Secondary Logo

Journal Logo

Spontaneous Bowel Perforation in Infants and Young Children: A Clinicopathologic Analysis of Pathogenesis

Chen, Jeng-Chang; Chen, Chiu-Chiang; Liang, Jin-Tung; Huang, Shiu-Feng*

Journal of Pediatric Gastroenterology and Nutrition : April 2000 - Volume 30 - Issue 4 - p 432-435
Original Article

Background: In children beyond the neonatal stage, spontaneous bowel perforation with no specific cause is rare. It has attracted little notice and does not fit into any established clinical category. This often results in treatment delays. To determine the underlying pathogenesis, a clinicopathologic analysis of this group of patients was performed.

Methods: From 1984 through 1997, 15 previously healthy children (10 boys and 5 girls) with bowel perforations were enrolled in this study. Children in whom the specific cause was established were excluded. The ages of the patients ranged from 3 months to 5 years, 5 months (average age: 2 years, 4 months). Data on clinical course, surgical and pathologic findings, and outcomes were analyzed.

Results: All the patients had high fever and acute, watery diarrhea with clinical dehydration. Abdominal distension developed and bowel perforation ensued 5 to 30 days after the onset of diarrhea. In all cases, the perforation consisted of either an isolated patch of bowel (n = 8) or patchy lesions in a segment (n = 7), which was confined to one of the following three anatomic watershed areas: the splenic flexure (n = 7); the lower sigmoid (n = 3); and the ileocecal region (n = 5). One patient died, and three had major sequelae including intra-abdominal abscess, adhesion ileus, and peristomal fistula. Specimens from 13 patients were submitted for histopathologic evaluation. All specimens had acute suppurative inflammation diffusely around the perforation site. Seven had areas of coagulation necrosis of the muscularis propria, especially in the inner circular muscle layer.

Conclusion: In acute diarrheal diseases, dehydration may cause bowel ischemia and lead to perforation. Physicians should be alert to the possibility of this abdominal catastrophe when progressive abdominal distension followed by acute diarrheal episodes occurs in children.

Department of Surgery and *Pathology, Chang Gung Children's Hospital, Chang-Gung University College of Medicine; and †Department of Surgery, National Taiwan University Hospital, National Taiwan University College of Medicine, Taipei, Taiwan

Received October 20, 1999; accepted December 29, 1999.

Address correspondence and reprint requests to Dr. Shiu-Feng Huang, Department of Pathology, Chang Gung Memorial Hospital, 199 Tung-Hwa North Road, Taipei 105, Taiwan.

Bowel perforations inevitably cause fecal peritonitis with life-threatening complications. Necrotizing enterocolitis (NEC) contributes to most perforations in neonates, especially in premature newborns (1–3). However, a rare entity of spontaneous intestinal perforation in the neonate has also been reported (4,5). This condition is not associated with NEC, the intestine has muscle wall deficiency, and the perforation sites tend to be distributed in watershed areas. It is usually accompanied by hemodynamic instability due to perinatal stresses, and ischemic insult is suggested as its cause (4,5).

Bowel ischemia in adults has been a well-recognized clinical event for more than a century (6,7). However, the study of this condition in children has been limited. Adults with bowel ischemia are usually more than 50 years of age and have underlying medical illnesses, such as cardiovascular disease, diabetes, hematologic disorders, or chronic renal failure, which predisposes them to the development of ischemic colitis (6,8,9). Based on severity, three types of ischemic colitis have been described: transient reversible colitis, ischemic stricture, and fulminant gangrene (10). In neonates, the clinical manifestations and physical findings are much more difficult to assess, and diagnosis of ischemic colitis remains a challenging task that has never been performed early in the ischemic process.

We identified a group of children beyond the neonatal stage, in whom bowel perforations had occurred without evidence of preceding events. It was difficult to fit these patients into any established clinical category. The purpose of this study was to determine the definite pathogenesis of these unspecified bowel perforations by reviewing the clinicopathologic finding of bowel perforations in this group of patients.

