This 4-month-old girl was admitted following a 1-day history of bilious emesis. She had been well until 1 day prior to admission, when she developed recurrent episodes of emesis and fever and became increasingly sleepy. At a local emergency department, she had an episode of bilious emesis. Prior to the start of this illness, she had been eating normally since the newborn period and had been having two formed bowel movements each day. Her diet consisted of a cow's milk formula, taking 5 to 6 ounces every 3 to 4 hours. However, her parents had changed formula multiple times in an attempt to determine which formula "she likes best."
She was the 4,300-gram product of a full-term pregnancy to a 25year-old GlPl female. The mother denied any history of alcohol, drug, or medication exposure during the pregnancy. It was a spontaneous vaginal delivery. During the first several hours of life, the baby started having multiple episodes of bilious emesis and was transferred to a tertiary care facility. During that time, she passed several meconium stools. An abdominal x-ray revealed dilated loops of bowel, and during the next week, she continued to have episodic vomiting. A barium examination of the small intestine was reportedly normal save for what was called gastroesophageal reflux, and eventually the vomiting stopped. She was started on metoclopramide (Reglan) and was discharged home after a 1-week hospitalization with a diagnosis of gastroesophageal reflux.
Developmental history was normal. The family was remarkable for a maternal uncle who was born with "dead bowel." Both parents are eastem European; there is no history of consanguinity.
On physical examination, she was a pale, ill-appearing child who required moderate stimulation to elicit a reaction. Pulse was 170, respiratory rate 56, blood pressure 118/61, and rectal temperature 39.2° C. Weight was in the 75th percentile, length in the 50th percentile, and head circumference in the 75th percentile. HEENT exam was unremarkable. Lungs were clear. S1 and S2 were normal without murmurs. The abdomen was distended and tympanitic to percussion, with decreased bowel sounds. There was no tenderness, organomegaly, or mass. Neurologic examination was unremarkable.
Pertinent laboratory evaluation on admission included hemoglobin 12.7 g/dL, white blood count 2,600/mm3 with 4% neutrophils, 21% immature neutrophils, 46% lymphocytes, 22% monocytes; and platelet count 389,000/mmp 3. Electrolytes and liver function testing were normal, save for an albumin of 2.5 g/dL.
Robert Listernick, MD, moderator: Let's go back to the newborn period. What should we be thinking about when faced with a newborn who has bilious emesis?
Karna Murthy, MD, neonatologist: Obviously, the foremost condition to worry about would be a malrotation and a midgut volvulus, as this has the potential to kill the child if not diagnosed immediately. Other possibilities would include lower intestinal obstructions such as or ileal atresia, necrotizing enterocolitis, or sepsis without an anatomic obstruction.
Lisa Abramson, MD, pediatric surgeon: I would add Hirschsprung disease to the list. In contrast to infants who have malrotation with volvulus, who tend to be critically ill, those with Hirschsprung disease generally have feeding intolerance, abdominal distention, and multiple dilated loops of bowel and present gradually over a period of days. While delayed passage of meconium is classic for children with Hirschsprung disease, it is not universally present.
Sri Pillai, MD, pediatrie surgeon: I'd also add that the presence of abdominal distention tends to suggest a more distal obstruction such as in the distal ileum or colon.
Timothy Sentongo, MD, pediatric gastroenterologist: Bilious emesis in a newborn infant is a surgical problem until proven otherwise. However, we shouldn't forget the medical causes of distal bowel obstructions, such as meconium ileus secondary to cystic fibrosis and hypothyroidism.
Dr. Listernick: What should be the radiographic approach to this patient?
Dr. Abramson: If it is clear that there is a distal obstruction, I would perform a barium enema. However, in this patient, I would start with a "limited" contrast examination of the upper intestine to rule out a malrotation and volvulus. The radiologist can instill a small amount of water-soluble contrast via nasogastric tube so that we can identify a normal duodenal sweep and the position of the ligament of Treitz. The "limited" amount of contrast would not preclude the performance of a subsequent barium enema if deemed necessary.
