| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 1998;116:633-640
© 1998 Mosby, Inc.
CARDIOPULMONARY SUPPORT AND PHYSIOLOGY |
San Francisco, Calif
From the Department of Cardiac Surgery, California Pacific Medical Center, San Francisco, Calif.
Received for publication Jan 30, 1998. Revisions requested March 24, 1998; revisions received May 7, 1998. Accepted for publication July 1, 1998. Address for reprints: David J. Farrar, PhD, Department of Cardiac Surgery, California Pacific Medical Center, 2351 Clay St, Room 5637, San Francisco, CA 94115.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1,2 the pathophysiologic mechanisms of end-organ failure and its recovery are not fully understood, and clinical markers and thresholds beyond which recovery is unlikely to occur are often lacking.
One of the crucial major organs determining patient survival during VAD support is the liver. However, the role of hepatic function in patients with a VAD and the limits of reversibility are poorly understood. In a state of acute circulatory failure, decreased hepatic blood flow and increased hepatic venous pressure can lead to cellular hypoxia 3 and a clinical picture similar to viral hepatitis, frequently called "ischemic hepatitis." 4,5 Common abnormalities include the elevation of prothrombin time, liver enzymes, and sometimes bile duct enzymes and bilirubin.
Because of the importance of patient selection, a number of publications have evaluated pre-VAD markers of survival in patients with significant end-organ failure, but with no consistent findings. 1,6-9 This article focuses on the importance of liver function in patients having a VAD. On the basis of a single-center experience, we analyze different liver function variables to assess which of them are predictive of patient survival and liver recovery.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Preoperative status and indication for support
All patients in this study were candidates for cardiac transplantation and were selected for VAD support at a stage of the disease where death or permanent end-organ damage was likely to occur before a donor heart was available. Hemodynamic guidelines for implantation were cardiac index less than 1.8 L/min per square meter, mean arterial pressure less than 70 mm Hg, and left atrial pressure greater than 20 mm Hg despite appropriate use of conventional therapies.
Before implantation, all patients were receiving maximal inotropic support and the average cardiac index was 1.6 L/min per square meter with an average pulmonary capillary wedge pressure of 28.9 mm Hg. Fifty percent of patients had prior cardiac arrests, 88% were supported with intra-aortic balloon pumps, and 69% were intubated and mechanically ventilated. More than half of the patients had substantial liver failure before implantation, with 18 patients (56%) having elevation of both transaminases and 19 (59%) having pathologic total bilirubin levels.
Mechanical circulatory support system
The Thoratec VAD System was used as a biventricular system in 28 patients (88%) and as a left VAD in 4 patients. The system consists of paracorporeally placed prosthetic ventricles with a 65-mL stroke volume connected to the heart and great vessels via cannulas crossing the chest wall and a pneumatic drive console. 10 The left VAD inflow was achieved from the left ventricular apex in 24 patients and from the left atrium in 8 patients; the right VAD inflow was from the right atrium; outflow of the VADs was through 14-mm preclotted polyester grafts to the ascending aorta and pulmonary trunk, respectively. The average support duration in the 11-year study period was 18.5 days with a maximum of 149 days.
The study was conducted under investigational device exemption regulations, with protocol and patient consent forms approved by the institutional review board of the hospital and the US Food and Drug Administration. Informed consent was obtained from either the patient or next of kin.
Factors analyzed
The following 30 preoperative factors were evaluated regarding their association with 2 end points (patient survival to discharge and liver recovery): age, sex, body surface area, weight, diagnosis, cardiac index (either measured by thermodilution or, in ventricular fibrillation, considered 0), pulmonary capillary wedge pressure, pre-VAD cardiac operations, preoperative mechanical ventilation, preoperative intra-aortic balloon pump, pre-VAD cardiac arrest, previous support with a different system such as a centrifugal pump or extracorporeal membrane oxygenation, and 18 laboratory values determined directly before implantation of the device (Table I). Additionally, several intraoperative and postoperative factors were analyzed: inflow cannulation site on the left side, univentricular or biventricular support, and reoperation for bleeding.
