Prognostic Significance of Increased Serum Bilirubin Levels Coincident With Cardiac Decompensation in Chronic Heart Failure

Circ J 28; 72: 364 369 Prognostic Significance of Increased Serum Bilirubin Levels Coincident With Cardiac Decompensation in Chronic Heart Failure Hisahito Shinagawa, MD; Takayuki Inomata, MD; Toshimi Koitabashi, MD; Hironari Nakano, MD; Ichiro Takeuchi, MD; Takashi Naruke, MD; Tsutomu Ohsaka, MD; Mototsugu Nishii, MD; Hitoshi Takehana, MD; Tohru Izumi, MD Background The aim of this study was to analyze the relationship between abnormal liver function tests (LFTs) coincident with heart failure (HF) exacerbation and subsequent long-term outcome in patients with chronic HF. Methods and Results The study population consisted of 183 consecutive patients admitted for HF exacerbation with left ventricular ejection fraction 4%. Cox proportional hazard analysis revealed that serum total bilirubin (T-Bil) levels on admission (hazard ratio 1.896, p<.1, 95%confidence interval 1.323 2.717), but not T-Bil at discharge or other LFTs, was an independent predictor of subsequent cardiac events after hospital discharge (cardiac death or readmission for HF exacerbation) The cardiac-event-free rates significantly decreased according to increasing tertiles of T-Bil stratified by the level of.7 and 1.2mg/dl (p<.1). T-Bil on admission had significant correlations with simultaneously-measured central venous pressure (CVP) (r=.42, p<.1) and cardiac index (CI) (r=.5, p<.1). The patients demonstrating high CVP together with low CI showed significantly increased T-Bil compared with any other group. Conclusions Increased T-Bil coincident with cardiac decompensation predicts a worse long-term prognosis of CHF, presumably through the potential liability to both congestion and tissue hypoperfusion simultaneously when HF deteriorates. (Circ J 28; 72: 364 369) Key Words: Acute exacerbation; Bilirubin; Heart failure; Liver function test; Prognosis Liver dysfunction has long been reported as 1 of the remarkable phenomena caused by heart failure (HF). 1 5 It is considered that both congestion and low perfusion of the liver, reflecting impaired hemodynamics, are the causative mechanisms. 2 5 This consistency in the mechanism of liver dysfunction and the pathogenesis contributing to severe HF implies that a liver function abnormality could have a prognostic impact on patients with chronic HF. The prognosis of chronic HF based on the serum levels of aspartate aminotransferase (AST) and bilirubin (Bil) has been reported, 1 but it is necessary to interpret the data from liver function tests (LFTs) according to the sampling time-points during the course of exacerbated HF, because abnormal LFTs are known to fluctuate and resolve rapidly coincident with compensation of HF. 3,5 The aim of this study was to analyze the relationship between the abnormal profile of LFTs coincident with HF exacerbation and subsequent long-term outcome of chronic HF. (Received June 11, 27; revised manuscript received November 5, 27; accepted November 13, 27) Department of Cardio-angiology, Kitasato University School of Medicine, Sagamihara, Japan Mailing address: Hisahito Shinagawa, MD, Department of Cardioangiology, Kitasato University School of Medicine, 1-15-1 Kitasato, Sagamihara 228-8555, Japan. E-mail: hisahito@med.kitasato-u.ac.jp All rights are reserved to the Japanese Circulation Society. For permissions, please e-mail: cj@j-circ.or.jp Methods Study Population Chronic HF patients who were admitted to Kitasato University Hospital for HF exacerbation with left ventricular ejection fraction (LVEF) 4% from January 1999 to December 22 were consecutively recruited. Patients with acute coronary syndrome, acute myocarditis, hepatobiliary diseases, chronic renal dysfunction (serum creatinine levels >2. mg/dl), active malignancy, or drug-induced liver dysfunction were excluded. From 188 recruits, the final study population comprised 183 patients (5 patients died in hospital) who would be followed after hospital discharge. Endpoints, defined as readmission for HF exacerbation or cardiovascular death (termed as cardiac events ) after hospital discharge, were retrospectively analyzed until September 25. HF exacerbation was judged by at least 2 experienced cardiologists according to the Framingham criteria, 6 together with other clinical data including chest X-rays and echocardiography. All patients gave informed consent and the study procedure was in compliance with the institutional guidelines of Kitasato University and the principles outlined in the Helsinki Declaration. Clinical Evaluation and Hemodynamic Measurements Biochemical laboratory data, New York Heart Association functional class (NYHA class), and echocardiography were evaluated on admission and at discharge. Arterial blood gas analysis was performed on admission. All clinical findings at discharge were sampled from the latest data to discharge when HF was adequately controlled. Plasma Circulation Journal Vol.72, March 28

Prognostic Significance of Bil for Chronic HF 365 Table 1 Patients Characteristics All patients (n=183) Patients characteristics according to T-Bil on admission Low <.7 mg/dl Intermediate.7 High 1.2 mg/dl (n=66) <1.2 mg/dl (n= 65) (n= 52) p value* Age, years (IQR) 66 (58 75) 67 (59 76) 66 (57 75) 66 (54 74).63 Male, n (%) 13 (71) 45 (68.1) 43 (66.1) 42 (8.7).18 Data on admission NYHA class II:III:IV, n 9:53:121 3:16:47 4:19:42 2:19:31.8 SBP, mmhg (IQR) 13 (118 146) 14 (124 163) 132 (119 146) 118 (16 14) <.1 DBP, mmhg (IQR) 8 (68 88) 91 (81 11) 78 (68 86) 77 (64 88).21 Heart rate, beats/min (IQR) 1 (84 118) 16 (88 12) 96 (8 114) 1 (78 119).16 LVEF,% (IQR) 29 (24 36) 3. (25.8 37.2) 29. (21. 36.) 27 (21.5 33.8).56 LVDd, mm (IQR) 61 (55 68) 59 (52 68) 6 (55 67) 65 (56 7).65 T-Bil, mg/dl (IQR).8 (.6 1.2).5 (.3.6).9 (.8 1.) 1.5 (1.3 1.7) <.1 D-Bil, mg/dl (IQR).4 (.3.4).2 (.2.3).3 (.3.4).7 (.5.9) <.1 AST, U/L (IQR) 36 (27 61) 34 (23 52) 39 (27 65) 4 (29 68).18 ALT, U/L (IQR) 29 (19 57) 26 (17 4) 31 (21 7) 35 (21 55).33 ALP, U/L (IQR) 25 (162 229) 176 (154 214) 188 (134 251) 221 (145 282).91 LDH, U/L (IQR) 57 (47 73) 547 (271 73) 567 (485 73) 622 (462 76).364 GTP, U/L (IQR) 71 (42 78) 38 (25 59) 61 (32 81) 73 (4 138).33 Creatinine, mg/dl (IQR) 1. (.8 1.3) 1.7 (.8 1.2) 1. (.8 1.25) 1.1 (.9 1.38).321 Serum sodium, mmol/l (IQR) 139 (137 142) 139 (137 142) 139 (138 142) 138 (137 141).5 BNP, pg/ml (IQR) 558 (36 96) 598 (32 9) 54 (35 85) 685 (559 1,26).6 Data at discharge NYHA class I:II:III:IV, n 58:13:17: 5 26:36:4: 19:38:6:2 13:29:7:3.28 LVEF, % (IQR) 37 (32 42) 37 (35 4) 37 (34 44) 36 (28 42).6 T-Bil, mg/dl (IQR).6 (.4.8).5 (.4.6).6 (.5.7).7 (.6.9).15 Creatinine, mg/dl (IQR) 1.1 (.8 1.2) 1.8 (.8 1.2) 1.14 (.8 1.2) 1.14 (.93 1.28) <.1 BNP, pg/ml (IQR) 174 (73 27) 162 (59 21) 15 (52 279) 194 (16 35).17 *p value among the tertiles stratified by T-Bil on admission. T-Bil, total bilirubin; IQR, interquartile range; NYHA class, New York Heart Association functional class; SBP, systolic blood pressure; DBP, diastolic blood pressure; LVEF, left ventricular ejection fraction; LVDd, left ventricular diastolic diameter; D-Bil, dilect birilubin; AST, aspartate aminotransferase; ALT, alanine aminotransferase; ALP, alkaline phosphatase; LDH, lactate dehydrogenase; GTP, glutamyl transferase; BNP, plasma B-type natriuretic peptide. B-type natriuretic peptide (BNP) was determined by sensitive noncompetitive immunoradiometric assays (Shionoria BNP, Shionogi, Osaka, Japan) using venous blood samples taken after 15 min of rest. Echocardiography was undertaken transthoracically using ProSound SSD-5 (ALOKA, Tokyo, Japan). M-mode images were obtained in the left parasternal long-axis views to measure each chamber dimension, and LVEF was calculated by the modified Simpson s method using biplane images from apical viewpoints. Right heart catheterization was performed on admission using Swan-Ganz catheters in 79 patients who were judged by individual chief physicians to require hemodynamic monitoring of central venous pressure (CVP), pulmonary artery wedge pressure (PAWP), and cardiac index (CI) determined by the Fick method. Statistical Analysis In order to investigate independent predictor(s) of cardiac events, clinical variables with a p-value <.5 in the univariate analysis were examined using multivariate analysis in a Cox proportional hazard model. The selection method was a forward selection with a significance level of.5 for entering an explanatory variable into the model. The final model only included significant variables with a p-value <.5 using the Wald test. The cardiac-event-free rates among the tertiles based on total Bil (T-Bil) levels on admission or at discharge were compared using Kaplan-Meier analysis with a log-rank test. Comparisons among groups of the patients characteristics were performed using a chisquared test for categorical variables, and the Mann-Whitney U-test or Kruskal Wallis rank test for continuous variables, followed by the Scheffe s method for multiple comparisons if necessary. Correlation coefficients between T-Bil and hemodynamic variables were tested with Spearman s rankcorrelation test. Continuous variables are presented as median and interquartile range or mean ± standard deviation. For all analyses, a 2-sided p-value <.5 was considered as statistically significant. The statistical package used was SPSS 11..1J for Windows (SPSS Inc, Chicago, IL, USA). Results Patients Characteristics and Subsequent Events During Follow-up The mean age of the patients was 65.4±14.2 years and 71% were men. Underlying disorders were ischemic heart disease (49%), dilated cardiomyopathy (29%), valvular heart disease (25%); a history of hypertension (45%), diabetes mellitus (38%), and hyperlipidemia (16%). Other characteristics are listed in Table 1. Cardiac events occurred in 44% (81/183) of the subjects during the follow-up period (29.5±2.8 months) and of these 25 were cardiac deaths (13.7%; 16 with HF exacerbation, 6 with lethal ventricular arrhythmias, 2 sudden cardiac deaths, and 1 acute myocardial infarction). Six patients (3.3%) died of other causes, such as pneumonia or malignancy, before the endpoint. Readmission for HF exacerbation occurred in 4% (73/183). Prognostic Significance of Serum T-Bil on Admission Among the clinical parameters on admission, univariate analysis showed that significant predictors were age, male sex, systolic blood pressure (SBP), diastolic blood pressure, heart rate, T-Bil, direct Bil (D-Bil), alkaline phosphatase Circulation Journal Vol.72, March 28

366 SHINAGAWA H et al. Table 2 Predictors of Subsequent Cardiac Events Univariate analysis Multivariable analysis HR (95%CI) p value HR (95%CI) p value Age, years 1.49 (1.29 1.7) <.1 1.5 (1.29 1.71) <.1 Male (vs Female) 1.725 (1.12 2.72).17 Data on admission NYHA class.283 SBP, mmhg.979 (.969.989) <.1.986 (.974.997).14 DBP, mmhg.974 (.959.99) <.1 Heart rate, beats/min.982 (.972.992) <.1 LVEF, %.342 LVDd, mm.165 T-Bil, mg/dl 1.948 (1.424 2.666) <.1 1.896 (1.323 2.717) <.1 D-Bil, mg/dl 4.634 (1.776 12.69).2 ALP, U/L 1.4 (1.1 1.6).9 GTP, U/L 1.5 (1.1 1.1).29 AST, U/L.784 ALT, U/L.761 LDH, U/L.948 Creatinine, mg/dl 1.649 (1.2 2.668).9 Serum sodium, mmol/l.716 BNP, pg/ml 1.1 (1. 