Hyperalimentation Associated Hepatotoxicity in the Newborn*

Transcription

1 ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 22, No. 2 Copyright 1992, Institute for Clinical Science, Inc. Hyperalimentation Associated Hepatotoxicity in the Newborn* JAN M. GOPLERUD, M.D. D ivision o f N eonatal-perinatal M edicine H ospital o f the U niversity o f P ennsylvania P hiladelphia, PA ABSTRACT Total parenteral nutrition (TPN) has become a mainstay of m odern neonatal care for the increasing population of prem ature infants who survive their initial pulmonary disease. As with other advances in neonatal therapy, hyperalim entation has associated complications and limitations, primary among them its toxicity to the liver. The basic pathologic lesion is bile cholestasis w hich is probably m ultifactorial in etiology. Amino acid solutions, excessive calorie-to-nitrogen ratios, and deficient trace elem ents and antioxidants have all been im plicated in this process. Total parenteral nutrition-cholestasis can progress to portal fibrosis and irreversible cirrhosis if long-term hyperalim entation is required. M ost at-risk for this iatrogenic condition are those prem ature infants less than 1500 g birth weight who are exposed to TPN for longer than two weeks. Enteral feedings providing as little as 10 percent of caloric intake are beneficial, and the prognosis for recovery is good once enteral feedings are established. History In 1968, W ilmore and D udrick first reported the successful long-term use of total parenteral nutrition (TPN) in an infant at the C hildren s Hospital of Philadelphia.11,29 This was followed in 1969 by Filler et al13 with a case series of 14 infants m anaged w ith TPN at the C hildren s Hospital in Boston. The infants involved prim arily had surgical com plications of th e gastrointestinal tract for * Send reprint requests to Jan M. Goplerud, M.D., Assistant Professor of Pediatrics, Division of Neonatal-Perinatal Medicine, Hospital of the University of Pennsylvania, 3400 Spruce Street 2nd Floor Maloney Building, Philadelphia, PA w hich they could not be fed. T he successful im plem entation of intravenous hyperalim entation with dextrose, amino acid s, and fat r e p re s e n te d a m ajor advance in neonatal care and survival. Thus began the inclusion of h y p eralimentation in the armamentarium of pediatric care, in itia lly for post-surgical patients. Its use quickly extended, how ever, to the increasing numbers of prem a ture infants who were surviving the pulm onary d ise a s e s of p re m a tu rity to require intravenous nutritional support until able to tolerate full enteral feeds. Then, in 1971, came a case report by Peden and his associates W itzleben and Skelton,20 of cholestasis and cirrhosis on post-mortem exam in a prem ature infant /92/ $00.90 Institute for Clinical Science, Inc.

2 80 GOPLERUD after long-term TPN. This m arked the beginning of the recognition of hyperalim entation-related hepatotoxicity and the first step in defining its pathophysiology and mechanisms. A series of case reports then quickly appeared confirm ing the association betw een hyperalim entation and cholestatic liver dysfunction.3,4,22 As had happened before with advances in neonatal care, a m ilestone in supportive care of the prem ature was followed by the recognition of its toxicity. Just as retro len tal fib ro p lasia and blin d n ess re s u lte d from th e u n m o n ito re d and excessive use of supplem ental oxygen and bronchopulm onary dysplasia was seen with the widespread use of mechanical ventilation in neonates, so hyperalim entation revealed its lim itations. Clinical Features T w en ty years of e x p e rie n c e w ith hyperalim entation in the neonatal intensive care setting have dem onstrated that the prem ature infant less than 1500 g, and usually 25 to 32 weeks gestational age, is most likely to develop TPN hepatotoxicity, especially if the duration of hyperalim