|
Statistical methods
Comparisons of continuous variables between survivors and nonsurvivors and between patients with and without liver recovery, respectively, were performed with a Mann-Whitney test. For comparisons of discrete variables, Pearson's 
2 test and Fisher's exact test were used. In addition, patients were divided into 3 groups on the basis of total and direct bilirubin values: normal, up to 3 times normal, and more than 3 times normal. For these groups, univariate logistic regression was carried out to identify odds ratios. Multivariate logistic regression of the data was not performed because of the small numbers of patients. In the analysis of laboratory factors, 1 patient had to be excluded because no preoperative tests were available. All statistical analyses were performed with the Crunch statistical data package (Crunch Software Corporation, Oakland, Calif) on a standard personal computer.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Association of preoperative factors with survival
The analysis of preoperative factors is shown in Tables I
and II. In general, the average values measured in nonsurvivors appeared different from those in the survivors, but as a result of high standard deviations these differences were not statistically significant. Only direct and indirect bilirubin values were significant, with group averages approximately 4 times higher in the nonsurvivors (2.8 ± 4.3 and 3.4 ± 5.1 mg/dL, respectively) compared with those of the survivors (0.7 ± 0.8 and 0.9 ± 0.9 mg/dL); the difference in total bilirubin values approached significance with a P value of .06 (Table I). Cardiac index between the groups was also significantly different, but with the survivors having a lower cardiac index. Leaving out patients in ventricular fibrillation before implantation (where cardiac index is assumed to be 0) and calculating only the values measured by thermodilution, the difference was no longer statistically significant (10 survivors with an average cardiac index of 1.4 ± 0.3 L/min per square meter versus 9 nonsurvivors with an average cardiac index of 1.9 ± 0.8 L/min per square meter).
|
). The differences between the groups for both variables were significant (see illustrations), although when total bilirubin is analyzed, survival remained at 69% to 70% until values of more than 3 times normal were achieved, at which point survival decreased to only 25%.
|
|
|
on the right side. Patients with high total, direct, and indirect bilirubin, AST, ALT, and lactate dehydrogenase (LD) levels had a significantly higher likelihood of dying of liver failure than did those without elevated levels. Weight and AST/LD ratio were approaching significance (ie, P value between .1 and .05).
Postimplantation course
The change of total bilirubin and AST in survivors and nonsurvivors during the first 3 weeks of support is depicted in Fig 3. In the majority of survivors, total bilirubin level rose slightly during the first few days and recovered to normal or near-normal values within 7 to 10 days of support; however, in most nonsurvivors it increased to highly pathologic values during this period. AST showed a gradual decline to normal values within approximately 10 days in survivors, whereas there was high variability and frequently initial increase in nonsurvivors.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
A low preoperative direct bilirubin level appears to be the best predictor of survival in our study. In the 3 groupings (Fig 2), the odds ratio for direct bilirubin level was 3.0, indicating that, with all other factors remaining equal, for each categoric increase in direct bilirubin, a patient was 3 times less likely to survive. In our study population, patients with a preoperative direct bilirubin level of more than 3 times normal had a survival of 33.3%, and patients having a total bilirubin level of more than 3 times normal had a survival of only 25%, compared with 82% and 70% survival-to-transplantation for patients with normal direct and total bilirubin levels, respectively. However, as the left sides of Figs 1 and 2 illustrates best, even patients with extremely elevated levels have a chance to survive. The 100% posttransplantation survivals (through hospital discharge) achieved with these patients indicate that preoperative liver dysfunction is no longer an issue in patients who survived with the aid of VAD support to qualify for heart transplantation.
A number of studies have previously been published on patient selection for VAD support—with the Thoratec VAD, 1 with other devices, 2,6,8,9,11 or with mixed device populations. 7,12,13 In some of these reports, a statistical evaluation of preoperative risk factors has been performed, but only the St Louis group, 7 the Columbia group, 9 and the Thoratec multicenter 1 results give statistically significant predictors for survival.
In the St Louis data, white blood count, platelet count, and preimplantation dialysis were significantly associated with survival. 7 These factors were not significant in our analysis (however, pre-VAD dialysis was not evaluated). The lack of significance of liver function tests in the St Louis series might be due to a different patient selection regimen resulting in fewer patients with severe liver failure, as the mean bilirubin level of the nonsurvivors was 1.2 mg/dL, compared with 5.8 mg/dL in the present study. Preoperative hepatic function in our patient population is also more severely compromised than in other studies reported in the literature, 2,14,15 implying that some patients in this report may not be representative of patients being selected for bridge to transplantation at other centers.
