Hemodynamic Exercise Testing
A Valuable Tool in the Selection of Cardiac Transplantation Candidates
Background Peak exercise oxygen consumption (V̇o2), a noninvasive index of peak exercise cardiac output (CO), is widely used to select candidates for heart transplantation. However, peak exercise V̇o2 can be influenced by noncardiac factors such as deconditioning, motivation, or body composition and may yield misleading prognostic information. Direct measurement of the CO response to exercise may avoid this problem and more accurately predict prognosis.
Methods and Results Hemodynamic and ventilatory responses to maximal treadmill exercise were measured in 185 ambulatory patients with chronic heart failure who had been referred for cardiac transplantation (mean left ventricular ejection fraction, 22±7%; mean peak V̇o2, 12.9±3.0 mL·min−1·kg−1). CO response to exercise was normal in 83 patients and reduced in 102. By univariate analysis, patients with normal CO responses had a better 1-year survival rate (95%) than did those with reduced CO responses (72%) (P<.0001). Survival in patients with peak V̇o2 of >14 mL·min−1·kg−1 (88%) was not different from that of patients with peak V̇o2 of ≤14 mL·min−1·kg−1 (79%) (P=NS). However, survival was worse in patients with peak V̇o2 of ≤10 mL·min−1·kg−1 (52%) versus those with peak V̇o2 of >10 mL·min−1·kg−1 (89%) (P<.0001). By Cox regression analysis, exercise CO response was the strongest independent predictor of survival (risk ratio, 4.3), with peak V̇o2 dichotomized at 10 mL·min−1·kg−1 (risk ratio, 3.3) as the only other independent predictor. Patients with reduced CO responses and peak V̇o2 of ≤10 mL·min−1·kg−1 had an extremely poor 1-year survival rate (38%).
Conclusions Both CO response to exercise and peak exercise V̇o2 provide valuable independent prognostic information in ambulatory patients with heart failure. These variables should be used in combination to select potential heart transplantation candidates.
Despite continuing advances in the therapy of heart failure, the mortality of this syndrome remains high. Heart transplantation, with 1-year survival rates now exceeding 80%, has emerged as an important adjunct to medical therapy and has been considered the treatment of choice for patients with end-stage heart disease.1 The resultant increase in referrals to heart transplantation centers has magnified the need for reliable prognostic markers as the gap widens between the number of potential transplant recipients and available donor organs.2 3 4
Recent American Heart Association consensus reports have recommended that peak exercise V̇o2 be used to help determine the timing of heart transplantation in ambulatory patients with heart failure.1 5 Specifically, it has been suggested that a peak exercise V̇o2 level of ≤14 mL·min−1·kg−1 be used as a key criterion for the acceptance of ambulatory patients for transplantation. This recommendation was based on several studies demonstrating that peak exercise V̇o2 is a valuable prognostic marker in patients with heart failure1 5 6 7 and on a study by Mancini et al8 that suggests that transplantation can be safely deferred in patients with peak exercise V̇o2 levels of >14 mL·min−1·kg−1.
However, the use of a specific peak exercise V̇o2 level as a selection criterion for heart transplantation has several key limitations. Peak exercise V̇o2 can be influenced by noncardiac factors such as muscle deconditioning, motivation, and obesity.9 10 The expected peak V̇o2 varies according to the age and sex of the individual.11 Furthermore, several groups have found no statistical difference in survival between patients with peak exercise V̇o2 levels of 10 to 14 mL·min−1·kg−1 and those with levels of 14 to 18 mL·min−1·kg−1.7 8 12
The exercise V̇o2 is widely thought to reflect the cardiac output (CO) response to exercise. Therefore, the prognostic value of peak exercise V̇o2 may be due to its relation to the exercise CO. If this is the case, then direct measurement of the CO response to exercise may provide a more accurate prognostic marker. The present study was undertaken to test this hypothesis. To this end, we evaluated both ventilatory and hemodynamic responses to maximal treadmill exercise in consecutive ambulatory patients being considered for heart transplantation at Vanderbilt University Medical Center. We then compared the relative prognostic values of peak exercise V̇o2, the CO response to exercise, left ventricular ejection fraction (LVEF), and resting hemodynamic parameters.
The study population consisted of ambulatory patients referred to the Vanderbilt Heart Failure and Transplantation Evaluation Clinic between December 1993 and December 1995. Patients were excluded from evaluation if the LVEF was ≥40% or if they achieved peak exercise V̇o2 levels of >20 mL·min−1·kg−1; prior reports1 5 6 7 8 have suggested that this group has an excellent prognosis. Patients were also excluded if they were dependent on inotropic or mechanical support or if exercise testing was not possible because of severe angina or an inability to walk on a treadmill. One hundred eighty-five patients (142 men and 43 women; mean age, 51.4±10.5 years) were included in the study. Cause of heart failure was coronary artery disease in 52% and dilated cardiomyopathy in 48%. At the time of initial evaluation, 32% of patients were in NYHA functional class II, 44% were in class III, and 24% were in class IV.
