Neuroblastomas arise from undifferentiated neural tissue of the adrenal medulla (40% to 70%), or extraadrenal sympathetic ganglia (30% to 60%), in the abdomen, thorax, pelvis, or head and neck (Birner, 1961; Beckwith and Perrin, 1963; Schneider et al., 1965; Janetschek et al., 1984; Ferraro et al., 1988). Neuroblastic nodules appear to be normal fetal structures within the adrenal gland that regress or differentiate throughout gestation. Greater numbers of these nodules are present early in gestation than at birth or during early infancy (Turkel and Habashi, 1974; Ikeda et al., 1981; Grosfeld et al., 1993). Neuroblastic nodules were found in 100% of the adrenal glands studied in second trimester abortuses (Turkel and Habashi, 1974; Ikeda et al., 1981) (Figure 113-1). Postnatal autopsy studies of newborn and young infants who died of unrelated causes reveal nodules to be present in 0.5% to 2.5% of carefully sectioned specimens (Beckwith and Perrin, 1963; Guin et al., 1969; Grosfeld et al., 1993). These figures contrast with the incidence of symptomatic neuroblastoma of 1 in 10,000 to 1 in 30,000 children. This calls into question the clinical significance of these early lesions and the need for treatment. It is also of interest that adrenal cysts are more commonly seen in fetal neuroblastoma, but are relatively uncommon in postnatal life (Turkel and Habashi, 1974). Cystic change may also represent a phase of normal adrenal development, thus accounting for prenatal detection of sonographically apparent but possibly clinically insignificant tumors (Tubergen and Heyn, 1970) (Figure 113-2). Although these tumors may represent rests of neural tissue with persistence into neonatal life, the potential for malignant transformation due to defective regression or differentiation must be considered (Turkel and Habashi, 1974; Grosfeld et al., 1993). Beckwith referred to these lesions as “neuroblastoma in situ” because these aggregates of neuroblasts exhibited mitotic figures and infiltration of the adult cortex (Beckwith and Perrin, 1963).

The results of mass infant screening programs in Japan and Canada using urinary vanillylmandelic acid (VMA) and homovanillic acid (HVA) support the view that neuroblastoma in situ may not represent clinically significant disease (Bessho et al., 1991; Murphy et al., 1991; Grosfeld et al., 1993; Woods et al., 2002). Screening infants at 6 months of age has resulted in a doubling of the incidence of neuroblastoma detected in infants. However, the expected fall in the incidence of more advanced stage neuroblastoma detected in older age groups was not seen. Furthermore, survival in the screened groups has approached 100%. The asymptomatic cases detected by screening may represent a subset of tumors destined to regress or differentiate without posing any clinical risk. This would explain both the high incidence in infants detected by screening programs and their excellent survival. The group detected by newborn screening may be the same subset of tumors detected prenatally with ultrasound examination.

In contrast, there is a subset of neuroblastoma diagnosed in utero or in newborns that does display an aggressive biologic behavior. Among the 300 cases of fetal or perinatal neuroblastoma reported, 83% were stage I or II, but there were also one each with stage III and IV disease (Fenart et al., 1983; Gadwood and Reynes, 1983;Sones, 1983; Newton et al., 1985; Atkinson et al., 1986; de Filippi et al., 1986; Giulian et al., 1986; Tovar et al., 1988; Kurtz and Hilbert, 1989; Pley et al., 1989; Forman et al., 1990; Hosoda et al., 1992; Suresh et al., 1993;Crombleholme et al., 1994a; Goldstein et al., 1994; Garmel et al., 1994b; Acharya et al., 1997; Granata et al., 2000; Sauvat et al., 2002; Nuchtern, 2006) and five with stage IVS disease (Hainaut et al., 1987; Ho et al., 1993). All children affected with stage IV disease died. In a review of congenital neuroblastoma reported between 1985 and 1991, Jennings et al. (1993) found 14 stillbirths, 44 neonatal deaths, and 2 late deaths, with only 10 survivors. Eight cases were associated with metastases to the placenta, and 1 had umbilical cord metastases. This subset of aggressive tumors associated with such poor outcomes may be quite different from the subset detected by the Japanese screening programs (Grosfeld et al., 1993) or by prenatal ultrasound examination.

The cause of neuroblastoma is unclear, although most cases appear sporadically (Behrman, 1992). At least one hypothesis is that antenatal factors play a role, with some adverse stimulus or stimuli occurring in late pregnancy. Preterm delivery is protective, while growth restricted term infants are at risk (Johnson and Spitz, 1985). There have been reports of congenital neuroblastoma in association with maternal phenytoin and phenobarbital use (Sherman and Roizen, 1976; Allen et al., 1980). The causative effect of phenytoin in neuroblastoma has been questioned and this is most likely a coincidental finding. In addition, some epidemiologic studies have suggested an association with maternal diabetes (Chow et al., 2007) and antenatal exposure to codeine (Cook et al., 2004).

