Introduction and History
Hirschsprung disease represents a common cause of intestinal obstruction in newborns. These patients do not have an actual mechanical obstruction but rather a serious disturbance in normal colonic peristalsis because of a lack of ganglion cells. The disease was named after Dr. Harald Hirschsprung, who presented what is considered a classic description of this condition at the Pediatric Congress in Berlin in 1886. His description, however, referred only to a congenital dilation of the colon, without a real understanding of the histologic features and pathogenesis.
The recognition that aganglionosis in the distal part of the colon is the main cause of symptoms in these patients led to the surgical treatment used presently. Swenson and Bill described the first such approach, consisting of resection of the aganglionic segment and pull through of a normal piece of colon to be connected to the anal canal. Variations of this basic principle were subsequently introduced into the literature, including the Duhamel procedure, the Rehbein approach in 1959, the Soave operation, the Boley modification, and the De La Torre transanal resection of aganglionic bowel. More recently, laparoscopic techniques have been added to the treatment approach.
Etiology, Pathophysiology, and Incidence
The absence of ganglion cells in the submucosal (Meissner) plexus and in the intermuscular (Auerbach) plexus in the distal intestine, as well as a marked increase in nerve fibers that extend into the submucosa, are the hallmarks of this condition. Patients also have an increase in acetylcholinesterase activity. The lack of ganglion cells affects the most distal part of the rectosigmoid in about 80% of cases—the so-called “typical cases” of Hirschsprung disease. The extent of the aganglionosis varies; it may reach anywhere from the descending colon to the transverse colon, where it is called “long-segment Hirschsprung disease,” and it may even affect the entire colon (“total colonic aganglionosis”). Reports have been made of patients with universal aganglionosis (total intestinal aganglionosis), a condition that thus far has been incompatible with life.
The fecal stasis that occurs in the proximal normal bowel does not produce a fecal impaction alone, as in cases of idiopathic constipation. These patients have other poorly understood functional abnormalities that may explain other more serious symptoms. The fecal stasis leads to bacterial overgrowth, which produces explosive diarrhea, abdominal distention, fever, and a very serious toxic condition. An inflammatory infiltrate of the intestinal mucosa occurs, the mucosa becomes ulcerated, and bacteria then traverse the intestinal epithelium. Abnormal bacteria proliferate, particularly Clostridium difficile . This condition is called “enterocolitis” and may occur from the first hours after the baby is born before he or she undergoes surgery or even after completion of a successful operation. Unknown factors contribute to this serious and mysterious condition. The babies become extremely sick, with development of an endotoxemia that is lethal if it is not treated aggressively and effectively.
The incidence of Hirschsprung disease has been reported to be 1 in 4000 to 1 in 7000. The disease has a definite male preponderance except for the long-segment type of Hirschsprung disease, in which the ratio seems to be 1:1 male to female. Offspring of male and female patients with long-segment Hirschsprung disease have an approximately 20% chance of being affected.
Traditionally, Hirschsprung disease was suspected in children who had constipation. However, during the past several decades the index of suspicion for this condition has increased, leading to earlier diagnosis. In the United States today, it is rather unusual to find an undiagnosed school-aged patient with this condition.
The most common manifestation of Hirschsprung disease during the newborn period is the lack of passage of meconium during the first 24 hours, which is highly suspicious for Hirschsprung disease. Subsequently, the abdomen becomes distended and the baby vomits. Symptoms progress as in any other case of colonic obstruction. The baby looks toxic, lethargic, dehydrated, and septic. Rectal stimulation with a thermometer or a finger provokes characteristic explosive bowel movements, followed by passage of large amounts of fetid gas, giving the baby temporary relief. However, a few hours later the symptoms recur.
A small percentage of patients may survive without medical help. In time, clinical signs of severe constipation, abdominal distension, and failure to thrive develop.
