- Email: [email protected]
Short bowel syndrome in infancy and childhood
Short Bowel Syndrome in Infancy and Childhood Analysis of Survival in 60 Patients
Jay L. GrosfekJ, MD, Indianapolis, Indiana Frederkk J. Rescoria, MD, Indianapolis, lndii Karen W. West, MD, Indianapolis, Indiana
Short bowel syndrome in infancy and childhood is usually the result of massive bowel loss due to intrauterine intestinal atresia, postnatal midgut volvulus, or necrotizing enterocolitis. Other causes include total colonic aganglionosis with extension of the aganglionic process to the mild or proximal small bowel, gastroschisis with volvulus or atresia, volvulus secondary to adhesive small bowel obstruction, trauma, and inflammatory bowel disease. Short bowel syndrome has been defined as the presence of less than 50 percent of the expected small bowel length at the conclusion of the initial operation [I]. Previous concepts concerning infant survival and bowel length were based on the study of Wilmore [2] which correlated a successful outcome with at least 15 cm of jejunum or ileum and an intact ileocecal valve and more than 40 cm of jejunum or ileum when the ileocecal valve had been resected. The present study reviews the clinical presentation, operative and medical management, morbidity, and mortality of 60 infants and children with short bowel syndrome. Herein, we also reassess survival in view of the development of adjunctive surgical procedures and current available nutritional support, including total parenteral nutrition and well-defined elemental diets [3-61. Patients and Methods Between January 1972 and December 1984,60 infants and children with short bowel syndrome were treated at the James Whitcomb Riley Hospital for Children, IndiFrom the Section of Pediatric Surgery, Department of Surgery, Indiana University Medical Center and the James Whitcomb Riley Hospital for Children, Indianapolis. Indiana. Requests for reprints should be addressed to Jay L. Crosfeld. MD, James Whitcomb Riley Hospital for Children, 702 Barnhill Drive (K21), Indianapolis, Indiana 46223. Presented at the 26th Annual Meeting of the Society for Surgery of the Alimentary Tract, New York, New York, May 14-15, 1965.
Volume 151, January 1986
ana University Medical Center, Indianapolis, Indiana. There were 40 boys and 20 girls. Twenty-eight infants were premature (gestational age 36 weeks or less), 26 were full-term infants, and 6 patients were older children. Causes included necrotizing enterocolitis in 24 infants, jejunoileal atresia in 13, malrotation with midgut volvulus in 6, total colonic aganglionosis with proximal small bowel extension in 7, gastroschisis in 5, volvulus related to adhesive bowel obstruction in 4, and trauma in 1. Cases of adhesive bowel obstruction included three infants with volvulus occurring after excision of a retroperitoneal tumor (Wilms’ tumor in two and teratoma in one) and a 14 year old patient who had previously undergone total proctocolectomy and ileostomy for ulcerative colitis. One instance of trauma was related to a shotgun wound with extensive mesenteric avulsion in a 6 year old child. Gangrenous or atretic bowel requiring resection was noted in 53 instances. Temporary enterostomies were performed in 40 patients in order to preserve bowel length in cases of questionable viability and to allow visualization of the mucosa at the temporary stoma site postoperatively. Second-look laparotomy at 24 to 48 hours was useful for preserving bowel length in two of four patients, one of whom had necrotizing enterocolitis and one midgut volvulus. A sterile Doppler probe was also employed intraoperatively. Delayed anastomosis was subsequently accomplished from 30 to 60 days later in these patients. Tapering proximal antimesenteric enteroplasty and primary anastomosis was performed in 13 infants with high jejunal atresia. The enteroplasty was usually accomplished over a 22 or 24 F. catheter using an autostapling device. Seven infants with total colonic aganglionosis and small bowel extension from the mid to the proximal small bowel underwent initial enterostomy in the segment of bowel with ganglion cells proved by biopsy. Two of these children had subsequent modified Soave procedures and one had an endorectal pull-through procedure with an antimesenteric aganglionic patch enteroplasty to increase the mucosal absorptive surface area. The mean length of remaining bowel was 58.4 cm, with a range of 13 to 120 cm in the infants and 120 to 150 cm in the older children. The ileocecal valve was resected in 33
41
Grosfeld et al
TABLE I
Short Bowel Syndrome: Results In 60 Patlents Results
n
%
Survivors Deaths TPN-related liver failure Sepsis Bronchopulmonary dvsolasia
51 9 5
85 15
3 1
. .
TPN = total parenteral nutrition.
and remained intact in 27. All patients were managed with total parenteral nutrition and early enteral feedings using highly defined elemental diets administered by hourly drip feeding. Home hyperalimentation was attempted when 50 percent of the total calories were enteral. Intestinal adaptation required from 3 to 14 months. Frequent setbacks were related to catheter sepsis, rotavirus infections, carbohydrate intolerance, and liver dysfunction. Gallstones developed in two patients and required subsequent cholecystectomy. Fifty-one of 60 patients survived (Table I). The shortest length of intestine in a survivor was 20 cm of proximal jejunum anastomosed to the splenic flexure. Weight gain was adequate (10th to 25th percentile), but longitudinal growth was harder to achieve, as some of the long-term survivors have short stature. There were nine deaths, including four infants with necrotizing enterocolitis, three with midgut volvulus, and two with gastroschisis complicated by atresia. In all but one patient, the ileocecal valve had been resected. Deaths were related to sepsis in three patients, bronchopulmonary dysplasia in one patient, and cholestasis associated with total parenteral nutrition that progressed to fatal liver disease in five patients. patients
Comments Short bowel syndrome continues to be a significant clinical problem, especially in the neonate. The development of sophisticated neonatal intensive care facilities has resulted in the survival of 80 to 85 percent of seriously ill premature infants [7]. Necrotizing enterocolitis, a highly lethal form of ischemic bowel injury, develops in approximately 1 to 2 percent of all patients admitted to these neonatal intensive care centers. Many of these infants require massive bowel resection, usually including the ileocecal valve, as the distal ileum and right side of the colon are most commonly involved by the ischemic process. In the present report, necrotizing enterocolitis was the most common cause of short bowel syndrome. Postoperatively, these premature infants are in fragile health, with immature enzyme systems, diminished absorptive capacities for protein, carbohydrate and fat, a reduced bile salt pool, and decreased resistance to infection compared with the full-term infant. The estimated small bow-
42
el length of the premature infant also differs significantly from that of the full-term infant. In an autopsy study, Touloukian and Smith [S] demonstrated that the small bowel length in infants varied directly with the gestational age. Small bowel length was only 114 t 21 cm in infants of 19 to 27 weeks gestation, compared with 248 f 40 cm for infants of more than 35 weeks gestation, indicating that the small bowel of the fetus doubles in length during the third trimester of pregnancy. This suggests that short residual segments of small bowel in the premature infant may eventually grow and function as well aa longer segments of bowel in the full-term infant. In infants with necrotizing enterocolitis and midgut volvulus where the viability of the entire small bowel is questionable, the use of an intraoperative Doppler probe and fluorescein dye may be helpful in preserving bowel length. In selected patients, closing the abdomen and vigorous intravenous fluid and antibiotic support has allowed a successful secondlook laparotomy 24 to 48 hours later. Two of four patients in this study had segments of bowel preserved by this technique. Whenever bowel viability is in question and anastomosis is a risk, the formation of temporary enterostomies is an attractive alternative. The second most common cause of short bowel syndrome in our study was intestinal atresia, usually due to an intrauterine volvulus or to multiple bowel atresias. This results in high jejunal atresia in which the only remaining bowel segment is the dilated proximal atretic loop. Although resection of the dilated loop is performed in infants with otherwise normal bowel length, in the infant with short bowel syndrome this is the only residual bowel and therefore, it must be preserved. Use of an antimesenteric tapering enteroplasty was successfully employed in 13 patients. This technique, originally described by Thomas [9], has also been utilized successfully by Howard and Otherson [IO], as well as by Weber et al [11]. Intestinal plication, an alternative technique, has recently been described by de Lorimier and Harrison [12]. An additional adjunctive surgical procedure is the use of an aganglionic antimesenteric patch enteroplasty in infants with total colonic aganglionosis with extension of the aganglionic process to the middle of the small bowel or jejunum. A portion of aganglionic ileum or colon can be used for the patch [13-161. The key to survival after massive bowel resection is the ability of the residual bowel to adapt, which depends on the extent and site of resection, adaptive changes in the remaining intestine, and the persistence of disease in the remaining bowel mucosa. Any bowel tissue affected by midgut volvulus or necrotizing enterocolitis may have sustained significant mucosal damage which interferes with healing and adaptation and delays tolerance of enteral feedings. Morphologic adaptive changes include in-
