The number of patients with advanced liver disease (ALD) who require orthotopic liver transplantation is rapidly increasing. The adequate selection of candidates for liver transplantation is critical since organs are in short supply, with more than 18,000 patients in the United States awaiting transplantation of the 5,000 cadaver organs available each year.1 Patients with ALD have unique hemodynamics, with high resting cardiac outputs and low systemic vascular resistance.2, 3 As orthotopic liver transplantation surgery induces marked perioperative hemodynamic alterations that could adversely affect the outcome of patients with coronary artery disease (CAD),4, 5 the cardiac evaluation is an important part in the screening of these patients. Although the optimal screening strategy for patients with ALD is not well established, dobutamine stress echocardiography has been explored as a noninvasive test for detecting CAD. Studies involving small numbers of patients have focused on identifying those at high risk for cardiac perioperative complications, but have yielded conflicting results.6–8 Real-time myocardial contrast echocardiography (RTMCE) is a recently developed technique that utilizes a low mechanical index that minimizes microbubble destruction and permits the evaluation of myocardial perfusion imaging (MPI) in real time.9–12 RTMCE has been shown to increase the sensitivity and accuracy of dobutamine stress, when compared to wall motion analysis alone, in detecting angiographically significant CAD following intravenous injection of contrast agents.9–12 In the present study, we sought to determine the value of dobutamine stress RTMCE using commercially available contrast agents for predicting prognosis in patients with ALD.
Although dobutamine stress echocardiography has been used for the preoperative evaluation of patients with advanced liver disease (ALD), no data exist regarding the value of myocardial perfusion imaging (MPI) with real-time myocardial contrast echocardiography (RTMCE) in this patient population. We sought to determine the value of MPI during dobutamine stress RTMCE for predicting prognosis in patients with ALD. We examined both wall motion and MPI in 230 patients with ALD who underwent dobutamine stress RTMCE using intravenous commercially available contrast agents (Optison, GE-Amersham, Princeton, NJ; or Definity, Bristol-Myers Squibb Medical Imaging, North Billerica, MA). The prognostic value of clinical variables, including the Model for End-Stage Liver Disease (MELD) score, and echocardiographic data were examined using a Cox Hazard model. The primary endpoint was mortality of all causes. Among the 85 patients who underwent orthotopic liver transplantation, 4 had abnormal MPI and 81 had normal perfusion. The hospital mortality rate was 50% (2/4) in patients with abnormal MPI and 2% (2/81) in patients with normal MPI (P = 0.01). Among patients with abnormal MPI, 1 died from myocardial infarction in the first postoperative day and the second 1 from hemorrhagic shock. During a median follow-up of 15 months, 53 (23%) patients died. The independent predictors of death were an age of ≥65 yr (RR = 2.2; 95% confidence interval (CI) = 1.1–4.4; P = 0.03), MELD score of ≥25 (RR = 3.2; 95% CI = 1.8–5.5; P < 0.0001), and abnormal MPI (RR = 2.4; 95% CI = 1.1–5.2; P = 0.02). The 2-yr mortality was 24% for patients with normal MPI and 45% for those with inducible MPI abnormalities (P = 0.003). In conclusion, MPI obtained by RTMCE appears to be a useful tool in predicting mortality in patients with ALD. Further studies are required to verify its independent value. Liver Transpl 12:592–599, 2006. © 2006 AASLD.
From January 2000 to January 2004, a total of 264 patients with ALD underwent dobutamine stress RTMCE as part of their initial screening for orthotopic liver transplantation at the University of Nebraska Medical Center. The exclusion criteria for dobutamine stress RTMCE were age of <18 years, hemodynamic instability at the time of the stress test, unstable angina, recent myocardial infarction, and contraindications to any drug used in the protocol.13, 14 Follow-up was not available in 34 patients. Thus, the final population consisted of 230 patients. The study was approved by the Institutional Review Board, and all patients gave informed written consent to participate.
