Sacrococcygeal teratoma (SCT) is defined as a neoplasm composed oftissues from either all three germ layers or multiple foreign tissues lacking an organ specificity arising in the sacrococcygeal region (Gross et al., 1951; Mahour et al., 1975). Because of the multiple cell lineages that characterize these tumors, it was previously suggested that SCT was of germ cell origin or a form of fetus in fetu (Theiss et al., 1960; Linder et al., 1975). Early theories suggested a “twinning accident” with incomplete separation during embryogenesis and abnormal development of one fetus (Waldhausen et al., 1963; Ashley, 1973; Cousins et al., 1980). In support of this theory, several authors have noted a family history of twinning in many SCT patients (Hickey and Layton, 1954; Grosfeld et al., 1976; Gross et al., 1987). However, more recently, SCT has been thought to arise from a totipotent somatic cell originating in Hensen’s node (Gross et al., 1987). This node is a caudal cell mass in the embryo that appears to escape normal inductive influences (Bale, 1984). Others hypothesize that SCT is derived from totipotent cells in reproductive gland anlage (Abbott et al., 1966).

SCT has been classified by the relative amounts of presacral and external tumor present [American Academy of Pediatrics Surgery Section (AAPSS) Classification (Table 115-1 and Figure 115-1)] (Altman et al., 1974). The utility of this classification scheme lies in the relationship between stage and timing of diagnosis, ease of resection, and malignant potential. Type I SCT is evident at birth, is usually easily resected, and has a low malignant potential. Similarly, types II and III SCT are recognized at birth, but resection may be difficult, requiring both an anterior and a posterior approach. In type IV SCT, the diagnosis may be delayed until it becomes symptomatic at a later age. Malignant transformation has frequently occurred by the time a type IV SCT is diagnosed.

SCT is one of the most common tumors in newborns; however, it is still rare, occurring in 1 in 23,000 to 1 in 40,000 livebirths (Schiffer and Greenberg, 1956; Altman et al., 1974; Tapper and Lack, 1983; Forrester and Merz, 2006). Females are four times more likely to be affected as males, however, malignant change is more frequently observed in males (Abbott et al., 1966; Conklin and Abell, 1967; Carney et al., 1972; Fraumeni et al., 1973; Altman et al., 1974).

Retrospective prenatal diagnosis of SCT was first made in the mid-1970s, and the first prospective prenatal diagnosis was reported by Horger and McCarter in 1979. They described a 13-cm complex mass at the caudal end of the fetus, with solid and cystic areas and bizarre internal echoes associated with polyhydramnios. This typical prenatal sonographic appearance has been confirmed by other authors (Figure 115-2) and approximately 60 cases of prenatally diagnosed SCT have been reported (Seeds et al., 1982; Grisoni et al., 1988; Bond et al., 1990). The most common clinical presentation is uterine size greater than dates, initiating an ultrasound examination (Seeds et al., 1982). To date, the earliest diagnosis of SCT that has been made is 12 3/7 weeks of gestation (Roman et al., 2004).

SCTs can grow at an unpredictable rate to tremendous dimensions. Several case reports note fetal tumors as large as 25 by 20 cm (Heys et al., 1967; Weiss et al., 1976). These tumors are generally exophytic (AAPSS type I), but may extend retroperitoneally displacing pelvic (type II) or abdominal structures (type III) (Litwiller, 1969).

Most SCTs are solid or mixed solid and cystic, consisting of randomly arranged irregularly shaped cysts (Seeds et al., 1982; Chervenak et al., 1985). Purely cystic SCT has also been described prenatally (Seeds et al., 1982; Hogge et al., 1987). Calcifications can be seen microscopically, although the majority are not visible on prenatal ultrasound examination. Most prenatally diagnosed SCTs are extremely vascular, which is easily demonstrated with the use of color flow Doppler studies (Figure 115-3). Three-dimensional power Doppler has been suggested to demonstrate the large vascular volume in SCT (Sciaky-Tamir et al., 2006). Polyhydramnios has been noted in most cases of prenatally diagnosed SCT, and–although the mechanisms for this are not known–it is likely secondary to renal hyperfiltration occurring as a result of high-output state (Chervenak et al., 1985).

Hepatomegaly, placentomegaly, and nonimmune hydrops have also been seen in association with SCT and appear to be secondary to high-output cardiac failure (Heys et al., 1967; Cousins et al., 1980; Gergely et al., 1980; Kapoor and Saha, 1989; Bond et al., 1990; Flake, 1993; Hedrick et al., 2004). High-output failure may be due to tumor hemorrhage or arteriovenous shunting within the tumor (Cousins et al., 1980; Flake et al., 1986; Alter et al., 1988; Schmidt et al., 1989; Bond et al., 1990). Some authors have attributed heart failure with subsequent hydrops to severe fetal anemia secondary to tumor hemorrhage (Alter et al., 1988). However, normal fetal hematocrits have also been reported, suggesting that congestive heart failure is more often due to high-output cardiac failure from arteriovenous shunting within the tumor (Schmidt et al., 1989). The demonstration ofheart failure or hydrops on ultrasound examination is usually a preterminal event (Flake et al., 1986; Kuhlmann et al., 1987; Bond et al., 1990).

