- Email: [email protected]
Biliary obstruction in infants with cystic fibrosis requiring Kasai portoenterostomy
Biliary
Obstruction By Stephen
in Infants With Cystic Fibrosis Kasai Portoenterostomy
K. Greenholz,
Baiya Krishnadasan, Clifford Sacramento, California
0 Cystic fibrosis is associated with an inspissated bile syndrome producing cholestasis secondary to plugging of macroscopically normal bile ducts. In extreme neonatal forms, with early profound intrahepatic cholestasis, the process can be associated with a marked decrease in ductal diameter, varying from hypoplasia to atresia. From 1990 to 1995 three infants were identified with cystic fibrosis, persistent jaundice, and complete absence of biliary excretion despite expectant and conservative treatment including choleretics and surgical biliary irrigation. Abdominal ultrasounds showed contracted gallbladders and no evidence of dilated ducts. Liver biopsy results in two infants showed portal fibrosis, paucity of bile ducts, and minimal inflammation. The third infant had moderate inflammation, bile duct replication, and plugging. Two infants had undergone intestinal resection followed by hyperalimenation for complications of meconium ileus in the newborn period. Surgical exploration was undertaken at 7 to 12 weeks of age. Gross findings were typical of biliary atresia with microgallbladders and nonpatency of the cystic duct. Cholangiograms failed to document ductal patency in two patients who were then treated with a Kasai portoenterostomy. The third infant had patent hypoplastic ducts and underwent only biliary irrigation. Although bile flow was transiently achieved, jaundice recurred, and at reexploration at 16 weeks of age a Kasai poroenterostomy was performed. Histological review of the biliary specimens showed microscopically patent ducts in two patients and proximal patency and distal atresia in the third. All the ducts had mural fibrosis with cystic changes. Bile drainage was achieved in each instance, although in one patient with hypoplastic ducts scant output of highly concentrated bile proved insufficient to arrest progressive liver failure. The subsequent two patients responded with resolution of hyperbilirubinemia and normalization of liver function. They remain free of biliary complications at 30 and 40 months postoperatively. This manifestation of cystic fibrosis in infants is suggested by prolonged jaundice unresponsive to choleretics, nondilated bile ducts and gallbladder on ultrasound, absent biliary excretion on nuclear scan, and characteristic liver biopsy. Exploration is warranted, and discovery of atrophic bile ducts may be best managed with reconstruction. Copyright o 1997 by W.6. Saunders Company
Marr, and Robert
Requiring Cannon
abnormalities, including hypoplasia and atresia,4.sand subsequentprogressive fatal liver disease.2,6-s To illustrate these issues, we report three infants with cystic fibrosis who eventually underwent Kasai portoenterostomy as treatment for persistentjaundice after failing to respondto conservative treatment. A comparative literature review is included. CASE REPORTS Case 1 A boy was born at 35 weeks’ gestation and required abdominal exploratton on the 11 th day of life for mecomum peritonitis wtth persistent obstructton. Ileal atresia with perforation was treated with resection and primary anastomosis. Feeds were started at age 21 days after 3 weeks of parenteral nutrition. Perststent jaundice (bihrubm, 8.5 mg/dL) was noted over the next 4 weeks, and the patient did not respond to oral cholerettcs (ursodeoxycholic acid and phenobarbital). Metabolic and mfectious work-ups were negative, and sweat chloride test confirmed the dtagnosts of cystic fibrosis. Liver scan showed absence of hepattc excretion at 24 hours. and abdommal ultrasound could not tdenttfy a gallbladder or dilated ducts. Percutaneous liver biopsy results showed portal zones with moderate fibrosts. btle duct replication. moderate inflammation. and bile plugging. Surgical exploratton was undertaken at 7 weeks of age Gross findings were similar to biliary atresia. wtth a small shrunken gallbladder contammg clear mucous and a very small tibrottc hepattc duct without an apparent lumen, Cholangiography vta the gallbladder failed to confirm ductal patency. A Kasai portoenterostomy with extertortzatton was performed. Transection of the hepatic duct at the liver htlum showed microscoptc patency of the left hepatic duct and obhteration of the right hepatic duct. Histological examination of these ducts showed architectural alteration with fibrosis. a markedly narrowed lumen (0.2 mm or less) and cysttc changes within the wall. Bile flow was achieved, although of relatively low volume (0 to 20 cm’/d) and very high concentratron (30 to 50 mg/dL). Multiple steroid boluses only transiently Improved output. Serum bilirubin gradually decreased to 5.2 mg/dL by 6 weeks after operation, but then began to increase. Bile output decreased to 1 to 5 cm3/d and did not change with ,iddittonal steroids. Serum bilirubm level Increased to 36 mg/dL 5 months after operation, and the parents declined consideration of hcer transplant. The infant died shortly thereafter.
Case 2 INDEX WORDS: Cystic fibrosis, biliary neonatal jaundice, inspissated bile.
atresia,
cholestasis,
P
ROLONGED NEONATAL jaundice in conjunction with a diagnosis of cystic fibrosis is generally presumed to be secondary to plugging of normal intraand extrahepatic bile ducts by abnormally thickened, highly viscus, biliary secretions.Relatively limited anecdotal experience has implied that conservative treatment will be followed by eventual spontaneousresolution-” However, there are equally compelling reports of intrahepatic cholestasis associated with extrahepatic biliary JournalofPediatncSurgery,
Vol32,No
2 (February),1997.pp
175-180
A boy was admitted to the hospital at 5 weeks of age for jaundice (serum bihrubm. 9.3 mg/dL) noted for 3 weeks Evaluation for mfecttous and metabolic causes was negative. but a sweat test dtag-
Presetlted ot the .SII~~ICY/I A ~socut~oti.
27th Amtnnl Surr Dte,qo.
Meettn,q Cul/jtirwc~.
qf the Amermn Pediatric Mu?. 20-23, 1996.