The American Journal of Surgery
Short Bowel Syndrome in Childhood
creases in bowel circumference, thickness of the bowel wall, average villus height, crypt depth, and rate of cell proliferation. An increase in the number of epithelial cells per centimeter of villus signifies hyperplasia of mucosal cells [ 17,181. Functional adaptation primarily depends on an increased number of absorptive cells that results in an increase in nutrient absorption and water and electrolyte transport. It brings about enhanced absorption per centimeter of residual gut of carbohydrates, protein, fat, electrolytes, water, calcium, vitamin B12, and bile acids, and an increase in the specific activity of a number of mucosal enzymes, including sodium and potassium ions, adenosine triphosphatase, enterokinase, and others. Regulation of bowel adaptation is affected by a number of mechanisms that probably stimulate intestinal hyperplasia. The most important stimulant is exposure of the intestinal lumen to an enteral nutrient [19-221. Although all infants with short bowel syndrome require total parenteral nutrition, postresection villus hyperplasia does not occur if enteral intake is denied. Cellular hyperplasia of the villus may also be stimulated by biliary and pancreatic secretions. An additional factor is the presence of an intact ileocecal valve [22]. Transit time, nutrient absorption, and survival may be dependent on the presence or absence of the ileocecal valve. A more rapid transit time, decreased absorption, and an increased mortality rate may be expected after massive distal small bowel resection, including the ileocecal valve [23-Z]. Acquired hypogammaglobulinemia has been clinically observed after massive ileal resection, and there is laboratory evidence of increased mortality after endotoxin challenge or administration of live Escherichia coli in rats with massive distal small bowel resection [26,27]. Wilmore [2] reported that infants with short gut syndrome required at least 40 cm of bowel in the absence of an ileocecal valve in order to survive. In infants with an intact ileocecal valve, those with 35 cm of residual bowel survived as did 50 percent of those with 15 to 25 cm, but there were no survivors with less than 15 cm of residual small bowel. Several other investigators have also reported successful outcomes in children with as little as 13 to 17 cm of small bowel in the presence of an ileocecal valve [B-30]. Ricour et al [31] have recently described two patients with 15 cm of small bowel without an ileocecal valve. Although both children survived, they were dependent on parenteral nutrition. In our study, one infant with 20 cm of small bowel with an anastomosis to the splenic flexure of the colon has survived for 11years on a full enteral diet. Rickham et al [32] reported excellent survival data in 49 children with short gut syndrome with residual intestinal lengths ranging from 26 to 75 cm, and suggested that the critical length of intestine was approximately 20 cm.
Volume 151, January 1986
The most significant factor in the improved survival rate of infants and children with short bowel syndrome has been the use of total parenteral nutrition, which delivers adequate calories in protein, carbohydrates, fat, and nutrients, allows growth, and provides time for intestinal adaptation to occur. The development of small Broviac or Hickman Silasti@ catheters has improved total parenteral nutritional care; the technique of administration can be taught to parents and can be carried out as a home hyperalimentation program when 50 percent of the required calories are tolerated by the enteral route. Catheter-related sepsis remains a constant threat, therefore, careful technique in catheter handling is important. Mineral deficiencies may develop during total parenteral nutrition due to excessive losses through an enterostomy or from diarrhea. These include calcium, phosphorus, zinc, copper, and magnesium deficiencies. Higher daily doses of vitamins, electrolytes, minerals, and intravenous fat solutions have greatly decreased the incidence of these nutritional deficiencies. Prolonged use of total parenteral nutrition, particularly in the premature infant, can result in severe cholestasis in 30 to 42 percent of all infants receiving total parenteral nutrition (33,341; in addition, neonates with short bowel syndrome may have an increased risk of development of cholestasis. Cooper et al [1] noted cholestasis related to total parenteral nutrition in 15 of 16 neonates with short bowel syndrome. Progressive jaundice resulting in a fatal form of poorly understood liver failure developed in three of these patients. Five infants in the present report also died from liver failure induced by total parenteral nutrition, making this unusual complication the most common cause of death in our series. Although the precise cause of total parenteral nutrition-related cholestasis is uncertain, several factors have been implicated, including prematurity, immature liver function, duration of total parenteral nutrition, and peritonitis with or without recurrent episodes of sepsis [35]. In addition, absence of enteral feedings may also contribute to progressive cholestasis. Reversal of cholestasis usually occurs when the patient progresses from total parenteral nutrition to complete enteral feedings. In the present series, however, when the bilirubin level was greater than 30 mg/dl, reversal did not occur, and a fatal outcome usually ensued, characterized by elevated alkaline phosphatase levels, hypoalbuminemia, hyperammonemia, and a coagulopathy that required daily administration of vitamin K and fresh frozen plasma. Bilirubin levels before death have increased to as much as 50 mg/dl. This is probably due to an inability of the premature infant to metabolize the 2.5 percent amino acid concentration commonly utilized in nutritional formulas for the neonatal period. Formation of a toxic bile salt may result in a canalicular injury and severe choles-
43
Grosfeld et al
tasis. Choline and taurine deficiences have also been suggested as a possible cause [36]. These observations have led to the development of new intravenous and enteral formulas containing taurine and the suggestion that the amino acid content of total parenteral nutrition solutions should not exceed 1.5 percent when meant for the premature infant. Cholelithiasis occurred in 2 of the 60 patients in this series and was managed by cholecystectomy in each instance. This is a well-recognized complication after extensive ileal resection and in patients managed with long-term total parenteral nutrition [37,38]. Its occurrence in our two patients was most likely related to a disturbance in the normal bile salt pool and decreased enterohepatic circulation as a result of bile salt losses in the stool. The incident of gallstones in the current series was lower than that reported in adults with short gut syndrome [38]; therefore, we have not performed routine cholecystectomy as a prophylactic measure in infants with short bowel syndrome. Despite the beneficial effects of total parenteral nutrition, a successful result will not be achieved without enteral nutrition. We usually begin early enteral feedings by a continuous slow drip technique through an orogastric or gastrostomy tube. Highly defined, easily digestible formulas of low osmolality are used. Formulas containing lactose are avoided due to lactase deficiency in most patients with short gut syndrome. Long chain fats are also avoided. Hyperosmolar feedings must be administered with great care to avoid injury to the hyperplastic mucosa, which may result in secretory diarrhea. Enteral intolerance to carbohydrates, including glucose, may require a low osmolar carbohydrate-free diet. Fructose may be added to increase the total caloric load. More complex substances can be added with time, including maltose, sucrose, and certain disaccharides, peptides, and medium chain triglycerides. Cholestyramine, a bile acid that binds resin, may decrease the frequency of stools and allow more absorption of water in the remaining colon. Adequate supplementation of vitamin B12, folit acid, other vitamin complexes, iron, sodium citrate, and potassium elixir are essential. The stool pH should be kept above 6, as an acidic fecal stream results in increased transit time. One milliequivalent of sodium bicarbonate with each feeding is useful in this regard. In older infants, loperamide (Imodium@) and diphenoxylate (LomotiF) may be useful adjuncts to decrease the number of stools. The period required for adaptation is quite variable and may take from 6 months to 2 years to be completed. Any cause of intestinal injury, such as sepsis, rotavirus infection, and hyperosmolar secretory diarrhea, will delay the process considerably. Patience, long-term diligence, and patient and parent sensitivity are distinct virtues in the management of this disorder. Assessment of weight in rela-