Controversy exists regarding the presence of associated anomalies and the need for chromosome analysis. The incidence of coexisting anomalies is 11% to 38%, primarily involving the nervous, cardiac, gastrointestinal, genitourinary, and musculoskeletal systems (Hickey and Layton, 1954; Schiffer and Greenberg, 1956; Carney et al., 1972; Fraumeni et al., 1973; Altman et al., 1974; Izant and Filston, 1975; Gonzalez-Crussi et al., 1978; Ein et al., 1980; Holzgreve et al., 1985; Kuhlmann et al., 1987; Werb et al., 1992). Several authors postulate that at least some of these anomalies are related to tumor development. Others have reported an increased incidence of spinal deformities (Ewing, 1940; Gruenwald, 1941; Alexander and Stevenson, 1946; Bentley and Smith, 1960; Wilson et al., 1963; Carney et al., 1972). Most authors agree with Berry et al.’s (1970) observation that local abnormalities such as rectovaginal fistula and imperforate anus are thought to be directly related to tumor growth during fetal development. Aneuploidy has not been reported with SCT and we do not recommend amniocentesis for karyotype analysis unless there are multiple anomalies, advanced maternal age, or fetal surgery is contemplated.

Fetal MRI has emerged as an adjunctive imaging modality that can provide important anatomical detail in cases of SCT (Avni et al., 2002; Hedrick et al., 2004; Nassenstein et al., 2006). MRI may be particularly useful in defining the pelvic component of SCT and impact on other pelvic structures (Garel et al., 2005). In cases in which fetal surgery is being considered, fetal MRI provides a broader field of view than ultrasound and may be helpful in operative planning. In cases in which SCT has a pelvic component or there is polyhydramnios, oligohydramnios, hydronephrosis or hydrocolpos, fetal MRI may provide additional information on the anatomical relationships not apparent on ultrasound alone (Danzer et al., 2006). Fetal MRI in cases of cystic SCT may be particularly helpful in excluding myelomeningocele from the differential diagnosis (Yoon and Park, 2005; Danzer et al., 2006).

The differential diagnosis of SCT includes lumbosacral myelomeningocele, which invariably demonstrates a spinal defect. Myelomeningoceles have a cystic or semicystic rather than a solid appearance and do not contain calcifications. Examination of the fetal brain is helpful in establishing the diagnosis, as most fetuses with lumbosacral myelomeningocele will have associated cranial findings. Rarer entities that mimic SCT include neuroblastoma, glioma, hemangioma, neurofibroma, cordoma, leiomyoma, lipoma, melanoma, and any of 50 tumors or malformations reported in the sacrococcygeal region (Table 115-2) (Lemire and Beckwith, 1982; Sebire et al., 2004; Tanaka et al., 2005).

Biochemical markers such as α-fetoprotein (AFP) and acetylcholinesterase are not reliable in distinguishing SCTs from other abnormalities (Holzgreve et al., 1987). It has been suggested, however, that AFP can be used to differentiate benign from malignant tumors, as marked elevations of AFP may reflect the presence of a malignant endodermal sinus component to the tumor (Tsuchilda et al., 1975; Grosfeld et al., 1976; Gonzalez-Crussi et al., 1978; Gonzalez-Crussi, 1982). AFP levels can be extremely high in normal newborns, limiting the utility of this marker to distinguish benign from malignant lesions (Ohama et al., 1997).

The antenatal natural history of prenatally detected SCT is not as favorable as that of SCT presenting at birth. Well-defined prognostic factors for SCT diagnosed postnatally, as outlined in the AAPSS classification system, do not necessarily apply to fetal cases (Altman et al., 1974; Bond et al., 1990) (see Table 115-1). While the mortality rate for SCT diagnosed in the newborn is at most 5%, the mortality rate for fetal SCT approaches 50% (Flake et al., 1986; Bond et al., 1990; Flake, 1993; Hedrick et al.,2004).

Most SCTs are histologically benign. The incidence of malignant elements present in fetal SCT has ranged from 7% to 30% (Hedrick et al., 2004; Heerema-McKenny et al., 2005). Malignancy appears to be more common in males, especially with solid versus complex or cystic tumors (Schey et al., 1977). The presence of histologically immature tissue does not necessarily signify malignancy (Carney et al., 1972; Gonzalez-Crussi, 1982). Calcifications occur more often in benign tumors but may also be seen in malignant tumors and are unreliable indicators of malignant potential (Hickey and Layton, 1954; Waldhausen et al., 1963; Grosfeld et al., 1976; Schey et al., 1977; Horger and McCarter, 1979). Although there is one reported case of malignant yolk sac differentiation in a fetal SCT, there has not been a case ofmetastatic teratoma in a neonate with a prenatally diagnosed SCT (Holzgreve et al., 1985; Flake, 1993).