1.1).2 Data at discharge NYHA class 1.993 (1.32 3.53).2 1.645 (1.55 2.516).28 LVEF.967 (.943.992).11 T-Bil.889 Creatinine 1.562 (1.7 2.43).42 BNP 1.3 (1.2 1.4) <.1 1.2 (1.1 1.3).1 HR, hazard ratio; CI, confidence interval. Other abbreviations as in Table 1. Cardiac-events-free survival rate (%) 1 8 6 4 2 Low Intermediate High n patients at risk T-Bil on admission Low (T-Bil <.7 mg/dl) Intermediate (.7 T-Bil <1.2 mg/dl) High (1.2 mg/dl T-Bil) 12 24 36 48 6 72 Time after hospital admission (months) 66 54 47 42 21 7 65 51 35 27 12 8 52 31 19 11 8 5 p=.36 p=.37 p<.1 Fig 1. Kaplan-Meier survival curves showed a significant difference of subsequent cardiac events among the tertiles stratified with serum total bilirubin (T-Bil) level on admission. (ALP), -glutamyltranspeptidase ( GTP), creatinine, and BNP. At discharge, NYHA class, LVEF, creatinine and BNP concentrations significantly predicted subsequent cardiac events (Table 2). Multivariate analysis revealed that the independent predictors of subsequent cardiac events were T-Bil on admission (hazard ratio 1.896, p<.1, 95% confidence interval 1.323 2.717), together with age, SBP on admission; NYHA class and BNP at discharge (Table2). Focusing on the contribution of the serum T-Bil level to the prognosis of HF, patients were divided into tertiles on the basis of T-Bil on admission or at discharge. The cardiacevent-free rates after hospital discharge significantly decreased according to increasing tertiles of T-Bil on admission (low level group: T-Bil <.7mg/dl (12. mol/l), n=66; intermediate:.7 <1.2 mg/dl (2.5 mol/l), n=65; high: 1.2 mg/dl, n=52) (p<.1; Fig 1). In the comparison of Circulation Journal Vol.72, March 28

Prognostic Significance of Bil for Chronic HF 367 CI (L min 1 m 2 ) 6 4 2 r= -.5 2 p <.1 y = 2.73 x + 4.66 y = -.72 x + 3.53 CVP (mmhg) 25 15 1 5 r=.42 p <.1 PAWP (mmhg) 1 2 3 4 4 3 2 1 y = 5.89 x + 9.5 r=.41 p <.1 PaO2 (mmhg) 1 2 3 4 25 2 15 1 5 r= -.1 p=.43 y = -1.98 x + 77.85 1 2 3 4 1 2 3 4 Fig 2. This chart indicates significant correlations between total bilirubin (T-Bil) on admission and cardiac index (CI), pulmonary artery wedge pressure (PAWP), and central venous pressure (CVP), respectively. 4 3 p<.1 p<.5 p<.5 2 p<.1 for all 1 Group A (n=31) B (n=23) C (n=11) D (n=14) CVP <8 CVP 8 CVP <8 CVP 8 (mmhg) CI 2.2 CI 2.2 CI <2.2 CI <2.2 (L min 1 m 2 ) Fig 3. Among the 4 groups (Group A-D) stratified with the levels of central venous pressure (CVP) and cardiac index (CI), the total bilirubin (T-Bil) level was significantly highest in Group D, in which systemic congestion and low perfusion coexisted. The lines represent the median, the boxes represent the interquartile range, and whiskers represent the 5th and 95th percentiles. the clinical characteristics among these 3 groups, there was no significant difference in the parameters including age, NYHA class, or BNP at discharge, other than SBP on admission (Table 1). Regarding the stratification based on T-Bil at discharge (low: <.5 mg/dl (8.6 mol/l), n=45; intermediate:.5 <.8 mg/dl (1.9 mol/l), n=97; high:.8 mg/dl, n=41), no significant difference in the cardiacevent-free rates after hospital discharge was seen among these tertiles (data not shown). Correlation of T-Bil Level and Hemodynamic Parameters T-Bil level on admission had significant correlations with simultaneously-measured CVP (r=.42, p<.1), PAWP (r=.41, p<.1) and CI (r=.5, p<.1) (Fig 2), but not with the arterial oxygen pressure (PaO2) (r=.1, p=.43; Fig 2). In this clinical setting, oxygen had already been administrated to 5% of patients (92/183) when arterial blood gas was analyzed. Furthermore, we analyzed 91 patients in whom oxygen was not given and there was no significant correlation (r=.1, p=.32) either. Based on the CVP and CI, patients were divided into Circulation Journal Vol.72, March 28

