Screening for coronary artery disease is common practice in the evaluation of liver transplantation candidates. However, it is unclear whether coronary screening influences transplantation eligibility. We sought to determine the association between screening stress myocardial perfusion imaging (MPI) results and the eligibility for liver transplantation. Within a retrospective cohort of liver transplantation candidates referred for screening stress MPI at a single institution from April 1998 to February 2004, we obtained the baseline characteristics, stress MPI results, transplantation eligibility, and transplantation denial criteria by chart review. Of 294 patients (39%) denied transplantation, the denial criteria were multifactorial for 91 (31%) of the candidates. Compared to candidates with low-risk stress MPI results, the odds of being denied transplantation were the same for candidates with intermediate-risk MPI results (odds ratio 0.93, 95% confidence interval 0.45 to 1.82) or high-risk MPI results (odds ratio 1.42, 95% confidence interval 0.54 to 3.73). This lack of association persisted in our analysis with additional stratification of stress MPI results into negative, positive–low-risk, positive–intermediate-risk, and positive–high-risk. In conclusion, the screening stress MPI results were not associated with liver transplantation eligibility. The large number of competing factors considered before transplantation listing and the low proportion of positive stress MPI results suggests that targeting screening to patients deemed otherwise acceptable for transplantation might increase the influence of stress MPI findings on transplantation eligibility.
Orthotopic liver transplantation (OLT) is an established therapy for fulminant and end-stage liver disease. In 2005, approximately 6,500 liver transplantations were performed in the United States, and the number has been increasing steadily for 15 years. The number of patients listed for liver transplantation was >15,000 in 2007. Previous studies have suggested that coronary artery disease (CAD) is associated with increased morbidity and mortality after transplantation. As a result, the current guidelines recommend CAD screening for patients with one or more traditional risk factors for CAD. However, >90% of CAD screening tests have negative findings in OLT candidates referred for preoperative evaluation. Additionally, the influence of CAD screening might be diluted by the large number of competing factors considered before listing a patient for transplantation.
Methods
We reviewed the data from 772 consecutive adult liver transplantation candidates at the University of Washington Medical Center (April 1998 to February 2004) who had been referred for stress myocardial perfusion imaging (MPI). The candidates referred for stress MPI evaluation were ≥40 years old and/or had one or more traditional risk factors for CAD (ie, hypertension, hyperlipidemia, diabetes mellitus, smoking, or family history). Nine of these patients had incomplete stress MPI and were excluded from the present analysis. The remaining 763 patients’ medical charts were reviewed. The institutional review board at the University of Washington approved the study.
Myocardial perfusion studies were performed using standard techniques with dual-isotope thallium-201/technetium-99m tetrofosmin or 2-day technetium-99m tetrofosmin. Stress MPI results were interpreted using the report in the medical record according to traditional criteria for noninvasive CAD testing. Low-risk MPI findings included studies read as normal or with a small myocardial perfusion defect at rest or with stress, intermediate-risk findings included those with a stress-induced moderate perfusion defect without left ventricular dilation or increased lung intake in the setting of thallium-201, and high-risk studies included the findings of a stress-induced large perfusion defect, stress-induced multiple perfusion defects of moderate size, or a large fixed perfusion defect with left ventricular dilation or increased lung uptake in the setting of thallium-201. Of the 763 patients, 132 had been on the transplantation waiting list for >1 year and had undergone repeat stress MPI according to the local protocol. In determining the effect of stress MPI results on OLT eligibility, we limited our analysis to the index evaluation.
The covariates of interest included a personal history of CAD (ie, history of CAD, myocardial infarction, coronary artery bypass grafting, percutaneous coronary intervention, or angina) and the risk factors for CAD (ie, a history of hypertension, diabetes mellitus, peripheral or cerebral vascular disease, smoking, family history of CAD, or a laboratory assessment of cholesterol ≥200 mg/dl at the index evaluation).
After completion of the pretransplantation evaluation, a letter documenting the OLT eligibility decision of the liver transplantation review committee was recorded in the medical record for all OLT candidates. This letter indicated the specific reasons (eg, ongoing alcohol use, illicit drug use, coronary disease) for the denial of transplantation. These individual denial criteria were identified from the transplantation review committee documentation of all candidates denied transplantation. In the patients who underwent liver transplantation, the post-transplantation events, including myocardial infarction, revascularization, or death, as documented in the medical record were collected through March 2006.
The baseline patient characteristics were summarized using categorical, parametric, or nonparametric descriptors, as appropriate. Differences in the baseline characteristics were evaluated using the one-way analysis of variance for continuous variables and chi-square tests or Fischer’s exact tests, as appropriate, for the categorical variables.
