Original Article
Gastric-Outlet Obstruction Induced by Prostaglandin Therapy in Neonates
N Engl J Med 1992; 327:505-510August 20, 1992
- Abstract
- Abstract
Background.
An infusion of prostaglandin E1 is widely used to maintain patency of the ductus arteriosus in neonates with congenital heart disease. After gastric-outlet obstruction was recognized in several infants who received prostaglandin E1, we studied the association between the drug and this complication.
Methods.
We evaluated all neonates who received prostaglandin E1, in our hospital between October 1, 1989, and September 30, 1991, for clinical, radiologic, or pathological evidence of acute gastric-outlet obstruction.
Results.
Of the 74 neonates evaluated, 65 had no signs of gastric obstruction and were considered normal; 5 had clinical and radiologic or pathological evidence of gastric obstruction consistent with the presence of antral mucosal hyperplasia. The remaining four neonates had clinical signs of gastric obstruction, but no radiologic or pathological examinations were performed.
The 5 neonates with antral hyperplasia had received prostaglandin E1 for longer periods (mean [±SD] duration, 569±341 hours) than the 65 normal neonates (54± 58 hours, P<0.001) or the 4 neonates with clinical signs of gastric obstruction (119±60 hours, P<0.05). The cumulative dose of prostaglandin E1 was higher in the neonates with antral hyperplasia (2982±1392 μg per kilogram of body weight) than in the normal neonates (279±270 μg per kilogram, P<0.001) or the neonates with signs of gastric obstruction (528±306 μg per kilogram, P<0.01). In two neonates with antral hyperplasia, the cessation of therapy lessened the gastric-outlet obstruction.
Conclusions.
The administration of prostaglandin E1 to neonates can cause gastric-outlet obstruction due to antral hyperplasia. Neonates who receive prostaglandin E1 at recommended doses for more than 120 hours should be closely monitored for evidence of antral hyperplasia. (N Engl J Med 1992;327:505–10.)
Media in This Article
Figure 1
Barium Contrast Roentgenography of the Stomach and Ultrasonography of the Antropyloric Region through the Antrum in Patient 1.Figure 2
Microscopical Sections of the Gastric Antrum from Patient 2.Article Activity
- Article
PROSTAGLANDIN E1 infusion is widely used to maintain patency of the ductus arteriosus in neonates with congenital heart disease, and also to treat persistent fetal circulation and pulmonary hypertension in newborns.1 , 2 An infusion rate of 0.05 μg per kilogram of body weight per minute is associated with a high probability of success in maintaining ductal patency1 and a low risk of serious cardiovascular, central nervous system, and respiratory complications.3 After gastric-outlet obstruction due to antral mucosal thickening had been identified in several neonates receiving prostaglandin E1, we studied the association between the drug and this complication. We describe a syndrome of severe gastric obstruction associated with the administration of prostaglandin E1 in neonates.
Case Reports
Patient 1
The first index patient was a baby girl born at 37 weeks of gestation (weight, 3 kg) in whom respiratory distress and cyanosis developed at 12 hours of age. Echocardiography and cardiac catheterization revealed a hypoplastic left ventricle and dysplastic pulmonary valve. The infant was treated with an intravenous infusion of prostaglandin E1 in a dose of 0.1 μg per kilogram per minute, beginning on the second day of life. An atrial septostomy was performed on day 4, with consequent hemodynamic improvement. Although the infant initially tolerated nasogastric feeding, vomiting and abdominal distention developed at 28 days of age. An upper gastrointestinal contrast study revealed a dilated stomach and delayed gastric emptying, with an apparently elongated pyloric canal suggestive of hypertrophic pyloric stenosis (Fig. 1Figure 1
Barium Contrast Roentgenography of the Stomach and Ultrasonography of the Antropyloric Region through the Antrum in Patient 1.). Ultrasonography, however, revealed thickening of the antral mucosa with no muscular thickening at the pyloric sphincter. A nasojejunal tube was placed, which overcame the obstruction and allowed normal feeding. The infusion of prostaglandin E1 was discontinued on day 39. The tube was withdrawn and oral feedings were begun 16 days later.Patient 2
The second index patient was a full-term baby girl born after a normal pregnancy and delivery (weight, 3 kg) who was noted to have cyanotic spells at 18 hours of age. An infusion of prostaglandin E1 was started at one day of age. During the next month her arterial blood oxygen saturation fluctuated between 28 percent and 100 percent in room air. Echocardiography revealed normal cardiovascular anatomy, but she continued to require mechanical ventilation and the infusion of prostaglandin E1. Cardiac catheterization showed normal anatomy and a large patent ductus arteriosus with bidirectional shunting. The mean systemic and pulmonary blood pressures ranged between 55 and 60 mm Hg. The cause of the persistent, severe pulmonary hypertension remained unexplained, and at the age of one month the patient still required high doses of prostaglandin E1 (0.1 to 0.2 μg per kilogram per minute). In addition, she received norepinephrine and dopamine for hypotension. An epigastric mass became palpable at one month of age. The infant died of shock and bradycardia one day later.
