- Top of page
- Materials and Methods
Patients with obstructive coronary artery disease (CAD) undergoing orthotopic liver transplantation (OLT) are at increased risk of poor outcomes. The accuracy of dobutamine stress echocardiography (DSE) to detect obstructive CAD is not well established in this population. We retrospectively identified patients with end-stage liver disease who underwent both DSE and coronary angiography as part of risk stratification prior to OLT. One hundred and five patients had both DSE and angiography, of whom 14 had known CAD and 27 failed to reach target heart rate during DSE. Among the remaining 64 patients (45 men; average age 61 ± 8 years) DSE had a low sensitivity (13%), high specificity (85%), low positive predictive value (PPV) (22%) and intermediate negative predictive value (NPV) (75%) for obstructive CAD. DSE as a screening test for obstructive CAD in OLT candidates has a poor sensitivity. The frequent chronotropic incompetence and low sensitivity in patients who achieve target heart rate, even in those with multiple cardiovascular disease risk factors, suggest that alternative or additional methods of risk stratification are necessary.
- Top of page
- Materials and Methods
Orthotropic liver transplantation (OLT) is becoming more frequent, most notably in patients over the age of 50 years (1). OLT in patients with coronary artery disease (CAD) is associated with significant morbidity and mortality. The presence of obstructive CAD for many years has been considered a contraindication to liver transplantation (2,3). Tiukinhoy-Laing et al. observed that in 161 potential OLT candidates >45 years old without known CAD, 26% had moderate-to-severe CAD on coronary angiography (4). Similarly, Carey et al. reported a 27% prevalence of CAD in potential OLT candidates >50 years of age (5).
An ideal preoperative screening modality for the detection of CAD in the end-stage liver disease population has yet to be determined. Davidson et al. reported that the predictive value of single-photon emission computed tomographic (SPECT) technetium-99 m sestamibi imaging with vasodilator stress in OLT candidates was poor (6). Limited studies of dobutamine stress echocardiography (DSE) as a screening tool in this population have yielded conflicting results (7–9).
The American College of Cardiology/American Heart Association (ACC/AHA) guidelines consider routine angiography not indicated (class III) in transplant candidates unless noninvasive testing reveals high risk for an adverse outcome (10). However, patients with cirrhosis and CAD without evidence of myocardial ischemia on noninvasive testing may also be at risk for acute coronary syndromes and postoperative heart failure. This may be due to the severe hemodynamic stress of transplant surgery combined with ventricular abnormalities due to the chronic hyperdynamic state often present in cirrhosis (11). Eimer et al. reported that 30% of patients developed acute heart failure after undergoing OLT despite a normal pretransplant DSE (12). Patients undergoing OLT are also subject to procoagulant and anticoagulant forces that can contribute to myocardial ischemia. In fact, nonobstructive CAD may result in myocardial infarction, during or soon after OLT (13,14). Rubin et al. reported that three of four patients who had evidence of postoperative myocardial ischemia and underwent coronary angiography had nonobstructive CAD (13). The authors concluded that a hypercoagulable state following OLT predisposes recipients to ischemic complications. Furthermore, structural heart abnormalities unrelated to CAD can also be present in cirrhotic patients that can impact their response to OLT. These include right ventricular enlargement and dysfunction, diastolic dysfunction, interatrial shunting (usually via a patent foramen ovale), pulmonary artery hypertension and left atrial enlargement (15,16).
We sought to evaluate the utility of DSE for the detection of obstructive CAD in a group of OLT candidates who were referred for coronary angiography.
Materials and Methods
- Top of page
- Materials and Methods
We identified 105 OLT candidates at Northwestern Memorial Hospital, who underwent both DSE and cardiac catheterization between January 2004 and June 2007. Patients underwent cardiac catheterization if the DSE was consistent with myocardial ischemia. In addition, cardiac catheterization was performed on patients still considered at high risk for CAD despite a DSE negative for ischemia based on the joint decision of all consulting services. High-risk patients were defined as having one or more of the following: (1) age ≥55 for men and ≥65 for women, (2) ≥2 cardiovascular risk factors, (3) echocardiographic evidence of ventricular enlargement or dysfunction, (4) symptoms suggestive of CAD including dyspnea or angina. The rationale for performing cardiac catheterization in these high-risk patients was to determine the presence, extent and severity of CAD in order to fully characterize their perioperative risk and to initiate appropriate therapy for CAD before considering OLT.