The present study analyzes data of a single center, which is a subset of the Thoratec multicenter data. 1 In the multicenter study, blood urea nitrogen and previous cardiac operation were found to be significant predictors of survival. In our center, blood urea nitrogen was not a significant factor, and although patients with previous operations had a lower survivorship, this difference did not reach statistical significance. However, bilirubin also showed a trend toward significance (P = .07) in the multicenter report. 1 Patients with a pre-VAD bilirubin concentration of more than 5 mg/dL had a survival of only 33%, suggesting a considerable effect of liver function on survival. Again, comparing the preoperative bilirubin levels, the patients in our subpopulation had comparably worse preoperative hepatic function (mean bilirubin averaging 5.8 mg/dL among nonsurvivors compared with 3.6 mg/dL in the multicenter group).
A consistent observation in all studies is that preoperative cardiac function (eg, cardiac index, left ventricular ejection fraction, pulmonary capillary wedge pressure) is not a significant predictor of survival, although central venous pressure greater than 16 mm Hg was included as a risk factor in the Columbia report. 9 In our analysis, patients with a lower cardiac index even appeared to have a better chance of survival. A possible explanation may be that the decision for support in patients with extremely low cardiac output or in ventricular fibrillation is being made immediately, whereas those patients with long-standing borderline cardiac function might remain in circulatory failure for too long, leading to more severe end-organ damage.
The majority (88%) of the analyzed patients at our center—and in fact all patients with impaired hepatic function—have been supported with biventricular VADs. Biventricular support has the potential not only to restore blood flow to the liver, which is commonly accepted as the main reason for ischemic hepatitis, 4 but also to directly relieve hepatic congestion, which is the other pathophysiologic mechanism leading to hepatic damage in cardiac failure. 3 Our results are consistent with the findings in another multicenter publication on the Thoratec VAD, 10 that patients requiring biventricular support are more severely ill and have slightly reduced, but still acceptable, survival.
Our results, although in only 4 of our patients, also suggest that patients who received VADs after support with other devices had reduced survival (only 1/4). These patients could have had increased total bilirubin levels related to hemolysis from the previous support. If so, that could have had some impact on the data.
Variables of hepatic function
It is appropriate to review the relevance of different liver function variables and to discuss their predictive value for liver recovery, because the study population seems to be one with a high degree of liver failure.
Bilirubin
In severe hepatocellular disease caused by congestive heart failure, particularly predominantly right-sided failure, a certain elevation of bilirubin concentration is common. However, high values of more than 5 mg/dL are rare and seen only in severe centrilobular necrosis, especially in acute failure. 3 Although classically one should expect mainly unconjugated (indirect) hyperbilirubinemia owing to the difficulty of the liver parenchyma to conjugate bilirubin, in most patients both direct and indirect bilirubin levels rise. The reason for the elevation of the direct fraction is a secretory abnormality of unknown biochemical mechanism. 16
In our study, the preoperative direct bilirubin fraction appears to be a better predictor of patient survival than total or indirect bilirubin. A possible reason might be the additional rise of the indirect fraction by hemolysis, which plays a role in hepatic congestion, but can also be increased by therapies such as intra-aortic balloon counterpulsation.
Transaminases
In isolated hepatic congestion, transaminases rise only slightly. However in ischemic hepatitis, with a decrease in blood flow, a more than 5-fold rise is frequent. This is strongly associated with the pathologic finding of centrilobular necrosis. 17 ALT is the more specific variable for liver recovery, because AST is also released from damaged myocardium, skeletal muscle, brain, and kidneys. However, in our patients AST and ALT both were significant predictors for liver recovery.
LD
Inasmuch as all cells contain LD as an essential enzyme, its rise is relatively nonspecific, although well recognized in ischemic hepatitis in particular. 4 In our study, it was not a significant predictor for survival, although it was a predictor for liver recovery. The LD isoenzyme 5 fraction might have been an even better marker for liver status, but no data were available.
AST/LD and ALT/LD ratio
Cassidy and Reynolds 18 reported that the AST/LD ratio and, perhaps even more important, the ALT/LD ratio can help differentiate between viral and ischemic hepatitis. The authors suggested that this could be due to less release of LD in viral hepatitis, because only hepatocytes in zone 1 were damaged, compared with all zones (although zone 3 in particular) in ischemic injury. We tested these ratios in our patient population with the underlying hypothesis that the ratios would be higher in more severe hepatocellular disease. Both ratios were indeed higher in patients whose livers did not recover, although the difference did not reach statistical significance.
