Cardiac Disease Evaluation and Management Among Kidney and Liver Transplantation Candidates: A Scientific Statement From the American Heart Association and the American College of Cardiology Foundation FREE

The challenges inherent in conducting accurate, clinically effective, and cost-effective cardiac evaluations among transplantation candidates relate to the large size of the target population, the prevalence of disease, the limited number of donated organs, and the often extended waiting periods between initial evaluation and transplantation surgery. According to Organ Procurement and Transplant Network (OPTN) records, nearly 85,000 candidates were on the waiting list for kidney transplantation in 2010, and ≈17,700 kidney transplantations (including 828 kidney-pancreas transplantations) were performed (1). Also in 2010, 16,000 people were awaiting liver transplantation and 6,000 received liver allografts (1). Marked shifts in the age composition of transplant waitlists toward older adults are also raising the average medical complexity and comorbidity burden among listed candidates. In 2011, 62% of kidney transplantation candidates were ≥50 years of age compared with 28.7% of kidney transplantation candidates in 1991 (1). A similar shift in age distribution has occurred among liver transplantation candidates; now, 77% are ≥50 years of age (1). Cardiovascular disease is a leading cause of morbidity and mortality among patients with end-stage failure of noncardiac organs before and after transplantation. Estimates of the cumulative incidence of myocardial infarction (MI) based on Medicare billing claims have ranged from 8.7% to 16.7% by 3 years after kidney transplant listing and from 4.7% to 11.1% after kidney transplantation (2). Observational data suggest particularly high frequencies of cardiovascular events in the first months after kidney transplantation (2). Cardiovascular diseases in aggregate make up the most common cause of death in patients with functioning allografts at all times after kidney transplantation, accounting for 30% of mortality overall, with highest rates in the peritransplantation period (6).

Guidelines and position papers by national organizations can serve as useful tools for informing cardiac evaluation practices before noncardiac surgery. However, the discrepancies among existing guidelines and the unique clinical characteristics of patients with end-stage organ failure raise questions about the applicability of available recommendations to transplantation candidates. In 2007, the American College of Cardiology (ACC) and American Heart Association (AHA) issued their most recent version of the “Guidelines on Perioperative Cardiovascular Evaluation and Care for Noncardiac Surgery” (7). The algorithm suggests consideration of further cardiac evaluation in symptomatic patients but does not encourage further testing for patients who have no cardiac symptoms with a functional capacity of ≥4 metabolic equivalent tasks (METS; i.e, ability to climb a flight of stairs), regardless of diabetic status, history of coronary artery disease (CAD), or other traditional cardiac risk factors. Consideration of noninvasive testing was given a Class IIb recommendation in asymptomatic patients with at least 1 to 2 clinical risk markers and poor functional capacity who require intermediate-risk noncardiac surgery if it will change management (Level of Evidence B), with the evidence grade reflecting lack of large randomized trials to support this strategy (Table 1). Similarly, the “2009 Appropriate Use Criteria for Cardiac Radionuclide Imaging” (8) by the American College of Cardiology Foundation (ACCF), along with key specialty and subspecialty societies, deemed that radionuclide imaging is appropriate for perioperative evaluation before intermediate-risk noncardiac surgery in asymptomatic patients only when at least 1 risk marker in present and the patient has poor (<4 METS) or unknown functional capacity (8). The ACC/AHA and ACCF guidelines were not written specifically for patients with end-stage organ failure, and the predictive value of the “absence of cardiac symptoms” may differ in transplantation candidates compared with the general population. These guidelines also take the perspective that noncardiac surgery will be performed shortly after the evaluation and that any management decisions will affect short-term (perioperative) outcomes. In contrast, cardiac evaluation and interventions in transplantation candidates should be considered from both the short- and long-term perspective.

