Author + information
- Received November 26, 2013
- Revision received February 5, 2014
- Accepted March 4, 2014
- Published online July 1, 2014.
- ∗Division of Cardiovascular Diseases, Department of Medicine, Mayo Clinic Rochester, Rochester, Minnesota
- †Division of Cardiology, Department of Internal Medicine, Gyeongsang National University Hospital, Jinju, Korea
- ‡Department of Cardiology, Institute of Clinical and Experimental Medicine–IKEM, Prague, Czech Republic
- ↵∗Reprint requests and correspondence:
Dr. Barry A. Borlaug, Division of Cardiovascular Diseases, Department of Medicine, Mayo Clinic Rochester, Mayo Clinic and Foundation, 200 First Street SW, Rochester, Minnesota 55905.
Objectives This study investigated the characteristics, evaluation, prognostic impact, and treatment of coronary artery disease (CAD) in patients with heart failure and preserved ejection fraction (HFpEF).
Background CAD is common in patients with HFpEF, but it remains unclear how CAD should be categorized, evaluated for, and treated in HFpEF.
Methods Clinical, hemodynamic, echocardiographic, treatment, and outcome characteristics were examined in consecutive patients with previous HFpEF hospitalizations who underwent coronary angiography.
Results Of the 376 HFpEF patients examined, 255 (68%) had angiographically-proven CAD. Compared with HFpEF patients without CAD, patients with CAD were more likely to be men, to have CAD risk factors, and to be treated with anti-ischemic medications. However, symptoms of angina and heart failure were similar in patients with and without CAD, as were measures of cardiovascular structure, function, and hemodynamics. Compared with patients without CAD, HFpEF patients with CAD displayed greater deterioration in ejection fraction and increased mortality, independent of other predictors (hazard ratio: 1.71, 95% confidence interval: 1.03 to 2.98; p = 0.04). Complete revascularization was associated with less deterioration in ejection fraction and lower mortality compared with patients who were not completely revascularized, independent of other predictors (hazard ratio: 0.56, 95% confidence interval: 0.33 to 0.93; p = 0.03).
Conclusions CAD is common in patients with HFpEF and is associated with increased mortality and greater deterioration in ventricular function. Revascularization may be associated with preservation of cardiac function and improved outcomes in patients with CAD. Given the paucity of effective treatments for HFpEF, prospective trials are urgently needed to determine the optimal evaluation and management of CAD in HFpEF.
- coronary artery disease
- diastolic heart failure
- heart failure
- heart failure with preserved ejection fraction
Approximately one-half of all patients with heart failure (HF) have heart failure with preserved ejection fraction (HFpEF) (1). In contrast to heart failure with reduced ejection fraction (HFrEF), there is no proven effective treatment for HFpEF (2). Accordingly, current studies and guidelines endorse treatment of commonly observed comorbidities (3–5). It has also recently been proposed that HFpEF represents a heterogeneous group of diseases that may respond differently to treatments (6). This heterogeneity may be minimized by subgrouping HFpEF patients according to the presence or absence of key comorbidities. Coronary artery disease (CAD) qualifies as a viable candidate for subclassification because it is common in HFpEF (1). CAD also plausibly explains the pathophysiology, because myocardial ischemia causes diastolic and systolic dysfunction (7–11), which are both common in patients with HFpEF (2,12).
However, because CAD and HFpEF are associated with common risk factors, such as aging and hypertension, it is also possible that CAD and HFpEF simply coexist in many patients without any mechanistic relationship. As such, it remains unclear whether HFpEF patients with CAD should be diagnostically grouped separately from those without CAD, how and when to evaluate for CAD in patients presenting with HFpEF, and how to manage CAD once it is identified, at least in the absence of an acute coronary syndrome.
As a first step toward better understanding of the implications of CAD in patients with HFpEF, we investigated the clinical, structural, functional, hemodynamic, and outcome characteristics in a rigorously phenotyped group of patients who were previously hospitalized for HFpEF, comparing those with angiographically-verified CAD with patients without significant CAD. To provide further insight into therapeutics, we then examined the associations of revascularization with survival and ventricular function in HFpEF patients with CAD.