44
tion to height, albumin level, transferrin level, skin fold measurement, and transit time, as well as monitoring the consistency, pH, and number of bowel movements are useful clinical guides. Occasionally, quantitative evaluation of carbohydrate, protein, and fat absorption may be useful, but these tests are time consuming, usually require hospitalization, and are also expensive. Despite the aforementioned techniques, some infants respond poorly to therapy; for some of them, adjunctive surgical procedures have been devised. Reversed small bowel segments that recirculate bowel loops and vagotomy and pyloroplasty have been attempted, but have not been uniformly successful and are presently not in use [39]. Prejejunal isoperistaltic colonic interposition has been used successfully to increase transit time in both experimental and clinical settings [40,41]. The use of an antiperistaltic colonic segment has also been attempted; however, bowel obstruction has been a frequent complication of this technique [42]. Bianchi [43] described a bowel-lengthening procedure in pigs which involved longitudinal division of the bowel, leaving half of the mesenteric blood supply with each half of the bowel. Boeckman and Traylor [44] reported successful use of this technique in a child with gastroschisis and short gut syndrome. Several additional procedures have been successful in the experimental laboratory setting and are awaiting clinical application. Replacement of the ileocecal valve as a nipple valve intussusception in either a jejunocolic or ileocolic position increased survival rates, transit time, favorably affected weight change and growth, and decreased the anaerobic bacterial counts in the small and large bowels in experimental studies [24,25]. In addition, Vinograd et al [45] recently reported the use of a submucosally tunnelled valve of ileum into the colon as a replacement for the ileocecal valve in dogs. Binnington and associates [46) reported the use of a colonic patch to increase the mucosal surface area of small bowel, and Lillemoe et al [47] demonstrated the feasibility of growing new intestinal mucosa on rectus muscle flaps attached to the antimesenteric surface of the small bowel in rabbits. Gladen and Kelly [48] and Layzell and Collin [49] showed an increased absorption of carbohydrate and water, increased body weight, decreased fecal fat, and decreased nitrogen losses in dogs with retrograde intestinal pacing. Allotransplantation of the small bowel has been performed in seven patients [50]. The first two transplants were performed in two children in 1964, however, both died. In 1967, Lillehei et al [51] reported a cadaveric donor transplant in a 46 year old woman with bowel infarction from mesenteric venous thrombosis. Unfortunately, she died 12 hours later from pulmonary complications. The longest survivor was a woman who received an incontinuity
The American Journal of Surgery
Short Bowel
graft from her HLA-identical sister [52]. She died from sepsis on the 176th postoperative day. The success of cyclosporine A in controlling both rejection and graft versus host disease in hepatic, renal, and cardiac transplantation has led to renewed interest in small bowel transplantation. Several centers are actively involved in small bowel transplantation research, however, clinical bowel transplantation remains an elusive goal and currently must be considered an experimental procedure.
This report concerns 60 infants and children with short bowel syndrome, most commonly caused by necrotizing enterocolitis in this study. Resection of atretic or gangrenous bowel was performed in 53 patients, tapering enteroplasty and primary anastomosis was performed in 13 patients, and temporary enterostomies were performed in 40 patients. Second-look laparotomy was useful in two of four cases of questionable bowel viability. The ileocecal valve was resected in 33 patients and remained intact in 27. The mean length of remaining bowel was 58.4 cm (range 13 to 150 cm). Seven patients with total aganglionosis and mid to proximal small bowel extension were managed with an initial enterostomy, whereas three had a pull-through procedure with an aganglionic patch enteroplasty. All patients received total parenteral nutrition and early enteral feedings. Home hyperalimentation was attempted when 50 percent of the calorie intake was enteral. Intestinal adaptation required from 3 to 14 months. Frequent setbacks were related to catheter sepsis, rotavirus infection, carbohydrate intolerance, and liver dysfunction. The overall survival rate was 85 percent, with mortality due to liver failure and sepsis associated with total parenteral nutrition. References 1. Cooper A, Floyd TF, Ross AJ Ill, Bishop HC, Templeton JM Jr, Ziegler MM. Morbidity and mortality of short-bowel syndrome acquired in infancy: an update. J Pediatr Surg 1984;19:711-8. 2. Wilmore DW. Factors correlating with a successful outcome following extensive Intestinal resection in newborn infants. J Pediitr 1972;80:88-95. 3. Postuma R, Moroz S, Friesen F. Extreme short-bowel syndrome in an infant. J Pediatr Surg 1983;18:284-8. 4. Aaronson IA, Bowie MD, Cywes S, Louw JH. Massive small bowel resection in a neonate. Arch Surg 1975110: 1485-90. 5. Bell MJ, Martin LW, Shubert WF, Pat-tin J, Burke J. Massive small bowel resection In an infant: long-term management and intestinal adaptation. J Pediatr Surg 1973;8: 197-204. 6. Kurz R, Sauer H. Treatment and metabolic findings in extreme short-bowel syndrome with 11 cm jejunal remnant. J Pediatr Surg 1983:18:257-63. 7. Cikrit 0, Mastandrea J, West KW, Schreiner RL, (jrosfeld JL. Necrotizing enterocolitis: factors affecting mortality in 101 surgical cases. Surgery 1984;96:848-55. 8. Touloukian RJ, Smith GJW. Normal intestinal length in preterm
Volume 151, January 1988
Syndrome in Childhood
infants. J Psdiatr Surg 1983;18:720-3. 9. Thomas CG. Jejunoplastyfor the correction of jejunal atresia. Surg Gynecol Dbstet 1969;129:545-6. 10. Howard ER, Dthersen HB. Proximal jejunoplasty in the treatment of jejunal atresia. J Pedlatr Surg 197$5:685-g. 11. Weber TR, Vane DW, Grosfeld JL. Tapering enteroplasty in infants with bowel atresia and short gut. Arch Surg 1982;117:684-8. 12. deLorimier AA, Harrison MG. Intestinal plication in the treatment of atresia. J Pediatr Surg 1983;18:734-7. 13. Kottmeier PK, Jongco B, Velcek FT, Friedman A, Klotz OH. Absorptive function of the aganglionic ileum. J Pediatr Surg 1981;16:275-8. 14. Kimura K, Nishijima E, Muraji T, Tsugawa C, Matsumoto Y. A new surgical approach to extensive aganglionosis. J Pedlatr Surg 1981;18:840-3. 15. Shandling B. Total colon aganglionosis: a new operation. J Pediatr Surg 1984; 19:503-5. 16. Boley SJ. A new operative approach to total aganglionosis of the colon. Surg Gynecol Dbstet 1984;159:481-4. 17. Jeejeebhoy KN. Therapy of the short-gut syndrome. Lancet 1983;1:1427-30. 18. Tilson MD, Wright HK. The effect of resection of the small intestine upon the fine structure of intestinal epithellum. Surg Gynecol Obstet 1972;134:992-4. 19. Weser E. Intestinal adaptation after small bowel resection. In: lsenberg JI, ed. Viewpoints on digestive disease. New Jersey: AGAS. 19781-4. 20. Feldman EJ, Dowling RH, McNaughton J, Peters TJ. Effects of oral versus intravenous nutrition on intestinal adaptation after small bowel resection in the dog. Gastroenterology 1976;70:712-9. 21. Morin CL, Ling V, Van Caillie M. Role of oral intake on intestinal adaptation after small bowel resection in growing rats. Pediatr Res 1978; 12:268-7 1. 22. Tilson MD. Pathophysiology and treatment of short bowel syndrome. Surg Clin North Am 1980;60: 1273-84. 23. Grieco GA, Reyes HM, Dstrovsky E. The role of a modified intussusception jejunocolic valve in short-bowel syndrome. J Pediatr Surg 1983; 18:354-a. 24. Myrvold H, Tindel MS, lsenberg HD, Stein TA, Sherer J, Wise L. The nipple valve as sphincter substitute for the ileocecal valve: prevention of bacterial overgrowth in the small bowel. Surgery 1984;96:42-7. 25. Careskey J, Weber TFt, Grosfeld JL. lleocecal valve replacement. Arch Surg 1981;118:618-22. 26. West KW, Malangoni MA, WeberTR, Gfosfekl JL. Theeffect of live E. coli and endotoxin on mortality following massive bowel resection. J Pediitr Surg 1981;18:846-9. 27. Malangoni MA, Cakmak 0, Grosfeld JL. Adverse effects of endotoxin following massive distal bowel resection. J Pediatr Surg 1979;14:708-12. 28. Postuma R, Moroz S, Friesen F. Extreme short bowel syndrome in an infant. J Pediatr Surg 1963;18:264-6. 29. Aaronson IA, Bowie MD, Cywes S, Louw JH. Massive small bowel resection in a neonate. Arch Surg 1975;llO: 1485-90. 30. Ball MI, Martin LW, Shubert WK. Partin J, Burke J. Massive small-bowel resection in an infant: long-term management and intestinal adaptation. J Pediatr Surg 1973;8: 197-204. 31. Ricour C. Duhamel JF, Arnaud-Battandier F, Collard Y, RevilIon Y, Nihoul-Fekete C. Enteral and parenteral nutrition in the short bowel syndrome in children. World J Surg 1985;9:310-5. 32. Rickham PP, Irving I, Shmerling OH. Long-term results following extensive small intestinal resection in the neonatal period. Prog Pediatr Surg 1977;10:65-75. 33. Postuma R, Trevenen CL. Liver disease in infants receiving total parenteral nutrition. Pediatrics 1979;63: 110-5. 34. Touloukian RJ, Seashore JH. Hepatic secretory obstruction with total parenteral nutrition in the infant. J Pediatr Surg 1975;10:353-60.
