Gastroschisis (Greek for belly cleft) is a full-thickness defect in the abdominal wall that occurs secondary to incomplete closure of the lateral folds during the 6th week of gestation (Moore and Stokes, 1953; Moore and Persaud, 1993). At birth, the eviscerated bowel characteristically has a thick edematous appearance described by Moore as a “peel” (Figure 63-1). The peel involves the serosa and is composed of fibrin and collagen. The peel in gastroschisis is thought to be caused by an inflammatory reaction as a result of exposure to amniotic fluid, combined with constriction at the abdominal wall defect (Amoury and Holder, 1977; Klein et al., 1983; Tibboel et al., 1986a, b; Amoury et al., 1988; Langer et al., 1990; Moore, 1992). Duhamel (1963) theorized that gastroschisis originates from a discrete teratogenic insult that results in an isolated defect in differentiation of the somatopleural mesenchyme. Others argue that gastroschisis is due to an in utero rupture of an umbilical cord hernia after completion of the infold in the anterior abdominal wall, but before complete closure of the umbilical ring (Shaw, 1975). In at least some cases, in utero rupture of a hernia of the cord that resulted in gastroschisis supports the argument for this being a cause (Glick et al., 1985). DeVries (1980) suggested that gastroschisis could be caused by a congenital weakness on the right side of the umbilical cord. From an umbilical cord hernia, premature atrophy or abnormal persistence of the right umbilical vein could predispose to disruption of the somatopleura at its junction with the body stalk. Because gastrointestinal defects such as atresia associated with gastroschisis are caused by vascular disruptions, Hoyme et al. (1983) suggested that gastroschisis may be caused by disruption of the right omphalomesenteric artery, which connects the yolk sac with the dorsal aorta. The fact that gastroschisis almost always occurs to the right of the umbilical ring is consistent with these latter theories (Torfs et al., 1990).

As second trimester maternal serum-α-fetoprotein (MSAFP) screening has become incorporated into routine prenatal care, more cases of gastroschisis are being detected prenatally. This is due to the association between elevated MSAFP levels and ventral abdominal wall defects (McKeown et al., 1953; Brock et al., 1979; Redford et al., 1985; Stiller et al., 1990; Killam et al., 1991). In one study, Larson et al. (1993) found a 58% incidence of major fetal congenital anomalies when extremely elevated MSAFP levels were detected between 15 and 20 weeks of gestation. Ten percent of these were due to abdominal wall defects. The sensitivity of MSAFP screening for the detection of abdominal wall defects varies depending on the type of abdominal wall defect, the geographical area, and the cutoff value of MSAFP used.

MSAFP screening has a higher sensitivity for detecting gastroschisis than omphalocele. In a population-based study of MSAFP screening, using patients from Maine and Rhode Island, Palomaki et al. (1988) found that the combined incidence of gastroschisis and omphalocele was 4.5 in 10,000 livebirths (excluding neural tube defects, twins, and autosomal chromosomal anomalies). At each cutoff of MSAFP, detection rates were higher for gastroschisis than for omphalocele. For example, at a cutoff of 2 multiples of the mean (MoM), the detection rate was more than 99% for gastroschisis and 78% for omphalocele. At cutoff values of 2.5 MoM and 3 MoM, the detection rates were more than 98% and 71%, and 96% and 65%, for gastroschisis and omphalocele, respectively. The median MSAFP value of the 20 cases of gastroschisis was 7 MoM and the median value for omphalocele was 4.1 MoM (Palomaki et al., 1988). Crandall et al. (1991) have shown that the higher the MSAFP level the greater the prevalence of adverse fetal outcome. A lower detection rate in omphalocele is likely due to the presence of an intact membrane covering the abdominal viscera in unruptured omphaloceles as opposed to the direct exposure of bowel to the amniotic fluid in gastroschisis.

Analysis of amniotic fluid α-fetoprotein and acetylcholinesterase pseudocholinesterase levels (ratio 0.13) is very sensitive in detecting gastroschisis (Goldfine et al., 1989). Human chorionic gonadotropin (hCG) levels have been measured in 16 pregnancies with abdominal wall defects (12 gastroschises, 4 omphaloceles) because this marker is already being increasingly used for fetal aneuploidy screening. It may have utility as an additional marker for the presence of abdominal wall defects. Using a cutoff value of 2.3 MoM, the detection rate using hCG was 31%, as compared with 81% for MSAFP. Using the two markers, the detection rate increased to 87.5%, by finding one additional case (Schmidt et al., 1993). More data are necessary to evaluate the usefulness of hCG levels to prospectively identify abdominal wall defects.

