Localized intestinal perforations after enteral administration of indomethacin in premature infants

Localized intestinal perforations after enteral administration of indomethacin in premature infants Four very-low-birth-weight infants developed localized intestinal perforations after enteral administration of indomethacin. The clinical picture and histologie findings were unlike those seen in necrotizing enterocolitis. (J PEDtATR 106:277, 1985)

G. Alpan, M.D., F. Eyal, M.D., I. Vinograd, M.D., R. Udassin, M.D., G. Amir, M.D., P. Mogle, M.D., and B. Glick, M.D. Jerusalem, Israel

INDOMETHAC1N is used to produce pharmacologic closure of the patent ductus arteriosus in premature neonates? -4 Reversible decline in renal function is the most common complication of indomethaein therapy?. 6 Gastrointestinal tract complications have also been attributed to indomethacin 7,8; Nagaraj et al. 9 have recently described a high incidence of adverse effects, including abdominal distention, bilious vomiting, bleeding, and necrotizing enterocolitis with perforation. W e report four patients with focal perforations of the intestinal tract that we believe were related to enteral administration of indomethacin. CASE R E P O R T S Clinical data are summarized in the Table; the histologic findings are described after the case reports because they were similar in all patients. Patient 1, This infant boy was delivered by cesarean section. Apgar scores were 6 and 9 at 1 and 5 minutes, respectively. The infant had respiratory distress syndrome and required mechanical ventilation. On day 2 of life, clinical signs of a patent ductus arteriosus became apparent , and a suspension of indomethacin was given via nasogastric tube three times, with marked respira-

From the Departments of Pediatrics, Pediatric Surgery, Pathology, and Radiology, Hadassah University Hospital, Mt. Scopus. Submitted for publication March 12, 1984; accepted July 20, 1984. Reprint requests: G. Alpan, M.D., Department of Pediatrics, Georgetown University Hospital, 3800 Reservoir Rd. N.W., Washington, D.C. 20007.

tory improvement. Throughout this time vital signs were normal, there had been no episodes of hypoxia, the infant had passed normal meconium stools, and radiograms showed normal bowel air distribution. Three days after the last dose of indomethacin, slight abdominal distention was noted. On the following day, distention was more marked and dark green residues were aspirated through the nasogastric tube. Abdominal radiograms showed mildly dilated bowel loops, but findings were otherwise unremarkable, even on retrospective analysis. Five days after the last dose of indomethacin, fresh blood was passed with meconium stools. Abdominal x-ray films, which had been taken every 6 hours, now showed a small amount of free intraperitoneal air. At no time were there laboratory results suggestive of necrotizing

NEC PDA RDS

Necrotizing enterocolitis Patent ductus arteriosus Respiratory distress syndrome

enterocolitis (e.g., thrombocytopenia, electrolyte imbalance). At laparotomy an isolated perforation was identified at the jejunal antimesenteric border 80 cm from the ileocecal valve. Distal to the perforation a small 4 to 5 cm segment of bowel was dilated, but the entire bowel, including the colon, was normal in appearance. A 10 cm section of bowel was reseeted and a primary anastomosis carried out. After surgery the infant continued to have respiratory failure with recurrence of patent ductus arteriosus, but no gastrointestinal tract complications were noted. He had also had massive intraventricular and intracerebral-hemorrhage, and died at 16 days of age. Patient2. This infant boy had Apgar scores of 1 and 9 at birth

The Journal of P E D I A T R I C S

277

278

Alpan et al.

The Journal of Pediatrics February 1985

Table. Clinical d a t a in infants with intestinal p e r f o r a t i o n

Patient

Gestational age (wk)

Weight (gin)

Age when drug first given (day)

1 2 3 4

27 28 29 30

980 990 1065 1100

2 4 1 2

(ms~ks)

Number of doses

Interval between doses (hr)

Onset of disease* (day)

0.25 0.3 0.2 0.2

3 6 3 9~"

8 12 24 8

3 2 4 5

Dose

*Onset of disease after last dose of indomethacin. tlndomethacin not administered on day 3 of life.

