Cover story

Update on intussusception

By Julian Orenstein, MD

The recently recognized link between intussusception and rotavirus vaccination has led to renewed awareness of this uncommon but potentially fatal condition. New diagnostic and treatment methods are challenging barium enema as the gold standard.

At least half of all the infants I see with intussusception are referred to the emergency department when their pediatrician suspects the diagnosis. It's the other half who are diagnostically challenging. Like office-based pediatricians, I must remember that a vomiting child who looks fine in the three to five minutes I spend with the family is not always "acute gastro." Intussusception usually comes to the top of the differential in the presence of a fairly distinctive triad of factors: vomiting without diarrhea; colicky, intermittent pain; and a positive guaiac test for occult blood. As with so many conditions that occur somewhat infrequently, the standard admonition that to make the diagnosis one must keep it in mind applies here. Thanks to the recent publicity about a link between intussusception and rotavirus vaccination (and the subsequent withdrawal of the vaccine), this is not such a tall order.

Management of intussuception has evolved rapidly during the past 10 years. Barium enema used to be the accepted diagnostic modality and, generally, the definitive treatment. Recent studies have shown, however, that air or saline enema reduction under ultrasound or fluoroscopic guidance is not only effective but does not increase the rate of perforation or the need for subsequent surgery. These techniques also substantially reduce the patient's exposure to radiation. Before reviewing the newer technologies, I will address the pathophysiologic basis for intussusception and its clinical presentations.

How it happens

Intussusception is prolapse of a segment of the intestinal tract (the intussusceptum) and subsequent telescoping of that segment distally.

Pathophysiology. The lead point of the intussusception is dragged by peristaltic contractions. Most commonly, the distal ileum at the ileocecal valve is prolapsed into the colon. The intussusceptum may sometimes traverse the entire length of the colon, causing a rectal prolapse with its mesenteric attachment. This condition initially causes venous obstruction, bowel edema, and mucosal bleeding, resulting in the early finding of occult rectal blood. With progressive edema and compromise of the bowel wall, the classic "currant jelly stool" may be seen, often about 24 hours after symptoms begin. If arterial flow becomes compromised, bowel necrosis, perforation, and gangrene develop. Intussusception does not spontaneously reduce itself in the vast majority of children. If intussusception goes unrecognized and untreated, it can progress to sepsis and death.

In children younger than 2 years with "idiopathic" intussusception, the usual lead point is thought to be a hypertrophied Peyer's patch (an elevated area of lymphoid tissue on the mucosa of the small intestine). The cause of the patch may itself be unknown or may represent local lymphadenopathy following resolution of an otherwise innocuous bout of gastroenteritis. In the child who is older than 3 years, a pathologic lead point such as a tumor or Meckel's diverticulum is increasingly likely to be the culprit. As with the Peyer's patch, the lead point usually is at the ileocecal site, but ileoileal, jejunoileal, or colocolic intussusceptions are also possible.

In children older than 2 years, any localized mass effect resulting from an intra-abdominal condition is more likely to become a lead point than in younger children. The list of diagnoses that can cause intussusception is almost limitless, therefore, and encompasses Meckel's diverticulum, a bowel wall hematoma secondary to blunt trauma, a hair ball (trichobezoar), and Henoch-Schönlein purpura. Intussusception also can be a postoperative complication.^1­3 Table 1 lists conditions that have produced intussusception. Whatever the cause, the child's symptoms usually are indistinguishable from those of the infant with the "classic" form of intussusception, though the degree of pain, vomiting, and rectal bleeding varies. When adults develop intussusception, the lead point often is neoplastic.

Intussusception and rotavirus vaccination. Most recently, intussusception has been associated with RotaShield vaccination for rotavirus, which had been added to the recommended childhood immunization schedule for 1999. In July of that year, however, the Centers for Disease Control and Prevention (CDC) revealed that 15 instances of intussusception following the first dose of vaccine had been reported to the vaccine adverse event reporting system (VAERS). In 13 of these children, the condition developed in the first week after the vaccine was administered.⁴ In quick succession, vaccine use was suspended pending a review of the data by the Advisory Committee on Immunization Practices (ACIP) of the CDC, and the American Academy of Pediatrics (AAP) issued a public health advisory and then a policy statement recommending that routine rotavirus vaccination be discontinued.⁵ Finally, the rotavirus vaccine was dropped from the schedule of recommended childhood immunizations for 2000.⁶ The manufacturer has voluntarily withdrawn RotaShield from the US market.

