Biegel JA, Yanofsky RA, Feldman GL, Brodeur GM. Familial predisposition to neuroblastoma does not map to chromosome band Ip36. Cancer Res 1996;56:3421-3425.
- Houlihan C, Jampolsky M, Shilad A, Principe D. Prenatal diagnosis of neuroblastoma with sonographic and magnetic resonance imaging. J Ultrasound Med 2004;23:547-550.
- Jennings RW, LaQuaglia MP, Leong K, Hendren WH, Adzick NS. Fetal neuroblastoma: prenatal diagnosis and history. J Pediatr Surg 1993;28:1168-1174.
- Isaacs H. Fetal and neonatal neuroblastoma: retrospective review of 271 cases. Fetal Pediatr Pathol 2007;26:177-184.
- Izbizky G, Elias D, Gallo A, Farios P, Sod R. Prenatal diagnosis of fetal bilateral adrenal carcinoma. Ultrasound Obstet Gynecol 2005;26:669-671.
- Froberg MK, Brown RE, Maylock J, Poling E. In utero development of a mediastinal teratoma: a second trimester event. Prenat Diagn 1994;14:884-887.
- Tapper D, Lock EE. Teratomas in infancy and childhood: a 54-year experience at the Children’s Hospital Medical Center. Ann Surg 1983;198:398-410.
- Wang R-M, Shih J-C, Ko T-M. Prenatal sonographic depiction of fetal mediastinal immature teratoma. J Ultrasound Med 2000;19:289-292.
- Takayasu H, Kitano Y, Kuroda T, Morikawa N, Tanaka H, Fujimo A, Muto M, Nosaki S, Tsutsumi S, Hayashi S, Sago H. Successful management of a large mediastinal teratoma complicated by hydrops fatalis. J Pediatr Surg 2010;45:e21-24.
- Crombleholme TM, Coleman B, Hendrick H, Liechty K, Howell L, Flake AW, Johnson M, Adzick NS. Cystic adenomatoid malformation volume ratio predicts outcome in prenatally diagnosed cystic adenomatoid malformation of the lung. J Pediatr Surg 2002;37:331-338.
- MacGillivray TE, Harrison MR, Goldstein RB, Adzick NS. Disappearing fetal lung lesions. J Pediatr Surg 1993;20:1321-1325.
- Adzick NS, Harrison MR, Glick PL, Golbus MS, Anderson RL, Mahony BS, Callen PW, Hirsch JH, Luthy DA, Filly RA, deLorimer A. Fetal cystic adenomatoid malformation: prenatal diagnosis and natural history. J Pediatr Surg 1985;20:483-488.
- Curran PF, Jelin EB, Rand L, Hirose S, Feldstein VA, Goldstein RB, Lee H. Prenatal steroids for microcystic congenital cystic adenomatoid malformations. J Pediatr Surg 2010;45:145-150.
- Morris LM, Lim FY, Livingston JC, Polzin WJ, Crombleholme TM. High risk fetal congenital pulmonary airway malformations have a variable response to steroids. J Pediatr Surg 2009;44:60-65.
- Kunisaki SM, Fauza DO, Bernewolt CE, Estroff JA, Myers LB, Bulich LA, Wong G, Levine D, Wilkins-Haug LE, Benson CB, Jennings RW. Ex utero intrapartum treatment with placement on extracorporeal membrane oxygenation for fetal mosaic masses. J Pediatr Surg 2007;42:420-425.
- VanLeduwen K, Teitelbaum DH, Hirschl RB, Austin E, Adelman SH, Polley TZ, Marshall KW, Coran AG, Nugent C. Prenatal diagnosis of congenital cystic adenomatoid malformation and its postnatal presentation, surgical indications, and natural history. J Pediatr Surg 1999;34:794-799.
- Crombleholme TM, Coleman B, Hedrick H, Liechty KW, Howell LJ, Flake AW, Johnson M, Adzick NS. Cystic adenomatoid malformation volume ratio predicts outcome in prenatally diagnosed cystic adenomatoid malformation of the lung. J Pediatr Surg 2002;37:331-338.
- Wilson RD. In utero therapy for fetal thoracic abnormalities. Prenat Diagn 2008;28:619-625.
- Bentur L, Canny G, Thoener P, Superina R, Babyn P, Levison H. Spontaneous pneumothorax in cystic adenomatoid malformation: unusual clinical and histologic features. Chest 1991;99:1292-1293.
- West D, Nicholson AG, Colquhoun I, Pollock J. Bronchioloalveolar carcinoma in congenital cystic adenomatoid malformation of the lung. Ann Thorax Surg 2007;83:68-689.
- Nasr R, Himidan S, Pastor AC, Taylor G, Kim DC. Is congenital cystic adenomatoid malformation a premalignant lesion for pleuropulmonary blastoma? J Pediatr Surg 2010;45:1086-1089.
- Ramos SG, Barbosa GH, Tavora FR, Jeudy J, Torres LA, Tone LG, Trad CS. Bronchioloalveolar carcinoma arising in a congenital pulmonary airway malformation in a child: case report with an update of this association. J Pediatr Surg 2007;42:E1-4.
- Mentzer SJ, Filler, RM, Phillips J. Limited pulmonary resections for congenital cystic adenomatoid malformation of the lung. J Pediatr Surg 1992;27:1410-1413.
- Lo A Y-S, Jones S. Lack of consensus among Canadian pediatric surgeons regarding the management of congenital cystic adenomatoid malformation of the lung. J Pediatr Surg 2008;43:797-799.
- Adzick N. Fetal thoracic lesions. Semin Pediatr Surg 1993;2:103-108.