At autopsy, the epigastric mass proved to be the duodenum, distended by copious bile-stained watery mucin; the stomach contained a moderate amount of tenacious white mucin. The remainder of the small bowel was slightly distended with watery bile-stained mucin. A small Meckel's diverticulum was present in the distal ileum. Microscopical examination of the fundus, body, and antrum of the stomach showed marked foveolar-cell hyperplasia with increased secretion of mucin (Fig. 2Figure 2
Microscopical Sections of the Gastric Antrum from Patient 2.). The mucin filling the stomach and the duodenum stained as neutral mucin, which is normal for both gastric mucin and Brunner's gland mucin. The antral mucosa was 3400 μm thick, three times thicker than in four 12-to-20-day-old infants who died with persistent pulmonary hypertension of the newborn and eight to nine times thicker than in three 14-to-36-day-old infants who died of sudden infant death syndrome. Both the mucosal and external layers of the gastric and intestinal muscle were normal except for their attenuation in the duodenum, which was consistent with the marked duodenal distention noted at autopsy (Fig. 2A). No abnormality was seen in the gastric glands in the fundus, body, or antrum of the stomach, except for mild dilatation of the gastric glands in the antrum (Fig. 2B). Apart from mucin-secreting cells, all gastric epithelial cells appeared to be normal. Goblet cells distended with acidic mucin were prominent throughout the ileum, especially in the distal ileum. Hyperplasia and hypersecretion of foveolar cells in antral-type ectopic gastric mucosa were present in the Meckel's diverticulum (Fig. 3Figure 3
Microscopical Sections of lleum and Meckel's Diverticulum from Patient 2.).The cause of the persistent pulmonary hypertension of the newborn in this infant was alveolar capillary dysplasia, a rare developmental anomaly of the pulmonary microcirculation.4 In lung sections, goblet cells in bronchial epithelium and mucous glands in bronchial walls appeared hyperplastic and hypersecretory.
Methods
To evaluate the association between prostaglandin E1 therapy and gastric-outlet obstruction due to antral hyperplasia, we reviewed the records of all neonates treated with prostaglandin E1 in the neonatal and pediatric intensive care units at the Hospital for Sick Children, Toronto, between October 1, 1989, and September 30, 1991. The neonates were identified from an existing pharmacy list. From the record of each infant who received prostaglandin E1 we collected the following data: maternal age, gestational age, birth weight, Apgar scores at one and five minutes, and all diagnoses. The details of prostaglandin E1 therapy were recorded, including duration, mean dose per hour, maximal dose per hour, and cumulative dose. Concomitant drug therapy was recorded for each infant. Evidence of gastrointestinal signs was sought for each patient.
A child was defined as having clinical signs of gastric obstruction if, while being fed through a nasogastric tube, he or she had vomiting or feedings had remained in the stomach for 48 hours. Antral hyperplasia was defined only on the basis of radiographic or autopsy findings. The radiographic features of antral hyperplasia included ultrasonographic evidence of marked gastric mucosal thickening without hypertrophy of gastric or pyloric muscle (Fig. 1).5 , 6 Pathologically, antral hyperplasia was defined as marked foveolar-cell hyperplasia with increased secretion of mucin (Fig. 2). The thickness of the antral mucosa at autopsy was measured by the method of Tytgat et al.7
Neonates receiving prostaglandin E1 were subsequently stratified into three groups: those who had no signs of gastric obstruction while being fed (normal infants); those with evidence of gastric obstruction due to antral mucosal thickening and no other causes of gastric obstruction (such as hypertrophic pyloric stenosis or pylorospasm); and those with clinical signs but no radiographic or pathological evidence of gastric obstruction, because such tests were not performed. These three groups were compared with respect to a large number of diagnostic and therapeutic characteristics by analysis of variance, Duncan's multiple-range test, or the chi-square test.