368 SHINAGAWA H et al. 4 groups (A: CVP <8mmHg, CI 2.2 L min 1 mm 2 ; B: CVP 8mmHg, CI 2.2L min 1 mm 2 ; C: CVP <8mmHg, CI <2.2L min 1 mm 2 ; D: CVP 8mmHg, CI <2.2L min 1 mm 2 ). T-Bil was significantly highest when systemic congestion and hypoperfusion coexisted, classified as group D, compared with any other group (group A vs D: p<.1, B vs D: p<.5, C vs D: p<.5; Fig 3). Furthermore, a similar relationship was found using PCWP instead of CVP (data not shown). Discussion Pathophysiological Mechanism of Increased Serum Bil Level in HF HF results in a broad range of abnormal LFTs, the most common of which is a mild elevation of serum Bil (3 7% of patients), with increases of other cholestatic markers such as ALP and GTP being milder and less frequent, up to 1 2%, as well as those of the serum transaminases such as AST and alanine aminotransferase (12 33%). 3,7,8 The pathomechanism responsible for the liver injury is considered to be derived from 2 hemodynamic alterations in HF: decreased hepatic blood flow originating from low cardiac output and increased hepatic venous pressure with subsequent atrophy of liver cells and edema of the peripheral area, both leading to hepatocellular hypoxia. 3,7 PaO2 itself is thought to be a minor factor in hepatic injury. 3 Several studies so far have shown that the serum Bil correlates with various hemodynamic parameters, such as right atrial pressure (RAP), 2,4,9 severity of tricuspid regurgitation (TR), 4,9 PAWP, 9 LVEF, 4 and CI, 9 whereas ALP and GTP only do so with RAP 3,8 and the severity of TR. 8 It has never been documented that the transaminases have a correlation with these parameters. 3,8 In our study, LFTs other than Bil did not have a significant correlation with hemodynamic parameters (data not shown). The differences might be explained by the different pathophysiological mechanisms of the increase in each LFT. Several metabolic processes of Bil, including the uptake of indirect Bil from blood, conjugation in hepatocytes 3 and secretion of D-Bil into bile, 1,11 are attenuated by hepatocellular hypoxia, leading to an enhanced increase of the serum T-Bil. Biliary obstruction caused by high hepatic venous pressure is also reported to be a causative mechanism when T-Bil is highly elevated. 8 Furthermore, hemolysis, which occurs mainly in the simply congested lungs of HF patients, leads to an increase in the serum indirect Bil. 2,7,12 It is certain that the levels of transaminases also correlate with hepatocellular hypoxia, but necrosis of the affected hepatocyte is required for an increase in the serum levels, 3 which is the reason why the cells swell dramatically in cases of acute decrease in cardiac output such as cardiogenic shock. 7 The increase in the level of ALP, as well as GTP, is derived mainly from obstruction of the biliary system, 13,14 not through hepatocellular hypoxia. Therefore, Bil is the most sensitive marker among the LFTs for severe HF, as an integral clinical indicator to reflect the multifarious conditions related to HF. Clinical Impact of Serum Bil Level on HF Exacerbation in Relation to Long-Term Prognosis Previous studies have reported conflicting results concerning the clinical significance of measuring the serum T- Bil level as a marker of hemodynamic parameters. It was shown in 1 study of patients with severe HF that the serum T-Bil level had significant correlations with CI, as well as RAP or PAWP, 7 whereas in another study of patients with stable HF the T-Bil correlated only with RAP. 2 In the present study, serum T-Bil levels at the time of HF decompensation had a clearer correlation with CI than either CVP or PAWP (r=.5,.42,.41, respectively) (Fig 2), which suggests that the serum T-Bil level coincident with cardiac decompensation more exactly reflects the hemodynamics, especially CI, compared with when HF is more controlled. Low cardiac output, more frequently coincident with exacerbated HF, may be the key to interpreting the present result that the serum T-Bil on admission, but not at discharge, was an independent predictor of subsequent cardiac events, because low cardiac output can predict HF prognosis better than concomitant pulmonary or systemic congestion. 