All patients discharged from St. Mary's Hospital at the Mayo Clinic with the primary diagnosis of HF (International Classification of Diseases-9th revision code 428) between January 1, 2004, and December 31, 2012, were identified. From this group, individuals who had undergone echocardiography were identified and cross-checked with the Mayo Clinic catheterization laboratory database to identify all patients with coronary angiography within 1 year of hospital discharge and echocardiography within 6 months before angiography. Data from the first angiogram were used for patients with >1 examination. HFpEF was defined by clinical diagnosis of decompensated HF according to the admitting physician and left ventricular ejection fraction (LVEF) ≥50% within 6 months of hospitalization. In addition to HF hospitalization, all HF patients had to fulfill the Framingham criteria and/or demonstrate elevated left heart filling pressures at catheterization (pulmonary capillary wedge pressure or left ventricular [LV] end-diastolic pressure; >15 mm Hg at rest or ≥25 mm Hg with exercise) in studies performed specifically for the evaluation of dyspnea (13). Patients with significant valvular disease (more than moderate left-sided regurgitation or more than mild stenosis); severe pulmonary disease; acute coronary syndrome (defined by ≥2 of the following: increasing cardiac enzymes, ischemic electrocardiographic changes, typical chest pain); primary renal, hepatic, or pulmonary vascular disease; high output HF; chest radiation; severe anemia (≤9.0 g/dl); constrictive pericarditis; and infiltrative, restrictive, or hypertrophic cardiomyopathies were excluded.
HFpEF patients were divided into those with and without significant anatomic CAD, defined by angiographic stenosis of >50% in ≥1 epicardial coronary artery with a visual reference lumen diameter of ≥2.5 mm, previous infarction, or any previous revascularization. All angiograms were interpreted by a single experienced interventional cardiologist (S.J.H.). Syntax score was calculated as previously described (14,15). Clinical, hemodynamic, stress testing, and echocardiographic data were abstracted from detailed chart review and compared in HFpEF patients with and without CAD. Ischemia on noninvasive stress testing was defined as ST-segment depression >2 mm, new regional wall motion abnormalities on echocardiography, or reversible perfusion defects on myocardial nuclear imaging.
Complete revascularization was defined as treatment of all >50% coronary stenoses in epicardial vessels by percutaneous intervention and/or coronary bypass grafting. Incomplete revascularization was defined as intervention on ≥1 significant stenosis, but with residual lesion(s) of >50% stenosis. The impact of the presence or absence of CAD and the impact of revascularization in HFpEF patients with CAD was assessed by follow-up echocardiography performed no sooner than 6 months after angiography and by assessing vital status ascertained through chart review and the Social Security Death Index.
Assessment of cardiovascular structure, function, and hemodynamics
Two-dimensional and Doppler echocardiography were performed to assess LV morphology and systolic and diastolic function according to American Society of Echocardiography guidelines by experienced sonographers and echocardiologists (16). Right and left heart catheterization were performed in the supine position via the jugular or femoral veins and femoral or radial arteries using fluid-filled catheters (13). Hemodynamic parameters including right and left heart filling pressures, pulmonary artery pressures, cardiac output, pulmonary and systemic arterial resistance, compliance, and elastance were determined as described previously (17).
Continuous variables were reported as mean ± SD or median (interquartile range [IQR]) and compared by analysis of variance, paired t test, or Mann-Whitney U test. Categorical variables were expressed as number (percent) and were compared by chi-square or Fisher exact test. Regression was used to adjust for potential confounding, in which the dependent variable was the normally distributed continuous (linear least-squares regression) or categorical (logistic regression) outcome variable of interest. The impact of the presence of CAD on survival and impact of revascularization in patients with CAD were assessed by the Kaplan-Meier method with Cox regression analysis to adjust for other univariate predictors of death. Univariate predictors were selected based on previously-published studies that showed an association with increased mortality in HFpEF (18,19) and sufficient availability of data in the sample population. In the primary treatment analysis, “revascularization” was considered complete in patients who received complete revascularization, whereas patients who did not undergo revascularization or had “incomplete revascularization” were included together in the comparator group (20).
During the 8-year study, there were 4,331 unique patients who were admitted with a primary diagnosis of HF who underwent both echocardiography and angiography within the protocol-specified timelines relative to hospitalization (Fig. 1). From this sample of HF patients, 52.6% had reduced ejection fraction (EF), and 47.4% had preserved EF. After exclusion of preserved EF patients with acute coronary syndrome, primary valvular heart disease, cardiomyopathies, and other exclusion criteria, 376 patients with HFpEF were identified, constituting the study population. Of this group, 255 (68%) had CAD and 121 (32%) did not have CAD (Table 1). Of HFpEF patients with CAD, 36% had 3-vessel disease, 36% had 2-vessel disease, and 28% had 1-vessel disease. Mean SYNTAX score in patients with CAD was 19 ± 14. Indications for angiography are provided in Online Table 1.