Address rqmnt requests to Stephrrr K Greerd~ol;. Pedtutrrc Surgery, 4301 X St, Room 2310, Sncrumento. C’oprr~ht o 1997 by W B Saunders Company
MD, Dwsron CA 95817.
oj
002-‘-3368/97/3202-0006$03.00/O 175
GREENHOLZ
176
nosed cystic fibrosis. Ultrasound could not demonstrate a gallbladder or bile ducts, and a liver scan showed no excretton. Ltver biopsy results showed pauctty of bile ducts, moderate portal tibrosis, and minimal inflammation. Persistence of jaundice, desptte oral cholerettcs, led to exploration at 8 weeks of age. A small fibrotic gallbladder had no visible lumen. but the common btle duct, although small. could be directly cannulated with a 27-gauge needle. Cholangiogram showed contrast withm the duodenum, although the ducts themselves were not clearly seen. Biliary irrigation with acetylcysteme and salme was performed. Histological examination of the gallbladder and cystic duct showed absence of a true lumen, and fibrous and cystic changes withm the wall. Durmg the first 10 days after surgery, serum bilirubm levels decreased from 9.3 mg/dL to 3.5 mg/dL, but then began to rise, reaching 9.4 mg/dL 1 week later. Two steroid pulses were ineffectual. It was decided to treat the patient conservatively with oral choleretics; however, Jaundice persisted, and repeat liver scan at 4 months of age showed no excretion The child was reexplored and underwent Kasat portoenterostomy without extertortzatton. The hepatic duct was stenotic but microscoptcally patent, wtth fibrous changes in the wall. Two weeks after surgery the serum bihrubm level was 4.2 mg/dL, and it was normal 2 weeks later. 40 months after surgery the pattent remains well wtthout evidence of liver disease.
Case 3 A baby girl requned laparotomy for meconium tleus wuh volvulus on the first day of life. A 15cm small bowel resectton with primary anastomosis was performed. Parenteral nutrition was required for 10 days and was followed by adequate feeding and discharge. Sweat test was positive for cysttc fibrosis. The infant was readmitted at 9 weeks of age for persistent jaundice (bilirubin, 5.3 mg/dL). Liver biopsy results
Fig 1. replacement
Serial
sections of the common by small ductules and fibrous
hepatic duct connective
in Case 3 demonstrating tissue (B).
ET AL
showed paucity of btle ducts, moderate fibrosis, and minimal inflammation Results of metabolic and infectious work-ups were negative. Liver scan showed no excretton at 24 hours. and the gallbladder and bile ducts were not identtfied by ultrasound. Conservative treatment with ursodeoxychohc acid was started. but after 3 weeks a repeat liver scan still showed no excretton. Thus, at 12 weeks of age the patient underwent surgical exploration. Gross findings were consistent with biliary atresia (fibrottc gallbladder and ducts), and Inability to identify a gallbladder lumen precluded cholangiography. Histological examination of the extrahepatic bihary tree followmg Kasai portoenterostomy showed mtcroscopic patency of the proximal 1 cm of the common hepatic duct (lumen, 0.5 mm). The bile duct then disappeared distally, and was replaced by numerous small ductules associated with prohferatmg fibrous connective tissue (Fig 11. Imtial bile flow was meager (
DISCUSSION
Hepatobiliary disease in cystic fibrosis is relatively common, increasesin incidence with age, and is often asymptomatic. Liver abnormalities will occur in 20% to 50% of children with cystic fibrosis with clinically apparent disease in 2% to 18%.’ The classic liver histology is that of focal biliary cirrhosis and consistsof inspissatedgranular, eosinophilic material within portal
proximal
microscopic
patency
(A) and
more
distal
atresia
with
BILIARY
OBSTRUCTION
IN CYSTIC
FIBROSIS
bile ductules, bile duct proliferation. inflammation, and periportal fibrosis.‘,9 These changes have been observed at autopsy in 11% of patients under 3 months of age, in 27% at 1 year of age, and in 12% at 24 years9 Extrahepatic biliary changes are also prevalent. Microgallbladders are seen in up to 30% of patients at autopsy and are radiologically detectable in 16% of children and adolescents.h,‘n Cystic duct atresia has also been frequently described.b Associated structural abnormalities extending to the hepatic and common bile ducts are rare. but include distal and diffuse stenosis, fibrosis, mucosal loss,“-I5 and frank atresia.JJ The pathophysiology remains conjectural. although several components of the disease have been clarified. Cystic fibrosis is caused by mutations involving the cystic fibrosis transmembrane conductance regulator (CFTR). Recent localization of this dysfunctional chloride channel to the intrahepatic biliary epithelium identifies the intrahepatic bile duct as the site of the primary defect for hepatobiliary complications.‘J.‘6 The subsequent decrease in chloride efflux reduces the secondary movement of sodium and water, leading to a decrease in bile volume, bile stasis. and an increase in bile acid concentration. Concomitant impaired secretion of mucins alters the protein composition of these fluids and disrupts the protective mucous layer coating the ducts. Further insult may be added by the fecal loss of taurine conjugated bile acids, leading to an increase in percentage of glycoconjugates. which are more hydrophobic and potentially more hepatotoxic. 9 ‘I “.Ix Recent evidence of an increase in certain HLA antigens in patients with liver disease and cystic fibrosis suggests that an autoimmune response may also contribute to the pathophysiology.“,‘The subsequent clinical manifestations of hepatobiliary disease in cystic fibrosis are quite variable.lJ This is presumably the result of differences in the contribut’on of each pathological factor. the relative severity of each, the timing of onset, and the impact of other, yet to be identified. components. Prolonged jaundice in the newborn with cystic fibrosis is a relatively rare result of this pathophysiological process. In addition to our three patients, we identified another 21 affected infants through a literature review. The early appearance of clinically significant symptoms in conjunction with advanced liver histological changes implies in some cases. onset in utero.“,‘(’ Forty-two percent of the overall group also had meconium ileus at birth, indicative of relatively severe, simultaneous accumulation of viscus secretions in both the biliary and intestinal tract, again in the antenatal period.‘” Yet despite similar points of onset, several distinct patterns of disease emerge. The first variable is the response of the infantile liver to cholestasis. Although the typical picture of bile duct
177
