Narkewicz, Michael R.*; Sondheimer, Henry M.†; Ziegler, James W.†; Otanni, Yvonne†; Lorts, Angela†; Shaffer, Elizabeth M.†; Horgan, J. Gerard‡; Sokol, Ronald J.*
The Fontan procedure, a relatively high-risk palliative cardiac operation, was first performed in 1968 for tricuspid atresia (1). Since its original description, the procedure has undergone several modifications and has been applied to other complex congenital cardiac lesions distinguished by the presence of a single functional ventricle (2). The rationale for the Fontan procedure rests on the premise that systemic venous blood can pass through the lungs without the assistance of a pumping right ventricle. Thus, systemic venous return is separated from pulmonary venous return by the creation of a direct cavopulmonary or atriopulmonary anastomosis, leaving the single ventricle as the systemic pumping chamber. Before the Fontan procedure, the natural history for patients with a single ventricle was poor, with most patients dying during adolescence, usually from failure of the systemic ventricle caused by longstanding volume overload and arterial hypoxemia (3). The Fontan procedure is specifically designed to remedy these two problems. By separating the pulmonary from the systemic circulation, it eliminates intracardiac mixing and the resultant arterial desaturation, allowing the systemic ventricle to function at a normal volume load (4). In low-risk patients, the modified Fontan procedure can be performed with an operative mortality of less than 10% (4,5), and actuarial survival is estimated at approximately 60% 10 years after repair (6). Most patients experience significant symptomatic and hemodynamic improvement after the Fontan procedure; however, cardiac transplantation often becomes necessary as cardiac function eventually deteriorates.
One of the consequences of the Fontan procedure is a significantly increased systemic venous pressure. Immediately postoperatively, a high right-sided intracardiac pressure (>20 mm Hg) is associated with increased morbidity and mortality (7–10). However, the long-term effects of a chronically elevated systemic venous pressure [averaging 12–14 mm Hg at rest (5)] after the Fontan procedure are not known, particularly the effect of this systemic venous hypertension on the liver (2,5,6,11–14). Chronic congestive heart failure leads to alterations of the hepatic microvasculature, causing chronic centrilobular ischemia and chronic hypoxia (15). Because standard liver blood tests do not quantitatively assess hepatic functional reserve or completely reflect the severity of chronic hepatic injury or fibrosis, the current study was undertaken to better characterize hepatic functional outcome after the Fontan procedure.
The primary objective of this study was to analyze markers of hepatic function in children who had undergone the Fontan procedure. The secondary objective of this study was to relate hepatic function to cardiac function and to the duration of hepatic congestion.
PATIENTS AND METHODS
Ultrapure galactose was obtained from Sigma (St. Louis, MO, U.S.A.) under investigational new drug status (approved by the US Food and Drug Administration), and a 20% sterile, pyrogen-tested solution in sterile water was prepared in the hospital pharmacy by filtration (0.22 micron), as previously described (16). The institutional review board approved the study protocol, and informed written consent was obtained from the parent or guardian of each subject.
This was a cross-sectional study of hepatic and cardiac function in children who had previously undergone the Fontan procedure and were aged more than 3 years. Consecutive subjects were approached during their follow-up visits and offered entry into the study (11 of 25 accepted). The Fontan procedure was performed for the following congenital heart defects in these patients: tricuspid atresia (n = 7), single ventricle (n = 3), and mitral stenosis with hypoplastic left ventricle (n = 1). Patients were excluded if 1) they had a history of other liver disease before performance of the Fontan procedure; 2) the family or subject was unable to comply with the requirements of the study; or 3) they had a history of alcohol use or intravenous drug abuse.
Patients were admitted to the Pediatric General Clinical Research Center, and an intravenous catheter was placed for blood sampling. The following data were obtained.
1. Historical data, including initial cardiac lesion; age at operation; type of operation; current cardiac medications; interim occurrence of jaundice, right upper quadrant abdominal pain, hematemesis, or ascites; and late postoperative Fontan complications [thromboembolic event (17–19), arrhythmia (20), and protein-losing enteropathy (21–23)].
2. Physical findings: a complete physical examination was performed recording weight and height, resting vital signs, presence or absence of ascites, liver span and texture, spleen size, anthropometrics, and resting pulse oximetry. Two investigators (M.R.N. and R.J.S.) independently confirmed the liver and spleen examination. Clinical liver disease was designated as moderate–severe if clinical signs of portal hypertension were present or if liver biopsy showed bridging fibrosis or cirrhosis; otherwise, none–mild clinical liver disease was assigned.