15 The significance of the serum T-Bil level depends on the type of HF, including disease severity and clinical stage; in other words, it is important to consider the HF status when the clinical significance of the serum T-Bil level is being evaluated. Another point is that the serum Bil at the time of acute decompensation may predict HF outcome in the remote future beyond the acute phase of HF exacerbation, despite the similar HF status at the starting point presumed by the plasma BNP, NYHA class and LVEF values at discharge. However, we can confirm that the serum Bil on admission predicts the patient s prognosis, even when we include the first hospitalization into the follow-up period, changing the starting point of follow-up from hospital discharge to admission (data not shown). Those results suggest that elevation of the serum T-Bil level represents not only the temporary severity of hemodynamics but also the latent cardiac status of chronic HF contributing to long-term prognosis. It seems that the serum T-Bil level at the time of HF exacerbation should be regarded as the marker of the hemodynamic response to deteriorated HF, which reflects cardiac pump/circulatory reserve, leading to differentiation of HF prognosis even among patients with similar HF status during the stable phase. Comparative Characteristics of Bil as an Indicator of HF The plasma BNP level is widely used in general practice as a valuable predictor of HF prognosis. 16,17 Although conflicting data exist, 18 the plasma BNP concentration has a good correlation with filling pressure, 19,2 but there are no data concerning the CI. 19 The serum T-Bil level, on the other hand, reflects the CI as well as congestion, as mentioned earlier. In clinical practice, for HF management it is generally easier to detect congestion than low cardiac output or hypoperfusion. The present study indicates that an increased serum T-Bil level could be a clue to a hypoperfused status in HF patients, leading to adequate therapy without using invasive clinical techniques. Study Limitations The study was carried out inside a single institute in both a retrospective and nonrandomized fashion. The acquired data might be biased because clinical judgment depended on the clinical status of each patient. Furthermore, while it is ideal that there should be no significant differences in clinical background among tertiles in practise this is impossible because the serum Bil levels are determined by pathological and hemodynamic changes. Based on the analysis of the present study, we paid attention to the clinical data on admission that were related to Circulation Journal Vol.72, March 28

Prognostic Significance of Bil for Chronic HF the prognosis of HF. Because the time of onset of acute cardiac decompensation leading to hospital admission is imprecise among individual patients, clinical parameters defined as within 24 h of admission might be estimated at different clinical time-points within the variation of HF status, leading to some biased interpretations. Finally, as the present study consisted of patients who had left ventricular systolic dysfunction (LVEF 4%) on admission, further study of patients with preserved systolic function is warranted. Acknowledgments This study was supported in part by grant-in-aid for scientific research from the Integrative Research Program of the Graduate School of Medical Sciences, Kitasato University. References 1. Batin P, Wickens M, McEntegart D, Fullwood L, Cowley J. The importance of abnormality of liver function tests in predicting mortality in chronic heart failure. Eur Heart J 1995; 16: 1613 1618. 