proliferation and inflammation is well described, there is an equal incidence of bile duct paucity and fibrosis (two of our three patients and three of seven patients with liver biopsies in the reviewed group). This may simply represent one stage, either early or late, in a continuum of changes, or may be indicative of suspension of bile duct differentiation and growth caused by an early antenatal insult with pericholangiolytic fibrosis during bile duct development.8 Regardless, limited numbers preclude attaching prognostic significance to this feature. In addition to these two histological patterns, further review of gross and microscopic findings, in conjunction with response to treatment. has suggested three clinical patterns of disease. The first consists of intra- and extrahepatic accumulation of inspissated bile within normal bile ducts and gallbladder. This is presumably indicative of less severe intrahepatic cholestasis. Ultrasound may show dilated ducts, and clearance with medical therapy can be expected. Persistence of jaundice warrants biliary irrigation, with optimism for resolution. The second pattern consists of more severe intrahepatic cholestasis, with small but otherwise normal bile ducts and perhaps a microgallbladder and atretic cystic duct. Again, conservative treatment or irrigation can result in resolution. The third pattern is severe intrahepatic cholestasis associated with extension of abnormalities to the hepatic and common bile duct, where marked stenosis, fibrosis. cystic changes, and atresia may be seen. These changes are presumably secondary to disuse atrophy and direct cytotoxic effect.“i5 Conservative therapy can be expected to lead to eventual fatal liver disease. All three of our patients belonged to this last category. All had microgallbladders without an apparent lumen (Fig 2), cystic duct atresia, and structural changes in the hepatic and common bile duct. Case 1 had only microscopic patency of the hepatic duct with significant fibrosis. Severe intrahepatic cholestasis did not resolve after reconstruction. Case 3 had proximal microscopic patency and distal atresia. Reconstruction led to resolution ofjaundice and reversal of liver histological changes. Case 2 had microscopic patency seen on initial cholangiogram, but the patient did not respond to irrigation. The markedly narrowed and fibrotic bile duct excised at reconstruction was presumably not sufficient to allow resolution of chronic intrahepatic cholestasis. Prompt improvement was noted after reconstruction provided a larger drainage conduit. It is notable that age of successful reconstruction was 12 and 16 weeks, respectively, well beyond that deemed necessary for success in isolated biliary atresia. Furthermore. a normal liver biopsy result 30 months later (Case 3) implies that provision of an adequate extrahepatic conduit may reverse cholestasis to the point of avoidance of liver toxicity, and suggests a potential contributing role for occult extrahepatic biliary
GREENHOLZ
178
Fig 2. complex wall.
Typical lumen
histology and notable
of a microgallbladder, fibrotic and cystic
with changes
an irregular within the
stenosis in the later development of focal biliary cirrhosis in untreated patients. Although controversial, documentation of ductal disease in older patients with liver disease has led to a similar conclusion.9~1z-14 Of the 21 patients within the review, nine had spontaneous resolution of cholestasis between 2 and 8 months of age with medical therapy only. I .4,7.’ X.20-24 Recommendations for a conservative management approach are based on several of these references. However, five patients were identified with common bile duct obstructive abnor-
ET AL
malities, two with atresia,4s5 and three with severe stenosis.5,7J The two patients with atresia underwent portoenterostomy. One operation was successful, the other patient died of postoperative complications. Failure to identify the intrinsic ductal obstruction in the three patients with severe stenosis resulted in eventual liver failure and death. Of the remaining seven patients, three have responded to surgical biliary irrigation,2*7,20 and four have died of respiratory complications with persistent jaundice and apparently normal bile ducts (although two had microgallbladders and cystic duct atresia).25 Thus, of the combined group, 33% (8 of 24) had anatomic justification for reconstruction. This experience leads to the following treatment recommendations. Persistent jaundice in the neonate requires evaluation for cystic fibrosis. If confirmed, then a trial of expectant medical therapy with choleretics is warranted. Although the urgency for diagnosis and treatment does not appear to be as extreme as in biliary atresia, lack of resolution by 2 to 3 months of age should prompt further evaluation including biliary excretion scan, abdominal ultrasound, and liver biopsy. Lack of excretion at 24 hours, a shrunken gallbladder without dilated ducts, and altered liver histology should suggest the possibility of ductal disease. The goal of subsequent exploration should be to document ductal patency (if technically feasible, ERCP would also be appropriate). A normal gallbladder and biliary tree should be irrigated to confirm flow into the duodenum and wash out inspissated bile. A microgallbladder and atretic cystic duct, although worrisome, do not necessarily imply common duct abnormality, and an attempt should be made to directly cannulate the common duct with a fine-gauge needle. Patency, even if small, allows further expectant management after irrigation. Nonpatency, or failure of irrigation after an additional 1 to 2 months, warrants reconstruction. Cholestasis and progressive liver failure unresponsive to reconstruction can be effectively treated with liver transplantation, with an overall outcome comparable to other chronic liver diseases.3
REFERENCES 1. Furuya RN, Roberts EA, Canny GJ, et al: Neonatal hepatitts syndrome with paucity of interlobular btle ducts m cystic fibrosis. J Pediatr Gastroenterol Nutr 12: 127-130, 1991 2. Evans JS, George DE, Mollit D: Biliary infusion therapy in the inspissated bile syndrome of cystic fibrosis. J Pediatr Gastroenterol Nutr 12:131-135, 1991 3. Colombo C, Battezzati PM, Podda M: Hepatobiliary disease in cysttc fibrosis. Semin Liver Dis 14:259-269, 1994 4. Perkins WG. Klein GL, Beckerman RC: Cysttc fibrosis mistaken for idtopathic biliary atresia. Clin Pediatr 24.107-109, 1985 5. Schwarz HP, Kraemer R, Thumheer U, et al: Liver mvolvement in cystic fibrosis. Helv Pediatr Acta 33:35 l-364, 1978 6. Roy CC, Weber AM, Morin CL, et al: Hepatobiliary disease in
