Resolution of Intractable Cholestasis Associated With Total Parenteral Nutrition Following Biliary Irrigation

By Arthur Cooper, Arthur d. Ross III, James A. O'Neill, Jr, Harry C. Bishop, John M. Templeton, Jr, and Moritz M. Ziegler

�9 Although the protracted use of total parenteral nutri- tion (TPN) in infants is known to predispose to cholestasis, which in certain instances may not be reversible, failure to understand the pathogenesis of this condition has hindered the development of an effective medical treatment. That surgical treatment may reverse this process in selected patients is the subject of the present report. Two prema- ture infants and one term infant, with short bowel syn- drome acquired in infancy, developed conjugated hyperbili- rubinemia following institution of balanced TPN. Infectious and metabolic etiologies for the jaundice were ruled out. No excretion of ~ 'Tc- labeled DISlDA from the biliary tree was demonstrable 24 hours following injection by nuclear scintigraphy, despite the absence of extrahepatic biliary obstruction by abdominal sonography. When all conven- tional medical methods of managing the problem failed, exploration was performed. Intraoperative cholangiogra- phy showed normal intrahepatic and extrahepatic biliary ducts in each instance. Irrigation of the biliary tree with normal saline solution was then performed. In all cases, liver biopsies were obtained that were consistent with TPN-associated cholestasis and in all cases, jaundice resolved promptly following operation. We conclude that intractable cholestasis associated with TPN, regardless of its etiology, is a functional abnormality of biliary excretion that may respond in selected patients to irrigation of the biliary tree with radiographic dye and/or normal saline solution. �9 1985 by Grune & Stratton, Inc.

INDEX WORDS: Total parenteral nutrition; cholestasis.

T HE P R OTRACTED USE of total parenteral nutrition (TPN) in infants is known to predispose

to cholestasis that in certain instances may not be reversible. Recent studies have focused on the imbal- ance of amino acids that results from long-term TP N infusion, 1'2 and on taurine deficiency 3'4 as probable causative factors, but as yet no specific etiology has been determined. Unfortunately, failure to understand its pathogenesis has hindered the development of an

From the Departments of Surgery, Babies" Hospital and Colum- bia University College of Physicians and Surgeons, New York; and The Children's Hospital of Philadelphia and University of Pennsyl- vania School of Medicine, Philadelphia.

Presented before the 16th Annual Meeting of the American Pediatric Surgical Association, Kohala Coast, Hawaii, May 1-4, 1985.

Address reprint requests to Arthur Cooper, MD, Department of Surgery, Babies" Hospital, Columbia-Presbyterian Medical Center, 3959 Broadway, New York, NY 10032.

�9 1985 by Grune & Stratton, Inc. 0022-3468/85/2006-0041 $03.00/0

effective medical treatment for this condition, which may affect, with varying degrees of severity, as many as half of all prematurely born low-birth-weight infants receiving TPN. That surgical treatment may reverse the process in selected patients is the subject of the present report.

MATERIALS AND METHODS

Two premature infants and one term infant with short bowel syndrome acquired during the newborn period developed conjugated hyperbilirubinemia 4, 2, and 10 weeks following institution of balanced TPN (20-25 g/kg/d dextrose, 2.0-2.5 g/kg/d crystalline amino acids, 1.0-3.0 g/kg/d fat emulsion). In all cases, enteral feedings were attempted, but in none of the cases could volumes or concentrations sufficient to support normal growth and development be tolerated. Brief clinical summaries of the three cases follow. Representative laboratory data are summarized in Table 1.

Case 1

This preterm female, born appropriate in size for her gestational age of 35 weeks at 2,390 grams, underwent resection of an apple peel deformity and repair of multiple jejunal, ileal, and colonic atresias on the first day of life. Unfortunately, progressive intraoperative acidosis necessitated termination of the procedure and creation of a Mickulicz jejunoileostomy before all repairs had been completed, leaving her with a functional short bowel syndrome. TPN has instituted near the end of the first postoperative week; jaundice developed in the immediate postoperative period but subsequently resolved, only to recur some four weeks following operation. Vigor- ous attempts were then made to wean the infant from TPN, but enteral feedings were insufficiently tolerated. TPN was therefore continued despite persistence of her jaundice.

Case 2

This preterm female, born appropriate in size for her gestational age of 30 weeks at 1250 grams, underwent small bowel resection for necrotizing enterocolitis at one month of age, leaving her with a functional short bowel syndrome due to a high ileostomy. TPN was instituted near the end of the first postoperative week; jaundice developed some two weeks later. Vigorous attempts were then made to wean the infant from TPN, but enteral feedings were insuffi- ciently tolerated. TPN was therefore continued despite persistence of her jaundice.