entation is greater than two weeks.21 With current survival rates, the population at risk is considerable. During a nine m onth period at the Hospital of the University of Pennsylvania, for example, in which 2613 live births occurred, 34 newborns less than 1500 g were admitted to the Neonatal Intensive Care Unit who subsequently received a minimum of 14 days of TPN, thus making them potential candidates for hyperalim entation hepatotoxicity. In a series of 62 infants reported by Beale et al,2 50 percent of neonates with a birth w eight less than 1000 g, and 18 percent of those 1000 to 1500 g in weight developed cholestasis. Of infants weighing 1500 to 2000 g at birth, only seven p ercen t developed cholestasis. C linically, the hepatotoxicity associated w ith hyperalim entation in the n ew b o rn is in sid io u s in o n se t; th e patients asymptomatic. H yperbilirubinem ia is the first abnorm ality detected, typically three to four weeks after initiation of TPN and beyond the usual seven to 10 day postnatal peak of unconjugated, physiologic h y p erb iliru b in em ia. Biochem ically, how ever, elevated serum bile and concentrations appear as early as five days6 after hyperalim entation is started and are present in 85 percent of infants after two w eeks of TPN.24 F ollowing the rise in serum bile acids, the most detectable biochemical event is an increase in serum conjugated bilirubin. Serum aminotransferases are frequently norm al early in the course of TPN - a sso c ia te d h e p a to to x ic ity a n d only increase after several weeks of cholestasis. Alkaline phosphatase concentrations are not particularly helpful in the diagnosis in the newborn period because of the norm ally high levels and the high concurrent incidence of bone disease in prem atures. Gamma peptidyl transferase, w hile a sensitive indicator of hepatic dysfu n c tio n, is n o t s p e c ific for T PN - associated hepatotoxicity and is elevated by d ru g th e ra p y, p a rtic u la rly a n ticonvulsants. 5'-Nucleotidase is a useful indicator of hepatic dysfunction during parenteral nutrition. H epatic synthetic function, reflected in serum album in and pre-album in concentrations, and p ro throm bin time remains normal with hyperalimentation-associated liver disease. Evaluation of Patients with Suspected TPN Cholestasis H yperalim entation-associated cholestasis is a diagnosis of exclusion. The evaluation of patients w ith suspected TPN cholestasis thus involves ruling out the various other causes of cholestatic liver disease in the newborn. An ultra

3 HYPERALIMENTATION ASSOCIATED HEPATOTOXICITY IN NEWBORN 81 sound of the biliary tract is a noninvasive study useful to detect structural abnormalities such as cysts and gallstones, and to detect the presence or absence of a gallbladder. Blood can be screened for alpha-1-antitrypsin phenotype, alpha-1- antitrypsin deficiency being a leading cause of neonatal cholestasis.19 Urine and serum samples can also be sent for amino acid analysis to look for evidence of m etabolic disorders, especially tyrosinem ia and galactosemia. Cystic fibrosis should be considered and excluded as a cause of neonatal cholestasis. A sweat chloride test for cystic fibrosis is difficult to accom plish in the prem ature infant, how ever, owing to inadequate sweat collection. Urinary tract infection can result in cholestasis and can be diagnosed on the basis of urine chem test and culture. Intrauterine-acquired infections, such as the TO R C H group, should be considered, particularly in the neonate who is small-for-gestational age and/or dysmorphic. T h ese in clu d e, Toxoplasm osis, O th e r, (sy p h ilis, p a rv o v iru s, H IV ), Rubella, Cytomegalovirus, and Herpes. To rule out biliary atresia, radioisotope scans (HIDA, PIPIDA) use an im inodiacetic acid agent to demonstrate hepatocyte function by the ability to excrete the isotope into the GI tract. Finally, liver biopsy may be necessary