cystic fibrosis: A survey of current issues and concepts. J Pedtatr Gastroenterol Nutr 1:469-478, 1982 7. Valman HB, France NE, Wallis PG: Prolonged neonatal Jaundice in cysttc fibrosis. Arch Dis Child 46:805-809, 1971 8. Shier KJ. Horn RC: The pathology of liver cnrhosis m patients with cystic fibrosis of the pancreas. Can Med J 89:645-651, 1963 9. Wtlliams SGJ, Westaby D, Tanner MS, et al. Liver and bihary problems m cystic fibrosis. Br Med Bull 48:877-892, 1992 10. Isenberg JN: Cystic fibrosis: It’s influence on the hver, biliary tree, and bile salt metabolism. Semm Liver Dts 2:302-313, 1982 11. Vitullo BB, Rochon L, Seemayer TA. et al: Intrapancreatic compresston of the common bile in cystic fibrosis. J Pedtatr 93: 10601061, 1978
BILIARY
OBSTRUCTION
IN CYSTIC
FIBROSIS
179
12 Gasken KJ, Waters DLM, Howman-Giles R. et al: Liver disease and common-bile-duct stenosis m cystic fibrosts. N Engl J Med 3 18 340-346, 1988 13. Bilton D, Fox R. Webb AK, et al: Pathology of common bile duct stenosrs m cystic fibroats. Gut 3 I :236-238, 1990 14. Tanner MS, Taylor CJ: Lrver disease m cystrc fibrosts. Arch DIS Chtld 72:281-283. 1995 15. Lmdblad A. Hultcrantz R, Strandvik B: Bile duct destruction and collagen deposrtion, Prominant ultrastructural feature of the liver m cystic fibrosts. Hepatology 16:372-38 1, 1992 16. Cohn JA. Strong TV. Prccrotto MR. et al: Localizatron of the cystrc iibrosts transmembrane conductance regulator in human brie duct eptthehal cells Gastroenterology 10.5: 1857-1864. 1993 17. Galabert C. Montet JC. Lengrand D. et al, Effects of ursodeoxycholic acrd on liver function m pattents wtth cystrc fibrosis and chrome cholestasis. J Pediatr 121 138-141, lY92 18 Tanner MS, Ltver and brhary problems m cystrc fibrosrs J R Sot Med 85 (suppl 19),20-24. 1992 19. Colombo C. Apostolo MG. Ferran M. et al, Analysts of nsk
factors for the development of liver drsease associated wrth cysttc fibrosis. J Pediatr 124:393-399. 1994 20. Dogan AS, Conway JJ, Lloyd-Sol1 JD Hepatobrhary scmtigraphy m children with cystic fibrosis and liver disease J Nucl Med 35.432-435. 1994 21. Rosenstein BJ, Oppenhermer EH: Prolonged obstructive Jaundice and giant cell hepatitis m an infant with cysttc fibrosis. J Pedtatr 91.1022-1023. 1977 22. Tulamo RC, Hendren WH: Prolonged obstructrve Jaundice: Report of a case m a neonate wrth mecomum ileus and JeJunal atresra. Am J Dis Child 115:74-79, 1968 23. Bemstem J. Vawter G. Harrts GBC: The occurrence of intestmal atresia m newborns with meconium tleus. Am J DIS Child 99:804-818. 1960 24 Taylor WF, Qaqundah BY: Neonatal Jaundtce associated wtth cystic fibrosis. Am J Dis Chrld 123:161-162, 1972 25 Gatzrmos CD. Jowitt RH. Jaundtce m mucovtscidosia (fibrocystic disease of pancreas). Report of four cases. Am J DIS Child 89:182-186. 1955
Discussion R.P. Altman (New York, NY): Were your two meconium ileus patients on long-term hyperalimentation? If so, were these manifestations of hyperalimentation cholestasis or manifestations of cystic fibrosis? I’ve not heard of biliary atresialike syndromes in patients with cystic fibrosis other than patients like two of yours with meconium ileus, intestinal obstructions, hyperalimentation. S.K. Greenholz (response): Meconium ileus is associated with prolonged jaundice in cystic fibrosis and it is presumably secondary to diffuse biliary and intestinal inspissation of secretions. beginning in the antenatal period. Of the entire group of infants with prolonged jaundice, of which there were 24 total, 42% also had meconium ileus. Many were not treated with hyperalimentation. Although there is the problem of the addition of hyperalimentation in members of that group, there certainly is a significant number that did not have this as a factor. Also if you look at the population of cystic fibrosis patients in general, ductal abnormalities are actually quite common, with 30% having microgallbladders. In infants and children, up to 16% have small gallbladders by radiological examination, and this is in the absence of meconium obstruction. Thus, the association of jaundice with meconium obstruction seems more likely to be related to simultaneous antenatal pathology than hyperalimentation. S.P Dunn (Philadelphia, PA): One of the problems I have with your conclusion is that you are now going to commit a number of the very cholestatic, very young, meconium ileus patients to operative exploration to determine if they have a normal extrahepatic biliary tree. In your literature review and in my experience, a number of these patients who are initially quite jaundiced from
prolonged hyperalimentation and not being fed, will clear their jaundice spontaneously proving that a significant portion of them did not need exploration. I would appreciate it if you could help us better understand who needs exploration and who does not. Quite a few of those kids have had very significant abdominal procedures in the past, and it is not going to be easy to get in there and prove that they have a patent extrahepatic biliary tree. S.K. Greenholz (response): I agree that is a problem. It is difficult to identify which of those patients will clear spontaneously and which will not. Of that group of 24 patients, nine patients cleared spontaneously. It is that relatively small group that has formed the basis for a general impression that these cases clear spontaneously. However, given the significant incidence of anatomic ductal abnormalities, I believe that the infant with cystic fibrosis and jaundice extending beyond 8 weeks of age requires evaluation. No excretion on scan. failure to respond to choleretics, shrunken gallbladder on ultrasound, and supportive liver biopsy results should lead to an attempt to document ductal patency. The issue is to perform this in the least morbid and most accurate way. Most recently we have successfully confirmed patency in a 2 month old with ERCP, and this will be our preferred approach in the future as our technical abilities improve. During a limited open exploration I am most concerned about mistaking gallbladder atresia alone for diffuse ductal atresia. However, I believe careful direct cannulation of the common duct (we use a 27-gauge needle) can differentiate between these entities, and the entire procedure can be perfomred with minimal morbidity. 7: Weber (St Louis, MO): Can you give us an idea of the incidence of this disorder? What is your total popula-
180
tion of cystic fibrosis patients from which these three patients were drawn? SK. Greenholz (response): I can’t give you that number. I can tell you that the time interval from which
GREENHOLZ
ET AL
these three patients were drawn was 5 years. We actually had a fourth patient who did not undergo reconstruction but whose duct was successfully cannulated, irrigated, and then went on to resolution.