Back to Top | Article Outline


From January 1984 through December 1997, 29 children more than 1 month of age with bowel perforations were identified. Patients with ruptured appendicitis (n = 10) were excluded from this study. In addition, patients with perforation caused by Pseudomonas sepsis (n = 5), amebic colitis (n = 1), Salmonella infection (n = 2), iatrogenic enema (n = 2), incarcerated inguinal hernia (n = 2), and intussusception (n = 2) were also excluded. Thus, 15 patients (10 boys and 5 girls) with bowel perforations of unspecified cause were enrolled in this study (Table 1). Their ages ranged from 3 months to 5 years, 5 months (average age: 2 years, 4 months). All were previously healthy children. The indications for laparotomy were evident peritonitis and a radiographic finding of pneumoperitoneum. Their clinical courses, surgical and pathologic findings, and outcomes were obtained for analysis.



Back to Top | Article Outline


All 15 patients had initial symptoms of vomiting, watery diarrhea, and fever. Associated respiratory symptoms were noted in five patients, and febrile convulsions in four. Their parents could not relate these manifestations to recent travel, prescription or nonprescription drugs, or any specific meal, such as raw milk or fast-food hamburgers. None of the patients had a recent history of trauma or accident. They had all been initially treated at home after consulting a pediatrician, who had made a preliminary diagnosis of acute viral gastroenteritis. However, most of the parents had not been adequately informed of the indispensability of oral rehydration therapy. Poor feeding and absence of stool passage with progressive abdominal distension developed in all patients.

Eleven patients were brought to the Emergency Department of National Taiwan University Hospital from home, and four were referred from other hospitals after tentative treatment. In physical examination at admission all were febrile (body temperature, 38–40.5°C) and had dehydration indicated by listlessness, tachypnea, tachycardia, little or no urination, wrinkled skin, sunken eyes, and dry mouth. The abdomen was distended with muscle guarding. The initial serum blood urea nitrogen (BUN) and creatinine (CRE) was examined in 8 patients and in 5 of them, results were abnormal (Table 1). Two of the three patients with normal data were referred from other hospitals after supportive treatment. Most of the patients had no data reporting initial blood pressure. One patient was in shock with unmeasurable blood pressure on admission (case 14). Bacterial cultures from blood and freshly passed stools at the time of admission yielded no significant pathogens in any of the patients. Stool examination for leukocytes, ameba, or parasites were negative in all patients. Examinations of fecal excretion for rotavirus in four patients and adenovirus in one patient were negative. The indirect hemagglutination antibody (IHA) and Widal tests were subsequently performed and no significantly increased titers were found.

Surgical examination found the abdomen was rigid with diminished or absent peristaltic sounds. Plain radiographs demonstrated pneumoperitoneum. As a result, emergent exploratory laparotomy was performed in each patient, and bowel perforation was revealed. The duration from initial symptoms until the day of operation ranged from 5 to 15 days, excluding the patient in case 3 for whom the duration was exceptionally long (30 days) due to delayed diagnosis. The perforation was located at the transverse colon in seven patients, sigmoid in three, cecum in two, and terminal ileum in three. The lesion involved a segment of bowel in a patchy manner in seven and only an isolated patch in eight. The affected bowel appeared dusky red with foci of subserosal ecchymotic discoloration. One patient underwent segmental ileal resection with reanastomosis, and the remainder had temporary enterostomy. The enterostomy was performed by exteriorizing the involved segment if feasible. Otherwise, it was fashioned proximal to the perforation to divert the fecal stream. Bowel continuity was re-established 3 to 6 months later.

One patient died of refractory shock consistent with overwhelming sepsis. Fourteen survived with three major morbidities: intraabdominal abscess, adhesion ileus, and ileostomy fistula. The serum BUN and CRE levels of these patients all returned to normal after treatment.

Surgical specimens around the perforation site were submitted for pathologic evaluations in 13 patients. All had acute suppurative inflammation of the bowel wall around the perforation site and diffusely over the peritoneum. In addition, seven patients had areas of coagulation necrosis of the muscle wall, located either at the margins of the perforation site or in the vicinity of the perforation (Figs. 1, 2). The inner circular muscle layer of the bowel was involved in all cases. The mucosa showed focal ulcer or diffuse necrosis. The areas of coagulation necrosis revealed no significant inflammatory cell infiltration, consistent with ischemic necrosis.