Case 3

This full-term adopted female underwent massive intestinal resec- tion for midgut volvulus with gangrene at one month of age. TPN was instituted near the end of the first postoperative week and was well tolerated; her jejunoileostomy was closed some three weeks later. Enteral feedings were slowly begun but were poorly tolerated. TPN was therefore continued as her oral diet was slowly advanced. Unfortunately, jaundice developed some ten weeks following opera- tion; vigorous attempts were subsequently made to wean the infant from TPN but enteral feedings could not be further advanced. TPN

772 Journal of Pediatric Surgery, Vol 20, No 6 (December), 1985: pp 772-774

TPN-ASSOCIATED CHOLESTASIS

Table 1. Laboratory Data of Three Patients Wi th Intractable Cholestasis Associated Wi th Total Parenteral Nutrition

773

Normal Values Case 1 Case 2 Case 3

Bilirubin Total < 1.0 mg/dL 6.8 12.1 14.4

Direct <0.4 mg/dL 8.9 6.0 10.0

Serum glutamic-oxaloacetic acid transaminase (SGOT) <40 IU 60 196 207

Serum glutamic-pyruvic acid transaminase (SGPT) 16-36 IU 53 96 195

Alkaline phosphatase 17-83 IU 285 224 312

was therefore continued in cyclic fashion, but her jaundice per- sisted.

All three infants then underwent complete evaluation for obstruc- tive jaundice. In all cases, congenital and acquired infectious etiolo- gies for the jaundice were ruled out by means of serologic tests for toxoplasmosis, rubella, cytomegalovirus, and herpesvirus, as well as multiple bacterial cultures of blood and urine. In all cases, a- l-antitrypsin levels and sweat chlorides were normal. Nuclear scintigraphic scans of the biliary tree were then obtained using 99mTc-labeled DISIDA; in none of the cases was excretion of the radioisotope demonstrable 24 hours following injection, confirming the results of earlier screening tests, which showed no visual evidence of bile in any of the 12 aliquots of duodenal drainage collected over a 24 hour period. Abdominal sonograms were then obtained that revealed no evidence of extrahepatic biliary obstruction, although abnormal sonographic textures were noted in the parenchyma of all three livers that were consistent with the clinical diagnoses of cholestasis; stones were also noted within the lumen of the gallblad- der in one instance (case 3). When all conventional methods of managing this problem had failed, including continual attempts at enteral refeeding, withdrawal of amino acids from the TPN solution, withholding of fat emulsion from the infusate, and administration of phenobarbital, exploration was performed.

RESULTS

Intraoperative cholangiography showed normal intrahepatic and extrahepatic biliary ducts in each instance. Hyperviscidity of the bile within the gallblad- der, obtained by aspiration to confirm the position of the cholangiographic catheter within the lumen of the gallbladder prior to injection of radiopaque dye, was noted in each case; accordingly, gentle irrigation of the biliary tree with several milliliters of normal saline solution to wash out both the radiographic dye and the hyperviscid bile was performed following the cholan- giogram. In each case, a liver biopsy was also obtained that subsequently confirmed the diagnosis of TPN- associated cholestasis established clinically at opera- tion; in two instances (cases 1 and 3), tube cholecystos- tomies were also performed, but drained only scant amounts of bile postoperatively. Finally, in all cases, the jaundice, which had persisted without improve- ment for 4, 6, and 8 weeks following its onset, progres- sively improved following operation, and completely resolved within two (cases 1 and 2) to four (case 3) weeks.

DISCUSSION

Intractable cholestasis is generally recognized to be the most serious metabolic complication of TPN in infants. First described histologically in 1971, 5 the lesion is characterized clinically by late onset of jaun- dice in association with conjugated hyperbilirubinemia and certain abnormalities of liver function, specifically mild to moderate increases in serum transaminases and moderate to marked increases in serum alkaline phosphatase. Prematurely born low-birth-weight in- fants are somewhat more susceptible to this lesion6; indeed, it may affect nearly 50% of such infants who receive TPN for protracted periods. Numerous etiolo- gies for the lesions have been proposed and later disproved; however, the fact that it most closely resem- bles a toxic liver injury, 7 together with the knowledge that biliary excretion of amino acids is saturable, 2 suggests that cholestasis may be due to a large hepatic load of potentially injurious amino acids. That bile acid independent flow (BAIF) is adversely affected by administration of commercially available amino acid mixtures ~ lends further support to this hypothesis, despite the fact that the development of cholestasis appears to be independent of the amount of protein intake, g since administration of such mixtures gener- ally results in an abnormal or imbalanced postprandial plasma amino acid profile, even in pharmacologic dose ranges. 9 The amino acid(s) responsible for this diminu- tion in BAIF have not yet been identified, but abnor- mally high levels of glycine are consistently detected both in the plasma of infants receiving "standard" TPN solutions 9 and in the perfusate of the isolated perfused rat liver systems that have been studied, ~ a change that is associated with significant increases in the hepatic uptake of this same amino acid. Interest- ingly, administration of Trophamine (American McGaw, Irvine, Calif), a new amino acid mixture designed specifically to mimic the postprandial plasma amino acid profile of the infant, has been associated in its clinical trials with a much lower rate of cholestasis than other solutions, ~~ but this finding has yet to be confirmed by large-scale experience.