Clinical characteristics in HFpEF patients with and without CAD
Compared with HFpEF patients without CAD, patients with CAD were slightly older and were more likely to be men; to have typical CAD risk factors, including hypertension, diabetes, dyslipidemia, and smoking history; and to be treated with anti-ischemic medicines, including beta-blockers, nitrates, statins, and aspirin (Table 1). However, none of these parameters effectively distinguished CAD from no CAD (all areas under the receiver-operating curve <0.7) (Online Table 2). There were no differences among HFpEF patients with or without CAD in body mass, atrial fibrillation, or use of other HF therapies, including inhibitors of the renin-angiotensin-aldosterone axis and diuretics.
Patients with and without CAD reported severe HF symptoms (>50% New York Heart Association functional class III or IV) with no group differences (Table 1). Intriguingly, the proportion of patients reporting any angina or severe angina (Canadian Cardiovascular Society class ≥II) was not different in HFpEF patients with or without CAD (Table 1). Anginal symptoms were similarly prevalent in patients with or without diabetes (35% vs. 37%; p = 0.4). Troponin T levels were assessed during HF hospitalization in 81 patients (22%) and were slightly higher in patients with CAD (Table 1), although troponin levels did not identify the presence of CAD in logistic regression analysis (p = 0.9) (Online Table 1). Compared with patients without CAD, HFpEF patients with CAD had more renal dysfunction and a trend for higher brain natriuretic peptide levels, although the latter difference was not observed after accounting for differences in renal function (p = 0.2).
Baseline ventricular structure and function
The LV chamber size, mass, stroke volume, and cardiac output were similar in patients with or without CAD (Table 2). LV mass and relative wall thickness were slightly greater and EF slightly lower in HFpEF patients with CAD compared with patients without CAD, although these differences were attenuated after adjusting for age and sex. LV diastolic function, estimated pulmonary artery systolic pressure (PASP), and arterial properties were similar in HFpEF patients with and without CAD, with the exception of echo-estimated LV filling pressures (E/e′ ratio), which were elevated in both groups but were significantly higher in HFpEF patients with CAD compared with patients without CAD.
Evaluation for ischemia
More than one-half of HFpEF patients underwent stress testing before angiography (57% vs. 53% of patients with and without CAD, p = 0.5) (Table 2). Treadmill electrocardiographic testing was performed in 16%, stress echocardiography in 39%, and nuclear testing in 45% of patients. Among patients who underwent stress testing, 70% with angiographically-proven CAD were found to have ischemia at the time of stress testing, with a 30% false negative rate (Fig. 2). Defining CAD using the more stringent criterion of stenosis ≥70% produced a similar 28% false negative rate. Conversely, nearly one-half (45%) of HFpEF patients with no significant anatomic CAD on angiography were found to have a positive test. False positive and false negative test rates were similar in patients presenting with or without angina (Fig. 2). Overall accuracy of stress testing to classify CAD was 66%, with no significant difference between the modalities (p = 0.18) (Online Table 3).
Approximately one-third and one-half of HFpEF patients with and without CAD underwent invasive hemodynamic assessment (Table 3). On average, HFpEF patients had systemic hypertension, elevated right and left heart filling pressures, mild pulmonary hypertension, preserved cardiac output at rest, and mild to moderate pulmonary vascular disease, but there were no differences noted in any hemodynamic parameters between HFpEF patients with or without CAD. A subset of patients underwent invasive exercise evaluation, which showed elevation in cardiac filling pressures and exercise-induced pulmonary hypertension with stress, but again there were no differences between patients with or without CAD. A smaller subset of patients received nitroprusside infusion, which also showed no discernible differences in central hemodynamic responses in patients with or without CAD.
Impact of CAD on ventricular function and mortality
Repeat echocardiography was performed in 218 patients (59% of patients with CAD, 55% of patients without CAD; p = 0.5) a median interval of 1,314 days (IQR: 655 to 1,947 days) after catheterization. Baseline characteristics were similar in patients who did or did not undergo repeat echocardiography (Online Table 4). Systolic function (LVEF) deteriorated in patients with CAD but not in patients without CAD (Figs. 3A and 3B). Compared with patients without CAD, HFpEF patients with CAD experienced a 4-fold greater decline in EF over time (−4.6 ± 10.3% vs. −1.0 ± 8.7%; p = 0.01) (Fig. 3C). Documented myocardial infarction occurred in 10 patients with CAD and 1 patient without CAD (p = 0.11). After excluding patients with known intercurrent infarction, EF deterioration remained significantly greater in patients with CAD (−3.3 ± 9.5% vs. −0.5 ± 9.4%; p = 0.02).