The contrast agents used for RTMCE were the commercially available albumin-encapsulated microbubble Optison (GE-Amersham, Princeton, NJ) or the lipid-encapsulated microbubble Definity (Bristol-Myers Squibb Medical Imaging, North Billerica, MA). Contrast agents were administered intravenously in doses of either a 0.2- to 0.3-mL bolus injection of Optison or a 0.1-mL bolus injection of Definity, followed by a 10-mL normal saline flush. The doses of contrast used for the assessment of MPI were the same doses recommended for enhancement of left ventricular border delineation during stress testing.15
RTMCE was performed using ultrasound scanners equipped with low-mechanical-index real-time pulse sequence schemes, which deploy either interpulse amplitude modulation or interpulse phase and amplitude modulation.16 Imaging was performed using pulse inversion Doppler (HDI 5000, Philips Medical Systems, Bothell, WA) in 164 (71%) patients, contrast pulse sequencing (Siemens Acuson Sequoia, Mountain View, CA) in 22 (10%) patients, or power modulation (Sonos 5500, Philips Medical Systems, Bothell, WA) in 44 (19%) patients. The systems were adjusted to achieve the optimal nonlinear signal at a mechanical index of ≤0.3 and a frame rate of ≥25 Hz. Time gain compensation and 2-dimensional gain settings were adjusted to suppress any nonlinear signals from tissue prior to contrast injection, and remained unchanged throughout the study. Contrast-enhanced images from apical views (4-, 2-, and 3-chamber) were obtained and digitized at rest and at maximal stress after the patients had achieved a test endpoint.
Patients were instructed to discontinue beta-blockers at least 24 hours before the stress test. Intravenous dobutamine was infused at a starting dose of 5 μg/kg/minute, followed by increasing doses of 10, 20, 30, and 40, up to a maximal dose of 50 μg/kg/minute, in 3- to 5-minute stages.13 Atropine (up to 2.0 mg) was injected in patients without symptoms or signs of myocardial ischemia not achieving 85% of the predicted maximal heart rate (220–age in years). The endpoints of stress tests were achievement of the target heart rate (85% of predicted maximal heart rate), maximal dobutamine/atropine doses, ST segment elevation of ≥2 mm at an interval of 80 msec after the J point in non-Q wave leads, sustained arrhythmias, severe chest pain, or intolerable adverse effects considered to be due to dobutamine or atropine.13, 14 Hypotension was defined as a fall of systolic blood pressure below 80 mm Hg or a reduction of ≥20 mm Hg from baseline. A hypertensive response was defined as a blood pressure of ≥230/120 mm Hg.
Baseline left ventricular ejection fraction was calculated by contrast-enhanced 2-dimensional images. The left ventricle was divided in 17 segments according to the joint recommendations of the American Society of Echocardiography.17 Both wall motion and MPI were analyzed during dobutamine stress RTMCE. All the tests were interpreted by an experienced observer (T.P. or F.X.), who was blinded to any angiographic data. MPI was analyzed during the 15-second period of myocardial contrast enhancement that typically appears following each bolus contrast injection.
Wall motion was defined as abnormal if there were new or worsening of preexisting wall motion abnormalities in ≥2 contiguous segments.13 MPI was considered abnormal when ≥2 segments exhibit a perfusion defect, defined as a transmural or subendocardial decrease of contrast enhancement.12, 18 Patients with normal MPI both at baseline and at peak stress were classified as having a normal MPI study. Patients with normal MPI at baseline and new perfusion defect during stress and those with baseline perfusion defect were classified as abnormal MPI studies. Attenuation from contrast or lung interference was considered present if the endocardial and epicardial borders of a segment could not be visualized, and thus were not distinguishable from surrounding tissues.18 The interobserver agreement of RTMCE in our laboratory for the analysis of MPI is 91%, and intraobserver agreement is 92%.9, 12
Quantitative Coronary Angiography
The value of RTMCE for detecting CAD was evaluated in patients who underwent coronary angiography within 3 months of RTMCE. Coronary angiography was performed at the discretion of the referring physician, using the Judkins' technique. Quantitative coronary angiographic analysis was performed by an experienced interventional cardiologist unaware of the results of RTMCE. Any visually evident stenosis was measured using a handheld electronic caliper (Tesa S.A., Renes, Switzerland) operated with custom-developed PC software. Measurements were expressed as the percent diameter narrowing using the diameter of the nearest normal-appearing region as the reference. Significant CAD was defined as >50% luminal diameter stenosis in ≥1 major coronary artery.