Before 1953, gastroschisis was not clearly distinguished from omphalocele (Moore and Persaud, 1993). Consequently, a reliable estimate of the frequency of gastroschisis before then is impossible. Several investigators have found that the prevalence of gastroschisis has increased over the past few decades in different geographic regions (Hwang and Kousseff, 2004). Roeper et al. (1987) documented in California that the rate of gastroschisis has increased from 0.006 in 1000 livebirths in 1968 to 0.089 in 1000 livebirths in 1977. Similarly, Florida, Sweden, Finland, and Spain have reported increased prevalence rates over the same period (Lindham, 1981; Hemmenki et al., 1982; Martinez-Frias et al., 1984; Hwang and Kousseff, 2004). The overall prevalence in Europe has been reported to be 0.07 in 1000 live-and stillbirths (Calzolari et al., 1993). Studies from British Columbia and Italy have not confirmed this trend (Baird and MacDonald, 1981; Calzolari et al., 1993).

Both older and more recent data from epidemiologic studies have consistently demonstrated that young maternal age is associated with an increased risk of gastroschisis (Colombani and Cunningham, 1977; Hemmenki et al., 1982; Martinez-Frias et al., 1984; Roeper et al., 1987). Goldbaum et al. (1990) compared 62 infants who had gastroschisis with 617 randomly selected, unaffected infants matched for the first year of birth in the state of Washington. They found that maternal age younger than 20 years was associated with a fourfold increased risk of having an infant with gastroschisis. A study by Werler et al. (1992) included patients from Boston, Philadelphia, and portions of Ontario and Iowa. They compared 76 cases of gastroschisis with 2582 malformed controls and found a strong inverse association with maternal age. Compared with women 30 years or older, the relative risks for maternal ages 25 to 29, 20 to 24, and younger than 20 were 1.7, 5.4, and 16, respectively (Werler et al., 1992).

In addition to young maternal age, maternal cigarette use has also been associated with an increased risk of gastroschisis. Goldbaum et al. (1990) were the first to suggest this relationship. In a prospective population-based study using 62,103 consecutive second trimester MSAFP samples, Haddow et al. (1993) found that women who smoked had a 2.1-fold greater risk of fetal gastroschisis than nonsmokers. Although their finding was not considered statistically significant, the trend was consistent with the report by Goldbaum et al. (1990). In contrast, Werler et al. (1992) found no association with cigarette use in early pregnancy when heavy (more than 15 cigarettes per day) and light (less than 15 cigarettes per day) were compared. However, when the results of these three studies are combined, the relative risk for having a baby with gastroschisis is 1.6 for women who smoke (Haddow et al., 1993).

Although reports are conflicting, the incidence of gastroschisis has also been associated with seasonal variation. These studies found an increased risk of gastroschisis in deliveries occurring during the first quarter of the year (Egenaes and Bjerkedal, 1982; Hemmenki et al., 1982; Goldbaum et al., 1990). However, three other studies found no seasonal variation (Paulozzi, 1986; Roeper et al., 1987; Haddow et al., 1993). Vascular disruption has been suggested by cases of gastroschisis observed in women who took medications with vasoactive properties during pregnancy (Van Allen, 1981). In a prospective screening study of maternal hair samples using gas chromatography with mass spectroscopy, Morrison et al. (2005) found a 25% incidence of periconceptual recreational drug use. The most commonly detected compounds were methamphetamine, including MDMA and MDEA, and cocaine. There is good evidence from animal data that methamphetamines cause fetal malformations including cardiac defects, facial clefts, eye abnormalities, skeletal malformations, kidney defects, and gastroschisis (Colado et al., 1997; Plessinger, 1998). Amphetamines and cocaine have similar maternal effects, as they are both central nervous system stimulants. They produce different fetal developmental effects (Little et al., 1988). Werler et al. (1992) reported on first trimester medication use in a case–control study of 76 cases of gastroschisis and 2142 matched controls. They found that the decongestant pseudoephedrine was associated with a greater than threefold risk of gastroschisis. Salicylates and acetaminophen were also associated with elevated risk, but these differences were not statistically significant. The authors also evaluated the use of these medications in pregnancies with other fetal anomalies that were presumed to have a vascular cause; no associations were found. Further studies are needed to clarify the role of vasoactive agents in the pathogenesis of gastroschisis.