3. Echocardiography was performed, and the following data were recorded: shortening fraction (when possible) and ejection fraction, Doppler inflow pattern of the systemic ventricle and flow patterns in the pulmonary artery and ascending aorta, evidence of atrioventricular (AV) valve dysfunction (stenosis or regurgitation), and evidence of residual atrial level communications. The diameter of the inferior vena cava was measured just below its junction with the right atrium. Dimensions were compared with established normal values (24). Echocardiograms were interpreted blinded to the patient's name, and diagnosis by a single coinvestigator (E.M.S.) to maintain consistency.
4. Blood tests included human immunodeficiency virus (HIV) antibody, hepatitis B surface antigen (HBsAg), anti-HBs, hepatitis C virus (HCV) antibody, complete blood count (CBC), blood urea nitrogen (BUN), creatinine, aspartate transaminase (AST), alanine transaminase (ALT), alkaline phosphatase, gamma-glutamyl transpeptidase (GGT), creatinine phosphokinase (CPK), total and direct bilirubin, cholylglycine, total protein, albumin, cholesterol, triglycerides, glucose, NH3, prothrombin time (PT), partial thromboplastin time (PTT), fibrinogen, and factors V, VII, and VIII.
5. A galactose elimination half-life and capacity were determined as previously described (16). Intravenous galactose was given as a 0.5 g/kg bolus dose of a sterile 20% solution, and blood samples were obtained for serum galactose concentration at 0, 20, 30, 40, 50, and 60 minutes. Plasma galactose concentration was measured in the Pediatric GCRC Core Laboratory (25). The galactose elimination capacity and elimination half-life were calculated by standard formulae (16,26). Galactose clearance from blood is primarily dependent on hepatic uptake of circulating galactose and subsequent metabolism in the hepatocyte. Extrahepatic (mainly renal) clearance can be accounted for by including a correction factor, and the resulting clearance of galactose represents the functional hepatic mass, which receives arterial or portal venous blood flow (16,26).
6. Doppler hepatic ultrasound with measurement of vessel diameter and flow patterns was performed by a single investigator (J.G.H.) using an Acuson 128 (Acuson, Mountain View, CA, U.S.A.) with pulse and color flow Doppler imaging. Three measurements of the main, right posterior, and left portal veins and the hepatic artery were performed, and mean and maximum velocity and direction of flow were recorded. Main and branch portal vein diameters were converted to Z-scores using previously published values (27). In other liver diseases, there is a correlation between a decrease in hepatic vascular velocities and the development of hepatic fibrosis (28). In addition, liver and spleen size, liver density, and the presence of ascites and varices (short gastric, left gastric, and paraumbilical) were noted.
Analysis of Data
Data from each patient were prospectively collected and recorded on individual data sheets. Comparisons were visually inspected using scattergrams with significant relationships sought between variables using linear regression analysis or by analysis of variance (ANOVA) or the Mann-Whitney test for nonparametric data. Data are presented as mean ± standard deviation. Statistical significance was taken at P ≤ 0.05.
Eleven children (4 boys, 7 girls) were evaluated at age 149 ± 51 months (range, 38–216 months) and at an interval of 100 ± 47 months (range, 9–176 months) after the Fontan procedure (Table 1). Hepatomegaly was present in six patients. Liver span correlated with time after Fontan procedure (R2 = 0.442, P < 0.03) but not with the age at time of the Fontan procedure. Four patients underwent a liver biopsy for clinical indications. Patients 8 and 10 had cirrhosis; patient 6 had extensive bridging fibrosis; and patient 11 had minimal pericentral and periportal fibrosis. Representative liver biopsies from patients 6 and 8 are shown in Figure 1. Four patients had moderate to severe liver disease (three by histopathology criteria; one with splenomegaly).
The echocardiographic findings of the 10 patients for whom data were available (Table 2) demonstrated a low shortening fraction in 8 patients (18–30%). The AV valve function was normal in six patients with only mild AV valve insufficiency in two patients and moderate regurgitation of the single AV valve in the other two. None of the patients had evidence of outflow obstruction. In all patients, the inferior vena cava (IVC) was dilated, with six of nine patients having more than 50% dilation compared with control subjects. IVC diameter was not related to any hepatic parameters except galactose elimination (see below).