and 5 minutes, respectively. He had mild RDS and received mechanical ventilation until 3 days of age, and then nursed in an oxygen hood. When he was 4 days of age PDA was diagnosed and indomethacin suspension was given through a nasogastric tube. Throughout this period there were no episodes of hypoxia, blood pressure was normal, and routine laboratory results yielded unremarkable findings, Daily radiograms revealed no abdominal abnormality. Two days after the last dose of indomethacin there was mild abdominal distention, and abdominal radiograms showed free intraperitoneal air. The WBC count was 2000//~1, and there was moderate metabolic acidosis (base excess - ! 0 mEq/L). Thrombocytes and serum electrolyte concentrations were normal. Blood cultures later grew Serratia marcescens and Enterobacter aerogenes. At laparotomy there was a perforation of the terminal ileum 10 cm proximal to the ileocecal valve. The perforation was clearly demarcated, and adjacent bowel was healthy in appearance. A 10 cm section of bowel were resected, and a double-barrel ileostomy was constructed. After surgery the infant required continued ventilatory support, with recurrence of PDA. At 1 month of age surgical ligation was carried out, with significant improvement. After 75 days, however, the patient died because of sepsis. Patient 3. This infant girl, the first of triplets, was delivered normally. Apgar scores were 7 and 9 at 1 and five minutes, respectively. The infant had mild respiratory insufficiency and was nursed in an oxygen hood. At 1 day of age signs of PDA developed, and indomethacin suspension was given through a nasogastric tube. Four days after the last dose of indomethacin abdominal distention was noted, and abdominal radiography showed free intraperit0neal air. Apart from a few episodes of apnea, which occurred on the second day of life and which terminated after administration of aminophylline, the infant had no signs or symptoms suggestive of abdominal or systemic disorders. Vital signs were normal, as were results of routine laboratory tests, including radiographs. At laparotomy an isolated perforation was identified at the jejunoileal junction, the rest of the bowel being normal in appearance. Ten centimeters of bowel were resected and an end-to-end anastomosis carried out. The infant's subsequent course was uneventful, and she was discharged in good general health. Follow-up at 2 years of age showed a normally developed child with no gastrointestinal tract complaints. Patient 4. This infant girl, the second of twins, was delivered by cesarean section. Apgar scores were 5 and 9 at 1 and 3 minutes,

respectively. The infant had moderately severe RDS and required mechanical ventilation. On the second day of life PDA was diagnosed, and she was given indomethacin via nasogastric tube. Signs of PDA recurred on the fourth and fifth days, and on each of these days further doses of indomethacin were administered. There was marked improvement in respiratory status, which enabled weaning from mechanical ventilation. The infant passed meconium normally and had no systemic or localized signs of an abdominal disorder. Five days after the last dose of indomethacin, abdominal distention was noted and the infant failed to pass stool. Vital signs were normal, and results of routine laboratory investigations were within normal limits, except for moderate metabolic acidosis (base excess - 1 1 mEq/L). Abdominal radiography showed dilated bowel loops in the right side of the abdomen, but otherwise there were no signs suggestive of NEC. A repeat radiogram, taken 6 hours later, showed free intraperitoneal air. At surgery, an isolated, clearly demarcated perforation was found in the terminal ileum 10 cm proximal to the ileocecal valve and extending to three fourths of the bowel circumference. The rest of the bowel was normal in appearance. Blood cultures, taken prior to surgery, grew Citrobacterfreundi. A 7 cm section of ileum was resected and a double ileostomy constructed. The infant's subsequent course was uneventful. Ten days after surgery gavage feeding was begun, and 3 months later ileal anastomosis was carried out. No gastrointestinal tract complications have been noted at 18 months follow-up, and the child is doing well. Pathologic examination. In all of the cases described, resected specimens showed a well-defined perforation with superficial mucosal ulceration, which was well circumscribed, around the perforation. The rest of the bowel was normal. Histologic examination revealed moderate to marked mucosal hemorrhagic necrosis with some submucosal hemorrhage but without significant inflammatory infiltrate. (Figs. 1 and 2). The changes seen in NEC were not observed, and the bowel adjacent to the perforation was unremarkable on microscopy. DISCUSSION T h e f o u r i n f a n t s d e s c r i b e d all h a d a c u t e , isolated, a n d localized small bowel p e r f o r a t i o n s within 3 to 5 days a f t e r nasogastric administration of indomethacin. None had f e a t u r e s p r e d i s p o s i n g to N E C o t h e r t h a n t h a t t h e y w e r e all V L B W p r e t e r m i n f a n t s . U m b i l i c a l a r t e r y c a t h e t e r s were i n s e r t e d in only t w o o f t h e s e patients, t h e o t h e r two b e i n g

Volume 106 Number 2

Fig. 1. Sharp border between necrotic and normal tissue demonstarted, as well as absence of inflammatory infiltrate (patient 1). Red blood cells appear as black dots. (H & E stain; •

monitored via right radial artery cannulas. None of the infants had been fed; it is our policy to withold enteral feeding in LBW infants until 3 weeks of age and until that time to provide peripheral intravenous alimentation. ~~ Moreover, none of these infants had clinical or laboratory signs suggestive of NEC prior to the occurrenc e of perforation. The time course of events suggested that the low WBC count in patient 2, the moderate metabolic acidosis in two patients, and the positive blood cultures (of enteral colonizers) in two patients were related to the perforation per se. Even retrospective review of all radiograms prior to perforation did not reveal signs suggestive of enterocolitis. Inspection of the bowel at surgery revealed a bowel healthy in appearance other than the localized perforation; in fact, in none of these infants was NEC suspected on observation at laparotomy. On pathologic examination, both gross and microscopic, the mucosa was normal except for the area immediately bordering the perforation; the lesions were not those seen in NEC.~t There are frequent exceptions to

Intestinal perforations after indomethacin

279

Fig. 2. Small intestine at edge of perforation, showing sharp border between necrotic and normal bowel (patient 2). (H & E stain; •

any single histologic finding described in NEC, but the sequence of negative findings on gross inspection and histologic examination, compounded by lack of clinical findings, suggests a different underlying disorder. Moreover, that this should occur in four infants after administration of indomethacin suggests a causal relationship. During the period when these four patients were seen, 101 neonates weighing