The primary study group consisted of 64 patients without a previous history of CAD who reached target heart rate (at least 85% of the maximum predictive heart rate) during the dobutamine infusion and therefore had an interpretable study. Patients who had a suboptimal heart rate response to dobutamine infusion (n = 27) were excluded from the primary study group as were those with a previous history of CAD (n = 14, all had undergone prior coronary artery bypass graft surgery). Institutional Review Board approval was obtained prior to review of medical records. Cardiovascular risk factors recorded were advanced age (men ≥ 55 years and women ≥ 65 years), tobacco use, diabetes mellitus, hypertension, hyperlipidemia and family history of premature CAD. Echocardiographic findings recorded were chamber sizes, presence of intracardiac and intrapulmonary shunting, diastolic dysfunction and ejection fraction.
DSE was performed using a previously described standard protocol (17). Intravenous dobutamine was infused incrementally at a dose of 5–10 μg/kg/min until a maximum dose (50 μg/kg/min) or study endpoint was achieved. The endpoints for dobutamine infusion termination included development of new wall motion abnormalities, attainment of 85% of age-predicted maximal heart rate or the development of significant adverse effects related to the dobutamine. Atropine was injected up to a maximum of 2.0 mg if a study endpoint was not achieved at the maximum dobutamine dose. Transthoracic echocardiographic images were obtained with commercially available ultrasound equipment. Four standard echocardiographic views were obtained with each acquisition: parasternal long axis, parasternal short axis, apical four-chamber and apical two-chamber views. Echocardiographic images were acquired at baseline, with each increment of dobutamine infusion and during the recovery phase. Any new or worsening dobutamine-induced wall motion abnormality seen was interpreted as a positive study. Studies were evaluated by board certified cardiologists trained in echocardiography not aware of cardiac catheterization results.
Cardiac catheterization was performed using standard Judkin's technique with multiple views. Obstructive CAD was defined, based on ACC/AHA consensus statements, as a visually estimated diameter stenosis of ≥ 70% in at least one of the three major coronary arteries in at least one view (18).
Results were expressed as mean (SD) unless otherwise specified. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated using the definitions of positivity as previously noted. Comparisons were made using chi-square tests for nominal variables. Unpaired t-tests were used for comparison of continuous variables. SPSS version 15.0 (SPSS, Inc., Chicago, IL) statistical software was used for analysis and p < 0.05 was considered significant for all analyses.
- Top of page
- Materials and Methods
We identified 105 OLT candidates who had both DSE and angiography (Figure 1). Fifty-three of the 105 patients identified were receiving β-blocking agents primarily for esophageal variceal bleeding prophylaxis. Thirty-one (58%) of those patients reached target heart rate. Of the 27 patients who did not achieve target heart rate, there were 22 (81%) who were taking β-blockers.
Figure 1. Patients included for analysis. DSE = dobutamine stress echocardiography; CABG = coronary artery bypass graft.
Download figure to PowerPoint
The 64 patients included in the primary study group had a mean age of 61 ± 8 years and 45 (70%) were men. Baseline characteristics are reported in Table 1. Viral infections accounted for 29 (45%) and alcohol for 12 (19%) patients in this end-stage liver disease population (Table 2). Echocardiographic findings are reported in Table 2. DSE was positive in 9 of the 64 patients (14%); angiography showed obstructive CAD in 16 (27%) (Table 3). The sensitivity of DSE was 13% and the specificity was 85%. Of these patients, 10 had 1-vessel disease, 3 had 2-vessel disease and 3 had 3-vessel disease (Table 4). Lowering the cutoff of a positive angiogram to a stenosis ≥ 50% had little impact on these results (Table 5).