During a median follow-up of 1,457 days (IQR: 692 to 2,366 days), there were 112 deaths. HFpEF patients with significant anatomic CAD had higher mortality compared with HFpEF patients without CAD (hazard ratio [HR]: 1.61, 95% confidence interval [CI]: 1.06 to 2.59; p = 0.026) (Fig. 4). Age, echo-estimated PASP, chronic kidney disease, atrial fibrillation, E/e′ ratio, hemoglobin, and sodium were also univariate predictors of death (Table 4). In multivariate analysis incorporating univariate predictors, the presence of CAD remained a significant predictor of increased risk of death (HR: 1.71; 95% CI: 1.03 to 2.98; p = 0.04).
Impact of revascularization in HFpEF patients with CAD
Of 255 HFpEF patients with significant CAD, 205 (80%) underwent revascularization (63% percutaneous intervention, 37% surgical bypass). Complete revascularization was performed in 102 patients, partial revascularization in 103 patients, and no revascularization in 50 patients. The clinical, echocardiographic, and hemodynamic characteristics, as well as CAD severity of patients who underwent complete revascularization, were not different from those who had incomplete and/or no revascularization (Online Tables 5 to 8). The presence and severity of angina and ischemia burden on stress testing were not different between patients who received complete, incomplete, or no revascularization. The most common documented reasons for not pursuing revascularization were uncertain relation to symptoms, indeterminate severity of lesions, and absence of angina (Online Table 9).
Repeat echocardiography was performed in 151 of the 255 patients with CAD a median of 1,219 days (IQR: 651 to 1,898 days) after catheterization. LVEF decreased on average in HFpEF patients with CAD (Figs. 3D and 3E), although patients who were not completely revascularized experienced a 2-fold greater decline in EF compared with patients who underwent complete revascularization (−2.7 ± 8.9% vs. −6.1 ± 11.1%; p = 0.04) (Fig. 3F). Longitudinal changes in EF were not different comparing patients with single-vessel disease with HFpEF patients without CAD (Online Fig. 1). The change in EF was not associated with mortality (p = 0.2).
During a median follow-up of 1,478 days (IQR: 708 to 2,371 days), there were 87 deaths among HFpEF patients with CAD. HFpEF patients who underwent complete revascularization had significantly improved survival compared with patients who did not undergo complete revascularization (Fig. 5A), with survival rates being similar to what was observed in HFpEF patients without CAD (Fig. 5B). Similar results were observed in a sensitivity analysis, in which revascularization was defined as treatment of stenoses of ≥70% severity (p = 0.03) (Online Fig. 2), comparing complete revascularization with partial or no revascularization separately (Online Fig. 3) and comparing surgical versus percutaneous revascularization (Online Fig. 4). Patients with multivessel disease or higher SYNTAX scores displayed better outcomes with revascularization compared with patients with single-vessel disease or low SYNTAX scores (Fig. 6). Survival in patients with single-vessel disease was not different from HFpEF patients without CAD (Online Fig. 5), and outcomes were similar in CAD patients with negative and positive stress tests (p = 0.5). Overall, differences in survival associated with revascularization status persisted after adjusting for other univariate predictors of death, including age, chronic kidney disease, atrial fibrillation, pulmonary artery pressure, previous myocardial infarction, and SYNTAX score (HR: 0.56; 95% CI: 0.33 to 0.93; p = 0.03) (Table 5, Online Table 10).