We evaluated the relation between stress MPI results and transplantation eligibility using chi-square tests in the unadjusted analysis and multiple logistic regression analysis in the fully adjusted model. Our adjusted model included covariates for age, gender, history of CAD, hypertension, diabetes, smoking, family history of CAD, cholesterol ≥200 mg/dl, and peripheral vascular or cerebral vascular disease. All covariates were included as categorical variables, with the exception of age, which was included as a continuous variable. Instead of ignoring missing data using complete-case analysis, we conservatively chose to impute the missing covariate values with chained equations. This method provides less biased estimates than a complete-case analysis. We used 10 imputation sets with the covariates and with bootstrap estimation of the regression coefficients for our regression analysis. In addition, a complete-case analysis was completed to evaluate the sensitivity of our results to the imputation method. All statistical analyses were conducted using Stata/IC, version 10.0 (StatCorp, College Station, Texas).
Results
Most (92%) patients underwent pharmacologic stress testing with dipyridamole. A small number of patients underwent an exercise stress test or pharmacologic stress test with dobutamine or adenosine. At the index evaluation, the MPI study findings were low risk in 710 patients (93%), intermediate risk in 36 patients (5%), and high risk in 17 patients (2%). The baseline patient demographics and characteristics are listed in Table 1 . The patients with high-risk stress MPI findings were older, more likely to have a history of CAD, hypertension, or congestive heart failure, or a family history of CAD. After stress MPI, coronary angiography was performed in 46 patients (7%) with low-risk MPI findings, 18 (50%) with intermediate-risk MPI findings, and 8 (47%) with high-risk MPI findings.
| Characteristic | Low-Risk MPI (n = 710) | Intermediate-Risk MPI (n = 36) | High-Risk MPI (n = 17) | p Value |
|---|---|---|---|---|
| Age (years) | 53 ± 7 | 55 ± 8 | 58 ± 8 | 0.04 |
| Model for end-stage liver disease score | 13.3 ± 4.8 | 14.0 ± 6.9 | 12.0 ± 4.9 | 0.39 |
| Men | 441 (62%) | 28 (78%) | 9 (53%) | 0.12 |
| History of CAD ⁎ | 13 (2%) | 2 (6%) | 6 (35%) | <0.01 |
| Diabetes mellitus | 172 (24%) | 12 (33%) | 7 (41%) | 0.12 |
| Hypertension | 128 (18%) | 13 (36%) | 6 (35%) | <0.01 |
| Smoker † | 443 (62%) | 22 (61%) | 10 (59%) | 0.93 |
| Family history of CAD ‡ | 120 (28%) | 10 (48%) | 6 (46%) | 0.06 |
| Total cholesterol ≥200 mg/dl § | 129 (20%) | 4 (14%) | 6 (40%) | 0.13 |
| Peripheral vascular disease or cerebral vascular disease | 8 (1%) | 2 (6%) | 0 (0%) | 0.15 |
| Congestive heart failure | 2 (<1%) | 1 (3%) | 1 (6%) | 0.02 |
⁎ Defined as documentation in the medical record of any of the following: history of CAD, myocardial infarction, coronary artery bypass grafting, percutaneous coronary intervention, or angina.
A total of 294 patients (39%) were denied transplantation after the index clinical assessment. The criteria for denial of transplantation were multifactorial in nearly 1/3 of the candidates. The most common components of denial criteria were psychosocial factors, active tobacco use, or illicit drug use ( Table 2 ). Of those denied transplantation, the criteria included CAD in 19 candidates (7%), and 6 (32%) of these patients had been known to have CAD before stress MPI. Coronary disease was the only criteria for denial in 7 candidates (2%), and 3 (43%) of these candidates had a history of CAD.
| Denial Criteria | Denied Transplantation (n = 292) |
|---|---|
| Psychosocial | 78 (27%) |
| Active tobacco use | 51 (17%) |
| Illicit drug use | 30 (10%) |
| Preserved hepatic function | 28 (10%) |
| Hepatocellular carcinoma | 26 (9%) |
| Alcohol use | 19 (7%) |
| Coronary disease | 19 (7%) |
| Obesity | 18 (6%) |
| Death before evaluation completed | 14 (5%) |
| Transferred care | 13 (4%) |
| Malignancy (nonhepatocellular carcinoma) | 10 (3%) |
| Other medical reason | 35 (12%) |
The liver transplantation eligibility status by MPI result is listed in Table 3 . In patients with low-risk stress MPI findings, 273 (39%) were denied transplantation compared to 13 (36%) with intermediate-risk stress MPI results and 8 (47%) with high-risk stress MPI results. Stress MPI results were not associated with transplantation eligibility in either the crude or adjusted analyses. Compared to the candidates with low-risk stress MPI results, the odds of being denied transplantation were the same for patients with intermediate-risk MPI or high-risk MPI results in the unadjusted and adjusted analyses. Similar results were seen in our analysis that was restricted to patients with complete ascertainment of covariates. CAD was the only criterion for the denial of transplantation in 3 candidates (1%) with low-risk MPI findings, 3 (25%) with intermediate-risk MPI findings, and 1 (13%) with high-risk MPI findings.