Table 1. Baseline characteristics
|Characteristic/variable||Study population (n = 64)|
|Mean age (SD), y||61 (8) |
|Men, no. (%)||45 (70%)|
|Ethnicity, no. (%)|
| White||42 (66%)|
| Black||5 (8%)|
| Other||17 (27%)|
|Cardiovascular risk factors|
| Hypertension||29 (45%)|
| Hyperlipidemia||13 (20%)|
| Diabetes mellitus||20 (31%)|
| Tobacco history||19 (30%)|
| Family history of CAD||13 (20%)|
|Mean ejection fraction (SD),%||59.7 (6.1) |
|Mean pulmonary artery pressure (SD), mm Hg||25.3 (10.2) |
|Etiology of end-stage liver disease|
| Viral hepatitis, no. (%)||29 (45%)|
| Alcoholic, no. (%)||12 (19%)|
| NASH, no. (%)|| 7 (11%)|
| Cryptogenic, no. (%)||3 (5%)|
| Other, no. (%)||13 (20%)|
|Complications of end-stage liver disease|
| Hepatic encephalopathy, no. (%)||30 (47%)|
| Esophageal varices, no. (%)||40 (63%)|
| Spontaneous bacterial peritonitis, no. (%)||11 (17%)|
| Hepatocellular carcinoma, no. (%)||19 (30%)|
|MELD, mean (SD)||16.3 (6.6) |
| Hemoglobin, mg/dL, mean (SD)||11.6 (1.8) |
| Platelets, mg/dL, mean (SD)||107.0 (68.9) |
| INR, mean (SD)||1.5 (0.3) |
| Blood urea nitrogen, mg/dL, mean (SD)||17.5 (11.1) |
| Creatinine, mean (SD)|
| mg/dL||1.5 (1.7) |
| Range, mg/dL||0.3–10.0|
| Total cholesterol, mg/dL, mean (SD)||141.3 (47.4) |
| HDL, mg/dL, mean (SD)||42.1 (17.7) |
| LDL, mg/dL, mean (SD)||80.7 (37.6) |
| Triglycerides, mean (SD)||90.1 (60.4) |
Table 2. Echocardiographic findings of the primary study group
|Characteristic||Study population (n = 64)|
|Left atrial enlargement||24 (38%)|
|Left ventricular hypertrophy||14 (22%)|
|Right atrial enlargement|| 8 (13%)|
|Right ventricular enlargement||3 (5%)|
|Intracardiac shunt||13 (20%)|
|Intrapulmonary shunt||20 (31%)|
|Diastolic dysfunction||14 (22%)|
|Pulmonary hypertension|| 15 (23%)1|
|Mean ejection fraction % (SD)||59.7 (6.1) |
Table 3. Correlation between DSE and coronary angiography (≥ 70% stenosis)
|Positive No.||Negative No.|
|Positive|| 2|| 7|| 9|
Table 4. Severity of coronary artery disease by angiography¶
|Patient||Coronary artery territory**|
| 1|| ||(Mid), (Distal)||(Proximal)|
| 2|| ||(Distal)|| |
| 3||(Distal)||(Proximal), (Mid)||(PDA)|
| 4||(Distal)||(Proximal), (Mid)|| |
| 5||(Mid)|| ||(Mid)|
| 6||(Proximal)|| |
| 7||(Mid)|| |
| 8||(Mid)|| |
| 9|| ||(Proximal)|| |
| 10|| ||(Proximal)|| |
| 11|| ||(Proximal)|
| 12|| ||(Proximal), (Mid)|
| 13|| ||(Mid)|
| 14|| ||(PLA)|
Table 5. Correlation between DSE and coronary angiography (≥ 50% stenosis)
|Positive No.||Negative No.|
|Positive|| 4|| 5|| 9|
The number of cardiovascular risk factors had little influence on the diagnostic utility of DSE. In patients with ≤2 risk factors (n = 41) the sensitivity was 13% and the specificity was 79%. The PPV was 13% and the NPV was 79%. However, DSE yielded different results in patients with ≥3 risk factors (n = 23), with a sensitivity of 13%, specificity of 100%, PPV of 100% and NPV of 68%. Patients with positive angiograms had a greater average number of risk factors compared to those with negative angiograms (2.9 vs. 1.9, p = 0.001). Of patients with obstructive CAD, half had ≥3 risk factors. There were 20 diabetic patients (six with obstructive CAD) in whom there was a sensitivity of 17% and specificity of 94%.