The fundamental basis of the latest ACC/AHA recommendations is grounded in understanding of the role of coronary revascularization before noncardiac surgery. The authors state, “Patients with asymptomatic ischemia…do not appear to be candidates for prophylactic preoperative coronary revascularization unless cardiac catheterization reveals high-risk surgical anatomy” (7). This statement is supported by 2 recent randomized trials that did not demonstrate benefit of percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG) for revascularization of asymptomatic CAD before major vascular surgery (9).

In 2005, the National Kidney Foundation published the “Clinical Practice Guidelines for Cardiovascular Disease in Dialysis Patients” within the Kidney Disease Outcomes Quality Initiative (NKF/KDOQI) (12). The section on CAD suggests more aggressive screening of patients with end-stage renal disease (ESRD) as part of the evaluation to determine candidacy for renal transplantation than would be suggested by ACC/AHA guidelines, although the statements were rated Level of Evidence C, this is, based on either weak evidence or the opinions of the working group. For example, this algorithm recommends that any patient on the kidney transplant waitlist with a history of diabetes mellitus or known CAD undergo noninvasive stress testing at baseline and then subsequently every 12 months until transplantation. There is a similar recommendation for transplantation candidates deemed at high risk per Framingham criteria (≥2 traditional risk factors, left ventricular ejection fraction [LVEF] ≤40%, or peripheral vascular disease) (12).

Other consensus-based recommendations for cardiac risk assessment before kidney transplantation have been offered. These include a 2007 report from an international collaboration of the NKF and the Transplantation Society called the Lisbon Conference (13), the 2001 American Society of Transplantation (AST) guidelines (14), and the 2000 European Renal Association-European Dialysis Transplant Association (ERA-EDTA) “European Best Practice Guidelines” (15). These 2 clinical practice guidelines are now >10 years old, were based on expert consensus panels, and were not the product of systematic review of the evidence. Although all these documents suggest that symptomatic patients should undergo further testing, they offer differing recommendations for asymptomatic patients. A comparison of these 5 clinical documents and their recommendations on testing asymptomatic patients for CAD before renal transplantation is summarized in (Table 2).

Several studies document heterogeneity in cardiac evaluation practices before kidney transplantation at the national level (Table 3). In a 1993 survey of directors of OPTN-participating centers, noninvasive stress testing was reported as the most common first approach to cardiac evaluation of asymptomatic patients, prompted by diabetes mellitus at 86% of responding centers, age (mean threshold, 52 years) at 67%, and multiple risk factors at 68% (18). Some centers used routine coronary angiography for patients with diabetes mellitus (15%), older age (7%; mean threshold, 57 years), or multiple risk factors (8%). A subsequent survey of OPTN centers about policies for patients on the deceased donor waiting list found that 8% of programs reported cardiac testing for all candidates, whereas 18% did not routinely order cardiac testing for any asymptomatic patient group; 59% screened patients with diabetes mellitus, 52% screened patients with a history of CAD, and 52% screened patients deemed to be high risk for cardiac events after transplantation given their age or obesity (19). Methods of screening were also variable: 40% pharmacological-nuclear, 33% exercise nuclear, 31% dobutamine stress echocardiography (DSE), and 15% cardiac catheterization. Cardiac surveillance policies among listed candidates also differ across centers. In a survey of 68 centers in 2005, 51% of program representatives indicated reliance on the initial cardiac evaluation and cardiac history, 7% used ACC/AHA criteria for noncardiac surgery in the general population to guide cardiac revaluation, and 32% applied a combination of ACC/AHA criteria, the initial cardiac evaluation, and cardiac history (20).