to docum ent both etiology and extent of hepatotoxicity. In one series of 47 newborns evaluated for TPN-associated cholestasis,12 five (10 percent) were found to have disorders other than hyperalim entation as the cause for th eir cholestasis. Pathology L iver Biopsy The initial lesion seen on liver biopsy and a constant feature of hyperalim entation-associated hepatotoxicity is bile stasis, both canalicular and hepatocellular. Cohen and O lsen19 identified canalicular cholestasis in 84 percent of newborns after as little as 10 days on hyperalim entation, followed by bile duct proliferation in 64 percent after three weeks of TPN. Next, there is a ballooning of hepatocytes and Kupffer cell hyperplasia, leading to lobular disarray of liver structure. D eposits of Iipofuscin pigm ent are present in the Kupffer cells and accumulate as parenteral n u tritio n continues. M ild-tosevere portal inflammation develops that is predom inantly lymphocytic, although eo sin o p h ils may also b e p ro m in en t. Finally, portal fibrosis appears. Fifty p ercent of infants reported by Dahms and H alpin10 had panlobular or pericellular fibrosis on initial liver biopsy. Irreversible changes such as moderate to severe portal fibrosis were seen only after 90 days of TPN and more advanced cirrhosis after even longer duration of TPN.9 Pathophysiology The pathogenesis of hyperalim entation-associated cholestasis is unknown, but appears to be m ultifactorial in etiology. It is important to rem em ber that the neonates who are prone to this disorder are prem ature and often have significant lung disease, hypoxia, hem odynam ic instability, and infections. In addition, they are repeatedly exposed to blood products w hich p u t them at risk for acquired hepatitis and other infections, and medications that affect liver function. There is an inherent immaturity of the enterohepatic circulation that contributes to bile stasis, resulting in a physiologic cholestasis. Thus, because of the lack of enteral feedings in m any of these neonates, there is a lack of stim ulation of enteric horm ones such as secretin, glucagon, gastrin, and m otilin,1 which play a role in the initiation and m aintenance of normal bile flow. While each of the components of hyperalim entation solutions has been implicated in the developm ent of hepatotoxicity, the evidence for amino

4 82 GOPLERUD acid solution toxicity is the m ost convincing. Both clinical27 and laboratory studies30 have dem onstrated the cholestatic potential of amino acids. Since cholestasis appears to be the prim ary hepatopathological lesion,4 amino acid solutions contribute to the earliest stage in the developm ent of TPN hepatotoxicity. D eficiency states of carnitine, m olybdenum, taurine, selenium, and vitamin E may also add to a predisposition to cholestasis with TPN. Specifically, carnitine deficiency has been associated with prolonged TPN, leading to inadequate fatty acid oxidation by im pairm ent of transport across the m itochondrial m em brane.7 M olybdenum is an ultratrace metal that is a co-factor in the enzyme system responsib le for th e d e g ra d a tio n of sulfurcontaining amino acids. Its deficiency may p o ten tia te th e hepatotoxicity of am ino acid so lu tio n s. T a u rin e is an essential amino acid-like substance that neonates have only a lim ited ability to synthesize. Since taurine is involved in the conjugation of bile acids, it may be a rate-limiting factor in bile flow, in which case its deficiency could contribute to cholestasis. Selenium, another trace element, is involved in antioxidant defense, as is vitamin E, both of which are known to be deficient in prem ature infants. The generation of oxygen free radicals, which are known to cause lipid peroxidation of m em branes,17 is one proposed m echanism of hyperalim entation associated hepatotoxicity.5 Lipofuscin pigment, w hich accu m u lates in K upffer cells throughout the duration