The majority of patients in this population had thrombocytopenia and elevated prothrombin times (PT) and international normalized ratios (INR). Despite this, there were only three minor complications related to cardiac catheterization, which were hematomas at the site of arterial access. There were no major adverse events and no reports of contrast-induced nephropathy. Pre- and post-angiography cretinine values were 1.5 ± 1.7 mg/dL versus 1.6 ± 1.8 mg/dL, in all patients, and 1.9 ± 0.7 mg/dL versus 2.0 ± 0.9 mg/dL (p = NS) in patients not on dialysis with a baseline creatinine greater than 1.5 mg/dL.
Of the patients in the primary study group, 21 candidates received an OLT as of December 2007. These patients were selected to undergo OLT based on a minimal degree of coronary atherosclerosis. Thus only patients with nonobstructive CAD (<70% lesions, 20 patients), or single vessel CAD (one patient) were transplanted. Two patients with obstructive CAD were revascularized, but have not yet been transplanted. Adverse postoperative cardiac outcomes were defined as any cardiac event within 30 days or cardiovascular-associated mortality within 1 year. Only one patient had an adverse cardiac outcome. This patient developed postoperative atrial fibrillation, myocardial ischemia and right heart failure and subsequently died 3 months after surgery. The preoperative findings in this patient were a positive DSE but negative coronary angiogram for the presence of CAD. Of the remaining 20 patients, there were 2 false-positive DSE, 1 false negative and 17 true negatives.
- Top of page
- Materials and Methods
In a group of OLT candidates at high risk for CAD, with the use of coronary angiography, we found that DSE has very poor sensitivity for the detection of obstructive CAD. When evaluating unselected patients for ischemic heart disease, DSE and SPECT are highly sensitive (80% and 86%, respectively) and specific (84% and 77%, respectively) when validated by coronary angiography (19). However, very few studies have included patients who all underwent cardiac catheterization. Candidates for OLT were once thought to be at low risk for cardiovascular complications based on the effects of liver failure, including arterial vasodilatation, low levels of LDL, and a relatively anticoagulated state (20). However, recent data indicate a greater prevalence of moderate-to-severe CAD in OLT candidates (4,5,21). This trend reflects the recognition of liver disease in an older patient population, the role of obesity in the development of cirrhosis and the role of nonalcoholic fatty liver disease, all resulting in a population with more cardiovascular risk factors now being considered for OLT (22,23).
Plotkin et al. reported that the mortality at 1 year after OLT was 40% in patients with known CAD. More than half of these patients succumbed to a perioperative cardiovascular related death. At 3 years, the mortality was 50%, and >80% of the survivors had significant morbidity (2). This greatly exceeds the accepted 25% mortality at 5 years, which has been demonstrated in other transplant series (24,25). Patients with CAD who are managed medically with or without revascularization have no differences in outcomes compared to those without CAD (26). An accurate preoperative risk assessment is necessary to identify which OLT candidates have underlying CAD and hence are at an increased risk of suffering perioperative or postoperative cardiovascular events.
The utility of other noninvasive testing for the detection of significant CAD in patients with liver disease, such as SPECT, has been addressed in several studies (6,27,28). It has been suggested that patients with cirrhosis have minimal coronary reserve and achieve little or no change in myocardial perfusion after an exogenous vasodilator such as adenosine is given, which may account for the limited utility of this modality (6). Zoghbi et al. examined SPECT imaging results in 87 subjects who underwent OLT (28). In this group, only 17% had >1 cardiovascular risk factor and only 8% demonstrated myocardial ischemia on SPECT imaging. Not surprisingly, SPECT imaging in this low-risk group was associated with a high NPV (99%) but a poor PPV for perioperative cardiac events. The two patients found to have obstructive CAD by angiography in this group underwent successful coronary revascularization to facilitate their liver transplant. This study does not comment on the patients undoubtedly denied OLT due to abnormal cardiovascular testing.