45
Grosfeld et al
35. Hodes JG, Grosfeld JL, Weber TR, Schreiner RL, Fitzgerald JF, Mirkin LD. Hepatic failure in infants on total parenteral nutrition: clinical and histopathologic observations. J Pediatr Surg 1982; 17:463-8. 36. Rigo J. Senterre J. Is taurine essential for the neonates? Biol Neonate 1977;32:73-6. 37. Pellerin D, Bertin P, Nihoul-Fekete C, Ricour C. Cholelithiasis and ileal pathoogy in childhood. J Pediatr Surg 1975;10:35-41. 38. Roslyn JJ, Pitt HA, Mann LL, Ament ME, DenBesten L. Gallbladder disease in patients on long-term parenteral nutrition. Gastroenterology 1983;84:148-54. 39. Mitchell A. Watkins RM, Collin J. Surgical treatment of the short bowel syndrome. Br J Surg 1984;7 1:329-33. 40. Hutchen NE, Mendez-Picon G, Salzberg AM. Prejejunal transposition of colon to prevent the development of short bowel syndrome in puppies with 90 percent small intestine resection. J Pediatr Surg 1973;8:771-7. 41. Garcia VG, Templeton JM, Eichelberger MR, Koop CE. Vinograd I. Colon interposition for the short bowel syndrome. J Pediatr Surg 1981;16:994-5. 42. Carner DV, Raju S. Failure of an antiperistaltic colon interposition to ameliorate short-bowel syndrome. Am Surg 1981;47:538-40. 43. Bianchi A. intestinal loop lengthening: a technique for increasing small intestinal length. J Pediatr Surg 1980; 15: 145-5 1. 44. Boeckman CR, Traylor R. Bowel lengthening for short gut syndrome. J Pediatr Surg 1981;16:996-7. 45. Vinograd BI, Merguerian P. Udassin R, Mogle P, Nissan S. An experimental model of a submucosally tunnelled valve for the replacement of the ileocecal valve. J Pediatr Surg 1984;19:726-31. 46. Binnington HB, Siegel BA, Kissane JM, Ternberg JL. A technique to increase mucosa surface area. J Pediatr Surg 1973;8:765-9. 47. Lillemoe KD, Berry WR. Harmon JW, Tai YH, Weichbrod RH, Cogen MA. Abdominal wall pedicle flaps to grow small bowel neomucosa. Surgery 1982;9 1:293-300. 48. Gladen HE, Kelly KA. Electrical pacings for short bowel syndrome. Surg Gynecol Obstet 1981;153:697-700. 49. Layzell T. Collin J. Retrograde electrical pacing of the small intestine: a new treatment for the short bowel syndrome. Br J Surg 1981;68:711-3. 50. Kirkman RL. Small bowel transplantation. Transplantation 1984;37:429-33. 51. Lillehei RC, ldezuki Y, Feemster JA, et al. Transplantation of stomach, intestines and pancreas: experimental and clinical observations. Surgery 1967;82:721-41. 52. Fortner JG, Sichuk G, Litwin SD, Beattie EJ. Immunological responses to an intestinal allograft with HLA-identical donor recipient. Transplantation 1972;14:531-5.
Diiussion Scott Boley (New York, NY): Dr. Rescorla, the number of patients in this study and the quality of the follow-up are impressive. You have made several major points worth emphasizing. First, the survival rate of these children with short bowel syndrome was far higher than most people would expect. Secondly, the cause of death in most of these children was not related to their primary short bowel syndrome but rather to its treatment with total parenteral nutrition; therefore, further improvements in the results will probably come from efforts to reduce the complications of parenteral nutrition. A third important
46
point is that in the subgroup of patients in your study with total colonic aganglionosis, where the problem is unique in that the shortage of bowel affects motility and absorption, bowel can be used for absorption even though there is abnormal motility by means of a patch enteroplasty or, as we have described it, using the right side of the colon as a reservoir. Dr. Rescorla, how would you treat a child who does not have enough bowel to realistically expect long-term survival? Jeffrey L. Ponsky (Cleveland, OH): Dr. Rescorla, does your use of the term “adaptation”of the bowel represent a modification by the patient of his diet or a true increase in the absorptive capacity of the bowel? Did you try to document this by any absorption studies? Cedric J. Priebe, Jr. (Stony Brook, NY): Dr. Rescorla, in regard to your excellent results, I think it is important to note that about 50 percent of your patients had an intact ileocecal valve, which is an important factor in slowing transit time. In patients who have lost their ileocecal valve, have you contemplated the use of other valves as described in the literature to help slow transit time and hopefully to improve intestinal absorption? All of your patients received total parenteral nutrition, which was the most important factor in their survival. We have been concerned about the use of home hyperalimentation in infants. Dr. Rescorla, you showed that when 50 percent of nutrition was being received by the enteral route, home hyperalimentation was attempted. How successful was this treatment, and what role did sepsis play in these patients? Liver failure associated with total parenteral nutrition is a significant problem that a number of centers are studying, but at present, no one has the complete solution as yet. The use of early enteral feedings by the drip feeding technique has helped to preserve villus height and enzymatic function, and we are learning more about these functions. We have also been plagued by rotavirus infection associated with this problem. Dr. Rescorla, how did you manage this infection in your series? Frederick J. Rescorla (closing): Dr. Boley, thank you for your comments concerning mortality ,related to total parenteral nutrition. In patients with total colonic aganglionosis, we have generally tried to use the right side of the colon for the patch enteroplasty, although one patient had a left colon patch. In response to your question concerning the infant with insufficient small bowel length to permit adequate survival, if this were a child with midgut volvulus or necrotizing enterocolitis, we would probably close the abdomen and then perform a second-look procedure if the child survived 24 to 48 hours. If the infant had atresia and a very short intestine of less than 15 cm in length, the decision would be left to the surgeon. Dr. Ponsky, in response to your questions regarding intestinal adaptation, we have generally observed increased absorption by clinical tolerance to feedings, and have not performed special tests.