of parenteral nutrition, is felt to be the product of lipid peroxidation. L ipofuscin was found in 90.3 percent of the livers of one series of infants on long-term TPN9 and as such is evidence of an oxygen free radical m echanism of liver damage. The general inadequacy of neonatal antioxidant defenses limits the ability to prevent or repair peroxidative injury, increasing the vulnerability to hepatotoxicity.18 M anagement of TPN-Associated Cholestasis Those patients most at risk for TPNassociated hepatotoxicity are exactly the ones least able to tolerate its withdrawal once evidence of toxicity presents. In lieu of stopping hyperalim entation, certain modifications of its adm inistration may attenuate its hepatotoxicity. R educing the protein infusion to two g per kg per day is suggested to minimize amino acid-related damage w hile continuing to m eet the nitrogen requirem ent in the grow ing p rem a tu re. In tra v en o u s fat em ulsions should be continued since currently available preparations are not associated with hepatotoxicity.27 A calorie-to-nitrogen ratio of less than 200:250 is recom m ended as is the initiation of enteral feeds, if possible. The prognosis for TPN cholestasis is good, if enteral feeds can be started and m aintained, even if they provide as little as 10 to 20 percent of total calorie intake. The worst clinical outcome has been reported for those infants u n a b le to to lerate any enteral feeds.23 Although liver enzyme changes persist for weeks to months after TPN is discontinued, the gradual normalization of bile follow and hepatic function does follow. More recently, glutam ine has been proposed as a trophic factor for the bowel that may facilitate recovery after enterocolitis14 or extensive intestinal resection.28 Adding glutam ine to TPN solutio n s m ay r e p r e s e n t a b e n e f ic ia l therapeutic intervention to prevent or alleviate TPN-associated liver dysfunction.16 Other data from TPN -dependent adult patients, as well as experimental animals, suggest that gut flora contribute to hyperalim entation-related hepatotoxicity. Studies using m etronidazole,8 15 polymyxin B,26 or oral gentam icin25 to alter intestinal bacterial flora have all demonstrated some improvement in parenteral nutrition-associated cholestasis.

5 HYPERALIMENTATION ASSOCIATED HEPATOTOXICITY IN NEWBORN 83 Specifically, each of these agents reduces the production of certain hepatotoxic by-products, such as enterotoxin, lithocholate, and tum or necrosis factor, of the bacterial overgrowth that occurs in the absence of enteral feedings. In summary, the hepatotoxicity associated with hyperalim entation in the new born is an iatrogenic condition for which preterm infants, and especially those requiring more than two weeks of TPN, are at greatest risk. It can be detected early by routine m onitoring of liver function tests, and is best m anaged by optimization of overall nutrition and growth, and earliest-possible im plem entation of enteral feeds. References 1. AynSLEY-GrEEN, A.: Plasma hormone concentrations during enteral and parenteral nutrition in the human neonate. J. Pediatr. Gastroenterol. Nutr. 2:S108-S112, B e a l e, E. F., N e l s o n, R. M., B u c c i a r e l l i, R. L., et al: Intrahepatic cholestasis associated with parenteral nutrition in premature infants. Pediatrics 64: , BENDA, G. I. M. and BABSON, S. G.: Peripheral intravenous alimentation of the small premature infant. J. Pediatr. 79: , BENJAMIN, D. R.: Hepatobiliary dysfunction in infants and children associated with long-term total parenteral nutrition. A clinico-pathologic study. Am. J. Clin. Pathol. 76: , B e r g e r, H. M., D e n O u d e n, A. L., and CALAM E, J. J.: Pathogenesis of liver damage during parenteral nutrition: Is lipofuscin a clue? Arch. Dis. Child. 60: , B l a c k, D. D., S u t t l e, A., W h i t i n g t o n, P. F., et al: The effect of short term parenteral nutrition on hepatic function in the human neonate: A prospective randomized study demonstrating alteration of hepatic canalicular function. J. Pediatr. 