Untreated patients with Hirschsprung disease who survive and reach school age have severe megacolon. These patients must be differentiated from patients with idiopathic constipation. Patients with Hirschsprung disease have a large amount of stool located in the very dilated proximal colon. However, the distal, aganglionic segment is usually empty. Patients with severe idiopathic constipation have a megarectum and fecal impaction located all the way down to the area of the anal canal. A huge fecal impaction can be felt upon examination with a finger, whereas patients with Hirschsprung disease are described as having an empty rectum. Patients with Hirschsprung disease characteristically do not soil their underwear. Only a very small percentage of all patients with Hirschsprung disease reach this age. Most undiagnosed patients die in the early stages of the condition because of a lack of treatment.
A high index of suspicion is the most important element for the early diagnosis and treatment of Hirschsprung disease. A plain abdominal radiograph shows very dilated loops of bowel. Unfortunately, during the newborn period, it is extremely difficult to differentiate small bowel from large bowel on a plain abdominal film. Affected patients have a dilated colon proximal to the aganglionic portion.
Babies with Hirschsprung disease do not have a severe degree of megacolon. However, as time goes by, the colon proximal to the aganglionic segment becomes more and more dilated, showing a dramatic contrast with the nondilated distal bowel and the transition zone. The clinician should not expect to find a conspicuous megacolon with a distal narrow segment in the newborn.
A contrast enema is extremely important for the diagnosis of this condition. Contrast enemas do not demonstrate a transition zone in all cases, and their value varies from institution to institution, depending to a large extent on the experience of the observer. A positive contrast study shows a nondilated distal large bowel with a dilated proximal segment. Sometimes the diagnosis is obvious, but sometimes the transition between dilated and nondilated sections is not well demarcated. In patients with total colonic aganglionosis, no colonic dilatation occurs; the dilatation mainly affects the small bowel.
The contrast study in these babies must be performed by an experienced pediatric radiologist. A catheter is introduced through the anus not more than 4 cm. Passing the catheter more than 4 cm will result in the injection of the contrast material directly into the dilated portion of the colon, bypassing the nondilated portion and the transition zone and therefore precluding a diagnosis. Failure to pass the contrast material in the 24 hours after the study is considered to be highly suggestive of Hirschsprung disease.
Occasionally, an unattended baby may experience a bowel perforation. The perforation is usually in the cecum, a presentation which in itself is suggestive of Hirschsprung disease. When a bowel perforation occurs, emergency surgery is necessary.
Inflation of a balloon in the rectum of a healthy person produces a drop in pressure in the anal canal. This drop in pressure is called the “rectoanal inhibitory reflex” and is present in all healthy persons. Anorectal manometry in children with symptoms suggestive of Hirschsprung disease is performed with the goal of eliciting such a reflex. A lack of relaxation of the anal canal is considered diagnostic of Hirschsprung disease. However, a significant degree of controversy exists about the value of this diagnostic test, and most pediatric centers and surgeons do not use it.
A rectal biopsy is the definitive diagnostic test in children with Hirschsprung disease; however, it requires interpretation by an experienced pediatric pathologist. Most surgeons and pediatric centers use a suction biopsy for the diagnosis of this condition. A suction biopsy can be performed at the bedside and is considered highly diagnostic. One of the limitations of this type of procedure is that the specimen may not include submucosa and therefore is not sufficient for an accurate diagnosis. Another limitation is performing the biopsy too distal in the rectum, where absence of ganglion cells is supposedly normal. Although the diagnosis of aganglionosis can be made with a suction rectal biopsy, the length of the abnormality cannot be determined with this biopsy.
A full-thickness rectal biopsy provides a much better specimen; however, induction of anesthesia is required to perform this biopsy. Full-thickness rectal biopsies are still performed by pathologists who are not familiar with using a suction biopsy to diagnose Hirschsprung disease and by surgeons who do not have access to the suction biopsy device. An increase in the activity of acetylcholinesterase is also considered diagnostic, but not all pathologists rely on this analysis.
A newborn baby who has abdominal distension, vomiting, dehydration, and explosive fetid bowel movements and who looks toxic and lethargic needs emergency management. Administration of intravenous fluids and antibiotics and—most importantly—rectal irrigations must be started promptly.