Obstruction By Stephen
in Infants With Cystic Fibrosis Kasai Portoenterostomy
K. Greenholz,
Baiya Krishnadasan, Clifford Sacramento, California
0 Cystic fibrosis is associated with an inspissated bile syndrome producing cholestasis secondary to plugging of macroscopically normal bile ducts. In extreme neonatal forms, with early profound intrahepatic cholestasis, the process can be associated with a marked decrease in ductal diameter, varying from hypoplasia to atresia. From 1990 to 1995 three infants were identified with cystic fibrosis, persistent jaundice, and complete absence of biliary excretion despite expectant and conservative treatment including choleretics and surgical biliary irrigation. Abdominal ultrasounds showed contracted gallbladders and no evidence of dilated ducts. Liver biopsy results in two infants showed portal fibrosis, paucity of bile ducts, and minimal inflammation. The third infant had moderate inflammation, bile duct replication, and plugging. Two infants had undergone intestinal resection followed by hyperalimenation for complications of meconium ileus in the newborn period. Surgical exploration was undertaken at 7 to 12 weeks of age. Gross findings were typical of biliary atresia with microgallbladders and nonpatency of the cystic duct. Cholangiograms failed to document ductal patency in two patients who were then treated with a Kasai portoenterostomy. The third infant had patent hypoplastic ducts and underwent only biliary irrigation. Although bile flow was transiently achieved, jaundice recurred, and at reexploration at 16 weeks of age a Kasai poroenterostomy was performed. Histological review of the biliary specimens showed microscopically patent ducts in two patients and proximal patency and distal atresia in the third. All the ducts had mural fibrosis with cystic changes. Bile drainage was achieved in each instance, although in one patient with hypoplastic ducts scant output of highly concentrated bile proved insufficient to arrest progressive liver failure. The subsequent two patients responded with resolution of hyperbilirubinemia and normalization of liver function. They remain free of biliary complications at 30 and 40 months postoperatively. This manifestation of cystic fibrosis in infants is suggested by prolonged jaundice unresponsive to choleretics, nondilated bile ducts and gallbladder on ultrasound, absent biliary excretion on nuclear scan, and characteristic liver biopsy. Exploration is warranted, and discovery of atrophic bile ducts may be best managed with reconstruction. Copyright o 1997 by W.6. Saunders Company
Marr, and Robert
Requiring Cannon
abnormalities, including hypoplasia and atresia,4.sand subsequentprogressive fatal liver disease.2,6-s To illustrate these issues, we report three infants with cystic fibrosis who eventually underwent Kasai portoenterostomy as treatment for persistentjaundice after failing to respondto conservative treatment. A comparative literature review is included. CASE REPORTS Case 1 A boy was born at 35 weeks’ gestation and required abdominal exploratton on the 11 th day of life for mecomum peritonitis wtth persistent obstructton. Ileal atresia with perforation was treated with resection and primary anastomosis. Feeds were started at age 21 days after 3 weeks of parenteral nutrition. Perststent jaundice (bihrubm, 8.5 mg/dL) was noted over the next 4 weeks, and the patient did not respond to oral cholerettcs (ursodeoxycholic acid and phenobarbital). Metabolic and mfectious work-ups were negative, and sweat chloride test confirmed the dtagnosts of cystic fibrosis. Liver scan showed absence of hepattc excretion at 24 hours. and abdommal ultrasound could not tdenttfy a gallbladder or dilated ducts. Percutaneous liver biopsy results showed portal zones with moderate fibrosts. btle duct replication. moderate inflammation. and bile plugging. Surgical exploratton was undertaken at 7 weeks of age Gross findings were similar to biliary atresia. wtth a small shrunken gallbladder contammg clear mucous and a very small tibrottc hepattc duct without an apparent lumen, Cholangiography vta the gallbladder failed to confirm ductal patency. A Kasai portoenterostomy with extertortzatton was performed. Transection of the hepatic duct at the liver htlum showed microscoptc patency of the left hepatic duct and obhteration of the right hepatic duct. Histological examination of these ducts showed architectural alteration with fibrosis. a markedly narrowed lumen (0.2 mm or less) and cysttc changes within the wall. Bile flow was achieved, although of relatively low volume (0 to 20 cm’/d) and very high concentratron (30 to 50 mg/dL). Multiple steroid boluses only transiently Improved output. Serum bilirubin gradually decreased to 5.2 mg/dL by 6 weeks after operation, but then began to increase. Bile output decreased to 1 to 5 cm3/d and did not change with ,iddittonal steroids. Serum bilirubm level Increased to 36 mg/dL 5 months after operation, and the parents declined consideration of hcer transplant. The infant died shortly thereafter.
Case 2 INDEX WORDS: Cystic fibrosis, biliary neonatal jaundice, inspissated bile.
atresia,
cholestasis,
P
ROLONGED NEONATAL jaundice in conjunction with a diagnosis of cystic fibrosis is generally presumed to be secondary to plugging of normal intraand extrahepatic bile ducts by abnormally thickened, highly viscus, biliary secretions.Relatively limited anecdotal experience has implied that conservative treatment will be followed by eventual spontaneousresolution-” However, there are equally compelling reports of intrahepatic cholestasis associated with extrahepatic biliary JournalofPediatncSurgery,
Vol32,No
2 (February),1997.pp
175-180
A boy was admitted to the hospital at 5 weeks of age for jaundice (serum bihrubm. 9.3 mg/dL) noted for 3 weeks Evaluation for mfecttous and metabolic causes was negative. but a sweat test dtag-
Presetlted ot the .SII~~ICY/I A ~socut~oti.
27th Amtnnl Surr Dte,qo.
Meettn,q Cul/jtirwc~.
qf the Amermn Pediatric Mu?. 20-23, 1996.