99: , BREMER, J.: Carnitine metabolism and functions. Physiol. Rev. 63: , C a p r o n, J. P., H e r v e, M. A., G i n e s t o n, J. L., et al: Metronidazole in prevention of cholestasis associated with total parenteral nutrition. Lancet 1: , C o h e n, C. and O l s e n, M. M.: Pediatric total parenteral nutrition. Arch. Pathol. Lab. Med. i 05: , DAHMS, B. B. and H a l p in, T. C.: Serial liver biopsies in parenteral nutrition associated cholestasis of early infancy. Gastroenterology 81: , D u d r ic k, S. J., W il m o r e, D. W., V a r s, H. M., et al: Long-term total parenteral nutrition with growth, development, and positive nitrogen balance. Surgery 64: , F a r r e l l, M. K. and Ba l is t r e r i, W. F.: Parenteral nutrition and hepatobiliary dysfunction. In: Clinics in Perinatology, vol. 13, no. 1, March 1986, pp F i l l e r, R. M., E r a k l is, A. J., R u b in, V. G., and Das, J. B.: Long-term parenteral nutrition in infants. New Engl. J. Med. 281: , Fox, A. D., D e P a u l o, J. A., Kr ip k e, S. A., et al: Glutamine supplemented elemental diets reduce endotoxemia in a lethal model of enterocolitis. Surg. Form 39:46-51, F r e u n d, H. R., M u g g i a - S u l l a m, M., L a F r a n c e, R., et al: A possible beneficial effect of metronidazole in reducing TPNassociated liver function derangements. J. Surg. Res. 38:356, F r e u n d, H. R.: Abnormalities of liver function and hepatic damage associated with total parenteral nutrition. Nutrition 7:1-6, H a l l i w e l l, B. and G u t t e r i d g e, J. M. C.: Lipid peroxidation, oxygen radicals, cell damage, and antioxidant therapy. Lancet 1: , Lubrano, R., F rediani, T., C it t i, G., Cardi, E., Mannarino, O., E lli, M., Co z z i, F., and GlARDINI, O.: Erythrocyte membrane lipid peroxidation before and after vitamin E supplementation in children with cholestasis. J. Pediatr. 715: , M o r o z, S. P., C u t z, E., C o x, D. W., et al: Liver disease associated w ith alpha-1- antitrypsin deficiency in childhood. J. Pediatr. 88:19-25, Peden, V. H., Witzleb en, C. L., and Skelto n, M. A.: Total parenteral nutrition. J. Pediatr. 78: , P e r e ir a, G. R., S h e r m a n, M. S., D ig ia c o m o, J., Z i e g l e r, M., R o t h, K., and Ja c o b o w s k i, D.: Hyperalimentation-induced cholestasisincreased incidence and severity in premature infants. Amer. J. Dis. Child. 135: , Postuma, R. and TREVENEN, C. L.: Liver disease in infants receiving total parenteral nutrition. Pediatrics 6 3 : , Ra g e r, R. and F in e g o l d, M. J.: Cholestasis in immature newborn infants: Is parenteral alimentation responsible? J. Pediatr. 86: , S o n d h e im e r, J. M., B r y a n, H., An d r e w s, W., et al: Cholestatic tendencies in premature infants on and off parenteral nutrition. Pediatrics 6 2 : , S p u r r, S., G r y l a c k, L. J., and M e h t a, N. R.: Hyperalimentation associated neonatal cholestasis: effect of oral gentamicin. J. Parenter. Enteral Nutr. 13:633, 1989.

6 84 GOPLERUD 26. Stokes, D. C., Sh e n o p, J. L., F ishm an, M., et al: Polymyxin B prevents lipodysaccharideinduced release of TNF from alveolar macrophages. J. Infect. Dis. 160:52, Vil e is is, R. A., I n w o o d, R. J., and H u n t, C. E.: Prospective controlled study of parenteral nutrition-associated cholestatic jaundice: Effect of protein intake. J. Pediatr. 96: , Wang, X. D., Jacobs, D. O., O D w yer, S. T., et al: Glutamine enriched parenteral nutrition prevents mucosal atrophy following massive small bowel resection. Surg. Form 39:44 46, Wilm ore, D. W. and D udrick, S.J.: Growth and development of an infant receiving all nutrients exclusively by vein. JAMA 203: , Zahavi, I., Sh a ffe r, E. A., and Ga l l, D. G.: Total parenteral nutrition-associated cholestasis: acute studies in infant and adult rabbits. J. Pediatr. Gastroenterol. Nutr. 4: , 1985.