The role of taurine is not yet clear; chronic deft-

774 COOPER ET AL

ciency may play a permissive role in the development of the severe, irreversible hepatic dysfunction, which occasionally develops from intractable cholestasis,l~ by decreasing the ratio of tauroconjugates of bile acids to the more cholestatic glycoconjugates,* but short- term administration of taurine during the first two weeks of life does not appear to affect the incidence of cholesta- sis. 12 It is, of course, noteworthy that the ratio of tauroconjugates to glycoconjugates of bile acids may also be lowered by increasing the circulating pool of glycine; this, of course, is precisely what the use of s tandard T P N solutions will do, and may represent an effect of glycine that is distinct from its role in decreasing BAIF.

Clearly, much remains to be learned before an effective medical t reatment for the intractable choles- tasis associated with T P N is available. However, evi- dence is presented in the current study that certain patients with TPN-associa ted cholestasis unresponsive to all forms of conventional medical management , in whom both infectious and metabolic causes have been ruled out, and who show no evidence of extrahepatic biliary obstruction despite the proven absence of bile flow, may respond to irrigation of the biliary tree with normal saline solution, after a cholangiogram has been obtained to confirm normal biliary ana tomy and the absence of obstructing choledocholiths. The mecha- nism of this action is unknown, but may be related to simple mechanical cleansing of hyperviscid secretions from the biliary tree, as in cases of the inspissated bile

syndrome, for which similar procedures have been previously used. ~3'14 In this regard, it is noteworthy that a recent histopathological study has also suggested that cholestasis itself may play a pr imary obstructive role in the progression of TPN-associated liver dis- ease. 15

Nevertheless we do not recommend that this proce- dure be used for every child with presumed TPN- associated cholestasis. Rather , it should be used as a last resort in patients who do not respond to conven- tional medical management , and only then in the course of a complete work-up of obstructive jaundice. In our opinion, the procedure of biliary irrigation should be viewed simply as a natural extension of the operative cholangiogram, designed to wash out the radiographic dye from the biliary tree together with the hyperviscid secretions with which it has become admixed. However, if it must be used, it should be performed within three months of onset of the jaun- dice, since after this t ime the syndrome may not be completely reversible. ~6

Finally, while we conclude that intractable cholesta- sis associated with TPN, regardless of its etiology, is a functional abnormali ty of biliary e~re t ion which may respond in selected patients to irrigation of the biliary tree with radiographic dye a n d / o r normal saline solu- tion, it is our belief that an effective medical t reatment for this condition is close at hand, and that the proce- dure will therefore be required only in exceptional cases.

REFERENCES

1. Graham MF, Tavill AS, Halpin TC, et al: Inhibition of bile flow in the isolated perfused rat liver by a synthetic parenteral amino acid mixture: Associated net amino acid fluxes. Hepatology 4:69-73, 1984

2. Preisig R, Rennert O: Biliary transport and cholestatic effects of amino acids. Gastroenterol 73:1240, 1977 (abstr)

3. Roy CC, Dorvil N, Tuchweber B, et al: Taurine prevents cholestasis induced by lithocholate acid sulfate in guinea pig. Pediatr Res 16:175A, 1982 (abstr)

4. Roy CC, Yousef IM, Dorvil N, et al: The pattern of bile acid conjugation: A critical determinant of the cholestatic potential of sulfolithocholate. Pediatr Res 17:199A, 1983 (abstr)

5. Peden VH, Witzleben CL, Skelton MA: Total parenteral nutrition. J Pediatr 78:180--181, 1971 (letter to editor)

6. Pereira GR, Sherman MS, DiGiacomo J, et al: Hyperalimen- tation-induced cholestasis. Am J Dis Child 135:842-845, 1981

7. Bernstein J, Chang C-H, Brough A J, et al: Conjugated hyper- bilirubinema in infancy associated with parenteral nutrition. J Pediatr 90:361-367, 1977

8. Vileisis RA, Inwood R J, Hunt CE: Prospective controlled study of parenteral nutrition-associated cholestatic jaundice: Effect of protein intake. J Pediatr 96:893-897, 1980

9. Winters RW, Heird WC, Dell RB, et al: Plasma amino acids in infants receiving parenteral nutrition, in Green HC, Holliday MA, Munro HN, (eds): Clinical Nutrition Update: Amino Acids. Chica- go, American Medical Association, 1977, pp 147-157

10. Heird WC: Personal communication

11. Cooper A, Betts JM, Pereira GR, et al: Taurine deficiency in the severe hepatic dysfunction complicating total parenteral nutri- tion. J Pediatr Surg 19:462-465, 1984

12. Cook RJ, Whitington EF: Effect of taurine supplementation on hepatocellular function. Pediatr Res 17:165A, 1983 (abstr)

13. Rickham PP, Lee EYC: Neonatal jaundice: Surgical aspects. Clin Pediatr (Phila) 3:197-208, 1964

14. Bernstein J, Braylan R, Brough A J: Bile-plug syndrome: A correctable cause of obstructive jaundice in infants. Pediatr 43:273- 276, 1969

15. Benjamin DR: Hepatobiliary dysfunction in infants and children associated with long-term total parenteral nutrition. Am J Clin Pathol 76:276-283, 1981

16. Cohen C, Olsen MM: Pediatric total parentaral nutrition: Liver histopathology. Arch Pathol Lab Med 105:152-156, 1981