The Model for End-Stage Liver Disease (MELD) score was used to determine the severity of liver disease. This is a survival model based on a composite of 3 laboratory variables, serum creatinine, serum bilirubin, and international normalized ratio for prothrombin time, and has been demonstrated to accurately stratify patients according to mortality risk in several independent cohorts of patients with cirrhosis.19–21 The MELD score was calculated according to the method of Malinchoc et al.19: R = 0.957 × log [creatinine (mg/dL)] + 0.378 × log [bilirubin (mg/dL)] + 1.12 × log international normalized ratio + 0.643 × cause of cirrhosis (0 for alcohol-induced cirrhosis and 1 for non-alcohol-induced cirrhosis).
Follow-up of the patients was obtained from review of the University of Nebraska Medical Center liver transplant database, review of the patients' hospital charts, and telephone interviews with patients conducted by trained personnel. The primary event was defined as mortality of any cause. Patients were included if they died at any time after RTMCE or had a minimum of 6 months of follow-up. Follow-up was completed in November 2004.
Continuous variables were expressed as mean and standard deviation and categorical variables as proportions. Two-tailed unpaired and paired Student t-tests were used for inter- and intragroup comparisons, respectively. Chi-square tests were used for comparison of proportions. Kaplan-Meier curves were used to estimate the distribution of time to death. Differences between times to death curves were compared with the log-rank test. The Cox Hazard model was used to estimate the relative risk of mortality for each variable, including clinical and echocardiographic characteristics. Age was analyzed as a qualitative variable because of its varying clinical significance at different cutoff values, and MELD score was dichotomized at <25 and ≥25. Orthotopic liver transplantation was modeled as a time-varying covariate, because patients were transplanted at different times following their dobutamine stress RTMCE. The variables identified as significant by univariate analysis were used for the determination of independent predictors of mortality in the multivariate analysis. The results of univariate and multivariate analysis were expressed as relative risk with their respective 95% confidence intervals (CI). A P value of <0.05 was considered significant.
The mean age of our study population was 56 ± 7 yr. There were 136 (59%) men. Risk factors for CAD were diabetes mellitus in 58 (25%), systemic hypertension in 69 (30%), hyperlipidemia in 23 (10%), and cigarette smoking in 69 (30%) patients. Four (1.7%) patients had a history of a previous myocardial infarction, 2 (0.9%) patients had previous percutaneous coronary intervention, and 5 (2.1%) patients had previous coronary artery bypass grafting surgery. The etiology of liver disease was hepatitis C virus in 86 (37%), alcoholic cirrhosis in 39 (17%), cryptogenic cirrhosis in 29 (13%), nonalcoholic steatohepatitis in 12 (5%), primary biliary cirrhosis in 15 (6%), autoimmune hepatitis in 11 (5%), primary sclerosing cholangitis in 9 (4%), and other etiologies in 29 (13%). The median MELD score was 16 (range, 6 to 40); 47 (20%) patients had a MELD score of ≥25. Eighty-five (37%) patients underwent orthotopic liver transplantation at a median of 6 months (range, 9 days to 35 months) after RTMCE, and 145 (63%) remained under medical therapy.
Dobutamine Stress RTMCE
All patients had normal resting left ventricular systolic function, with a mean left ventricular ejection fraction of 65 ± 5%. The contrast agent Optison was used in 175 (76%) patients, with a mean cumulative dose of 2.6 ± 0.5 mL, and Definity was used in 55 (24%) patients, with a mean cumulative dose of 1.0 ± 0.4 mL. The mean maximal dose of dobutamine was 35 ± 9 μg/kg/minute. Atropine was injected in 198 (86%) patients with a mean cumulative dose of 0.8 ± 0.6 mg. Target heart rate was achieved in 205 (89%) patients. Chest pain occurred in 23 (10%) patients. One patient had a brief episode of supraventricular tachycardia, and 3 (1.3%) patients had atrial fibrillation during stress. Hypotension was observed in 49 (21%) patients. No death or myocardial infarction occurred during or immediately after the dobutamine stress tests. The stress echocardiogram was normal for MPI in 212 (92%) and abnormal in 18 (8%) patients. Wall motion abnormalities during stress were detected in only 5 patients (2%). All patients with abnormal wall motion also had abnormal MPI.