The fate of the rotavirus vaccine was not a complete surprise to those who had followed its development. During prelicensure studies, several infants developed intussusception, but the overall frequency did not reach statistical significance. The vaccine's packet insert included a warning that intussusception was a potential adverse reaction, and a mechanism was established for close postlicensure surveillance. The number of cases of intussusception reported to VAERS was higher than expected, however. Although investigators have identified no causal effects between the vaccine and intussusception, the pathologic specimens of affected children showed lymphoid hyperplasia, which matches the understood pathophysiology of intussusception: The vaccine generates an immune reaction, which can act as a lead point.

Identifying intussusception

Intussusception is most common between the ages of 3 and 12 months. Boys are affected twice as often as girls, though it is not understood why. Because gastrointestinal infection may give rise to a lead point, it is most common in the summer (enterovirus season) and winter (rotavirus season).

Classic presentations. In infant ileo-colic intussusception, the typical triad of vomiting, intermittent colicky abdominal pain, and "currant jelly stools" is present in just 20% of cases. The pain, characterized by inconsolable crying, possibly high-pitched or in some way markedly different from the baby's normal cry, usually lasts several minutes and is followed by periods of more normal behavior. The pain precedes vomiting in about three quarters of infants. Because the baby may recently have had an uncomplicated gastroenteritis, the parent (or physician) may be lulled into thinking that the recurrence of vomiting is caused either by a too-early return to milk products or by the failure of the gastroenteritis episode to resolve completely. As the disease process progresses, however, the child becomes more irritable and the bouts of pain last longer. At this point, the pain is almost always accompanied by vomiting, often nonbilious. If the patient has still not been treated or receives a presumptive diagnosis of acute gastroenteritis, grossly bloody stools and bilious emesis develop, anywhere from 12 to 48 hours after onset of symptoms.

Occasionally, in a child younger than 1 year, the clinical picture may be dominated by neurologic symptoms—lethargy, apnea, miosis, or seizures—rather than pain and vomiting.^7,8 Ultimately, however, the appearance of abdominal distention and grossly bloody stools clarifies the diagnosis.

The physical examination may at first appear to be wholly normal—another reason the diagnosis may be overlooked. The infant acts irritable or lethargic briefly during the colicky episodes, which are accompanied by apnea in rare instances, but otherwise does not seem ill. About one in 10 infants has fever. In about half of infants with intussusception, the abdominal exam reveals a mass, which often is described as sausage-shaped. The finding of a sausage-shaped mass in the right upper quadrant with absence of bowel in the right lower quadrant, known as Dance's sign, is pathognomonic of intussusception. Other common findings are abdominal tenderness ( in 60% of patients) and distension (10%). Signs of peritoneal irritation such as rebound guarding may be present as well.

A recent review of "currant jelly stool" indicates that this description of stool is more colorful than it is accurate.⁹ A chart review of 107 proven cases of intussusception showed that the patients' grossly bloody stools generally were described as "mucusy," "bloody," or "red." Overall, about 40% of patients have occult stool blood and another 40% grossly bloody stool. The remaining 20% test guaiac negative for blood but have a pattern of pain and vomiting that leads to consideration of the correct diagnosis. In these children, intussusception probably has just recently developed.

An adequate history and physical should generate suspicion of intussusception, which then leads to appropriate radiologic studies, discussed below. Lab studies are limited to complete blood count, electrolytes, and cultures of stool, urine, and blood to evaluate the child for serious dehydration and possible infectious causes. No specific lab tests can confirm the diagnosis or rule it out.

Presentations in special populations. Neonates, both full-term and premature, may develop intussusception, but only rarely. In term babies, ileocolic intussusception develops from a pathologic lead point; symptoms resemble those in older infants. In premature infants, the clinical picture resembles that of necrotizing enterocolitis, and therefore the intussusception may be advanced when it finally is identified.¹⁰ Older children and adults may develop intussusception from a pathologic lead point anywhere along the small or large bowel, causing ileoileal or colocolic intussusception. In intussusception associated with Henoch-Schönlein purpura, for example, ultrasound has shown ileoileal intussusceptions that resolve spontaneously.¹¹

Differential diagnosis. Faced with a child with a combination of fever, vomiting, and abdominal pain, pediatricians often think first of acute viral or bacterial gastroenteritis, which, as mentioned, frequently precedes intussusception. Since occult or obvious rectal blood may be present in either intussusception or gastroenteritis, its presence doesn't help distinguish the two. But if pain is a significant component of gastroenteritis, its presence should prompt consideration of a radiographic study to rule out intussusception. The pediatrician should consider the many conditions that produce pain and vomiting in the infant or child who appears ill or dehydrated. Table 2 lists these conditions.

Confirmation and management

Several advances in the past several years, both radiologic and surgical, have led to changes in the approach to imaging and management of intussusception.