This is the first study to thoroughly examine the clinical, structural, functional, hemodynamic, and prognostic implications of CAD and its treatment in patients with HFpEF. We studied patients with unequivocal, rigorously adjudicated HF characterized by a previous hospitalization, in which alternative etiologies, including acute coronary syndrome, valvular heart disease, cardiomyopathy, and pericardial disease, were excluded. The presence of significant CAD, ascertained anatomically using the gold standard of coronary angiography, was observed in two-thirds of patients. Compared with HFpEF patients without CAD, patients with CAD were more likely to be men, to have typical atherosclerotic risk factors, and to be treated with anti-ischemic medications. However, dyspnea and angina symptoms were similar, as were invasively-measured hemodynamics and most indexes of cardiovascular structure and function. Noninvasive stress testing poorly classified the presence or absence of anatomic CAD among patients with and without angina. Over a median follow-up of 4 years, HFpEF patients with CAD experienced greater deterioration in systolic function and significantly worse survival compared with patients without CAD. However, HFpEF patients with CAD who underwent complete revascularization experienced less reduction in LVEF and had improved survival compared with patients who had incomplete or no revascularization, particularly among patients with more severe CAD. We conclude that despite numerous clinical, structural, and hemodynamic similarities, important differences in natural history and response to treatment justify the diagnostic separation of HFpEF patients according to the presence or absence of CAD. The failure of symptoms and noninvasive testing to adequately identify or exclude CAD in patients with HFpEF raises questions regarding its optimal assessment in this population. Although prospective trials are needed, the current exploratory data support the hypothesis that revascularization of CAD in patients with HFpEF might be effective to improve both ventricular function and survival in this population.
Community-based studies have shown that CAD, diagnosed based on a history of myocardial infarction, revascularization, or electrocardiographic changes, is common in HFpEF, and is present in 40% to 50% of patients (1,19,21–23). The prevalence of angiographically-ascertained CAD was higher in the present study (68%). Although this higher prevalence is due in part to referral bias, it is also possible that previous studies that relied on clinical criteria might have underappreciated the burden of CAD in HFpEF. We observed that the presence of CAD was associated with greater reduction in LVEF over time, confirming and extending a recent study by Dunlay et al. (24). In contrast, reduction in EF to <50% was distinctly uncommon in HFpEF patients without CAD (Fig. 3A). The worsening ventricular function did not appear to be completely explainable by clinically-apparent intercurrent myocardial infarction, suggesting that CAD may adversely affect ventricular function in HFpEF through a combination of acute and chronic ischemic effects.
Despite the common presence of CAD in HFpEF, data regarding its prognostic implications and optimal treatment are sparse and somewhat conflicting. A study from the CASS (Coronary Artery Surgery Study) registry showed that the presence of HF in patients with CAD and EF >45% was associated with increased risk of death (25). However, 2 more recent studies observed no excess risk in HFpEF patients with CAD (22,23), although CAD was defined clinically rather than angiographically. Importantly, most previous studies of HFpEF have not rigorously subphenotyped patients to exclude alternative etiologies of the clinical syndrome of HF. The present data show, in a carefully defined, homogenous, well-described HFpEF cohort, that the presence of CAD is associated with increased risk of death, even after adjusting for other independent markers of risk. Changes in LV function and outcome were similar for HFpEF patients with no CAD and patients with single-vessel disease, suggesting that the adverse impact of CAD on HFpEF might be more related to multivessel disease; this is similar to what has been reported in HFrEF (26). The differences observed between HFpEF patients with and without CAD in the present study in ventricular function and in outcome provide justification for the subcategorization of HFpEF patients according to the presence or absence of CAD in both clinical practice and research.
It is notable that HFpEF patients with and without CAD did not differ in clinically meaningful ways in terms of anginal symptoms, laboratory results, and cardiovascular structure, function, and hemodynamics. Demographic characteristics and comorbidities were clearly different in patients with and without CAD, with a greater prevalence in men and more atherosclerotic risk factors in the CAD group, as expected. However, in receiver-operating curve analysis, none of these factors effectively distinguished patients with CAD from those without CAD (Online Table 2). Importantly, 30% of patients with anatomically-proven CAD had a negative stress test result, suggesting that a substantial number of HFpEF patients might not receive potentially effective therapies if stress imaging alone were relied upon to exclude CAD. The common misclassification of the presence or absence of anatomically-defined CAD by stress testing observed in the present study suggests that there might be previously unrecognized limitations of stress testing in this population, although the rates of misclassification noted might be inflated by higher pre-test probability for CAD on average among referring cardiologists. Further study is required to identify the optimal diagnostic assessments for CAD in patients presenting primarily with the clinical syndrome of HFpEF.
No treatment has been shown to improve survival in HFpEF (2), leading many authorities to emphasize treatment of commonly-observed comorbidities such as CAD (3–5). However, currently-available data regarding optimal management of CAD in HFpEF are scarce. An early study from the CASS registry showed that survival was similar in patients with HF, CAD, and EF >45% treated medically and with revascularization (25), although both medical and revascularization options have changed dramatically since that era. In a retrospective, observational series of patients admitted for acute pulmonary edema, Kramer et al. (27) found that revascularization of CAD was not associated with a reduction in recurrent episodes of edema, although the sample size was small, and there were very few deaths (27). In the present study, which had a much larger sample and longer duration of follow-up, complete revascularization was associated with lower mortality, with outcomes that were not different than the HFpEF group without CAD.