Heart rate increased from 75 ± 13 beats/minute at baseline to 146 ± 11 beats/minute at peak stress (P < 0.001), and the rate-pressure product increased from 9,474 ± 2,366 mm Hg/minute at baseline to 18,549 ± 4,997 mm Hg/minute at peak stress (P < 0.001). There was no significant change of mean systolic blood pressure from rest to peak stress (126 ± 23 mm Hg vs. 127 ± 32 mm Hg; P = 0.6), whereas diastolic blood pressure decreased (70 ± 11 mm Hg vs. 63 ± 15 mm Hg; P < 0.001) from rest to peak stress. At baseline, no differences were observed in the resting blood pressure or rate-pressure product between the patients with MELD scores of <25 and ≥25. At peak stress, patients with MELD scores of ≥25 had significantly lower levels of systolic blood pressure (118 ± 33 mm Hg vs. 129 ± 32 mm Hg; P = 0.03), diastolic blood pressure (58 ± 14 mm Hg vs. 65 ± 15 mm Hg; P = 0.005), and rate-pressure product (17,091 ± 5,244 mm Hg/minute vs. 18,910 ± 4,883 mm Hg/minute; P = 0.03) than patients with MELD scores of <25.
Quantitative Angiographic Findings
Seventeen patients underwent quantitative coronary angiography within 3 months of the dobutamine stress testing. CAD (> 50% stenosis) was detected in 7 (41%) of them. All patients with CAD at quantitative angiography had inducible perfusion defects, while only 2 had wall motion abnormalities during dobutamine stress RTMCE.
Outcome of Patients with ALD
Among the 85 patients who underwent orthotopic liver transplantation, 4 had abnormal MPI and 81 had normal perfusion. The hospital mortality rate was 50% (2/4) in patients with abnormal MPI and 2% (2/81) in patients with normal MPI (P = 0.01). Among patients with abnormal MPI, 1 died from hemorrhagic shock and 1 died from acute myocardial infarction in the anteroseptal wall and cardiogenic shock in the first postoperative day. In this latter patient, preoperative dobutamine stress RTMCE demonstrated a perfusion defect in the anteroapical region of the left ventricle. Quantitative coronary angiography revealed a 30% lesion in the left anterior descending coronary artery, and the patient was considered of low risk for liver transplantation. Among the 81 patients with normal MPI, 2 died in the postoperative period because of surgical complications.
During the median follow-up of 15 months (maximum, 46 months) after dobutamine stress RTMCE, 53 (23%) patients died. Two patients had cardiac and 51 had noncardiac death, mostly attributed to end-stage liver disease complications. Deaths occurred at a median of 5 months (range, 4 days to 26 months) after the stress test. The clinical and echocardiographic characteristics of patients who died are presented in Tables 1 and 2, respectively. The positive and negative predictive values for MPI to detect death were 44 and 79%, respectively.
|Variables||No death (n = 177)||Death (n = 53)|
|Age ≥65 years||16 (9%)||10 (19%)*|
|Male gender||108 (61%)||28 (53%)|
|Hypertension||51 (29%)||18 (34%)|
|Cigarette smoking||55 (31%)||14 (26%)|
|Hyperlipidemia||17 (10%)||6 (11%)|
|Diabetes mellitus||48 (27%)||10 (20%)|
|Etiology of liver disease|
|Alcoholic||29 (16%)||10 (19%)|
|Hepatitis C virus||69 (39%)||17 (32%)|
|Primary biliary cirrhosis||13 (7%)||2 (4%)|
|Primary sclerosing cholangitis||6 (4%)||3 (5%)|
|Cryptogenic cirrhosis||20 (11%)||9 (17%)|
|Autoimmune||9 (5%)||2 (4%)|
|Others||31 (18%)||10 (19%)|
|MELD score ≥25||27 (15%)||20 (38%)*|
|Variables||No death (n = 177)||Death (n = 53)|
|Left ventricular ejection fraction (%)||65 ± 5||64 ± 5|