Results
Eighty-three neonates were treated with prostaglandin E1. Four were transferred to other hospitals after the cardiac diagnosis was established, and five could not be assessed because no attempts were made to feed them and no ultrasonographic or pathological studies were performed. The study group therefore consisted of 74 neonates (Table 1Table 1
Diagnoses and Concomitant Medications in 74 Neonates Who Received Prostaglandin E1.). Five of the 74 were classified as having antral hyperplasia: 3 by ultrasonography, 1 by ultrasonography and autopsy, and 1 by autopsy only. Four other neonates had clinical signs of gastric obstruction. Sixty-five of the neonates were considered normal for the purposes of this analysis.The five neonates with gastric obstruction secondary to antral hypertrophy did not differ from the other two groups in maternal age, birth weight, Apgar scores at one and five minutes, frequency of the various diagnoses, or frequency of concomitant drug therapy (Table 2Table 2
Characteristics of the Study Neonates with Antral Hyperplasia, Neonates with Only Clinical Signs of Gastric Obstruction, and Normal Neonates.*). The neonates in the group with antral hyperplasia received prostaglandin E1 for a significantly longer period than the normal neonates. The mean rate of administration of prostaglandin E1 was not different among the three groups. The cumulative dose of prostaglandin E1 was significantly higher in the group with antral hyperplasia than in the other two groups, and the group with clinical signs of gastric obstruction received significantly more prostaglandin E1 than the normal group.In two patients with antral hyperplasia, the cessation of therapy resulted in a decrease in gastric mucosal thickness, as documented by ultrasonography. These two neonates were successfully treated with a nasojejunal tube and had no evidence of further obstruction.
Discussion
During the past decade prostaglandin E1 has been widely used to maintain the patency of the ductus arteriosus in infants who depend on the ductus for oxygenation.1 , 2 , 8 Simultaneously, there has been an increase in the use of prostaglandins among adults to prevent and treat peptic disease and stress ulcers, as well as to induce uterine contraction during labor.9 10 11
Infusions of prostaglandin E1 cause respiratory depression in about 10 percent of neonates, cardiovascular effects in 20 percent, generalized flushing in 10 percent, and central nervous system effects (fever, lethargy, irritability, and myoclonic jerks) in 17 percent.3 The gastrointestinal tract has not been recognized as a major site of serious adverse effects of prostaglandin therapy, although diarrhea is a well-recognized uncommon side effect that usually responds to dose reduction. Prostaglandin E2 was suspected of causing necrotizing enterocolitis among infants with symptomatic heart disease,12 , 13 but larger series revealed this complication to be rare.3
Prostaglandin E1 and E2 have complex effects on the gastrointestinal tract. They induce proliferation of the gastrointestinal mucosa, especially in the gastric antral region, where gastric pits (foveolae) are normally longest. These mucosal changes are caused predominantly by elongation and dilatation of the gastric pits.14 15 16 17 In a controlled study among adults, oral therapy with prostaglandin E2 resulted in mucosal thickening that was most conspicuous in the gastric antrum and that gradually regressed after therapy ended.7 Tytgat et al. demonstrated no increase in proliferative activity in foveolar epithelium and concluded that prostaglandin E2 retards the senescence and exfoliation of cells.7 In that study, the dose of prostaglandin E2 (as corrected for body weight) that was administered to the normal subjects was only 2 to 5 percent of that usually given to neonates. Moreover, although no pharmacokinetic comparison has been undertaken, it is conceivable that sick neonates eliminate prostaglandin more slowly than normal adults. The stimulating effects of prostaglandin E1 on the gastrointestinal mucosa may therefore be much more intense in neonates.