The American Journal of Surgery
Jay L. GrosfekJ, MD, Indianapolis, Indiana Frederkk J. Rescoria, MD, Indianapolis, lndii Karen W. West, MD, Indianapolis, Indiana
Short bowel syndrome in infancy and childhood is usually the result of massive bowel loss due to intrauterine intestinal atresia, postnatal midgut volvulus, or necrotizing enterocolitis. Other causes include total colonic aganglionosis with extension of the aganglionic process to the mild or proximal small bowel, gastroschisis with volvulus or atresia, volvulus secondary to adhesive small bowel obstruction, trauma, and inflammatory bowel disease. Short bowel syndrome has been defined as the presence of less than 50 percent of the expected small bowel length at the conclusion of the initial operation [I]. Previous concepts concerning infant survival and bowel length were based on the study of Wilmore [2] which correlated a successful outcome with at least 15 cm of jejunum or ileum and an intact ileocecal valve and more than 40 cm of jejunum or ileum when the ileocecal valve had been resected. The present study reviews the clinical presentation, operative and medical management, morbidity, and mortality of 60 infants and children with short bowel syndrome. Herein, we also reassess survival in view of the development of adjunctive surgical procedures and current available nutritional support, including total parenteral nutrition and well-defined elemental diets [3-61. Patients and Methods Between January 1972 and December 1984,60 infants and children with short bowel syndrome were treated at the James Whitcomb Riley Hospital for Children, IndiFrom the Section of Pediatric Surgery, Department of Surgery, Indiana University Medical Center and the James Whitcomb Riley Hospital for Children, Indianapolis. Indiana. Requests for reprints should be addressed to Jay L. Crosfeld. MD, James Whitcomb Riley Hospital for Children, 702 Barnhill Drive (K21), Indianapolis, Indiana 46223. Presented at the 26th Annual Meeting of the Society for Surgery of the Alimentary Tract, New York, New York, May 14-15, 1965.
Volume 151, January 1986
ana University Medical Center, Indianapolis, Indiana. There were 40 boys and 20 girls. Twenty-eight infants were premature (gestational age 36 weeks or less), 26 were full-term infants, and 6 patients were older children. Causes included necrotizing enterocolitis in 24 infants, jejunoileal atresia in 13, malrotation with midgut volvulus in 6, total colonic aganglionosis with proximal small bowel extension in 7, gastroschisis in 5, volvulus related to adhesive bowel obstruction in 4, and trauma in 1. Cases of adhesive bowel obstruction included three infants with volvulus occurring after excision of a retroperitoneal tumor (Wilms’ tumor in two and teratoma in one) and a 14 year old patient who had previously undergone total proctocolectomy and ileostomy for ulcerative colitis. One instance of trauma was related to a shotgun wound with extensive mesenteric avulsion in a 6 year old child. Gangrenous or atretic bowel requiring resection was noted in 53 instances. Temporary enterostomies were performed in 40 patients in order to preserve bowel length in cases of questionable viability and to allow visualization of the mucosa at the temporary stoma site postoperatively. Second-look laparotomy at 24 to 48 hours was useful for preserving bowel length in two of four patients, one of whom had necrotizing enterocolitis and one midgut volvulus. A sterile Doppler probe was also employed intraoperatively. Delayed anastomosis was subsequently accomplished from 30 to 60 days later in these patients. Tapering proximal antimesenteric enteroplasty and primary anastomosis was performed in 13 infants with high jejunal atresia. The enteroplasty was usually accomplished over a 22 or 24 F. catheter using an autostapling device. Seven infants with total colonic aganglionosis and small bowel extension from the mid to the proximal small bowel underwent initial enterostomy in the segment of bowel with ganglion cells proved by biopsy. Two of these children had subsequent modified Soave procedures and one had an endorectal pull-through procedure with an antimesenteric aganglionic patch enteroplasty to increase the mucosal absorptive surface area. The mean length of remaining bowel was 58.4 cm, with a range of 13 to 120 cm in the infants and 120 to 150 cm in the older children. The ileocecal valve was resected in 33
41
Grosfeld et al
TABLE I
Short Bowel Syndrome: Results In 60 Patlents Results
n
%
Survivors Deaths TPN-related liver failure Sepsis Bronchopulmonary dvsolasia
51 9 5
85 15
3 1
. .
TPN = total parenteral nutrition.
and remained intact in 27. All patients were managed with total parenteral nutrition and early enteral feedings using highly defined elemental diets administered by hourly drip feeding. Home hyperalimentation was attempted when 50 percent of the total calories were enteral. Intestinal adaptation required from 3 to 14 months. Frequent setbacks were related to catheter sepsis, rotavirus infections, carbohydrate intolerance, and liver dysfunction. Gallstones developed in two patients and required subsequent cholecystectomy. Fifty-one of 60 patients survived (Table I). The shortest length of intestine in a survivor was 20 cm of proximal jejunum anastomosed to the splenic flexure. Weight gain was adequate (10th to 25th percentile), but longitudinal growth was harder to achieve, as some of the long-term survivors have short stature. There were nine deaths, including four infants with necrotizing enterocolitis, three with midgut volvulus, and two with gastroschisis complicated by atresia. In all but one patient, the ileocecal valve had been resected. Deaths were related to sepsis in three patients, bronchopulmonary dysplasia in one patient, and cholestasis associated with total parenteral nutrition that progressed to fatal liver disease in five patients. patients
Comments Short bowel syndrome continues to be a significant clinical problem, especially in the neonate. The development of sophisticated neonatal intensive care facilities has resulted in the survival of 80 to 85 percent of seriously ill premature infants [7]. Necrotizing enterocolitis, a highly lethal form of ischemic bowel injury, develops in approximately 1 to 2 percent of all patients admitted to these neonatal intensive care centers. Many of these infants require massive bowel resection, usually including the ileocecal valve, as the distal ileum and right side of the colon are most commonly involved by the ischemic process. In the present report, necrotizing enterocolitis was the most common cause of short bowel syndrome. Postoperatively, these premature infants are in fragile health, with immature enzyme systems, diminished absorptive capacities for protein, carbohydrate and fat, a reduced bile salt pool, and decreased resistance to infection compared with the full-term infant. The estimated small bow-
42
el length of the premature infant also differs significantly from that of the full-term infant. In an autopsy study, Touloukian and Smith [S] demonstrated that the small bowel length in infants varied directly with the gestational age. Small bowel length was only 114 t 21 cm in infants of 19 to 27 weeks gestation, compared with 248 f 40 cm for infants of more than 35 weeks gestation, indicating that the small bowel of the fetus doubles in length during the third trimester of pregnancy. This suggests that short residual segments of small bowel in the premature infant may eventually grow and function as well aa longer segments of bowel in the full-term infant. In infants with necrotizing enterocolitis and midgut volvulus where the viability of the entire small bowel is questionable, the use of an intraoperative Doppler probe and fluorescein dye may be helpful in preserving bowel length. In selected patients, closing the abdomen and vigorous intravenous fluid and antibiotic support has allowed a successful secondlook laparotomy 24 to 48 hours later. Two of four patients in this study had segments of bowel preserved by this technique. Whenever bowel viability is in question and anastomosis is a risk, the formation of temporary enterostomies is an attractive alternative. The second most common cause of short bowel syndrome in our study was intestinal atresia, usually due to an intrauterine volvulus or to multiple bowel atresias. This results in high jejunal atresia in which the only remaining bowel segment is the dilated proximal atretic loop. Although resection of the dilated loop is performed in infants with otherwise normal bowel length, in the infant with short bowel syndrome this is the only residual bowel and therefore, it must be preserved. Use of an antimesenteric tapering enteroplasty was successfully employed in 13 patients. This technique, originally described by Thomas [9], has also been utilized successfully by Howard and Otherson [IO], as well as by Weber et al [11]. Intestinal plication, an alternative technique, has recently been described by de Lorimier and Harrison [12]. An additional adjunctive surgical procedure is the use of an aganglionic antimesenteric patch enteroplasty in infants with total colonic aganglionosis with extension of the aganglionic process to the middle of the small bowel or jejunum. A portion of aganglionic ileum or colon can be used for the patch [13-161. The key to survival after massive bowel resection is the ability of the residual bowel to adapt, which depends on the extent and site of resection, adaptive changes in the remaining intestine, and the persistence of disease in the remaining bowel mucosa. Any bowel tissue affected by midgut volvulus or necrotizing enterocolitis may have sustained significant mucosal damage which interferes with healing and adaptation and delays tolerance of enteral feedings. Morphologic adaptive changes include in-