Neuroblastoma is one of the most common tumors of infancy and childhood, with a clinical incidence of between 1 in 10,000 and 1 in 30,000 individuals (Janetschek et al., 1984; Fowlie et al., 1986; Goodman et al., 1999). Gurney et al., estimated the rate of neuroblastoma in the U.S. at 58 per 1,000,000 infants per year (Gurney et al., 1997). Neuroblastoma may be more common in white children than in other ethnic groups, and more common in male than in female children (Miller et al., 1968; Turkel and Habashi, 1974; Askin and Geschickter, 1985; Behrman, 1992). It is estimated that 16% of infant neuroblastomas are diagnosed in the first month of life and 41% within the first three months (Goodman et al., 1999).

Since the first case of prenatally detected neuroblastoma was reported by Fenart et al., 300 cases of neuroblastoma have been suspected or diagnosed by prenatal ultrasound examination (Nuchtern, 2006; Fenart et al., 1983). All have been visualized during the third trimester of pregnancy (Fenart et al., 1983; Gadwood and Reynes, 1983; Sones, 1983; Janetschek et al., 1984; Newton et al., 1985; Atkinson et al., 1986; deFilippi et al., 1986; Fowlie et al., 1986; Giulian et al., 1986; Hainaut et al., 1987; Ferraro et al., 1988; Tovar et al., 1988; Kurtz and Hilbert, 1989; Pley et al., 1989; Forman et al., 1990; Hosoda et al., 1992; Liyanage and Katoch, 1992; Ho et al., 1993; Jaffa et al., 1993; Jennings et al., 1993; Suresh et al., 1993; Crombleholme et al., 1994a; Garmel et al., 1994b; Goldstein et al., 1994; Acharya et al., 1997; Granata et al., 2000; Sauvat et al., 2002). The sonographic findings described by these studies were quite variable. The primary tumor is often small (only a few millimeters or centimeters in greatest diameter). Ninety percent of cases involve the adrenal gland, creating difficulties in distinction between the mass itself and the upper pole of the ipsilateral kidney. Cystic and solid areas within the mass are typically seen, which may be related to hemorrhage and necrosis of the tumor (Atkinson et al., 1986). Purely cystic lesions have also been reported in fetal neuroblastoma and may indicate a more favorable prognosis (Atkinson et al., 1986; Kurtz and Hilbert, 1989; Crombleholme et al., 1994a; Garmel et al., 1994b; Hamada et al., 1999; Petit et al., 2001; Lopez Alvarez-Buhilla et al., 2002; Tanaka et al., 2003; Athanassiadou et al., 2005). Occasionally, calcifications can be seen within the tumor (Potter, 1961; Fowlie et al., 1986; Giulian et al., 1986). Calcifications within neuroblastomas are often described as microcalcifications with acoustic shadowing (Kurtz and Hilbert, 1989; Gorincour et al., 2003). The tumor is usually well encapsulated and maydisplace the kidney inferiorly and laterallybut preserves the renal outline (Figure 113-3A and B). If arising in the sympathetic ganglia, the mass may be seen in the chest, cervical region, or intra-abdominal paravertebral locations. Cervical lesions sufficiently large to compromise the fetal airway have been reported (Gorincour et al., 2003), as have large retroperitoneal masses presenting as flank masses that are readily palpable on physical examination (Nagasako et al., 2004).

Occult neuroblastomas are rarely metastatic, but prenatal detection of liver metastases has been reported (Liyanage and Katoch, 1992; Jaffa et al., 1993; Nagasako et al., 2004). A suprarenal mass associated with hepatomegaly is highly suggestive of the diagnosis of neuroblastoma. Likewise, if a liver mass is seen on prenatal ultrasound examination, one needs to carefully examine all neural crest regions, especially in the renal and suprarenal areas, to rule out a primary tumor locus (Jaffa et al., 1993). Several cases have been reported with neuroblastoma metastatic to the placenta and umbilical cord. Careful imaging of the placenta and umbilical cord is also indicated because of reports of metastases to these areas. Fetoplacental metastases have been also reported in monochorionic twins with a primary tumor detectable in only one twin, with both having simultaneous evidence of metastatic disease (Adaletli et al., 2006).

Occasionally, ultrasound examination may also reveal polyhydramnios and/or hydrops (Falkinburg and Kay, 1953; Forman et al., 1990). Although the underlying mechanism is unknown, numerous hypotheses have been proposed, including hepatomegaly with subsequent mechanical obstruction of the umbilical vein or vena cava (Moss and Kaplan, 1978; Van der Slikke and Balk, 1980), compromised liver function with resultant hypoproteinemia (Adzick and Harrison, 1994); metastatic involvement of the placenta with placentomegaly and hydrops (Janetschek et al., 1984), tumor infiltration of bone marrow with subsequent anemia and heart failure (Moss and Kaplan, 1987), or arrhythmia and heart failure due to catecholamine release (Moss and Kaplan, 1987).

Malformations have occasionally been noted in association with neuroblastoma in children, including microcephaly, hydrocephaly, absence of the corpus callosum, cleft lip and palate, tracheoesophageal fistula, heart defects, skeletal abnormalities, genitourinary abnormalities, polydactyly, and single umbilical artery (Potter and Parrish, 1942; Kouyoumdjian and McDonald, 1951; Bodian, 1963; Berry et al., 1970; Andersen and Hariri, 1983). A consistent pattern ofassociated anomalies has not been demonstrated however, and the majority have been isolated cases of fetal neuroblastomas (Miller, 1966; Sy and Edmonson, 1968; Berry et al., 1970).