Focal foveolar hyperplasia of the stomach alone has been previously described in only two infants.18 , 19 It is unclear whether these two infants had received prostaglandin therapy and whether the pathologic changes occurred only in the stomach or also in other areas of the gastrointestinal tract, since autopsies were not performed. Ménétrier's disease may produce similar pathologic features in the stomach, but it also results in eosinophilic infiltration and has not been described in neonates.20 , 21
The association between prostaglandin E1 therapy and antral hyperplasia could reflect the severity of the cardiac lesion and cyanosis rather than a specific effect of prostaglandin E1. However, low cardiac output and cyanosis in babies and children with cyanotic heart disease result mainly in intestinal ischemia. Initially, the superficial mucosa is sloughed off; submucosal hemorrhage and edema, then ulceration and perforation, follow.18 Although babies and children with chronic severe hypoxia usually have preserved intestinal function, mild intermittent episodes of ischemia may result in progressive circumferential scarring and the formation of strictures, which may eventually lead to partial obstruction. The areas of the gut most severely affected are usually the cecum and terminal ileum. Other zones, such as the splenic flexure and proximal descending colon, are also susceptible to ischemic damage.18 Such sequelae of ischemia are clinically, radiologically, and pathologically different from those we have described in infants receiving prostaglandin.
We found an association between the duration of prostaglandin E1 therapy and its cumulative dose and the development of antral hyperplasia leading to gastric obstruction in neonates. Similar pathologic changes in animals and adults receiving prostaglandins increase the likelihood of a direct causal effect. The apparent regression of the antral mucosal hyperplasia in two infants in our study after the cessation of therapy further implicates prostaglandin E1 as the cause. In conclusion, newborns receiving prostaglandin E1 at the recommended dose of 0.05 μg per kilogram per minute for more than 120 hours should be monitored for evidence of antral mucosal thickening and gastric obstruction. Gastric obstruction caused by prostaglandin E1 may be reversible, and placing a nasojejunal tube may thus obviate the need for surgery in these infants, who are at very high risk of complications.
Dr. Koren is a Career Scientist of the Ontario Ministry of Health.
Source Information
From the Departments of Radiology (N.P., P.B.), Critical Care (O.D., G.B.), and Pathology (M.M.S.), the Divisions of Clinical Pharmacology and Toxicology (D.S., G.K.) and Neonatology (J.H.), the Department of Pediatrics (J.H., G.K.), and the Research Institute (G.K.), Hospital for Sick Children, Toronto; and the Departments of Radiology (N.P., P.B.), Pathology (M.M.S.), Pediatrics (J.H.), and Pharmacology (D.S., G.K.), University of Toronto, Toronto. Address reprint requests to Dr. Koren at the Division of Clinical Pharmacology, Hospital for Sick Children, 555 University Ave., Toronto, ON M5G 1×8, Canada.
- References
References
1
Heymann
CrossRef | Web of Science | Medline2
Thanopoulos
CrossRef | Web of Science | Medline3
Lewis
CrossRef | Web of Science | Medline4
Cullinane 5
Swischuk
Medline6
O'Keeffe
Web of Science | Medline7
Tytgat
Web of Science | Medline8
Lock
CrossRef | Web of Science | Medline9
Feldman
CrossRef | Web of Science | Medline10
Zinner
CrossRef | Web of Science | Medline11
Lange
Web of Science | Medline12
Leung 13
Miller
Web of Science | Medline14
Helander
Web of Science | Medline15
Reinhart
Web of Science | Medline16
Goodlad
CrossRef | Web of Science | Medline17
Gilger M, Jensen C, Kesscer B, Nanjundiah P, Klish WJ. Nutrition, growth and the gastrointestinal system: basic knowledge for the pediatric cardiologist. In: Garson AJR, Bricker JT, McNamara DG, eds. The science and practice of pediatric cardiology. Philadelphia: Lea & Febiger, 1990: 2358–9.
18
McAlister
CrossRef | Web of Science | Medline19
Katz
CrossRef | Web of Science | Medline20
Menetrier's disease. In: Whitehead R. Mucosal biopsy of the gastrointestinal tract. 4th ed. Philadelphia: W.B. Saunders, 1990:112–5.
21
Chouraqui
Web of Science | Medline
- Citing Articles (20)
Citing Articles
1
Louisa W. Chiu, Oliver S. Soldes. 2011. Congenital Anomalies and Surgical Disorders of the Stomach. , 269-276.
CrossRef2
Ranjan Dohil, Eric Hassall. 2011. Gastritis, Gastropathy, and Ulcer Disease. , 277-292.