The American Journal of Surgery
Short Bowel Syndrome in Childhood
creases in bowel circumference, thickness of the bowel wall, average villus height, crypt depth, and rate of cell proliferation. An increase in the number of epithelial cells per centimeter of villus signifies hyperplasia of mucosal cells [ 17,181. Functional adaptation primarily depends on an increased number of absorptive cells that results in an increase in nutrient absorption and water and electrolyte transport. It brings about enhanced absorption per centimeter of residual gut of carbohydrates, protein, fat, electrolytes, water, calcium, vitamin B12, and bile acids, and an increase in the specific activity of a number of mucosal enzymes, including sodium and potassium ions, adenosine triphosphatase, enterokinase, and others. Regulation of bowel adaptation is affected by a number of mechanisms that probably stimulate intestinal hyperplasia. The most important stimulant is exposure of the intestinal lumen to an enteral nutrient [19-221. Although all infants with short bowel syndrome require total parenteral nutrition, postresection villus hyperplasia does not occur if enteral intake is denied. Cellular hyperplasia of the villus may also be stimulated by biliary and pancreatic secretions. An additional factor is the presence of an intact ileocecal valve [22]. Transit time, nutrient absorption, and survival may be dependent on the presence or absence of the ileocecal valve. A more rapid transit time, decreased absorption, and an increased mortality rate may be expected after massive distal small bowel resection, including the ileocecal valve [23-Z]. Acquired hypogammaglobulinemia has been clinically observed after massive ileal resection, and there is laboratory evidence of increased mortality after endotoxin challenge or administration of live Escherichia coli in rats with massive distal small bowel resection [26,27]. Wilmore [2] reported that infants with short gut syndrome required at least 40 cm of bowel in the absence of an ileocecal valve in order to survive. In infants with an intact ileocecal valve, those with 35 cm of residual bowel survived as did 50 percent of those with 15 to 25 cm, but there were no survivors with less than 15 cm of residual small bowel. Several other investigators have also reported successful outcomes in children with as little as 13 to 17 cm of small bowel in the presence of an ileocecal valve [B-30]. Ricour et al [31] have recently described two patients with 15 cm of small bowel without an ileocecal valve. Although both children survived, they were dependent on parenteral nutrition. In our study, one infant with 20 cm of small bowel with an anastomosis to the splenic flexure of the colon has survived for 11years on a full enteral diet. Rickham et al [32] reported excellent survival data in 49 children with short gut syndrome with residual intestinal lengths ranging from 26 to 75 cm, and suggested that the critical length of intestine was approximately 20 cm.
Volume 151, January 1986
The most significant factor in the improved survival rate of infants and children with short bowel syndrome has been the use of total parenteral nutrition, which delivers adequate calories in protein, carbohydrates, fat, and nutrients, allows growth, and provides time for intestinal adaptation to occur. The development of small Broviac or Hickman Silasti@ catheters has improved total parenteral nutritional care; the technique of administration can be taught to parents and can be carried out as a home hyperalimentation program when 50 percent of the required calories are tolerated by the enteral route. Catheter-related sepsis remains a constant threat, therefore, careful technique in catheter handling is important. Mineral deficiencies may develop during total parenteral nutrition due to excessive losses through an enterostomy or from diarrhea. These include calcium, phosphorus, zinc, copper, and magnesium deficiencies. Higher daily doses of vitamins, electrolytes, minerals, and intravenous fat solutions have greatly decreased the incidence of these nutritional deficiencies. Prolonged use of total parenteral nutrition, particularly in the premature infant, can result in severe cholestasis in 30 to 42 percent of all infants receiving total parenteral nutrition (33,341; in addition, neonates with short bowel syndrome may have an increased risk of development of cholestasis. Cooper et al [1] noted cholestasis related to total parenteral nutrition in 15 of 16 neonates with short bowel syndrome. Progressive jaundice resulting in a fatal form of poorly understood liver failure developed in three of these patients. Five infants in the present report also died from liver failure induced by total parenteral nutrition, making this unusual complication the most common cause of death in our series. Although the precise cause of total parenteral nutrition-related cholestasis is uncertain, several factors have been implicated, including prematurity, immature liver function, duration of total parenteral nutrition, and peritonitis with or without recurrent episodes of sepsis [35]. In addition, absence of enteral feedings may also contribute to progressive cholestasis. Reversal of cholestasis usually occurs when the patient progresses from total parenteral nutrition to complete enteral feedings. In the present series, however, when the bilirubin level was greater than 30 mg/dl, reversal did not occur, and a fatal outcome usually ensued, characterized by elevated alkaline phosphatase levels, hypoalbuminemia, hyperammonemia, and a coagulopathy that required daily administration of vitamin K and fresh frozen plasma. Bilirubin levels before death have increased to as much as 50 mg/dl. This is probably due to an inability of the premature infant to metabolize the 2.5 percent amino acid concentration commonly utilized in nutritional formulas for the neonatal period. Formation of a toxic bile salt may result in a canalicular injury and severe choles-
43
Grosfeld et al
tasis. Choline and taurine deficiences have also been suggested as a possible cause [36]. These observations have led to the development of new intravenous and enteral formulas containing taurine and the suggestion that the amino acid content of total parenteral nutrition solutions should not exceed 1.5 percent when meant for the premature infant. Cholelithiasis occurred in 2 of the 60 patients in this series and was managed by cholecystectomy in each instance. This is a well-recognized complication after extensive ileal resection and in patients managed with long-term total parenteral nutrition [37,38]. Its occurrence in our two patients was most likely related to a disturbance in the normal bile salt pool and decreased enterohepatic circulation as a result of bile salt losses in the stool. The incident of gallstones in the current series was lower than that reported in adults with short gut syndrome [38]; therefore, we have not performed routine cholecystectomy as a prophylactic measure in infants with short bowel syndrome. Despite the beneficial effects of total parenteral nutrition, a successful result will not be achieved without enteral nutrition. We usually begin early enteral feedings by a continuous slow drip technique through an orogastric or gastrostomy tube. Highly defined, easily digestible formulas of low osmolality are used. Formulas containing lactose are avoided due to lactase deficiency in most patients with short gut syndrome. Long chain fats are also avoided. Hyperosmolar feedings must be administered with great care to avoid injury to the hyperplastic mucosa, which may result in secretory diarrhea. Enteral intolerance to carbohydrates, including glucose, may require a low osmolar carbohydrate-free diet. Fructose may be added to increase the total caloric load. More complex substances can be added with time, including maltose, sucrose, and certain disaccharides, peptides, and medium chain triglycerides. Cholestyramine, a bile acid that binds resin, may decrease the frequency of stools and allow more absorption of water in the remaining colon. Adequate supplementation of vitamin B12, folit acid, other vitamin complexes, iron, sodium citrate, and potassium elixir are essential. The stool pH should be kept above 6, as an acidic fecal stream results in increased transit time. One milliequivalent of sodium bicarbonate with each feeding is useful in this regard. In older infants, loperamide (Imodium@) and diphenoxylate (LomotiF) may be useful adjuncts to decrease the number of stools. The period required for adaptation is quite variable and may take from 6 months to 2 years to be completed. Any cause of intestinal injury, such as sepsis, rotavirus infection, and hyperosmolar secretory diarrhea, will delay the process considerably. Patience, long-term diligence, and patient and parent sensitivity are distinct virtues in the management of this disorder. Assessment of weight in rela-
44