Serum AST, ALT, and GGTP were only mildly elevated (highest: AST, 50 IU/L; ALT, 42 IU/L; GGTP, 103 IU/L) and were not correlated with age at or interval since the Fontan procedure. Bilirubin, albumin, ammonia, and cholesterol were normal in virtually all patients. Markers of hepatic synthetic function are shown in Table 3. PT and factor V levels were abnormal in 8 and 9 of 11 patients, respectively. The PT was significantly higher in the subjects with moderate–severe liver disease (16.3 ± 1.1 seconds, n = 4) than in those with none–mild disease (14.7 ± 0.3 seconds, n = 7, P = 0.003). However, there was no difference between the factor V levels of the subjects with moderate–severe liver disease (0.45 ± 0.15, n = 4) compared with those with none–mild disease (0.56 ± 0.19, n = 7, P = 0.64). Galactose elimination capacity (GEC) was abnormal (<3.5 mL/kg/min) in only 2 of 10 patients, but GEC was inversely related to interval since Fontan procedure (r2 = 0.65, P = 0.005, Fig. 2). Galactose elimination half-life was abnormal (>12 minutes) in all 10 patients, but there was no relationship between galactose elimination half-life and age at or interval since Fontan procedure. Galactose elimination half-life tended to be greater (P = 0.07) in the group with an IVC >50% dilated (17.5 ± 1.8 minutes, n = 6) compared with the group with an IVC dilated <50% (15.1 ± 1.2 minutes, n = 4).
Doppler ultrasound investigation of portal and splenic venous circulation (Table 4) showed a wide variation in size of the main, left, and right portal veins. The left portal vein was significantly larger than the right (P = 0.005). There was a strong trend to a larger left portal vein Z-score in the patients with moderate–severe liver disease (8.2 ± 7.0, n = 4) compared with those with none–mild disease (2.9 ± 1.9, n = 7, P = 0.054). The main portal vein mean velocities and the sum of the main and left and right portal vein mean velocities were normal in all but one subject. There were no significant relationships between the portal or splenic flow velocities or diameters and any other cardiac or hepatic variables. Thus, other than the size of the left portal vein, Doppler ultrasound investigation of portal and splenic circulations was not helpful in identifying patients with significant hepatic fibrosis or liver dysfunction.
In adults, chronic congestive heart failure and systemic venous engorgement produce a spectrum of histologic and functional abnormalities in the liver, ranging from dilation of sinusoids with centrilobular necrosis to replacement of centrilobular hepatocytes with fibrous tissue and cirrhosis (29). Cardiac cirrhosis refers to the irreversible, end-stage phase of hepatic damage as a result of long-standing hepatic venous congestion from cardiac failure. In an autopsy study of 605 adults who died with chronic cardiac decompensation, 10% were noted to have histologic changes in the liver consistent with cardiac cirrhosis (30).
There is little information available regarding the long-term consequences of chronic hepatic congestion in children. Matsuda et al. (31) and Jenkins et al. (32) described a high incidence of acute liver dysfunction after the Fontan procedure, caused predominantly by low cardiac output, which did not correlate with the presence of abnormal standard liver blood tests obtained approximately 1 year after surgery. Moreover, serum aminotransferase levels were rarely elevated during long-term follow-up evaluation after the Fontan procedure (13,14). However, factors V (14) and VII (13) were low in 24% to 43% of patients during long-term follow-up evaluation after the Fontan procedure. Stanton et al. (33) described a patient who died suddenly 21 months after surgery, presumably from an arrhythmic event, who was found to have histologic evidence of cirrhosis. Lemmer et al. (11) reported a 13-year-old patient who developed cirrhosis 5 years after performance of a modified Fontan procedure for tricuspid atresia. Despite the concern for the potential deleterious consequences of the high central venous pressure on the liver, no study has examined the effect on quantitative tests of hepatic function in this clinical setting. It is important to define the scope of this problem for the following reasons. The indications for the Fontan procedure as corrective surgery have expanded beyond tricuspid atresia to include an increasing number of complex congenital cardiac defects. In addition, the operation is now being performed at an earlier age. Thus, there may be an increasing number of children who are at risk for the development of liver disease from chronically elevated systemic venous pressure. Orthotopic cardiac transplantation is now an option for patients who have undergone the Fontan procedure who experience late complications, or, in some cases, as an alternative treatment. Because cardiac transplantation risk is greatly increased if cirrhosis is established, cirrhosis is considered a relative contraindication to orthotopic cardiac transplantation at some centers. Liver biopsy in the setting of high right-sided cardiac pressures may carry an increased risk of bleeding from the higher pressures in the hepatic vein and the frequently associated coagulopathy. In addition, transjugular liver biopsy is usually precluded by the post-Fontan anatomy. Thus, it would be optimal to determine objective, noninvasive criteria that precede the development of cirrhosis so that intervention with cardiac transplantation could be offered to prevent further hepatic injury.