Tamar Ben-Ami, MD, pediatrie radiologist: A water-soluble contrast enema can be both diagnostic and therapeutic in cases of lower bowel obstruction due to either a meconium plug or meconium ileus.
Dr. Listernick: Is the sensitivity of these tests 100% for diagnosing either a malrotation, a distal small bowel obstruction, or Hirschspning disease?
Dr. Ben-Ami: Nothing is 100%! Rarely, a midgut volvulus can be missed if the obstruction is incomplete or if the bowel is twisted in such a way that the jejunum has moved to its normal position on the left side of the abdomen. The radiographic literature suggests that the transition zones in as many as 10% of cases of Hirschsprung disease are missed by contrast enemas.
Marieta Reynolds, MD, pediatrie surgeon: Certainly there are other causes of intestinal obstruction that may be missed by a quick examination of the upper intestines followed by an enema, such as a Meckel's band or an internal hernia. However, time may be limited in evaluating a sick infant with a possible bowel obstruction and there may not be time to perform serial x-rays following the contrast as it travels through the small intestines.
Dr. Sentongo: I'd also point out that the diagnosis of gastroesophageal reflux in the newborn period, particularly with a history of bilious emesis, is somewhat suspect.
Dr. Ben-Ami: The plain films showed massive dilation of the small intestines with multiple air-fluid levels. A peculiar finding is that one gets a sense that the entire colon is denser than normal; the resident on call that night assumed that there was residual contrast from a previous barium examination which, in fact, the patient had months earlier (Figure 1, see page 474). There was clearly dense material in the colon.
Dr. Listernick: What was the thought process when this infant was admitted?
Dr. Abramson: This infant appeared sick with a distended abdomen. She received vigorous fluid resuscitation that made her look and feel much better. We were concerned about an acute distal obstruction such as intussusception. We have seen rare cases of appendicitis in infants that are often difficult to diagnose. Although this could have been a late diagnosis of Hirschsprung disease, this was somewhat lower down on our diagnostic list. In retrospect, I don't believe that we fully appreciated the significance of the colon's appearance, but we believed that the appropriate diagnostic study was a water-soluble contrast enema.
Dr. Ben-Ami: On the enema, the colon appeared to be normal in diameter; there was no evidence of intussusception. The cecum was normally positioned in the right lower quadrant. Although there was no colonic transition zone characteristic of Hirschsprung disease, the distal small bowel was normal in caliber, whereas the proximal small bowel was markedly dilated. This finding raised the possibility of total colon Hirschsprung disease. In addition, even though the original reading stated that the colon appeared normal, the flexures are rounded and ill-defined, also highly suggestive of total colon Hirschsprung disease (Figure 2, see page 477). Other suggestive findings of this disease that were not present in this patient include a microcolon or massive reflux of barium into either the small bowel or the appendix.
Dr. Abramson: When we showed the x-ray findings to the mother and mentioned the possibility of Hirschsprung disease, she instantly remembered that that is what her brother had had as an infant!
Dr. Listernick: If that is what she has, why did it take four months for her to come to medical attention, and why had she been gaining weight and having normal bowel movements?
Figure 1. Dense material throughout the colon seen on a plain film of the abdomen.
Dr. Reynolds: We know that infants with Hirschsprung disease who are breastfed may present later in the first year of life. Perhaps this child's multiple formula changes were actually an indication that she wasn't tolerating her feedings particularly well.
Dr. Listernick: How sensitive and specific for the presence of Hirschsprung disease is the history of the failure of passage of meconium by 48 hours of life?
Dr. Reynolds: Any child who fails to pass meconium by age 48 hours should be evaluated for Hirschsprung disease. However, there are certainly a number of children who have a normal history of neonatal bowel movements who ultimately are found to have Hirschsprung disease. Of course, parents' recall of bowel movements during the first 2 days of life may be sketchy at best.
Dr. Judah Jona, MD, pediatric surgeon: It is erroneous to assume that the longer the segment of involved colon the more severe the disease at presentation, as evidenced by this case.
Dr. Listernick: What is the approach when Hirschsprung disease is suspected?