tion to height, albumin level, transferrin level, skin fold measurement, and transit time, as well as monitoring the consistency, pH, and number of bowel movements are useful clinical guides. Occasionally, quantitative evaluation of carbohydrate, protein, and fat absorption may be useful, but these tests are time consuming, usually require hospitalization, and are also expensive. Despite the aforementioned techniques, some infants respond poorly to therapy; for some of them, adjunctive surgical procedures have been devised. Reversed small bowel segments that recirculate bowel loops and vagotomy and pyloroplasty have been attempted, but have not been uniformly successful and are presently not in use [39]. Prejejunal isoperistaltic colonic interposition has been used successfully to increase transit time in both experimental and clinical settings [40,41]. The use of an antiperistaltic colonic segment has also been attempted; however, bowel obstruction has been a frequent complication of this technique [42]. Bianchi [43] described a bowel-lengthening procedure in pigs which involved longitudinal division of the bowel, leaving half of the mesenteric blood supply with each half of the bowel. Boeckman and Traylor [44] reported successful use of this technique in a child with gastroschisis and short gut syndrome. Several additional procedures have been successful in the experimental laboratory setting and are awaiting clinical application. Replacement of the ileocecal valve as a nipple valve intussusception in either a jejunocolic or ileocolic position increased survival rates, transit time, favorably affected weight change and growth, and decreased the anaerobic bacterial counts in the small and large bowels in experimental studies [24,25]. In addition, Vinograd et al [45] recently reported the use of a submucosally tunnelled valve of ileum into the colon as a replacement for the ileocecal valve in dogs. Binnington and associates [46) reported the use of a colonic patch to increase the mucosal surface area of small bowel, and Lillemoe et al [47] demonstrated the feasibility of growing new intestinal mucosa on rectus muscle flaps attached to the antimesenteric surface of the small bowel in rabbits. Gladen and Kelly [48] and Layzell and Collin [49] showed an increased absorption of carbohydrate and water, increased body weight, decreased fecal fat, and decreased nitrogen losses in dogs with retrograde intestinal pacing. Allotransplantation of the small bowel has been performed in seven patients [50]. The first two transplants were performed in two children in 1964, however, both died. In 1967, Lillehei et al [51] reported a cadaveric donor transplant in a 46 year old woman with bowel infarction from mesenteric venous thrombosis. Unfortunately, she died 12 hours later from pulmonary complications. The longest survivor was a woman who received an incontinuity
The American Journal of Surgery
Short Bowel
graft from her HLA-identical sister [52]. She died from sepsis on the 176th postoperative day. The success of cyclosporine A in controlling both rejection and graft versus host disease in hepatic, renal, and cardiac transplantation has led to renewed interest in small bowel transplantation. Several centers are actively involved in small bowel transplantation research, however, clinical bowel transplantation remains an elusive goal and currently must be considered an experimental procedure.
This report concerns 60 infants and children with short bowel syndrome, most commonly caused by necrotizing enterocolitis in this study. Resection of atretic or gangrenous bowel was performed in 53 patients, tapering enteroplasty and primary anastomosis was performed in 13 patients, and temporary enterostomies were performed in 40 patients. Second-look laparotomy was useful in two of four cases of questionable bowel viability. The ileocecal valve was resected in 33 patients and remained intact in 27. The mean length of remaining bowel was 58.4 cm (range 13 to 150 cm). Seven patients with total aganglionosis and mid to proximal small bowel extension were managed with an initial enterostomy, whereas three had a pull-through procedure with an aganglionic patch enteroplasty. All patients received total parenteral nutrition and early enteral feedings. Home hyperalimentation was attempted when 50 percent of the calorie intake was enteral. Intestinal adaptation required from 3 to 14 months. Frequent setbacks were related to catheter sepsis, rotavirus infection, carbohydrate intolerance, and liver dysfunction. The overall survival rate was 85 percent, with mortality due to liver failure and sepsis associated with total parenteral nutrition. References 1. Cooper A, Floyd TF, Ross AJ Ill, Bishop HC, Templeton JM Jr, Ziegler MM. Morbidity and mortality of short-bowel syndrome acquired in infancy: an update. J Pediatr Surg 1984;19:711-8. 2. Wilmore DW. Factors correlating with a successful outcome following extensive Intestinal resection in newborn infants. J Pediitr 1972;80:88-95. 3. Postuma R, Moroz S, Friesen F. Extreme short-bowel syndrome in an infant. J Pediatr Surg 1983;18:284-8. 4. Aaronson IA, Bowie MD, Cywes S, Louw JH. Massive small bowel resection in a neonate. Arch Surg 1975110: 1485-90. 5. Bell MJ, Martin LW, Shubert WF, Pat-tin J, Burke J. Massive small bowel resection In an infant: long-term management and intestinal adaptation. J Pediatr Surg 1973;8: 197-204. 6. Kurz R, Sauer H. Treatment and metabolic findings in extreme short-bowel syndrome with 11 cm jejunal remnant. J Pediatr Surg 1983:18:257-63. 7. Cikrit 0, Mastandrea J, West KW, Schreiner RL, (jrosfeld JL. Necrotizing enterocolitis: factors affecting mortality in 101 surgical cases. Surgery 1984;96:848-55. 8. Touloukian RJ, Smith GJW. Normal intestinal length in preterm
Volume 151, January 1988
Syndrome in Childhood
infants. J Psdiatr Surg 1983;18:720-3. 9. Thomas CG. Jejunoplastyfor the correction of jejunal atresia. Surg Gynecol Dbstet 1969;129:545-6. 10. Howard ER, Dthersen HB. Proximal jejunoplasty in the treatment of jejunal atresia. J Pedlatr Surg 197$5:685-g. 11. Weber TR, Vane DW, Grosfeld JL. Tapering enteroplasty in infants with bowel atresia and short gut. Arch Surg 1982;117:684-8. 12. deLorimier AA, Harrison MG. Intestinal plication in the treatment of atresia. J Pediatr Surg 1983;18:734-7. 13. Kottmeier PK, Jongco B, Velcek FT, Friedman A, Klotz OH. Absorptive function of the aganglionic ileum. J Pediatr Surg 1981;16:275-8. 14. Kimura K, Nishijima E, Muraji T, Tsugawa C, Matsumoto Y. A new surgical approach to extensive aganglionosis. J Pedlatr Surg 1981;18:840-3. 15. Shandling B. Total colon aganglionosis: a new operation. J Pediatr Surg 1984; 19:503-5. 16. Boley SJ. A new operative approach to total aganglionosis of the colon. Surg Gynecol Dbstet 1984;159:481-4. 17. Jeejeebhoy KN. Therapy of the short-gut syndrome. Lancet 1983;1:1427-30. 18. Tilson MD, Wright HK. The effect of resection of the small intestine upon the fine structure of intestinal epithellum. Surg Gynecol Obstet 1972;134:992-4. 19. Weser E. Intestinal adaptation after small bowel resection. In: lsenberg JI, ed. Viewpoints on digestive disease. New Jersey: AGAS. 19781-4. 20. Feldman EJ, Dowling RH, McNaughton J, Peters TJ. Effects of oral versus intravenous nutrition on intestinal adaptation after small bowel resection in the dog. Gastroenterology 1976;70:712-9. 21. Morin CL, Ling V, Van Caillie M. Role of oral intake on intestinal adaptation after small bowel resection in growing rats. Pediatr Res 1978; 12:268-7 1. 22. Tilson MD. Pathophysiology and treatment of short bowel syndrome. Surg Clin North Am 1980;60: 1273-84. 23. Grieco GA, Reyes HM, Dstrovsky E. The role of a modified intussusception jejunocolic valve in short-bowel syndrome. J Pediatr Surg 1983; 18:354-a. 24. Myrvold H, Tindel MS, lsenberg HD, Stein TA, Sherer J, Wise L. The nipple valve as sphincter substitute for the ileocecal valve: prevention of bacterial overgrowth in the small bowel. Surgery 1984;96:42-7. 25. Careskey J, Weber TFt, Grosfeld JL. lleocecal valve replacement. Arch Surg 1981;118:618-22. 26. West KW, Malangoni MA, WeberTR, Gfosfekl JL. Theeffect of live E. coli and endotoxin on mortality following massive bowel resection. J Pediitr Surg 1981;18:846-9. 27. Malangoni MA, Cakmak 0, Grosfeld JL. Adverse effects of endotoxin following massive distal bowel resection. J Pediatr Surg 1979;14:708-12. 28. Postuma R, Moroz S, Friesen F. Extreme short bowel syndrome in an infant. J Pediatr Surg 1963;18:264-6. 29. Aaronson IA, Bowie MD, Cywes S, Louw JH. Massive small bowel resection in a neonate. Arch Surg 1975;llO: 1485-90. 30. Ball MI, Martin LW, Shubert WK. Partin J, Burke J. Massive small-bowel resection in an infant: long-term management and intestinal adaptation. J Pediatr Surg 1973;8: 197-204. 31. Ricour C. Duhamel JF, Arnaud-Battandier F, Collard Y, RevilIon Y, Nihoul-Fekete C. Enteral and parenteral nutrition in the short bowel syndrome in children. World J Surg 1985;9:310-5. 32. Rickham PP, Irving I, Shmerling OH. Long-term results following extensive small intestinal resection in the neonatal period. Prog Pediatr Surg 1977;10:65-75. 33. Postuma R, Trevenen CL. Liver disease in infants receiving total parenteral nutrition. Pediatrics 1979;63: 110-5. 34. Touloukian RJ, Seashore JH. Hepatic secretory obstruction with total parenteral nutrition in the infant. J Pediatr Surg 1975;10:353-60.