Address rqmnt requests to Stephrrr K Greerd~ol;. Pedtutrrc Surgery, 4301 X St, Room 2310, Sncrumento. C’oprr~ht o 1997 by W B Saunders Company
MD, Dwsron CA 95817.
oj
002-‘-3368/97/3202-0006$03.00/O 175
GREENHOLZ
176
nosed cystic fibrosis. Ultrasound could not demonstrate a gallbladder or bile ducts, and a liver scan showed no excretton. Ltver biopsy results showed pauctty of bile ducts, moderate portal tibrosis, and minimal inflammation. Persistence of jaundice, desptte oral cholerettcs, led to exploration at 8 weeks of age. A small fibrotic gallbladder had no visible lumen. but the common btle duct, although small. could be directly cannulated with a 27-gauge needle. Cholangiogram showed contrast withm the duodenum, although the ducts themselves were not clearly seen. Biliary irrigation with acetylcysteme and salme was performed. Histological examination of the gallbladder and cystic duct showed absence of a true lumen, and fibrous and cystic changes withm the wall. Durmg the first 10 days after surgery, serum bilirubm levels decreased from 9.3 mg/dL to 3.5 mg/dL, but then began to rise, reaching 9.4 mg/dL 1 week later. Two steroid pulses were ineffectual. It was decided to treat the patient conservatively with oral choleretics; however, Jaundice persisted, and repeat liver scan at 4 months of age showed no excretion The child was reexplored and underwent Kasat portoenterostomy without extertortzatton. The hepatic duct was stenotic but microscoptcally patent, wtth fibrous changes in the wall. Two weeks after surgery the serum bihrubm level was 4.2 mg/dL, and it was normal 2 weeks later. 40 months after surgery the pattent remains well wtthout evidence of liver disease.
Case 3 A baby girl requned laparotomy for meconium tleus wuh volvulus on the first day of life. A 15cm small bowel resectton with primary anastomosis was performed. Parenteral nutrition was required for 10 days and was followed by adequate feeding and discharge. Sweat test was positive for cysttc fibrosis. The infant was readmitted at 9 weeks of age for persistent jaundice (bilirubin, 5.3 mg/dL). Liver biopsy results
Fig 1. replacement
Serial
sections of the common by small ductules and fibrous
hepatic duct connective
in Case 3 demonstrating tissue (B).
ET AL
showed paucity of btle ducts, moderate fibrosis, and minimal inflammation Results of metabolic and infectious work-ups were negative. Liver scan showed no excretton at 24 hours. and the gallbladder and bile ducts were not identtfied by ultrasound. Conservative treatment with ursodeoxychohc acid was started. but after 3 weeks a repeat liver scan still showed no excretton. Thus, at 12 weeks of age the patient underwent surgical exploration. Gross findings were consistent with biliary atresia (fibrottc gallbladder and ducts), and Inability to identify a gallbladder lumen precluded cholangiography. Histological examination of the extrahepatic bihary tree followmg Kasai portoenterostomy showed mtcroscopic patency of the proximal 1 cm of the common hepatic duct (lumen, 0.5 mm). The bile duct then disappeared distally, and was replaced by numerous small ductules associated with prohferatmg fibrous connective tissue (Fig 11. Imtial bile flow was meager (
DISCUSSION
Hepatobiliary disease in cystic fibrosis is relatively common, increasesin incidence with age, and is often asymptomatic. Liver abnormalities will occur in 20% to 50% of children with cystic fibrosis with clinically apparent disease in 2% to 18%.’ The classic liver histology is that of focal biliary cirrhosis and consistsof inspissatedgranular, eosinophilic material within portal
proximal
microscopic
patency
(A) and
more
distal
atresia
with
BILIARY
OBSTRUCTION
IN CYSTIC
FIBROSIS
bile ductules, bile duct proliferation. inflammation, and periportal fibrosis.‘,9 These changes have been observed at autopsy in 11% of patients under 3 months of age, in 27% at 1 year of age, and in 12% at 24 years9 Extrahepatic biliary changes are also prevalent. Microgallbladders are seen in up to 30% of patients at autopsy and are radiologically detectable in 16% of children and adolescents.h,‘n Cystic duct atresia has also been frequently described.b Associated structural abnormalities extending to the hepatic and common bile ducts are rare. but include distal and diffuse stenosis, fibrosis, mucosal loss,“-I5 and frank atresia.JJ The pathophysiology remains conjectural. although several components of the disease have been clarified. Cystic fibrosis is caused by mutations involving the cystic fibrosis transmembrane conductance regulator (CFTR). Recent localization of this dysfunctional chloride channel to the intrahepatic biliary epithelium identifies the intrahepatic bile duct as the site of the primary defect for hepatobiliary complications.‘J.‘6 The subsequent decrease in chloride efflux reduces the secondary movement of sodium and water, leading to a decrease in bile volume, bile stasis. and an increase in bile acid concentration. Concomitant impaired secretion of mucins alters the protein composition of these fluids and disrupts the protective mucous layer coating the ducts. Further insult may be added by the fecal loss of taurine conjugated bile acids, leading to an increase in percentage of glycoconjugates. which are more hydrophobic and potentially more hepatotoxic. 9 ‘I “.Ix Recent evidence of an increase in certain HLA antigens in patients with liver disease and cystic fibrosis suggests that an autoimmune response may also contribute to the pathophysiology.“,‘The subsequent clinical manifestations of hepatobiliary disease in cystic fibrosis are quite variable.lJ This is presumably the result of differences in the contribut’on of each pathological factor. the relative severity of each, the timing of onset, and the impact of other, yet to be identified. components. Prolonged jaundice in the newborn with cystic fibrosis is a relatively rare result of this pathophysiological process. In addition to our three patients, we identified another 21 affected infants through a literature review. The early appearance of clinically significant symptoms in conjunction with advanced liver histological changes implies in some cases. onset in utero.“,‘(’ Forty-two percent of the overall group also had meconium ileus at birth, indicative of relatively severe, simultaneous accumulation of viscus secretions in both the biliary and intestinal tract, again in the antenatal period.‘” Yet despite similar points of onset, several distinct patterns of disease emerge. The first variable is the response of the infantile liver to cholestasis. Although the typical picture of bile duct
177