Dr. Abramson: If the infant is not acutely ill, we perform a barium enema to try to document a transition zone. If one is found, or if we clinically suspect Hirschsprung disease even if the enema is normal, we will perform multiple bedside rectal suction biopsies, which requires no sedation. Assuming there is adequate tissue depth, the finding of the absence of ganglion cells confirms the diagnosis.
Dr. Listernick: What is the sensitivity and specificity of this procedure for diagnosing Hirschsprung disease?
Pauline Chou, MD, pediatrie pathologist: An inexperienced surgeon might biopsy the bowel too low, near the pectinate line, the transition zone in the anal canal between the simple columnar epithelium of the rectum and the stratified epithelium of the anal canal. This area normally has a low number of ganglion cells, and Hirschsprung disease might be mistakenly diagnosed. If the surgeon doesn't take a deep enough sample, or if a large lymphoid follicle is biopsied, similarly erroneous results may occur. A good pathologist should always recognize that the sample is inadequate or obtained from a site below the pectinate line.
One should never make the diagnosis of Hirschsprung disease on a frozen section because it's difficult to distinguish between endothelial cells and ganglion cells. The Diff Quick stain, a fast, easy technique that highlights ganglion cells, has been proven extremely useful, allowing us to guide the surgeons in the operating room as to where the transition between abnormal and normal colon begins.
Figure 2. A barium contrast enema showed no evidence of intussusception. The cecum was normally positioned in the right lower quadrant, and the distal small bowel was normal in caliber. However, the proximal small bowel was markedly dilated and the flexures rounded and ill-defined. These finding raised the possibility of total colon Hirschsprung disease.
Dr. Abramson: At surgery, we found grossly dilated proximal ileum. We performed multiple biopsies of the colon and distal ileum, which confirmed total colon Hirschsprung disease with distal ileal involvement.
The treatment of total colon Hirschsprung disease is a multistage procedure. First, we identified a normal segment of ileum that could be used as a stoma for the diverting ileostomy. In the future, we will create a ganglionated colonie patch, called a Kimura patch, in which abnormal agangl ionic colon is attached to normal ileum to create a rectal reservoir. This reservoir prevents the children from having constant diarrhea throughout the day. Finally, this combined segment is "pulled through" to create a functioning rectum. AH the stages generally are performed within several months of the initial surgery.
Dr. Listernick: What is the functional outcome of these children?
Dr. Reynolds: There's no question that the morbidity of children with total colon Hirschsprung disease is greatly increased over those who have the more typical sigmoid short aganglionic segment. The former have more frequent problems with diarrhea, dehydration, and enterocolitis.
Dr. Listernick: What should we make of the family history of Hirschsprung disease?
Barbara Burton, MD, geneticist: A great deal! There are several genes that predispose one to the occurrence of Hirschsprung disease, and a parent can be a silent carrier of one of those genes. Approximately 12% of people with Hirschsprung disease have a chromosomal disorder, most commonly Trisomy 21, It is seen in a number of single gene disorders, such as Bardet-Biedl syndrome, cartilage-hair hypoplasia, Waardenburg syndrome type 4, congenital central hypoventilation syndrome, and multiple endocrine neoplasia type 2, which is caused by a mutation in the RET proto-oncogene. Mutations in RET are estimated to occur in as many as 40% of cases of short-segment Hirschsprung disease and as many as 80% of total colonic aganglionosis. All individuals who have Hirschsprung disease should be examined for other manifestations of these above syndromes. It might also be reasonable to screen all affected patients for RET mutations.
Dr. Listernick: Why would the mother not express the gene if she were a carrier?
Dr. Burton: There might be other genes that modify its expression. This clearly occurs in some families in whom multiple members, but not all genotypically abnormal individuals, are affected.
Katrin Carlson, PhD, cytogeneticist: RET is a tyrosine kinase receptor expressed by neural crest cells. Gain of function mutations in RET may lead to multiple endocrine neoplasia syndromes type IIA and IIB, whereas loss of function mutations are often found in individuals with Hirschsprung disease. However, there is clearly some overlap, as some patients who have multiple endocrine neoplasia syndrome also develop Hirschsprung disease.
Dr. Listernick: Thank you, everybody.