Although the Fontan procedure has been performed for more than 30 years in children with complex congenital heart diseases, the long-term effects of elevated right-sided cardiac pressure on the developing liver have not been delineated. In this pilot study, we observed that abnormalities of standard biochemical indices of hepatic injury are uncommon in children after the Fontan procedure, confirming previous reports (13,14). One of the major findings of this study was the frequency of abnormal coagulation in children who were studied at least 5 years after the Fontan procedure. Our patients frequently had prolonged PT and a reduction in factor V levels. Previous studies have suggested that intrahepatic microscopic thromboses contribute to the pathology of hepatic injury and fibrosis during hepatic congestion (19). However, the prolonged PT and low factor V levels were unlikely to be caused by consumption because factor VII and VIII levels were normal in almost all subjects, as noted previously by Kaulitz et al. (14). Thus, factor V synthesis appears to be dysregulated during hepatic congestion and may serve as an early biochemical marker of hepatic dysfunction after the Fontan procedure. The mechanism underlying this dysregulation requires further investigation.
An important clinical issue in children who have undergone the Fontan procedure is the development of significant hepatic fibrosis or cirrhosis and its impact on timing for a possible cardiac transplantation. Because cardiac cirrhosis would further reduce hepatic blood flow in a congested liver, we investigated indirect markers of hepatic blood flow. Galactose elimination half-life and GEC decreased as the interval since the Fontan procedure lengthened. These data suggest that GEC may be a useful marker for the gradual impairment of hepatic function that occurs with cardiac cirrhosis because galactose elimination depends on hepatocyte mass and hepatic blood flow.
Liver biopsy is the most accurate current means of quantifying hepatic fibrosis. In this study, liver biopsies were only performed for clinical indications (primarily evaluation for orthotopic cardiac transplantation). Ideally, we would have preferred to have liver histology for all patients; however, the potential increased risk of bleeding and lack of access for transjugular liver biopsy precluded biopsy for research purposes alone on ethical grounds.
Doppler ultrasound evaluation of portal and splenic venous blood flow did not demonstrate reduced flow, suggesting that the effect of hepatic injury after the Fontan procedure is likely a decrease of hepatocyte mass. Although there was a trend to a larger left portal vein diameter in subjects with clinical liver disease, left portal vein diameter did not appear to be as useful a discriminant function as PT or GEC. The temporally related increase of PT in conjunction with the decrease in GEC suggest that hepatocyte loss of function may be the earliest hepatic abnormality that develops within 5 years of the Fontan procedure.
In conclusion, routine biochemical indices of hepatic injury may underestimate the extent of congestive hepatopathy and hepatic fibrosis after the Fontan procedure. In contrast, our data suggest that PT and GEC may be useful markers of hepatic dysfunction and fibrosis in these patients. Sequential determination of PT may be a useful part of the routine monitoring after the Fontan procedure. Further, prospective investigation of serial galactose clearance testing will be necessary to determine if this test is a useful predictor of progressive hepatic dysfunction and fibrosis after the Fontan procedure (34).
1. Fontan F, Baudet E. Surgical repair of tricuspid atresia. Thorax 1971; 26:240–8.
2. Castaneda AR. From Glenn to Fontan, a continuing evolution. Circulation 1992; 86(suppl II):II-80–3.
3. Dick M, Fyler DC, Nadas AS. Tricuspid atresia: clinical course in 101 patients. Am J Cardiol 1975; 36:327–37.
4. Mayer JEJ, Bridges ND, Lock JE, et al. Factors associated with marked reduction in mortality for Fontan operations in patients with single ventricle. J Thorac Cardiovasc Surg 1992; 103:444–51.
5. Mair DD, Rice MJ, Hagler DJ, et al. Outcome of the Fontan procedure in patients with tricuspid atresia. Circulation 1985; 72:II88–92.
6. Driscoll DJ, Offord KP, Feldt RH, et al. Five- to fifteen-year follow-up after Fontan operation. Circulation 1992; 85:469–96.
7. Sanders SP, Wright GB, Keane JF, et al. Clinical and hemodynamic results of the Fontan operation for tricuspid atresia. Am J Cardiol 1982; 49:1733–40.
8. Mair DD, Hagler DJ, Puga FJ, et al. Fontan operation in 176 patients with tricuspid atresia. Results and a proposed new index for patient selection. Circulation 1990; 82:IV164–9.
9. Mair DD, Hagler DJ, Julsrud PR, et al. Early and late results of the modified Fontan procedure for double inlet left ventricle: the Mayo Clinic experience. J Am Coll Cardiol 1991; 18:1729–32.
10. Cohen AJ, Cleveland DC, J D, et al. Results of the Fontan procedure for patients with univentricular heart. Ann Thorac Surg 1991; 52:1266–70.