proliferation and inflammation is well described, there is an equal incidence of bile duct paucity and fibrosis (two of our three patients and three of seven patients with liver biopsies in the reviewed group). This may simply represent one stage, either early or late, in a continuum of changes, or may be indicative of suspension of bile duct differentiation and growth caused by an early antenatal insult with pericholangiolytic fibrosis during bile duct development.8 Regardless, limited numbers preclude attaching prognostic significance to this feature. In addition to these two histological patterns, further review of gross and microscopic findings, in conjunction with response to treatment. has suggested three clinical patterns of disease. The first consists of intra- and extrahepatic accumulation of inspissated bile within normal bile ducts and gallbladder. This is presumably indicative of less severe intrahepatic cholestasis. Ultrasound may show dilated ducts, and clearance with medical therapy can be expected. Persistence of jaundice warrants biliary irrigation, with optimism for resolution. The second pattern consists of more severe intrahepatic cholestasis, with small but otherwise normal bile ducts and perhaps a microgallbladder and atretic cystic duct. Again, conservative treatment or irrigation can result in resolution. The third pattern is severe intrahepatic cholestasis associated with extension of abnormalities to the hepatic and common bile duct, where marked stenosis, fibrosis. cystic changes, and atresia may be seen. These changes are presumably secondary to disuse atrophy and direct cytotoxic effect.“i5 Conservative therapy can be expected to lead to eventual fatal liver disease. All three of our patients belonged to this last category. All had microgallbladders without an apparent lumen (Fig 2), cystic duct atresia, and structural changes in the hepatic and common bile duct. Case 1 had only microscopic patency of the hepatic duct with significant fibrosis. Severe intrahepatic cholestasis did not resolve after reconstruction. Case 3 had proximal microscopic patency and distal atresia. Reconstruction led to resolution ofjaundice and reversal of liver histological changes. Case 2 had microscopic patency seen on initial cholangiogram, but the patient did not respond to irrigation. The markedly narrowed and fibrotic bile duct excised at reconstruction was presumably not sufficient to allow resolution of chronic intrahepatic cholestasis. Prompt improvement was noted after reconstruction provided a larger drainage conduit. It is notable that age of successful reconstruction was 12 and 16 weeks, respectively, well beyond that deemed necessary for success in isolated biliary atresia. Furthermore. a normal liver biopsy result 30 months later (Case 3) implies that provision of an adequate extrahepatic conduit may reverse cholestasis to the point of avoidance of liver toxicity, and suggests a potential contributing role for occult extrahepatic biliary
GREENHOLZ
178
Fig 2. complex wall.
Typical lumen
histology and notable
of a microgallbladder, fibrotic and cystic
with changes
an irregular within the
stenosis in the later development of focal biliary cirrhosis in untreated patients. Although controversial, documentation of ductal disease in older patients with liver disease has led to a similar conclusion.9~1z-14 Of the 21 patients within the review, nine had spontaneous resolution of cholestasis between 2 and 8 months of age with medical therapy only. I .4,7.’ X.20-24 Recommendations for a conservative management approach are based on several of these references. However, five patients were identified with common bile duct obstructive abnor-
ET AL
malities, two with atresia,4s5 and three with severe stenosis.5,7J The two patients with atresia underwent portoenterostomy. One operation was successful, the other patient died of postoperative complications. Failure to identify the intrinsic ductal obstruction in the three patients with severe stenosis resulted in eventual liver failure and death. Of the remaining seven patients, three have responded to surgical biliary irrigation,2*7,20 and four have died of respiratory complications with persistent jaundice and apparently normal bile ducts (although two had microgallbladders and cystic duct atresia).25 Thus, of the combined group, 33% (8 of 24) had anatomic justification for reconstruction. This experience leads to the following treatment recommendations. Persistent jaundice in the neonate requires evaluation for cystic fibrosis. If confirmed, then a trial of expectant medical therapy with choleretics is warranted. Although the urgency for diagnosis and treatment does not appear to be as extreme as in biliary atresia, lack of resolution by 2 to 3 months of age should prompt further evaluation including biliary excretion scan, abdominal ultrasound, and liver biopsy. Lack of excretion at 24 hours, a shrunken gallbladder without dilated ducts, and altered liver histology should suggest the possibility of ductal disease. The goal of subsequent exploration should be to document ductal patency (if technically feasible, ERCP would also be appropriate). A normal gallbladder and biliary tree should be irrigated to confirm flow into the duodenum and wash out inspissated bile. A microgallbladder and atretic cystic duct, although worrisome, do not necessarily imply common duct abnormality, and an attempt should be made to directly cannulate the common duct with a fine-gauge needle. Patency, even if small, allows further expectant management after irrigation. Nonpatency, or failure of irrigation after an additional 1 to 2 months, warrants reconstruction. Cholestasis and progressive liver failure unresponsive to reconstruction can be effectively treated with liver transplantation, with an overall outcome comparable to other chronic liver diseases.3
REFERENCES 1. Furuya RN, Roberts EA, Canny GJ, et al: Neonatal hepatitts syndrome with paucity of interlobular btle ducts m cystic fibrosis. J Pediatr Gastroenterol Nutr 12: 127-130, 1991 2. Evans JS, George DE, Mollit D: Biliary infusion therapy in the inspissated bile syndrome of cystic fibrosis. J Pediatr Gastroenterol Nutr 12:131-135, 1991 3. Colombo C, Battezzati PM, Podda M: Hepatobiliary disease in cysttc fibrosis. Semin Liver Dis 14:259-269, 1994 4. Perkins WG. Klein GL, Beckerman RC: Cysttc fibrosis mistaken for idtopathic biliary atresia. Clin Pediatr 24.107-109, 1985 5. Schwarz HP, Kraemer R, Thumheer U, et al: Liver mvolvement in cystic fibrosis. Helv Pediatr Acta 33:35 l-364, 1978 6. Roy CC, Weber AM, Morin CL, et al: Hepatobiliary disease in