Plain film. As in the past, the first study is a plain film of the abdomen. Initial X-rays may be normal but as the intussusception progresses, a picture of small bowel obstruction without a localizing mass can be seen. A more distinctive pattern emerges when the intussusceptum crosses the hepatic flexure or is in the transverse colon: The lower margin of the liver, in the right upper quadrant, is lost or indistinct, and dilated loops of small bowel can be seen occupying the left half of the abdomen. In fewer than half of patients with intussusception, a "colon cut-off" sign can be seen as a distinct mass bordering on an otherwise normal distal colon (Figure 1). Free air on plain film indicates perforation and directly contraindicates further imaging procedures. Surgery should be performed immediately.

Ultrasound. If plain film fails to make the diagnosis clear, ultrasound is the preferred next imaging study and makes the diagnosis in more than 90% of patients. Typical ultrasound findings are a longitudinal, layered oval mass and a series of concentric rings on cross-sectional views. Under optimal circumstances, ultrasound combined with color-flow Doppler can detect ischemia and impending perforation, which should lead directly to a surgical procedure without attempting radiographic guided reduction.

Barium enema used to be the gold standard for diagnosing and treating intussusception (Figure 2). A successful reduction could be relied on in 70% to 90% of typical cases. Higher reduction rates were generally associated with cases that presented within 24 hours; reduction rates were lower in more advanced cases. If there is no evidence of shock or perforation, the presence of symptoms for more than 48 hours is not a contraindication for a radiographic reduction. It used to be believed that barium enema should not be used in children with symptoms for more than 72 hours, but some observers now think that in the absence of any complicating factors even this limitation may be unnecessary.¹² It also has been discovered that repeated or "delayed" attempts at hydro- or aerostatic reduction can be successful if with each succeeding try the intussusceptum can be moved a little further back into place.¹³ When repeated attempts achieve no further movement, surgery is indicated.

Other reduction materials. During the past 10 to 15 years air, oxygen, saline, and aqueous contrast materials have shown rates of success that are similar to those of barium enema (70% to 90%).^14,15 As experience with these different materials accumulated, it became evident that failure to achieve a reduction was related to generated intracolonic pressures rather than to the type of hydro- or aerostatic agent used. To be successful, pressures of about 60 mm Hg of whatever agent is used are required to push back the intussusceptum and about 100 mm Hg to achieve a definitive reduction. If pressures of 120 mm Hg are required, perforation is more likely—not only because of the pressures generated, but because of the likelihood that the bowel is already compromised or necrotic. One slight disadvantage of air as a contrast medium is that demonstration of air reflux into the small bowel is not always as clearcut as it is with barium or saline.

Radiologic reduction. Recently, ultrasound guidance of hydrostatic reduction of intussusception has been demonstrated to be as effective as traditional air or barium enema reduction under fluoroscopy.¹⁵ Should these findings be replicated, ultrasound-guided saline reduction may well replace the traditional barium enema as the gold standard procedure. This procedure is more widely accepted and used in Europe than in the United States.

Before the radiologist attempts to reduce an intussusception, however, two tasks are critical: Notify the surgeon, whose services may be required if the radiologic reduction is unsuccessful, and replace fluid deficits. This may mean resuscitating a child in borderline (compensated) shock. The volume of fluid lost through vomiting can be substantial and can, by itself, lead to hemodynamic instability. Internal hemorrhage and fluid loss secondary to bowel necrosis may contribute to this fluid deficit and must be considered if a child cannot be adequately resuscitated. A child in shock must be surgically managed and should not undergo radiographic enema. Before attempting a radiologic reduction, the radiologist may administer prophylactic antibiotics to cover enteric flora because of the risk of perforation. Should the intussusception recur, a radiographic reduction is again appropriate.

Surgical intervention. Adverse outcomes of radiologic reduction are bowel perforation, which occurs 1% to 3% of the time, and an unsuccessful reduction, which happens 10% to 30% of the time. Both these circumstances require surgical intervention, so radiologic reduction should be conducted at a facility where an appropriately trained surgeon and a radiologist experienced in performing such reductions are available. At laparotomy, the surgeon most often is able to reduce the intussusception manually. In rare instances, necrotic or gangrenous bowel must be resected, along with any pathologic lead points that may be identified.¹⁶ Another surgical option is laparoscopy to verify radiographic reduction that contrast reflux has failed to reveal. If the intussusception remains, laparoscopic instruments may be used to reduce the bowel directly. The appendix is often removed when the intussusception is reduced surgically.

After the reduction. Once the intussusception is reduced, the child is no longer at significant risk for complications such as bowel obstruction, necrosis, perforation, peritonitis, sepsis, or shock. Intussusception does recur, however, in 4% to 10% of cases, usually within 24 hours of the initial reduction.¹⁷ Some of these recurrences may be attributable to an incomplete reduction. Other illness or complications are not associated with intussusception reduction.