2. Sherlock S. The liver in heart failure: Relation of anatomical, functional, and circulatory changes. Br Heart J 1951; 13: 273 293. 3. Dunn GD, Hayes P, Breen KJ, Schenker S. The liver in congestive heart failure: A review. Am J Med Sci 1973; 265: 174 189. 4. Lau GT, Tan HC, Kritharides L. Type of dysfunction in heart failure and its relation to the severity of tricuspid regurgitation. Am J Cardiol 22; 9: 145 149. 5. Naschitz JE, Slobodin G, Lewis RJ, Zuckermqn E, Yashurun D. Heart disease affecting the liver and liver diseases affecting the heart. Am Heart J 2; 14: 111 12. 6. Ho KL, Prinsly JL, Kannel WB, Levy D. The epidemiology of heart failure: The Framingham Study. J Am Coll Cardiol 1993; 22(Suppl): 6A. 7. Sherlock S, Dooley J. The hepatic artery and hepatic veins: The liver in circulatory failure. In: Sherlock S, Dooley J, editors. Diseases of the liver and biliary system, 11 th edn. Oxford: Blackwell Science; 22; 199 23. 8. Rosenberg PM, Lawrence SF. The liver in circulatory failure. In: Schiff ER, Sorrell MF, Maddrey WC, editors. Shiff s diseases of the liver, 9 th edn. Philadelphia: Willis C. Maddrey; 23; 1328 134. 369 9. Kubo SH, Walter BA, John DH, Clark M, Cody RJ. Liver function abnormalities in chronic heart failure: Influence of systemic hemodynamics. Arch Intern Med 1987; 147: 1227 123. 1. Reese AJ. The effect of hypoxia on liver secretion studied by intravital fluorescence microscopy. Br J Exp Pathol 196; 41: 527 535. 11. Shorey J, Schenker S, Combes B. Effect of acute hypoxia on hepatic excretory function. Am J Physiol 1969; 216: 1441 1452. 12. Kramer MR, Marshall SE, Tiroke A, Lewiston NJ, Starnes VA, Theodore J. Clinical significance of hyperbilirubinemia in patients with pulmonary hypertension undergoing heart-lung transplantation. J Heart Lung Transplant 1991; 1: 317 321. 13. Poynard T, Imbert-Bismut F. Laboratory testing for liver disease. In: Boyer TD, Wright TL, Manns MP, editors. Zakim and Boyer s hepatology: A text book of liver disease, 5 th edn. New York: Elsevier; 26; 235 237. 14. Richman SM, Delman AJ, Grob D. Alteration in indices of liver function in congestive heart failure with particular reference to serum enzymes. Am J Med 1961; 3: 211 225. 15. Nohria A, Tsang SW, Fang JC, Lewis EF, Jarcho JA, Mudge GH, et al. Clinical assessment identifies hemodynamic profiles that predict outcomes in patients admitted with heart failure. J Am Coll Cardiol 23; 41: 1797 184. 16. McCullough PA, Nowak RM, McCord J, Hollander JE, Herrmann HC, Steg PG, et al. B-type natriuretic peptide and clinical judgment in emergency diagnosis of heart failure: Analysis from Breathing Not Properly (BNP) Multinational Study. Circulation 22; 16: 416 422. 17. Berger R, Huelsman M, Strecker K, Bojic A, Moser P, Stanek B, et al. B-type natriuretic peptide predicts sudden death in patients with chronic heart failure. Circulation 22; 15: 2392 2397. 18. Parsonage WA, Galbraith AJ, Koerbin GL, Potter JM. Value of B- type natriuretic peptide for identifying significantly elevated pulmonary artery wedge pressure in patients treated for established chronic heart failure secondary to ischemic or idiopathic dilated cardiomyopathy. Am J Cardiol 25; 95: 883 885. 19. Maeda K, Tsutamoto T, Wada A, Hisanaga T, Kinoshita M. Plasma brain natriuretic peptide as a biochemical marker of high left ventricular end-diastolic pressure in patients with symptomatic left ventricular dysfunction. Am Heart J 1998; 135: 825 832. 2. Yasue H, Yoshimura M, Sumida H, Kikuta K, Kugiyama K, Jougasaki M, et al. Localization and mechanism of secretion of B-type natriuretic peptide in comparison with those of A-type natriuretic peptide in normal subjects and patients with heart failure. Circulation 1994; 9: 195 23. Circulation Journal Vol.72, March 28