|Dobutamine dose (μg/kg/minute)||35 ± 9||35 ± 9|
|Atropine dose (mg)||0.9 ± 0.6||0.7 ± 0.6|
|Abnormal wall motion||1 (0.6%)||4 (7.5%)*|
|Abnormal myocardial perfusion||10 (6%)||8 (15%)*|
|Heart rate (beats/minute)||75 ± 12||74 ± 14|
|Systolic blood pressure (mm Hg)||127 ± 22||122 ± 22|
|Diastolic blood pressure (mm Hg)||71 ± 11||69 ± 11|
|Rate-pressure product (mm Hg/minute)||9,588 ± 2,381||9,093 ± 2,295|
|Heart rate (beats/minute)||147 ± 10||144 ± 13|
|Systolic blood pressure (mm Hg)||128 ± 32||121 ± 31|
|Diastolic blood pressure (mm Hg)||64 ± 15||62 ± 16|
|Rate-pressure product (mm Hg/minute)||18,789 ± 4,998||17,724 ± 4,955|
|% Predicted maximal heart rate||89 ± 6||89 ± 8|
|Hypotension during stress||37 (21%)||12 (23%)|
|RR (95% CI)||P value||RR (95% CI)||P value|
|Age ≥65 years||2.1 (1.1–4.2)||0.03||2.2 (1.1–4.4)||0.03|
|Diabetes mellitus||0.7 (0.4–1.4)||0.34|
|Known CAD*||1.8 (0.5–5.7)||0.34|
|MELD score ≥25||3.0 (1.7–5.2)||0.0001||3.2 (1.8–5.5)||<0.0001|
|Liver transplantation||0.8 (0.4–1.7)||0.61|
|Abnormal wall motion||6.4 (2.3–17.9)||0.004|
|Abnormal MPI||2.4 (1.1–5.0)||0.02||2.4 (1.1–5.2)||0.02|
Figure 1 presents echocardiographic images of a 51-yr-old man with ALD secondary to hepatitis C virus and alcohol abuse. Dobutamine stress RTMCE revealed no inducible wall motion abnormalities, but there was an inducible perfusion defect in the apical and lateral wall. Although coronary angiography was indicated, the patient did not undergo the procedure and was not listed for orthotopic liver transplantation because of limited social support. The patient died 5 months after stress test.
By univariate analysis, MELD scores of ≥25, patient age, abnormal wall motion, and abnormal MPI were significantly associated with outcome. Figure 2 depicts the event-free survival curves according to the results of MPI, and Figure 3 according to the MELD score. Using the Cox multivariate model, a MELD score of ≥25, patient age of ≥65 yr, and abnormal MPI were all independent predictors of death (Table 3). After adjusting for MPI, the analysis of wall motion no longer had predictive value. The risk of death for patients with an abnormal MPI was 2.6 times the risk of death for patients with a normal MPI study, after adjusting for age and MELD score. The 2-yr event-free survival rate for patients with normal MPI by RTMCE was 76% (95% CI = 68–82%), while for patients with abnormal MPI it was 55% (95% CI = 30–74%).
The prevalence of CAD in patients with end-stage liver disease is estimated to range from 2.5 to 27%.5, 22, 23 Identification of patients at higher risk for cardiac events is an important step in determining the management strategy of these patients. Those with advanced CAD not amenable to interventional therapy or with poor cardiac function are not candidates for orthotopic liver transplantation, since patients with underlying cardiovascular disease have a worse perioperative outcome.4 Although the incidence of myocardial ischemic events is low, a complete cardiac evaluation including noninvasive stress testing is usually recommended, especially for patients with moderate risk for CAD. Dobutamine stress echocardiography6–8 and single-photon emission computed tomographic technetium-99m Sestamibi24–26 have been used for detecting significant CAD and evaluating perioperative risk in patients with ALD. However, the optimal screening method in this population has not been defined. Furthermore, data regarding the value of noninvasive stress testing in predicting outcome in this patient population are scarce.