FIG. 1.

FIG. 1.

FIG. 2.

FIG. 2.

Back to Top | Article Outline


Pediatric bowel perforations, most centered on NEC occurring in the neonatal period, have been reported in several articles (1–3). However, the bowel perforations in this series occurred beyond the neonatal stage and in previously healthy children. Although these cases were encountered sporadically, they all had the unique symptom of preceding watery diarrhea.

Acute diarrheal diseases are not usually related to bowel perforation unless they are caused by invasive organisms such as Salmonella, Shigella, Campylobacter jejuni, or Entamoeba histolytica. In this series, stool examinations for the presence of mucus, gross or occult blood, leukocytes, and parasites in association with cultures of freshly passed stools and serologic tests (IHA and Widal test) found no evidence of invasive microorganisms. A careful history also excluded possible noninfectious causes of acute diarrhea. Accident or trauma that might lead to perforation also was not reported. The clinical manifestations in the patients in these series resembled viral gastroenteritis, a condition that accounts for 70% to 80% of acute diarrheal diseases in young children (11) and is unquestionably one of the most prevalent and serious clinical syndromes around the world. Typically, the pathogens responsible for acute viral diarrheal diseases remain unidentified during the disease course, and most patients can be treated simply by nonprescription medications and adequate hydration, regardless of a specific diagnosis (12). Thus, it is intriguing to attempt to determine under which circumstances this usually self-limited condition evolves into the catastrophe of abdominal perforation, as in the patients of this series.

Analysis of the data obtained for this series of patients revealed that perforations had a predilection for three areas: the transverse colon near the splenic flexure, the lower sigmoid, and the ileocecal region. These areas correspond with three anatomically vulnerable points in the colonic vasculature that have been termed watershed areas, including the splenic flexure (Griffith's point), rectosigmoid region (Sudeck's point), and ileocecal region (13). These areas have poor or absent anastomoses of marginal arteries, which makes them vulnerable to ischemic injury (6). Grossly, the involved segments of the bowel display localized patchy lesions rather than whole-segment damage, indicating that the pathogenesis involves nonocclusive ischemic injury. It has been reported that localized nonocclusive ischemic injuries classically involve the watershed areas of the colon (14).

This finding was supported by data in this series, in which seven patients showed definite ischemic necrosis of the bowel. The six others had more marked suppurative inflammation because of perforation, which could have masked the original ischemic change. Because the circular muscle and mucosa layers represent two areas with terminal blood supply, derived from two branches, leaving the submucous plexus (13), the finding of coagulation necrosis in circular muscle layers or mucosal necrosis should indicate intramural ischemia. Nevertheless, the mucosa has a much higher regenerative capacity than muscle (4), and mucosal ulceration is therefore not necessarily a universal finding in cases of ischemic bowel. Areas of coagulation necrosis in the muscularis propria, especially in the inner circular muscle should be a more reliable diagnostic finding of ischemia. Nonocclusive ischemic injuries may occur as a result of systemic hypoperfusion secondary to heart failure, shock, sepsis, or aortic surgery (8,15) or of localized diminution of blood flow secondary to functional and physiological alterations. Previous investigators have found that about 70% of cases of ischemic colitis occur spontaneously without any antecedent recognized hemodynamic event—that is, any evident decrease in blood pressure (8,15). Infants and young children have higher total body water contents and are sensitive to dehydration. An acute fluid loss of 10% of body weight may lead to circulatory impairment as well as classic signs of dehydration, and a loss of 15% may lead to circulatory collapse and a moribund state (16).

Previous reports have emphasized the importance of oral rehydration therapy from the beginning of diarrhea to prevent development of life-threatening dehydration and death (17). The parents of the patients in this series were poorly informed regarding fluid requirements, and oral rehydration solution was rarely used. In this series, acute diarrhea did not account for evident hemodynamic collapse except in the patient in case 15. However, it may have accounted for circulatory impairment of viscera from clinical dehydration and resulted in ischemic injuries of the bowel in areas with watershed colonic vasculature. Therefore, the presumption was that ischemic injuries may result from sympathetic vasoconstriction and thereby cause shunting of splanchnic blood flow to more vital organs in response to dehydration due to diarrhea, fever, vomiting, and poor intake, as well as inadequate fluid replacement.