All patients had a history of chronic heart failure for >6 months. Almost all patients were receiving digoxin (94%), an ACE inhibitor (92%), and diuretics (97%). Initial evaluation included estimation of LVEF by echocardiography and/or radionuclide ventriculography (mean, 22±7%).
All patients gave written informed consent before participation in the hemodynamic exercise studies.
Exercise Hemodynamic Protocol
All patients underwent hemodynamic exercise testing with concomitant respiratory gas analysis in a Special Procedures Room located in The Vanderbilt Cardiology Outpatient Facility. Patients were studied after optimization of medical therapy by the Vanderbilt Heart Failure Service. Before this study, all patients had performed at least one cardiopulmonary exercise test and therefore were familiar with the procedure.
After arrival in the procedures room, a 7F Swan-Ganz catheter was inserted via the right internal jugular vein. After each patient rested in the semirecumbent position for 20 to 30 minutes, resting hemodynamic measurements were obtained, including pulmonary artery, pulmonary wedge, and right atrial pressures. Thermodilution CO values were obtained in triplicate through the injection of 10-mL boluses of iced saline and the use of Baxter Edwards thermodilution computer. Arterial blood pressure was measured with cuff sphygmomanometry. Pulmonary artery samples were taken for measurement of plasma lactate and hemoglobin O2 saturation.
Each patient then stood on a Marquette treadmill and was connected to a MedGraphics CardioO2 System via a disposable Pneumotach mouthpiece. Arterial O2 saturation was monitored noninvasively via a pulse oximeter. Hemodynamic recordings and blood samples were repeated. After 3 minutes of resting data acquisition, maximum symptom-limited exercise testing was performed according to a 3-minute Naughton protocol. At the end of each exercise stage and at peak exercise, hemodynamic measurements and blood samples were taken again. In addition, at each stage of exercise and at peak exercise, the patient was asked to rate the level of dyspnea and leg fatigue according to the Borg scale.13 This scale rates the level of perceived symptoms on a scale of 6 (none) to 20 (severe). After exercise, the Swan-Ganz catheter was removed, and the patient was discharged to home. The total procedure time was ≈1 hour.
The hemodynamic exercise procedures were associated with no complications that required additional therapy. In addition, there were no associated tachyarrhythmias necessitating electrical or chemical cardioversion.
Arterial hemoglobin O2 saturation was measured with a Ciba-Corning 2500 Co-Oximeter. Blood lactate concentrations were determined with a calorimetric test using Ektachem Clinical Chemistry slides and a Johnson and Johnson Ektachem 950IRC System. The normal range at rest for this technique is 3 to 12 mg/dL.
The following hemodynamic variables were calculated using standard formulas: mean arterial pressure, arteriovenous O2 difference, cardiac index, stroke volume and stroke work index, and systemic and pulmonary vascular resistances. Thermodilution CO values were obtained with the patient in the resting supine position, whereas Fick CO values were calculated during upright exercise.
The CO response to exercise was defined as normal or reduced. The normal CO response was derived from data previously reported by Higginbotham et al.14 These investigators measured the CO response to upright exercise in normal subjects and calculated a regression curve and 95% confidence limits for the relation between cardiac index and O2 consumption. We converted this regression curve to CO by multiplying cardiac index by 1.8 m2, an average body surface area. This conversion resulted in the following equation for the lower limit of normal for CO: 5×V̇o2 (L/min)+3 L/min.
To test this definition of the lower limits of normal CO during exercise, studies by Damato et al,15 Becklake et al,11 and Julius et al16 were examined. These investigators studied 172 normal subjects, 65 of whom were women. Damato et al15 provided individual data points for O2 consumption and CO. The linear regression line noted above fell just below the normal data points. The regression line fell below all of the normal regression lines described by Becklake et al.11 Julius et al16 reported mean CO levels±1 SEM versus O2 consumption for the normal population. The linear regression line noted above was located 2 to 3 SEMs below the normal mean values. These observations suggest that the equation of 5×V̇o2 (L/min)+3 L/min provides a valid reflection of the lower limits of normal CO during exercise.
To determine whether a patient's CO response was normal or reduced, serial CO measurements made during exercise were compared with the normal level. In the vast majority, all CO measurements fell either below or above the normal cutoff level. In the rare event that some points fell above and some below this level, the output response was defined on the basis of the location of the majority of output measurements.
The anaerobic threshold was defined using three criteria: the point after which the respiratory gas exchange ratio (V̇co2/V̇o2) consistently exceeded the resting ratio, the point at which the ventilatory equivalent for oxygen (Ve/V̇o2) was minimal followed by a progressive increase in Ve/V̇o2, and the V̇o2 after which a nonlinear rise in Ve occurred relative to V̇o2.8 If these three criteria identified different anaerobic thresholds, threshold values were averaged.