We do not know much about the pathophysiology of the enterocolitis that may lead to death in these children. However, stasis seems to be the most important predisposing factor, and therefore decompressing the colon leads to dramatic improvement; this maneuver can be performed on a temporary basis by passing a rectal tube. Sometimes the characteristics of the meconium and the fecal matter may interfere with the drainage of gas and liquid from the colon through the tube. Therefore, the tube should be irrigated with small amounts of saline solution (10 to 15 mL) and moved back and forth. This technique produces a spectacular decompression of the colon with a dramatic improvement of the patient’s symptoms. Sometimes the babies are so sick that one cannot consider performing a contrast enema because of the risk of perforation. The rectal irrigations and administration of antibiotics and intravenous fluids allow improvement so that other diagnostic and therapeutic procedures can be performed. When the baby recovers from the acute enterocolitis, a rectal biopsy is performed, establishing the definitive diagnosis.
Traditionally, patients with Hirschsprung disease were treated in three stages. The first stage consisted of a diverting colostomy, which decompressed the colon and allowed the baby to recover and go home. Most surgeons create a colostomy in the right transverse colon (or an ileostomy was created in cases of total colonic aganglionosis). The second stage was resection of the aganglionic segment and pull through of the normal bowel to be anastomosed to the anal canal with the technique of choice. The third stage was closure of the colostomy after the baby recovered completely from the main pull-through procedure.
A two-stage repair was subsequently devised. In the first operation, a colostomy was created immediately proximal to the transition zone, called a “leveled colostomy.” In the second and definitive stage, surgeons performed the pull through, pulling down what used to be the colostomy into the anal canal and leaving the patient without a protective colostomy.
The most recent approach involves performing the main pull through during the first few days, weeks, or months of life as a primary procedure without a protective colostomy. This approach avoids two operations (a colostomy and colostomy closure) and is feasible without adding extra morbidity. However, a colostomy is still an extremely valuable option for some babies with Hirschsprung disease. For instance, a very sick baby who has other aggravating factors and/or enterocolitis still could be a candidate for this type of preliminary procedure. In general, most surgeons take these babies to the operating room for a pull-through procedure with a plan for a colostomy if necessary. When the technical circumstances of a pull-through procedure are not satisfactory and the surgeon has any questions about the viability of the pulled-through bowel in terms of blood supply or any other type of technical difficulty, he or she can always protect the pull through with a proximal colostomy. Also, in hospitals, cities, or countries with inadequate conditions and infrastructure, a surgeon can save the baby’s life with a colostomy.
To perform the primary neonatal pull through, one should expect to have a newborn baby in excellent clinical condition with a colon that has been completely decompressed. Several days are usually required to prepare such a baby for surgery, using parenteral nutrition and colonic irrigations.
The Swenson procedure was the first to be designed specifically to treat Hirschsprung disease and was used for many years by Dr. Swenson himself. A few surgeons still perform the original Swenson operation. For this type of pull-through operation, a total body preparation is performed so the surgeon can approach the abdomen or the perineum several times either by turning the patient from the supine into the prone position or simply by lifting the legs up to approach the perineum and then down to approach the abdomen. Transabdominally, the aganglionic portion of the colon is resected, including the most dilated portion of the bowel ( Fig. 18-1, A ). In a case of typical Hirschsprung disease, only the splenic flexure must be mobilized. On the other hand, when dealing with long-segment type disease, it may be necessary to mobilize the right colon to obtain enough length for the pull-through procedure. The aganglionic portion of the colon below the peritoneal floor is dissected in a very precise manner, staying as close as possible to the rectal wall down to the level of the levator ani muscle. The dissection of the rectum includes the ligation of the middle hemorrhoidal vessels and the use of cautery in the perirectal vasculature ( Fig. 18-1 ). Special care is taken to preserve the anal canal above the dentate line. The aganglionic segment of the colon is resected and normal colon is pulled through ( Fig. 18-1, B and C ) and anastomosed by a conventional, transanal, hand-sewn technique. The basic principles of this procedure are still observed in most modern operations. Dr. Swenson developed an extraordinary experience with this operation with good outcomes. However, other surgeons reported damage of important pelvic nerves, and thus other procedures were designed, such as those created by Duhamel and Soave.