This sample is subject to referral bias because of the requirement for angiography. The prevalence of CAD would be expected to be lower in a randomly-selected population of patients, and we cannot determine how many patients were admitted for HFpEF who did not have an angiogram. The operating characteristics reported for stress testing in this study were affected by the catheterization laboratory referral population, in which, presumably, the pre-test probability of CAD was, on average, higher among ordering physicians. All patients were required to have been hospitalized for HF, and these results might not apply to the larger ambulatory population of HFpEF patients who never required hospitalization. The retrospective, observational nature of this study did not permit conclusions regarding the causal effects of CAD or revascularization on LV function or outcome, or on the potential impact of CAD on the pathophysiology of HFpEF. It is possible that complete revascularization identified a healthier subset of patients or one that was better treated, although medication use, symptoms, ischemia burden, LV function, CAD severity, and other characteristics did not differ in patients who did or did not receive complete revascularization (Online Tables 4 to 7). This study did not assess the impact of revascularization on symptoms, because there was marked variability in follow-up duration and completeness of documentation of symptoms at subsequent visits. This study did not assess the impact of CAD or revascularization on recurrent HF hospitalizations. Follow-up echocardiography was not performed at consistent time points and was obtained only at the discretion of ordering cardiologists, and survival bias might also affect the longitudinal changes in LV function, although one would expect this to only bias the results toward the null.
CAD is common in patients with HFpEF, and noninvasive diagnosis may be less accurate in this cohort than has been previously recognized. Although symptoms, ventricular structure, function, and hemodynamics are similar in patients with and without CAD, important and significant differences in outcome and response to treatment are present that suggest that HFpEF should be nosologically subcategorized according to the presence or absence of CAD. The presence of CAD is associated with worse outcome in HFpEF independent of other predictors, and complete revascularization may be associated with improved survival and less deterioration in LV function over time. Prospective trials are needed to determine the optimal techniques to identify and treat CAD in patients with HFpEF, a disease for which no current proven treatment exists.
The authors thank Dr. Robert L. Frye for his very helpful critique.
For supplemental figures and tables, please see the online version of this article.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- coronary artery disease
- confidence interval
- heart failure
- heart failure with preserved ejection fraction
- heart failure with reduced ejection fraction
- hazard ratio
- interquartile range
- left ventricular
- left ventricular ejection fraction
- pulmonary artery systolic pressure
- Received November 26, 2013.
- Revision received February 5, 2014.
- Accepted March 4, 2014.
- American College of Cardiology Foundation
- Borlaug B.A.,
- Paulus W.J.
- Yancy C.W.,
- Jessup M.,
- Bozkurt B.,
- et al.
- Ather S.,
- Chan W.,
- Bozkurt B.,
- et al.
- Shah S.J.
- Muller O.,
- Rorvik K.
- Mann T.,
- Goldberg S.,
- Mudge G.H. Jr..,
- Grossman W.
- Kass D.A.,
- Midei M.,
- Brinker J.,
- Maughan W.L.
- Borlaug B.A.,
- Olson T.P.,
- Lam C.S.,
- et al.
- Borlaug B.A.,
- Nishimura R.A.,
- Sorajja P.,
- Lam C.S.,
- Redfield M.M.
- Lang R.M.,
- Bierig M.,
- Devereux R.B.,
- et al.
- Schwartzenberg S.,
- Redfield M.M.,
- From A.M.,
- Sorajja P.,
- Nishimura R.A.,
- Borlaug B.A.
- Komajda M.,
- Carson P.E.,
- Hetzel S.,
- et al.
- Kereiakes D.J.
- Lee D.S.,
- Gona P.,
- Vasan R.S.,
- et al.
- Tribouilloy C.,
- Rusinaru D.,
- Mahjoub H.,
- et al.
- Dunlay S.M.,
- Roger V.L.,
- Weston S.A.,
- Jiang R.,
- Redfield M.M.
- Judge K.W.,
- Pawitan Y.,
- Caldwell J.,
- Gersh B.J.,
- Kennedy J.W.
- Felker G.M.,
- Shaw L.K.,
- O'Connor C.M.