Following successful reduction of intussusception and 24 hours of observation, the child should be discharged if no contraindications develop. The child's pediatrician should closely follow up during the next two to three days. In children for whom a pathologic lead point is suspected, for example a neonate or child older than 3 years, a computed tomography scan or Meckel scan should be performed to look for masses that may themselves require surgical intervention. This should be done while the child is hospitalized for postreduction observation.

Keeping costs down

Developments in radiologic and surgical management of intussusception have been coupled with advances in analyzing associated costs of care. A study comparing treatment, total costs, and lengths of stay for intussusception in New York showed that hospitals that provided care for large numbers of children or that had a training program in pediatrics or pediatric radiology delivered the most cost-effective care. The mean cost and length of stay for intussusception at a children's hospital, for example, were $1,530 and 2.2 days compared with $2,308 and 2.8 days at a hospital that was not exclusively for children.¹⁸ Investigators concluded that hospitals specializing in pediatric care are more likely to perform radiologic reduction successfully than nonspecialized hospitals and argued that referral of suspected cases of intussusception to a hospital with experienced pediatric radiologists and surgeons would improve overall outcomes and minimize surgical interventions.

Looking ahead

New radiographic approaches to intussuception that emphasize evaluation and reduction using ultrasound may be expected to supplant more traditional techniques. The new methods have results just as good as those of more traditional means, are not associated with greater rates of complications or surgery, and reduce exposure to radiation. When surgery is required, laparoscopic approaches may be used increasingly in the future, avoiding the need for more invasive procedures.

REFERENCES

1. Conners GP, Weber CE, Emmens RW: Intussusception following a baby walker injury. J Emerg Med 1999; 17(2):269

2. Dalshaug GB, Wainer S, Hollaar GL: The Rapunzel syndrome (trichobezoar) causing atypical intussusception in a child: A case report. J Pediatr Surg 1999;34(3):479

3. de Vries S, Sleeboom C, Aronson DC: Postoperative intussusception in children. Br J Surg 1999;86(1):81

4. Intussusception among recipients of rotavirus vaccine—United States, 1998­1999. MMWR 1999; 48(27):577

5. AAP Committee on Infectious Diseases: Possible association of intussusception with rotavirus vaccination. Pediatrics 1999;104:575

6. What's new in the 2000 immunization schedule. Contemporary Pediatrics 2000;17(1):32

7. Birkhahn R, Fiorini K, Gaeta TJ: Painless intussusception and altered mental status. Am J Emerg Med 1999;17(4):345

8. Conway EE: Central nervous system findings and intussusception. Pediatr Emerg Care 1993;9:15

9. Yamamoto LG, Morita SY, Boychuk RB, et al: Stool appearance in intussusception: Assessing the value of the term "currant jelly" Am J Emerg Med 1997;15(3):293

10. Wang NL, Yeh ML, Chang PY, et al: Prenatal and neonatal intussusception. Pediatr Surg Int 1998; 13(4):232

11. Swischuk LE, John SD, Swischuk PN: Spontaneous reduction of intussusception: Verification with US. Radiology 1994;192(l):269

12. Okuyama H, Nakai I-L, Okada A: Is barium enema reduction safe and effective in patients with a long duration of intussusception? Pediatr Surg Int 1999;15(2):105

13. Sandler AD, Ein SK, Connolly B, et al: Unsuccessful air-enema reduction of intussusception: Is a second attempt worthwhile? Pediatr Surg Int 1999;15(3-4):214

14. del-Pozo G, Albillos JC, Tcjedor D, et al: Intussusception in children: Current concepts in diagnosis and enema reduction. Radiographics 1999;19(2):299

15. Gonzalez-Spinola J, Del Pozo G, Tejedor D, et al: Intussusception: The accuracy of ultrasound-guided saline enema and the usefulness of a delayed attempt at reduction. J Pediatr Surg 1999;Jun;34(6):1016

16. Poddoubnyi IV, Dronov AF, Blinnikov 0I, et al: Laparoscopy in the treatment of intussusception in children. J Pediatr Surg 1998;33(8):1194

17. Daneman A, Alton DJ, Lobo E, et al: Patterns of recurrence of intussusception in children: A 17-year review. Pediatr Radiol 1998;28(12):913

18. Chamberlain JK, Huang ZJ, Guagliardo MF, et al: The effects of volume and specialty care on improved outcomes of intussusception. Presented at the 1999 Annual Meeting of the Ambulatory Pediatric Association, San Francisco, May 3

THE AUTHOR is Associate Professor of Pediatrics, University of Virginia, Charlottesville, and Associate Professor of Emergency Medicine, George Washington University, Washington, DC. He is an attending physician in the Emergency Department, Inova Fairfax Hospital, Falls Church, VA.

Julian Orenstein. Update on intussusception.

Contemporary Pediatrics

2000;3:180.