CrossRef3
Lisa M. Kohr, Nancy J. Braudis. 2010. Growth and Nutrition. , 1285-1298.
CrossRef4
Leyla H. Alparslan, Barbara N. Weissman. 2009. Imaging Findings of Drug-Related Musculoskeletal Disorders. , 264-279.
CrossRef5
GYULA TÁLOSI, MÁRTA KATONA, SÁNDOR TÚRI. (2007) Side-effects of long-term prostaglandin E 1 treatment in neonates. Pediatrics International 49:3, 335-340
CrossRef6
2006. Prostaglandins. , 2955-2964.
CrossRef7
TILMAN HUMPL, SYED H.E. ZAIDI, JAMES Y. COE, JENNIFER RUSSELL, YASUFUMI KANEDA, HAMID MASSAELI, LEE N. BENSON, MARLENE RABINOVITCH. (2005) Gene Transfer of Prostaglandin Synthase Maintains Patency of the Newborn Lamb Arterial Duct. Pediatric Research 58:5, 976-980
CrossRef8
Kellie Rohrer, Robyn Murphy, Caroline Thresher, Nabil Jacir, Kerry Bergman. (2005) Infantile myofibromatosis: a most unusual cause of gastric outlet obstruction. Pediatric Radiology 35:8, 808-811
CrossRef9
John C. Bucuvalas, Frederick C. Ryckman, Susan Krug, Maria P. Alonso, William F. Balistreri, Uma Kotagal. (2001) Effect of treatment with prostaglandin E 1 and N -acetylcysteine on pediatric liver transplant recipients: A single-center study. Pediatric Transplantation 5:4, 274-278
CrossRef10
Yuka Osaki, Shinji Oana, Kumi Yasukawa, Ken Fujimori, Masaru Terai, Hiroo Niimi. (2000) CATCH22 syndrome with gastroesophageal reflux. Pediatrics International 42:4, 367-368
CrossRef11
Michael S. McLeary, Robert D. Thomas, Lionel W. Young. (2000) Imaging of congenital and acquired gastric abnormalities in children. Academic Radiology 7:2, 117-127
CrossRef12
Eric L. Carter, Maria C. Garzon. (2000) Neonatal Urticaria Due to Prostaglandin E1. Pediatric Dermatology 17:1, 58-61
CrossRef13
Ranjan Dohil, Eric Hassall, Gareth Jevon, James Dimmick. (1999) Gastritis and Gastropathy of Childhood. Journal of Pediatric Gastroenterology & Nutrition 29:4, 378-394
CrossRef14
Vahid Master, Geoff Davidson, Lloyd Morris, James Martin, Declan Kennedy, Roger Byard, John Freeman. (1998) Focal Foveolar Cell Hyperplasia Presenting as Recurrent Emesis in a Young Infant. Journal of Pediatric Gastroenterology & Nutrition 26:2, 222-225
CrossRef15
RH Regev, T Dolfin, S Katz. (1996) Transient gastric outlet obstruction and tetralogy of Fallot: a true association or a coincidental finding?. Acta Paediatrica 85:4, 514-515
CrossRef16
A. K. Kalloghlian, H. H. Frayha, M. M. A. deMoor. (1996) Cortical hyperostosis simulating osteomyelitis after short-term prostaglandin E1 infusion. European Journal of Pediatrics 155:3, 173-174
CrossRef17
P. Babyn, N. Peled, D. Manson, O. Dagan, M. M. Silver, G. Koren. (1995) Radiologic features of gastric outlet obstruction in infants after long-term prostaglandin administration. Pediatric Radiology 25:1, 41-43
CrossRef18
Debra Beneck. (1994) Hypertrophic Gastropathy in a newborn: A Case Report and Review of the Literature. Fetal & Pediatric Pathology 14:2, 213-221
CrossRef19
PeterJ.F.M. Merkus, AdriH. Cromme-Dijkhuis, SimonG.F. Robben, John Hess. (1993) Prostaglandin E1 and gastric outlet obstruction in infants. The Lancet 342:8873, 747
CrossRef20
Milla, Peter J., . (1992) Gastric-Outlet Obstruction in Children. New England Journal of Medicine 327:8, 558-560
Full Text