The diagnosis of abdominal wall defects during the first trimester is difficult because it is normal for the midgut to be herniated into the umbilical cord. Cyr et al. (1986) documented the events of bowel migration by abdominal sonographic examination of 10 normal first trimester fetuses, as well as by pathologic examination of several embryo specimens. The midgut that will normally form the small bowel, cecum, and ascending and proximal transverse colon is connected to the yolk sac. This connection is reduced to a narrow yolk stalk as the amniotic cavity expands and the yolk sac is pulled away from the embryo. The mesentery suspending the midgut then rapidly elongates, creating a U-shaped loop of midgut that herniates into the umbilical cord. The loop of intestine then rotates 90 degrees in a counterclockwise direction about the axis of the superior mesenteric artery. As these loops of bowel return to the abdominal cavity by 11 weeks of gestation, the loops rotate counterclockwise another 180 degrees to complete the bowel rotation. Cyr et al. (1986) have suggested that ultrasound examination should be performed at 14 weeks of gestation because the bowel should be entirely intra-abdominal by 11 weeks, and this allows for errors in estimating gestational age. However, in 20% of fetuses, the bowel was still outside the abdomen at 12 weeks (Green and Hobbins, 1988). The size of the herniated gut may be helpful in distinguishing between physiologic and pathologic bowel herniation. The dimensions of bowel herniation have been reported to be 1 cm in the study by Cyr et al. (1986) and 7 mm in greatest dimension in the study by Bowerman (1993).

Vaginal sonography affords a closer look at the fetus in early gestation and may be helpful in distinguishing the contents of the herniation. In a series of 61 fetuses studied by vaginal sonography, by 12 weeks of gestation the midgut herniation no longer persisted (Timor-Tritsch et al., 1989). The earliest diagnosis of gastroschisis was 12 weeks 3 days (Guzman, 1990).

Gastroschisis is a full-thickness defect in the anterior abdominal wall almost invariably located to the right of the intact umbilical cord, measuring 2 to 3 cm in diameter (Figure 63-2) (Fonkalsrud, 1980). Color Doppler studies assist in demonstrating normal umbilical cord insertion with herniation of intestine to the right of the umbilical cord. Nyberg et al. (1993) have summarized the sonographic features of gastroschisis. In addition to the small abdominal wall defect located to the right of a normal umbilical cord insertion site, there is a variable amount of bowel protruding through the defect, floating in the amniotic fluid, which may be disproportionally large relative to the small size of the abdominal cavity (Figure 63-3).

Despite the frequent use of antenatal ultrasound examination in obstetric care over the past two decades, there is still little information regarding the accuracy of routine ultrasound examination in the detection of abdominal wall defects (Hill et al., 1985; Rosendahl and Kivinen, 1989; Ewigman et al., 1993). Walkinshaw et al. (1992) reported on their experience in the United Kingdom for more than a 4-year period, extending from 1984 to 1988, during which 115 cases of anterior abdominal wall defects were found in 202,488 livebirths and intrauterine deaths beyond 22 weeks of gestation. They found that routine scanning identified 60% of the defects, with a false-positive rate of 5.3%. Gastroschisis and omphalocele were accurately distinguished in 79.3% of cases on initial diagnosis and 84.5% of cases after referral for further evaluation. Many factors contribute to the less than 100% detection of abdominal wall defects, including the quality of the ultrasound equipment, the experience of the sonographer, and the defect itself (Paidas et al., 1994). It is possible that one form of gastroschisis, due to rupture of a hernia of the cord as originally described by Moore (1962), would not be detected by routine ultrasound examination because it appears late in pregnancy. Knott and Colley (1987) have described two similar cases of gastroschisis not detected by antenatal ultrasound examination that were felt to occur as a result of late gestational rupture of a hernia of the cord.

The differential diagnosis of gastroschisis should include omphalocele, ruptured omphalocele, hernia of the cord, and limb–body wall complex. Gastroschisis is distinguished from omphalocele by having no membrane surrounding the herniated loops of intestine. In contrast, omphalocele has a peritoneo-amniotic membrane covering the defect and the size is significantly larger than the defect in gastroschisis. Ruptured omphalocele may be confused with gastroschisis, as the loops of intestine are observed to be floating free in the amniotic cavity. However, extracorporeal liver may be seen in ruptured omphalocele, but is never seen in gastroschisis. The limb–body wall complex may have an abdominal wall defect but is usually easily distinguished from gastroschisis by a short umbilical cord and numerous other associated structural anomalies.