cystic fibrosis: A survey of current issues and concepts. J Pedtatr Gastroenterol Nutr 1:469-478, 1982 7. Valman HB, France NE, Wallis PG: Prolonged neonatal Jaundice in cysttc fibrosis. Arch Dis Child 46:805-809, 1971 8. Shier KJ. Horn RC: The pathology of liver cnrhosis m patients with cystic fibrosis of the pancreas. Can Med J 89:645-651, 1963 9. Wtlliams SGJ, Westaby D, Tanner MS, et al. Liver and bihary problems m cystic fibrosis. Br Med Bull 48:877-892, 1992 10. Isenberg JN: Cystic fibrosis: It’s influence on the hver, biliary tree, and bile salt metabolism. Semm Liver Dts 2:302-313, 1982 11. Vitullo BB, Rochon L, Seemayer TA. et al: Intrapancreatic compresston of the common bile in cystic fibrosis. J Pedtatr 93: 10601061, 1978
BILIARY
OBSTRUCTION
IN CYSTIC
FIBROSIS
179
12 Gasken KJ, Waters DLM, Howman-Giles R. et al: Liver disease and common-bile-duct stenosis m cystic fibrosts. N Engl J Med 3 18 340-346, 1988 13. Bilton D, Fox R. Webb AK, et al: Pathology of common bile duct stenosrs m cystic fibroats. Gut 3 I :236-238, 1990 14. Tanner MS, Taylor CJ: Lrver disease m cystrc fibrosts. Arch DIS Chtld 72:281-283. 1995 15. Lmdblad A. Hultcrantz R, Strandvik B: Bile duct destruction and collagen deposrtion, Prominant ultrastructural feature of the liver m cystic fibrosts. Hepatology 16:372-38 1, 1992 16. Cohn JA. Strong TV. Prccrotto MR. et al: Localizatron of the cystrc iibrosts transmembrane conductance regulator in human brie duct eptthehal cells Gastroenterology 10.5: 1857-1864. 1993 17. Galabert C. Montet JC. Lengrand D. et al, Effects of ursodeoxycholic acrd on liver function m pattents wtth cystrc fibrosis and chrome cholestasis. J Pediatr 121 138-141, lY92 18 Tanner MS, Ltver and brhary problems m cystrc fibrosrs J R Sot Med 85 (suppl 19),20-24. 1992 19. Colombo C. Apostolo MG. Ferran M. et al, Analysts of nsk
factors for the development of liver drsease associated wrth cysttc fibrosis. J Pediatr 124:393-399. 1994 20. Dogan AS, Conway JJ, Lloyd-Sol1 JD Hepatobrhary scmtigraphy m children with cystic fibrosis and liver disease J Nucl Med 35.432-435. 1994 21. Rosenstein BJ, Oppenhermer EH: Prolonged obstructive Jaundice and giant cell hepatitis m an infant with cysttc fibrosis. J Pedtatr 91.1022-1023. 1977 22. Tulamo RC, Hendren WH: Prolonged obstructrve Jaundice: Report of a case m a neonate wrth mecomum ileus and JeJunal atresra. Am J Dis Child 115:74-79, 1968 23. Bemstem J. Vawter G. Harrts GBC: The occurrence of intestmal atresia m newborns with meconium tleus. Am J DIS Child 99:804-818. 1960 24 Taylor WF, Qaqundah BY: Neonatal Jaundtce associated wtth cystic fibrosis. Am J Dis Chrld 123:161-162, 1972 25 Gatzrmos CD. Jowitt RH. Jaundtce m mucovtscidosia (fibrocystic disease of pancreas). Report of four cases. Am J DIS Child 89:182-186. 1955
Discussion R.P. Altman (New York, NY): Were your two meconium ileus patients on long-term hyperalimentation? If so, were these manifestations of hyperalimentation cholestasis or manifestations of cystic fibrosis? I’ve not heard of biliary atresialike syndromes in patients with cystic fibrosis other than patients like two of yours with meconium ileus, intestinal obstructions, hyperalimentation. S.K. Greenholz (response): Meconium ileus is associated with prolonged jaundice in cystic fibrosis and it is presumably secondary to diffuse biliary and intestinal inspissation of secretions. beginning in the antenatal period. Of the entire group of infants with prolonged jaundice, of which there were 24 total, 42% also had meconium ileus. Many were not treated with hyperalimentation. Although there is the problem of the addition of hyperalimentation in members of that group, there certainly is a significant number that did not have this as a factor. Also if you look at the population of cystic fibrosis patients in general, ductal abnormalities are actually quite common, with 30% having microgallbladders. In infants and children, up to 16% have small gallbladders by radiological examination, and this is in the absence of meconium obstruction. Thus, the association of jaundice with meconium obstruction seems more likely to be related to simultaneous antenatal pathology than hyperalimentation. S.P Dunn (Philadelphia, PA): One of the problems I have with your conclusion is that you are now going to commit a number of the very cholestatic, very young, meconium ileus patients to operative exploration to determine if they have a normal extrahepatic biliary tree. In your literature review and in my experience, a number of these patients who are initially quite jaundiced from
prolonged hyperalimentation and not being fed, will clear their jaundice spontaneously proving that a significant portion of them did not need exploration. I would appreciate it if you could help us better understand who needs exploration and who does not. Quite a few of those kids have had very significant abdominal procedures in the past, and it is not going to be easy to get in there and prove that they have a patent extrahepatic biliary tree. S.K. Greenholz (response): I agree that is a problem. It is difficult to identify which of those patients will clear spontaneously and which will not. Of that group of 24 patients, nine patients cleared spontaneously. It is that relatively small group that has formed the basis for a general impression that these cases clear spontaneously. However, given the significant incidence of anatomic ductal abnormalities, I believe that the infant with cystic fibrosis and jaundice extending beyond 8 weeks of age requires evaluation. No excretion on scan. failure to respond to choleretics, shrunken gallbladder on ultrasound, and supportive liver biopsy results should lead to an attempt to document ductal patency. The issue is to perform this in the least morbid and most accurate way. Most recently we have successfully confirmed patency in a 2 month old with ERCP, and this will be our preferred approach in the future as our technical abilities improve. During a limited open exploration I am most concerned about mistaking gallbladder atresia alone for diffuse ductal atresia. However, I believe careful direct cannulation of the common duct (we use a 27-gauge needle) can differentiate between these entities, and the entire procedure can be perfomred with minimal morbidity. 7: Weber (St Louis, MO): Can you give us an idea of the incidence of this disorder? What is your total popula-
180
tion of cystic fibrosis patients from which these three patients were drawn? SK. Greenholz (response): I can’t give you that number. I can tell you that the time interval from which
GREENHOLZ
ET AL
these three patients were drawn was 5 years. We actually had a fourth patient who did not undergo reconstruction but whose duct was successfully cannulated, irrigated, and then went on to resolution.