45
Grosfeld et al
35. Hodes JG, Grosfeld JL, Weber TR, Schreiner RL, Fitzgerald JF, Mirkin LD. Hepatic failure in infants on total parenteral nutrition: clinical and histopathologic observations. J Pediatr Surg 1982; 17:463-8. 36. Rigo J. Senterre J. Is taurine essential for the neonates? Biol Neonate 1977;32:73-6. 37. Pellerin D, Bertin P, Nihoul-Fekete C, Ricour C. Cholelithiasis and ileal pathoogy in childhood. J Pediatr Surg 1975;10:35-41. 38. Roslyn JJ, Pitt HA, Mann LL, Ament ME, DenBesten L. Gallbladder disease in patients on long-term parenteral nutrition. Gastroenterology 1983;84:148-54. 39. Mitchell A. Watkins RM, Collin J. Surgical treatment of the short bowel syndrome. Br J Surg 1984;7 1:329-33. 40. Hutchen NE, Mendez-Picon G, Salzberg AM. Prejejunal transposition of colon to prevent the development of short bowel syndrome in puppies with 90 percent small intestine resection. J Pediatr Surg 1973;8:771-7. 41. Garcia VG, Templeton JM, Eichelberger MR, Koop CE. Vinograd I. Colon interposition for the short bowel syndrome. J Pediatr Surg 1981;16:994-5. 42. Carner DV, Raju S. Failure of an antiperistaltic colon interposition to ameliorate short-bowel syndrome. Am Surg 1981;47:538-40. 43. Bianchi A. intestinal loop lengthening: a technique for increasing small intestinal length. J Pediatr Surg 1980; 15: 145-5 1. 44. Boeckman CR, Traylor R. Bowel lengthening for short gut syndrome. J Pediatr Surg 1981;16:996-7. 45. Vinograd BI, Merguerian P. Udassin R, Mogle P, Nissan S. An experimental model of a submucosally tunnelled valve for the replacement of the ileocecal valve. J Pediatr Surg 1984;19:726-31. 46. Binnington HB, Siegel BA, Kissane JM, Ternberg JL. A technique to increase mucosa surface area. J Pediatr Surg 1973;8:765-9. 47. Lillemoe KD, Berry WR. Harmon JW, Tai YH, Weichbrod RH, Cogen MA. Abdominal wall pedicle flaps to grow small bowel neomucosa. Surgery 1982;9 1:293-300. 48. Gladen HE, Kelly KA. Electrical pacings for short bowel syndrome. Surg Gynecol Obstet 1981;153:697-700. 49. Layzell T. Collin J. Retrograde electrical pacing of the small intestine: a new treatment for the short bowel syndrome. Br J Surg 1981;68:711-3. 50. Kirkman RL. Small bowel transplantation. Transplantation 1984;37:429-33. 51. Lillehei RC, ldezuki Y, Feemster JA, et al. Transplantation of stomach, intestines and pancreas: experimental and clinical observations. Surgery 1967;82:721-41. 52. Fortner JG, Sichuk G, Litwin SD, Beattie EJ. Immunological responses to an intestinal allograft with HLA-identical donor recipient. Transplantation 1972;14:531-5.
Diiussion Scott Boley (New York, NY): Dr. Rescorla, the number of patients in this study and the quality of the follow-up are impressive. You have made several major points worth emphasizing. First, the survival rate of these children with short bowel syndrome was far higher than most people would expect. Secondly, the cause of death in most of these children was not related to their primary short bowel syndrome but rather to its treatment with total parenteral nutrition; therefore, further improvements in the results will probably come from efforts to reduce the complications of parenteral nutrition. A third important
46
point is that in the subgroup of patients in your study with total colonic aganglionosis, where the problem is unique in that the shortage of bowel affects motility and absorption, bowel can be used for absorption even though there is abnormal motility by means of a patch enteroplasty or, as we have described it, using the right side of the colon as a reservoir. Dr. Rescorla, how would you treat a child who does not have enough bowel to realistically expect long-term survival? Jeffrey L. Ponsky (Cleveland, OH): Dr. Rescorla, does your use of the term “adaptation”of the bowel represent a modification by the patient of his diet or a true increase in the absorptive capacity of the bowel? Did you try to document this by any absorption studies? Cedric J. Priebe, Jr. (Stony Brook, NY): Dr. Rescorla, in regard to your excellent results, I think it is important to note that about 50 percent of your patients had an intact ileocecal valve, which is an important factor in slowing transit time. In patients who have lost their ileocecal valve, have you contemplated the use of other valves as described in the literature to help slow transit time and hopefully to improve intestinal absorption? All of your patients received total parenteral nutrition, which was the most important factor in their survival. We have been concerned about the use of home hyperalimentation in infants. Dr. Rescorla, you showed that when 50 percent of nutrition was being received by the enteral route, home hyperalimentation was attempted. How successful was this treatment, and what role did sepsis play in these patients? Liver failure associated with total parenteral nutrition is a significant problem that a number of centers are studying, but at present, no one has the complete solution as yet. The use of early enteral feedings by the drip feeding technique has helped to preserve villus height and enzymatic function, and we are learning more about these functions. We have also been plagued by rotavirus infection associated with this problem. Dr. Rescorla, how did you manage this infection in your series? Frederick J. Rescorla (closing): Dr. Boley, thank you for your comments concerning mortality ,related to total parenteral nutrition. In patients with total colonic aganglionosis, we have generally tried to use the right side of the colon for the patch enteroplasty, although one patient had a left colon patch. In response to your question concerning the infant with insufficient small bowel length to permit adequate survival, if this were a child with midgut volvulus or necrotizing enterocolitis, we would probably close the abdomen and then perform a second-look procedure if the child survived 24 to 48 hours. If the infant had atresia and a very short intestine of less than 15 cm in length, the decision would be left to the surgeon. Dr. Ponsky, in response to your questions regarding intestinal adaptation, we have generally observed increased absorption by clinical tolerance to feedings, and have not performed special tests.
The American Journal of Surgery