The prenatal history of SCT is quite different from the postnatal natural history. Flake et al. (1986) reviewed 27 cases of prenatally diagnosed SCT. Five cases were electively terminated and 15 of the remaining 22 died, either in utero or shortly after delivery. The majority of these patients presented between 22 and 34 weeks of gestation with a uterus large for gestational age secondary to severe polyhydramnios. The presence of hydrops and/or polyhydramnios was associated with intrauterine fetal death in seven of seven cases. The International Fetal Medicine and Surgery Society reported a mortality rate of 52% among cases of prenatally diagnosed SCT (Bond et al., 1990). When SCT was seen in association with placentomegaly or hydrops, all affected fetuses died in utero. The indication for ultrasound examination was also found to be a predictive factor. If SCT was an incidental finding, the prognosis was favorable at any gestational age. However, if the ultrasound examination was performed for maternal indications, 22 of 32 fetuses died. In addition, diagnosis prior to 30 weeks was associated with a poor outcome.

Sheth et al. (1988) also reported significant perinatal mortality associated with SCT, with only 6 survivors among 15 cases diagnosed prenatally. Three off our cases associated with hydrops were rapidly fatal. The sole survivor was salvaged by emergency cesarean section at 35 weeks. This series was unusual because three cases had severe obstructive uropathy and secondary renal dysplasia. A more favorable outcome was reported by Gross et al. (1987) in which 8 of 10 fetuses with prenatally diagnosed SCT survived. However, no fetus had hydrops or placentomegaly, and the two nonsurvivors were electively terminated.

Hydrops in SCT is usually, but not always, fatal. Nakoyama et al. (1991) reported survival in two fetuses with SCT presenting with hydrops at 27 and 30 weeks of gestation. In addition, Robertson et al. (1995) were able to salvage a hydropic fetus at 26 weeks of gestation by staged resection of the SCT in the neonatal period (Figure 115-4). In this case, acute rapid growth of the SCT led to polyhydramnios and preterm delivery. After delivery, the newborn was noted to be in a high-output state from shunting through the tumor. In a staged resection, the tumor was initially devascularized by ligation of both internal iliac arteries. Twenty-four hours later, the external portion of the mass was resected. The infant subsequently underwent resection of the intrapelvic portion of the tumor at 3 months of age, and did well.

Hedrick et al. (2004) reviewed their experiences with 30 cases of prenatally diagnosed SCT and reported 4 terminations, 5 fetal deaths, 7 neonatal deaths, and only 14 survivors (47%). Among the 26 patients continuing the pregnancy, 81% experienced obstetric complications including polyhydramnios (n = 7), oligohydramnios (n = 4), preterm labor (n = 13), pre-eclampsia (n = 4), gestational diabetes (n = 1), HELLP syndrome (n = 1), and hyperemesis (n = 1).

Sonographic features of SCT such as size, AAPSS classification, solid or cystic composition, or presence or absence of calcifications have not been predictive of either fetal survival or future malignant potential (Altman et al., 1974; Flake, 1993). One exception to this may be the unilocular cystic form of SCT, which has a relatively favorable prognosis because of benign histology and limited vascular and metabolic demand (Horger and McCarter, 1979; Mintz et al., 1983). The growth of the SCT in relation to the size of the fetus is also unpredictable and may increase, decrease, or stabilize as gestation proceeds. However, a rapid phase of tumor growth usually precedes the development of placentomegaly and hydrops.

Highly vascular lesions are more likely to undergo rapid tumor growth and to be associated with the development of placentomegaly and hydrops. The prenatal mortality, unlike postnatal mortality, is not due to malignant degeneration, but to complications of tumor mass or tumor physiology (Flake et al., 1993). The tumor mass may result in malpresentation or dystocia, which in turn may result in tumor rupture and hemorrhage during delivery. Dystocia has been reported in 6% to 13% of cases in postnatal series (Giugiaro et al., 1977; Musci et al., 1983; Gross et al., 1987). SCTs may also spontaneously rupture in utero leading to significant fetal anemia or death (Sy et al., 2006). The most important benefit of prenatal diagnosis is prevention of dystocia by elective or emergency cesarean section. Tumor mass effect may also result in uterine instability and preterm delivery because of uterine distention (Flake et al., 1986; Bond et al., 1990). Massive polyhydramnios is frequently seen in large fetal SCT, which also predisposes to uterine irritability and preterm delivery.

SCT may occur in twins further complicating the prenatal management. In Hedrick et al.’s series, 10% of the cases occurred in twin gestations (Hedrick et al., 2004). The presence of SCT in a twin gestation increases the risk of preterm delivery. Because SCT is associated with an increased risk of fetal death, intrauterine demise of a monochorionic twin with SCT places the surviving unaffected co-twin at risk of adverse neurologic outcome (Ayzen et al., 2006).