11. Lemmer JH, Coran AG, Behrendt DM, et al. Liver fibrosis (cardiac cirrhosis) five years after modified Fontan operation for tricuspid atresia. J Thorac Cardiovasc Surg 1983; 86:757–60.
12. Behrendt DM, Rosentahl A. Cardiovascular status after repair by Fontan procedure. Ann Thorac Surg 1980; 29:322–30.
13. Cromme-Dijkhuis AH, Hess J, Hahlen K, et al. Specific sequelae after Fontan operation at mid- and long-term follow-up. Arrhythmia, liver dysfunction, and coagulation disorders. J Thorac Cardiovasc Surg 1993; 106:1126–32.
14. Kaulitz R, Luhmer I, Bergmann F, et al. Sequelae after modified Fontan operation: postoperative haemodynamic data and organ function. Heart 1997; 78:154–9.
15. Sherlock S. The liver in circulatory failure. In: Schiff L, Schiff ER, eds. Diseases of the Liver.
Vol. 2. Philadelphia: JB Lippincott; 1993:1431–7.
16. Narkewicz MR, Smith D, Gregory C, et al. Effect of ursodeoxycholic acid therapy on hepatic function in children with intrahepatic cholestatic liver disease. J Pediatr Gastroenterol Nutr 1998; 26:49–55.
17. Fletcher SE, Case CL, Fyfe DA, et al. Clinical spectrum of venous thrombi in the Fontan patient. Am J Cardiol 1991; 68:1721–2.
18. Dobell AR, Trusler GA, Smallhorn JF, et al. Atrial thrombi after the Fontan operation. Ann Thorac Surg 1986; 42:664–7.
19. Cromme-Dijkhuis AH, Henkens CM, Bijleveld CM, et al. Coagulation factor abnormalities as possible thrombotic risk factors after Fontan operations. Lancet 1990; 336:1087–90.
20. Peters NS, Somerville J. Arrhythmias after the Fontan procedure. Br Heart J 1992; 68:199–204.
21. Hess J, Kruizinga K, Bijleveld CM, et al. Protein-losing enteropathy after Fontan operation. J Thorac Cardiovasc Surg 1984; 88:606–9.
22. Rothman A, Snyder J. Protein-losing enteropathy following the Fontan operation: resolution with prednisone therapy. Am Heart J 1991; 121:618–9.
23. Rychik J, Piccoli DA, Barber G. Usefulness of corticosteroid therapy for protein-losing enteropathy after the Fontan procedure. Am J Cardiol 1991; 68:819–21.
24. Hanseus K, Bjorkhem G, Lundstrom NR. Dimensions of cardiac chambers and great vessels by cross-sectional echocardiography in infants and children. Pediatr Cardiol 1988; 9:7–15.
25. Hjelm M, de Verdier CH. Colorimetric method with galactose oxidase. In: Bergmeyer H, Bergmeyer J, Grabl M, eds. Methods of Enzymatic Analysis.
Vol. 6. Deerfield Beach, FL: Verlag Chemie, 1984:281–8.
26. Lindskov J. The quantitative liver function as measured by the galactose elimination capacity. II. Prognostic value and changes during disease in patients with cirrhosis. Acta Med Scand 1982; 212:303–8.
27. Vocke AK, Kardorff R, Ehrich JH. Sonographic measurements of the portal vein and its intrahepatic branches in children. Eur J Ultrasound 1998; 7:121–7.
28. Wanek EA, Horgan JG, Karrer FM, et al. Portal venous velocity in biliary atresia. J Pediatr Surg 1990; 25:146–8.
29. Dunn GD, Hayes P, Breen KJ, et al. The liver in congestive heart failure: a review. Am J Med Sci 1973; 265:174–89.
30. Kotin P, Hall E. “Cardiac” or congestive cirrhosis of liver. Am J Pathol 1951; 27:561–71.
31. Matsuda H, Covino E, Hirose H, et al. Acute liver dysfunction after modified Fontan operation for complex cardiac lesions. Analysis of the contributing factors and its relation to the early prognosis. J Thorac Cardiovasc Surg 1988; 96:219–26.
32. Jenkins JG, Lynn AM, Wood AE, et al. Acute hepatic failure following cardiac operation in children. J Thorac Cardiovasc Surg 1982; 84:865–71.
33. Stanton RE, Lurie PR, Lindesmith GG, et al. The Fontan procedure for tricuspid atresia. Circulation 1981; 64:II140–6.
34. Zaramella MG, Vallini R, Tiribelli C. Estimation of the hepatic blood “flow” by galactose plasma clearance in patients with liver disease. Liver 1985; 5:129–33.
© 2003 Lippincott Williams & Wilkins, Inc.