Survey responses are limited by nonresponse rates, and reported policies may differ from actual practices. A retrospective study of the U.S. Renal Data System (USRDS) registry used billing claims as measures of cardiac evaluation services in Medicare beneficiaries transplanted in 1991 to 2004 (21). Forty-six percent of the sample received noninvasive stress testing or angiography at some time before transplantation (65% of high risk, defined as diabetes mellitus, prior ischemic heart disease, or ≥2 other coronary risk factors, and 20% of lower risk). There was substantial heterogeneity in cardiac evaluation frequency according to patient-level factors even within risk groups. After adjustment for patient traits and consistent within risk profile–stratified samples, transplantation without cardiac evaluation was more likely for black people, women, and patients in certain geographic regions. Race-related practice variables were notable because in the lower-risk group transplanted without cardiac evaluation, black patients faced higher risks of post-transplantation MI than nonblack patients. Black race has previously been identified as an independent predictor of failure to complete the pretransplantation evaluation (22) and of reduced access to coronary angiography and revascularization in populations without kidney disease (23). Thus, some of the observed practice variation may reflect access barriers rather than appropriate determinations of low clinical risk. An important limitation of studies based on billing data, however, is that claims may not distinguish screening tests from tests performed because of cardiac symptoms.

A recent single-center study examined the approach to the asymptomatic kidney transplantation candidate, which defines the majority of potential candidates at the time of referral (24). In this study, medical charts were reviewed to quantify the hypothetical frequency of recommended testing per practice recommendations and the observed results of cardiac testing (primarily DSE or myocardial perfusion scintigraphy [MPS]) among 204 consecutive patients who were determined to be free of an active cardiac condition by a cardiologist at the time of transplantation evaluation. Active cardiac conditions were defined according to the ACC/AHA definition including significant valvular disease, decompensated heart failure, significant arrhythmias, and unstable coronary syndromes. If followed precisely, the ACC/AHA guidelines recommended testing in only 20% of patients, whereas the KDOQI guidelines would have resulted in 100% of patients being tested. Among the 178 patients who underwent stress tests, the prevalence of ischemia was similar among those for whom testing was and was not recommended per the ACC/AHA guidelines, 10.3% versus 9.4%, respectively. The relatively low use of coronary revascularization after pretransplantation cardiac evaluation also raises concern for the clinical and cost-effectiveness of pretransplantation cardiac evaluation as applied. Several registry-based and single-center observational studies have found that only 2.9% to 9.5% of patients who receive pretransplantation cardiac stress testing or angiography proceeded to angioplasty or surgical bypass (21).

Given the variation between practice and prior guidelines in patients being evaluated for solid-organ transplantation, it is important to determine whether evidence can resolve the basis of this difference. The role for stress testing in the absence of symptoms has been called into question among patients undergoing noncardiac surgery because randomized studies of coronary revascularization before vascular surgery have failed to show a consistent benefit (9). Outside the perioperative setting, PCI has failed to demonstrate benefit for the risk of major adverse cardiovascular events (MACEs) in a randomized trial among stable patients with CAD (29), including a subgroup with chronic kidney disease (CKD) at trial enrollment (30). Evidence suggests that the only asymptomatic patients in whom coronary revascularization may be helpful are the minority found to have occult high-risk coronary anatomy such as significant left main disease or severe proximal 3-vessel disease, especially in the presence of reduced left ventricular systolic function (31). However, noninvasive cardiac testing of transplantation candidates might yield findings that call into question the appropriateness of transplantation or identify high-risk coronary lesions associated with long-term benefit from revascularization. This report evaluates the state of evidence regarding cardiac risk evaluation and management in kidney transplantation and liver transplantation candidates, considering data specific to these populations and the appropriateness of extrapolations when data from these populations are lacking. This article focuses on cardiac disease; issues related specifically to the evaluation of carotid or peripheral vascular disease are beyond the scope of this document.