The treatment of these patients in this series mainly consisted of temporary enterostomy for fecal diversion followed by re-establishment of bowel continuity later. Because of the debilitated condition of patients at the time of operation and the unclear cause of the bowel perforation, primary anastomosis would carry the risk of leakage. Therefore, a two-stage approach was adopted.

The results of this study suggest that children with ischemic bowel perforation are at lower risk of death (7%) in comparison with the higher than 50% mortality rate in elderly patients (8,9). The association of concurrent medical illnesses in the aged may contribute to this disparity.

The results of this study suggest that dehydration from acute diarrhea may cause bowel ischemia leading to perforation. Physicians should be alert to the possibility of development of this complication when progressive abdominal distension followed by acute diarrheal episodes is present in children. Because no specific predictive factor for bowel perforation was noted in this series of 15 patients, we were unable to determine specific factors that may be suggestive of increased risk of perforation. The need for adequate rehydration as a means of prevention should be emphasized.

Back to Top | Article Outline


1. Grosfeld JL, Molinari F, Chaet M, et al. Gastrointestinal perforation and peritonitis in infants and children: Experience with 179 cases over ten years. Surgery 1996; 120:650–6.
2. St-Vil D, LeBouthillier G, Luks FI, Bensoussan AL, Blanchard H, Youssef S. Neonatal gastrointestinal perforations. J Pediatr Surg 1992; 27:1340–2.
3. Tan CEL, Kiely EM, Agrawal M, Brereton RJ, Spitz L. Neonatal gastrointestinal perforation. J Pediatr Surg 1989; 24:888–92.
4. Huang SF, Vacanti J, Kozakewich H. Segmental defect of the intestinal musculature of a newborn: Evidence of acquired pathogenesis. J Pediatr Surg 1996; 31:721–5.
5. Meyer CL, Payne NR, Roback SA. Spontaneous, isolated intestinal perforations in neonates with birth weight < 1,000 g not associated with necrotizing enterocolitis. J Pediatr Surg 1991; 26:714–7.
6. Bower TC. Ischemic colitis. Surg Clin North Am 1993; 73:1037–53.
7. Boley SJ. Colonic ischemia: 25 years later. Am J Gastroenterol 1990; 85:931–4.
8. Guttormson NL, Bubrick MP. Mortality from ischemic colitis. Dis Colon Rectum 1989; 32:469–72.
9. Parish KL, Chapman WC, Williams LF Jr. Ischaemic colitis: An ever-changing spectrum? Ann Surg 1991; 57:118–21.
10. Marston A, Pheils MT, Thomas ML, Morson BC. Ischaemic colitis. Gut 1966; 7:1–15.
11. Merrick N, Davidson B, Fox S. Treatment of acute gastroenteritis: Too much and too little care. Clin Pediatr 1996; 35:429–35.
12. Brownlee HJ Jr. Family practitioner's guide to patient self-treatment of acute diarrhoea. Am J Med 1990; 88(Suppl 6A):27–9.
13. Siddharth P, Ravo B. Colorectal neurovasculature and anal sphincter. Surg Clin North Am 1988; 68:1185–200.
14. Mitsudo S, Brandt LJ. Pathology of intestinal ischaemia. Surg Clin North Am 1992; 72:43–63.
15. Longo WE, Ballantyne GH, Gusberg RJ. Ischaemic colitis: Patterns and prognosis. Dis Colon Rectum 1992; 35:726–30.
16. Finberg L. Dehydration in infants and children. N Engl J Med 1967; 276:458–60.
17. Bhattacharya SK, Bhattacharya MK, Manna B, et al. Risk factors for development of dehydration in young children with acute watery diarrhoea: A case-control study. Acta Paediatr 1995; 84:160–4.

Bowel perforation; Dehydration; Diarrhea; Ischemia

© 2000 Lippincott Williams & Wilkins, Inc.