The conventional wisdom regarding abdominal wall defects is that gastroschisis, unlike omphalocele, is not associated with chromosomal abnormalities. This discrepancy regarding the presence of associated chromosomal abnormalities has provided further impetus to distinguish these two entities by sonography. Reports in the literature have confirmed that chromosomal abnormalities are rare or absent in gastroschisis (Mayer et al., 1980; Mann and Ferguson-Smith, 1984; Sermer et al., 1987; Romero et al., 1988; Lewinsky et al., 1990; Sipes et al., 1990a). In the 17-year experience at the Children’s Hospital in Columbus, Ohio, chromosomal analysis was available in 128 of 144 cases of gastroschisis and there was only 1 case of chromosomal abnormality (trisomy 18) (King et al., 1980; Caniano et al., 1990). Nicolaides et al. (1992) did not find any chromosomal abnormalities in 26 cases of gastroschisis detected during an 8-year period. Abdullah et al. (2007) found only four cases (0.1%) of aneuploidy among 4344 infants with gastroschisis in the United States. In the absence of sonographically detectable associated anomalies, the risks for fetal aneuploidy are probably comparable to the risks due to maternal age alone. If additional fetal abnormalities are detected sonographically, chromosomal evaluation should be recommended.

In contrast to omphalocele, gastroschisis is usually not associated with extra gastrointestinal abnormalities. This is in part responsible for the better outcome observed in gastroschisis. In a survey of 13 years of the National Inpatient Sample Database and 3 years of the KIDs’ Inpatient Database, Abdullah et al. found among 4344 cases of gastroschisis reported only 72 (1.7%) cases of pulmonary defects (including agenesis, hypoplasia, and bronchopulmonary dysplasia) and 379 cardiac defects, of which 190 were atrial septal defects (ASDs) and 214 patent ductus arteriosus cases that would not be detectable on prenatal ultrasound examination. There were also 138 cases of undescended testicles (6.8%) and 58 cases of hydronephrosis (1.3%) (Abdullah et al., 2007). However, in a report from Kunz et al. (2005) in a study of California hospital discharge data from 1972 to 1997, 25 of 621 infants with gastroschisis were found to have congenital heart disease (4% incidence). This group also found a significant increase of congenital heart disease if the gastroschisis was complicated by bowel atresia, or if the infant was African American. Gibbin et al. found an incidence of abnormal cardiac findings of 15%, but of the four, one was persistent pulmonary hypertension, two were supraventricular tachycardia, and the last was peripheral pulmonic stenosis (Gibbin et al., 2003). However, gastrointestinal anomalies can commonly be seen in gastroschisis, and occur in 20% to 40% of cases (King et al., 1980; Mayer et al., 1980; DeLorenzo et al., 1987; Nicolaides et al., 1992; Novotny et al., 1993). These gastrointestinal abnormalities may be secondary to the gastroschisis. These include malrotation, atresia, “Christmastree deformity,” volvulus, and infarction.

Even in the rare cases in which they are present, nongastrointestinal abnormalities are not usually life-threatening. It is interesting that the data registry report from Abdullah et al. (2007) had a lower than expected 8.1% incidence.

Fetuses with gastroschisis are at risk for a number of complications that directly affect survival during the newborn period. Obstetric complications include intrauterine growth restriction (IUGR), which can affect up to 77% of fetuses (Carpenter et al., 1984; Molenaar and Tibboel, 1993). This may be due to nutritional deprivation rather than a constitutional limitation (Gutenberger et al., 1973). In an analysis of biometric data Royner and Richards (1977) found that IUGR was predicted in 43% of infants, but was present in only 23% at birth. This group thought that the prevalence of IUGR is increased in gastroschisis, but is overestimated with prenatal ultrasonography, primarily because of the smaller-than-average abdominal circumference. Carroll et al. (2001) suggested a novel explanation for the frequent observation of growth restriction among fetuses with gastroschisis; protein loss into the amniotic cavity. In a small series of 12 fetuses with prenatally diagnosed gastroschisis were compared with 29 control infants without gastroschisis. They found that fetuses with gastroschisis had significantly lower serum total protein and significantly higher amniotic fluid total protein, and α-fetoprotein.

Despite the predisposition to IUGR, a more important factor in neonatal outcome is prematurity. Puligandla et al. (2004) found that in a retrospective series of 113 cases of gastroschisis, infants with IUGR had similar outcomes to non-IUGR infants. Factors that were more important with regard to neonatal outcome were prematurity and the presence of atresia (Puligandla et al., 2004).

Other obstetric complications include preterm labor, which occurs in more than one-third of cases due to associated polyhydramnios (Mayer et al., 1980; Kirk and Wah, 1983; Carpenter et al., 1984; Caplan and MacGregor, 1989; Molenaar and Tibboel, 1993). Abnormalities of amniotic fluid volume, both polyhydramnios and oligohydramnios, can accompany gastroschisis (Bair et al., 1986; Crawford et al., 1992). Mercer et al. (1988) reported that in their series of 22 cases of gastroschisis, marked amniotic fluid staining occurred in 73%. The association of amniotic fluid staining with fetal distress is controversial (Carpenter et al., 1984; Crawford et al., 1992).