The AHA Writing Committee on Cardiac Disease Evaluation and Management Among Kidney and Liver Transplantation Candidates conducted a comprehensive review of the literature relevant to perioperative cardiac evaluation of potential kidney or liver transplant recipients, including the prevalence of CAD in these populations; incidence of MACEs before and after transplantation; accuracy of clinical risk markers, symptoms, and noninvasive testing before and after transplant listing for detecting active cardiac conditions and CAD; and clinical outcomes of revascularization and the medical management of atherosclerosis. Each section was assigned to a lead author and coauthor. Literature searches were conducted in the following databases: PubMed, MEDLINE, and the Cochrane Library (including the Cochrane Database of Systematic Reviews and the Cochrane Controlled Trials Register). Searches were limited to the English language, the years 1990 through March 2010, and human subjects. Related-article searches were conducted in MEDLINE to find additional relevant articles. Finally, committee members recommended applicable articles outside the scope of the formal searches. Interval drafts were discussed during conference calls and 2 face-to-face meetings. Recommendations included an evaluation of the strength of the evidence for or against a particular procedure or treatment in terms of the magnitude of effect (evidence class) and estimate of certainty (evidence level) (Table 1). Recommendations were subjected to formal, anonymous voting. The volume of text devoted to cardiac evaluation and management issues for kidney and liver transplantation candidates, respectively, reflects the relative sizes of the target populations; the number of patients awaiting and receiving kidney transplants is >4 times the number of patients awaiting and receiving liver allografts. Correspondingly, a substantially larger number of publications with liver transplantation candidates.

The most compelling goal of preoperative cardiac risk evaluation is to reduce the morbidity and mortality of cardiovascular disease. Any test used to screen a population is associated with false-positive and -negative results that may diminish utility. False-positive results in particular may lead to patient and physician anxiety and the possibility of additional and often unnecessary testing or invasive procedures. Screening asymptomatic patients should be used only if the benefits of screening outweigh the harms. In asymptomatic patients, screening for CAD would be of value if the results of testing lead to management changes that reduce the occurrence of patient-level outcomes. Screening should also be cost-effective. For organ transplantation, cardiac evaluation could also be used to deny transplantation to high-risk patients, provided that it can be shown that patients with severe cardiovascular disease have sufficiently short life expectancy to make transplantation a poor use of scarce donated organs. However, studies have shown that survival is generally improved by transplantation compared with remaining on the transplant waiting list, even among high-risk patients (32). Thus, the burden of proof in using screening to determine transplantation candidacy from a patient-centered perspective is to demonstrate that denying transplantation on the basis of test results is in the best interest of the patient. Alternatively, society may decide that cardiovascular evaluation results can help in guiding allocation of organs to recipients who are most likely to benefit in the long term. However, nationally agreed-on allocation priorities of the OPTN use waiting time and a liver failure severity metric (Model for End-Stage Liver Disease score) as the dominant criteria for achieving fairness in kidney and liver allocation, respectively; allocation schemes that seek to maximize net societal benefit from donated allografts have not been adopted. Although prognostic information from noninvasive cardiac testing of asymptomatic patients may be useful for adjusting center performance metrics such as post-transplantation mortality for the “case mix” of each center's recipients, this approach would incur substantial expense compared with the use of information available from the history and physical examination.

Some patients have been shown to undergo renal transplantation safely despite clinical markers of high cardiovascular risk. Jeloka et al. (35) retrospectively classified 429 renal transplant recipients as high cardiovascular risk (n=61) and low cardiovascular risk (n=368). The high-risk group included patients with a history of angina, MI, or significant CAD found on cardiac catheterization. Outcomes of interest were post-transplantation cardiac events (MI, angina, new arrhythmias, heart failure, and/or sudden cardiac death) and overall survival. The distribution of events among the high-risk and low-risk groups was 31.3% versus 6.5%, respectively (p=0.001). Five-year survival in the high-risk group was 82.8% compared with 93.1% in the low-risk group (p=0.004). Among the subgroup who underwent coronary revascularization before transplantation (n=28; 25% PCI, 75% CABG), 43% subsequently experienced a cardiac event. The authors contended that in selected high-risk patients, overall 5-year survival after renal transplantation was actually quite good, superior to the expected 5-year survival with continued dialysis.