Biliary tract enzymes
The reason for the rise of alkaline phosphatase and gamma-glutamyl transferase levels in heart failure is thought to be pressure-induced intrahepatic biliary obstruction, and it is therefore mainly associated with elevated hepatic venous pressures. 3 Both enzymes are rather mildly elevated even in severe ischemic hepatitis and had no relevant predictive ability for survival or liver recovery in our patients.
Protein and albumin
Serum albumin levels reflect long-term hepatic function. They often remain within normal limits in ischemic hepatitis 19 and do not correlate well with the degree of actual hepatic damage. 3 Serum albumin levels can also be acutely influenced by many factors (ie, the altered fluid status in heart failure or iatrogenic factors). Protein and albumin therefore did not show good predictive ability.
Prothrombin time
Prothrombin time is usually prolonged earlier than partial thromboplastin time because of a deficiency of factor VII, the vitamin K–dependent coagulation factor with the shortest half-life. Although it is widely used clinically as a marker for severity of liver damage, in our patients the difference between survivors and nonsurvivors was not significant. Also, no association between preoperative prothrombin time and incidence of reoperation for bleeding was found.
Importance of patient selection and early VAD support
In conclusion, pre-existing severe liver failure was prevalent in our patient population being mechanically bridged to transplantation with the Thoratec VAD. Preoperative liver function was identified as the most important predictor of survival, with total bilirubin and, in particular, its direct portion being the best predictors. However, a few patients with highly impaired liver function recovered during VAD support, and thus no threshold value was found that should absolutely lead to exclusion of a patient. In our experience, we have found that biventricular support is highly recommended in extremely ill patients such as those presented in this report. The encouraging survivals in patients without liver failure (eg, 82% survival to transplantation for patients with normal direct bilirubin levels) reinforces the importance of early VAD intervention to restore circulatory support before the onset of irreversible organ failure.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Kirsch, E. Vermes, C. Radu, B. Streich, K. Nakashima, A. Mekontso-Dessap, and D. Loisance Impact of preoperative hemodynamic support on early outcome in patients assisted with paracorporeal Thoratec((R)) ventricular assist device. Eur. J. Cardiothorac. Surg., August 1, 2008; 34(2): 262 - 267. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. C. McGee Jr., P. M. McCarthy, and N. Moazami Temporary Mechanical Circulatory Support Card. Surg. Adult, January 1, 2008; 3(2008): 507 - 534. [Full Text]
|
|
|
|
|
|
Y. An, Y.-B. Xiao, and Q.-J. Zhong Open-heart surgery in patients with liver cirrhosis Eur. J. Cardiothorac. Surg., June 1, 2007; 31(6): 1094 - 1098. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Hayashida, T. Shoujima, H. Teshima, Y. Yokokura, K. Takagi, H. Tomoeda, and S. Aoyagi Clinical outcome after cardiac operations in patients with cirrhosis Ann. Thorac. Surg., February 1, 2004; 77(2): 500 - 505. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Williams, J. Casher, N. Joshi, T. Hankinson, M. Warren, M. Oz, Y. Naka, and D. Mancini Insertion of a left ventricular assist device in patients without thorough transplant evaluations: a worthwhile risk? J. Thorac. Cardiovasc. Surg., August 1, 2003; 126(2): 436 - 441. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. W. Miller Patient selection for the use of ventricular assist devices as a bridge to transplantation Ann. Thorac. Surg., June 1, 2003; 75(90060): S66 - 71. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Moazami and P. M. McCarthy Temporary Circulatory Support Card. Surg. Adult, January 1, 2003; 2(2003): 495 - 520. [Full Text]
|
|
|
|
|
|
F. D. Pagani, K. D. Aaronson, F. Swaniker, and R. H. Bartlett The use of extracorporeal life support in adult patients with primary cardiac failure as a bridge to implantable left ventricular assist device Ann. Thorac. Surg., March 1, 2001; 71 (2007): S77 - S81. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. D. Pagani, K. D. Aaronson, D. B. Dyke, S. Wright, F. Swaniker, and R. H. Bartlett Assessment of an extracorporeal life support to LVAD bridge to heart transplant strategy Ann. Thorac. Surg., December 1, 2000; 70(6): 1977 - 1985. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A K Mahmood, J M Courtney, S Westaby, M Akdis, and H Reul Critical review of current left ventricular assist devices Perfusion, September 1, 2000; 15(5): 399 - 420. [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