As a stress agent, dobutamine affords a hemodynamic target (heart rate) that is not available with other pharmacologic stress agents. It has been reported that factors associated with chronotropic incompetence (inability to meet target heart rate) include male gender, tobacco history and lower resting systolic blood pressure (29). Thus the low systolic blood pressure frequently present in OLT candidates may contribute to an inability to meet target heart rate. In addition, many patients with liver failure, including 50% of those in our study, are treated with β-blockers, which may reduce the sensitivity of DSE by decreasing heart rate responses (30). The clearance of β-blockers is reduced in liver disease and their activity may be present for several days after discontinuation. The sensitivity of DSE increases as the maximal heart rate achieved increases (31). In our population, 26% of patients did not meet target heart rate, the majority of whom (81%) were on β-blockers. Therefore, the common use of β-blocking agents in this population may make DSE an impractical screening tool.
Some studies have suggested that a coronary artery stenosis ≥50% be used as a cutoff for obstructive CAD in this population (30,32). However, our data show that DSE does not perform better when this cutoff is employed. Based on the high morbidity and mortality seen in transplanted patients with CAD, these values indicate that DSE may be of limited use as a screening test for obstructive CAD in this population.
There are several limitations to our study. Although this is the largest sample size directly comparing DSE to angiography, this is a retrospective study. There is an inherent selection bias in terms of which patients were selected to undergo angiography. However, one would expect that including patients with abnormal DSE or significant CAD risk factors that we would be more likely to discover and more accurately assess CAD. Therefore, the pretest probability in selected patients is higher and posttest probability after a normal DSE was still high given the significant cardiovascular risk factors present in this population. Patients with known CAD were excluded from the study. However, the goal of this study was to identify patients without known CAD. Ultimately, the fact remains that a normal DSE does not indicate absence of obstructive CAD.
It is possible that prognosis and outcomes are related not only to the severity of CAD, but also to the presence or absence of ischemia on DSE. It is also plausible that the lack of ischemia on DSE, irrespective of the degree of CAD, may confer a good prognosis. However, the issue of outcomes related to the study could not be thoroughly addressed as a result of the relatively small sample size. In the setting of severe CAD, management strategies are complex, and patients are often no longer considered for OLT due to the perceived high operative risk. In a small cohort study, Axelrod et al. demonstrated the safety and efficacy of combined OLT and coronary artery bypass grafting. Four of five patients were alive and doing well at 1 year and one patient was noted to be alive at 4 years (33).
In conclusion, an increasing number of patients are undergoing OLT. CAD is relatively common in this population. OLT in patients with CAD is associated with poor outcomes. The knowledge of the presence of obstructive CAD, not only ischemia, may lead to the modification of medical therapies or revascularization, which can lead to a more stable perioperative patient. Alternatively, patients found to be at prohibitive operative risk may be declined for OLT. DSE, like SPECT, appears to have a relatively low sensitivity for detecting significant obstructive CAD in this population. However, stress echocardiography with doppler can provide important structural and functional (i.e. ejection fraction, pulmonary artery pressure, ischemia) information, which may have prognostic implications for OLT candidates. Given the relatively small sample size of this study, our results will need to be confirmed, not only with a larger number of patients, but also at other centers. In addition, the role for other novel imaging modalities for CAD detection, such as cardiac CT angiography, will need to be explored. Our data demonstrate that coronary angiography can be performed safely in this relatively coagulopathic population. Accordingly, at this time there is no reliable screening method for obstructive CAD, thus coronary angiography should be considered in addition to noninvasive imaging in OLT candidates at high risk for CAD to more completely assess their perioperative risk.