Previous studies have yielded conflicting results regarding the value of wall motion responses to dobutamine in detecting CAD in patients with ALD. Plotkin et al.6 demonstrated that dobutamine stress echocardiography was highly efficacious in detecting CAD in patients with end-stage liver disease undergoing orthotopic liver transplantation. Donovan et al.8 studied 165 with ALD who underwent dobutamine stress echocardiography and found a sensitivity of 75% and specificity of 57% in detecting CAD. However, Williams et al.27 studied 61 patients undergoing liver transplantation and found that 56% of dobutamine stress tests were nondiagnostic because of inadequate heart rate response. The authors concluded that in patients with low to moderate risk of cardiac disease, dobutamine stress echocardiography was a poor predictor of major perioperative events. One possible explanation for the variable results with wall motion analysis is the reduced rate-pressure product achieved at peak dobutamine stress. We found that patients with ALD and higher MELD scores achieved a significantly lower rate-pressure product at peak stress. This lower level of demand stress may prevent the induction of transient wall motion abnormalities in the patients with underlying CAD.28, 29
Although the lower rate-pressure product response to dobutamine may reduce the sensitivity of wall motion analysis, perfusion imaging should still identify occult CAD in this setting.28, 29 However, only a few studies have assessed the value of myocardial perfusion in patients with ALD.24–26 Nonetheless, these studies have suggested that a perfusion assessment may be a more accurate method of detecting occult CAD in this patient population. Zoghbi et al.24 studied 339 patients with ALD under evaluation for liver transplantation. Eighty-seven (26%) patients underwent single-photon emission computed tomographic imaging; of them, 91% had normal perfusion. Patients with a negative perfusion study by single-photon emission computed tomography had a 1% perioperative cardiac event rate and were at low risk for cardiac events at 2-yr follow-up.
This is the first study to demonstrate that MPI during dobutamine stress echocardiography may be an independent predictor of mortality, along with the MELD score and age. Priority for liver transplantation is currently based on the MELD score, a mathematical function that includes bilirubin, creatinine, and international normalized ratio and is predictive of mortality.1, 30 The survival benefit of liver transplantation has been shown to depend on the severity of disease, in a way that survival benefit increased with increasing MELD score. The finding that abnormal MPI was still an independent predictor of mortality after consideration of the MELD score illustrates that patients with reduced myocardial blood flow reserve and ALD are an even higher risk patient population, with a corresponding worse prognosis. This was true even when angiography did not find a significant stenosis. Although a higher MELD score appears to indicate patients most likely to benefit from transplantation, patients with high MELD scores and abnormal MPI studies had a 50% perioperative mortality in our study. Conversely, a negative MPI study conferred both a low perioperative event rate (2%) and a significantly better 2-yr event-free survival.
A contrast defect with MPI indicates reduced coronary blood flow reserve. Although the reason for most contrast defects is occult CAD, microvasculature alterations may have contributed to the reduced blood flow reserve. Reduced blood flow reserve has been associated with worse outcomes in patients with hypertension, as well as in patients post cardiac transplantation, even in the absence of significant abnormalities at coronary arteriography.31, 32 Although the pathophysiological basis for why a reduced blood reserve contributes to increased mortality cannot be elucidated from this study, we postulate that the reduced myocardial blood flow reserve leads to cardiac output failure during periods of extensive metabolic demand that occur postoperatively. The reduced flow reserve may also indicate a stenosis more vulnerable to thrombosis following plaque rupture during the hypercoagulable, high catecholamine states that exist postoperatively. This is exemplified in the 4 patients with positive abnormal MPI studies that went on to liver transplantation, in which 3 had no significant (>50% diameter) narrowings at angiography. Two of these died postoperatively. The discrepancy between angiography and MPI could indicate the better potential for MPI with real-time echocardiography to detect patients who have reduced blood flow reserve from both epicardial and microvascular abnormalities. Since the number of patients in which death occurred is small, however, we will need larger multicenters to explore the role of the microvasculature in predicting patients who are at higher risk for complications.
The presence of intrapulmonary shunting with intravenous saline contrast has been shown to identify patients with ALD who will have a better outcome.33 In the present study, we only used an intravenous second-generation contrast agent that has excellent transpulmonary capillary passage. This produced opacification of left cardiac chambers and the analysis of myocardial perfusion in all patients, but could not identify those patients who had intrapulmonary shunting. Therefore, the independent influence of intrapulmonary shunting on the prognosis of these patients could not be determined.
Although we demonstrated that abnormal MPI may be more predictive of outcome than wall motion during dobutamine stress echocardiography in patients with ALD, multicenter studies are still necessary to confirm this.
The assessment of myocardial perfusion during dobutamine stress RTMCE appears to be a useful tool in predicting mortality in patients with ALD. A normal MPI is associated with a low perioperative mortality following orthotopic liver transplantation, and significantly better overall mortality at 2-yr follow-up. Conversely, abnormal MPI during dobutamine stress RTMCE is an independent predictor of mortality in patients with ALD. However, further studies are required.