All data are expressed as mean±SD. Differences between groups were evaluated by ANOVA or χ2 analysis. Probability of survival was analyzed by the life table method from the time of initial hemodynamic exercise testing. Survival curves were compared using the log-rank test and Wilcoxon analysis. Outcomes were defined as sudden death if the event occurred out of hospital without symptomatic deterioration within the preceding 24 hours or as progressive heart failure in patients with progressive symptomatic or hemodynamic deterioration. Cardiac transplantation was treated as a censored observation (ie, withdrawn at the time of transplantation). Multivariate analysis was performed using the Cox proportional hazards model to identify independent predictors of survival. Variables that were assessed included age, sex, functional class, ejection fraction, pulmonary wedge pressure, central venous pressure, resting CO, CO response to exercise, and peak V̇o2. A value of P<.05 was considered significant.
The clinical, cardiopulmonary, and resting hemodynamic characteristics of the study population are summarized in Table 1⇓. The mean peak V̇o2 was 12.9±3.0 mL·min−1·kg−1. Eighty-three of the patients (45%) had a normal CO response to exercise, whereas 102 (55%) had a reduced CO response. The characteristics of these two groups are given in Table 2⇓. Patients with normal CO responses had higher resting CO values, lower pulmonary wedge pressures, and higher peak exercise V̇o2. However, peak V̇o2 was markedly depressed in both groups (13.8±3.2 versus 12.1±2.6 mL·min−1·kg−1, P=NS, normal CO response versus reduced CO response). The respiratory gas exchange ratio at peak exercise was comparable in both CO response groups, suggesting similar exercise intensity.
The mean follow-up period was 307±192 days. Information was obtained on outcomes for all subjects at the end of the follow-up period. Cumulative survival for the entire group was 83% at 1 year (Fig 1⇓). There were 32 deaths during the follow-up period; the cause of death was progressive heart failure in 31% of patients and sudden death in 69%. Thirty-six patients underwent transplantation, whereas 26 were listed as status 1 for urgent transplantation (ie, requiring inotropic or mechanical support in hospital) secondary to deterioration of cardiovascular status during follow-up. Twenty-three of these patients (88%) successfully underwent transplantation, whereas 1 died while waiting in the hospital. The clinical, cardiopulmonary, and hemodynamic characteristics of survivors versus nonsurvivors are given in Table 3⇓.
Univariate and Multivariate Analyses
Results of analyses for univariate and multivariate predictors of survival are shown in Table 4⇓. By univariate analysis, only the CO response to exercise, peak exercise V̇o2 (when dichotomized at a V̇o2 of 10 mL·min−1·kg−1), and pulmonary wedge pressure were found to correlate with survival. Patients with a normal CO response to exercise had a survival of 95% at 1 year, whereas survival in those with a reduced CO response was 72% (P<.0001) (Fig 2⇓). Only 7 patients in the normal CO response group died: all from sudden death. Conversely, 25 patients died in the reduced CO group: 44% from progressive heart failure and 56% from sudden death.
Patients with a peak V̇o2 of >10 mL·min−1·kg−1 had a 1-year survival rate of 89%, whereas those with peak V̇o2 of ≤10 mL·min−1·kg−1 had a 52% 1-year survival rate (P<.001) (Fig 3⇓). The 1-year survival rate for patients with a peak exercise V̇o2 of >14 mL·min−1·kg−1 (88%) was not significantly different from that of patients with a V̇o2 of ≤14 mL·min−1·kg−1 (79%, P=NS).
Multivariate analysis identified exercise CO response and peak V̇o2 (only when dichotomized at a level of 10 mL·min−1·kg−1) as the only independent predictors of outcome in this population. In a stepwise analysis, the exercise CO response was the most powerful prognostic variable (P<.0009; risk ratio, 4.3). Peak V̇o2 dichotomized at 10 mL·min−1·kg−1 was the second most important predictor (P<.001; risk ratio, 3.3). Peak V̇o2 was not an independent predictor when analyzed as a continuous variable or dichotomized at the level of 14 mL·min−1·kg−1. Although resting pulmonary wedge pressure was a univariate predictor of survival (17±9 mm Hg in survivors versus 23±9 mm Hg in nonsurvivors; P<.003), it was not an independent prognostic marker. None of the other variables assessed were found to be predictors of survival in univariate or multivariate analyses.
In patients with a reduced CO response to exercise, two subgroups with significantly different prognoses could be defined by their peak V̇o2 levels. In patients with a peak V̇o2 of >10 mL·min−1·kg−1, 1-year survival rate was 81%, whereas patients with a peak V̇o2 of ≤10 mL·min−1·kg−1 had a 1-year survival rate of only 38% (P<.0009). In patients with a normal CO response, there was no significant difference in the survival of patients based on peak V̇o2, although the relatively few patients with a peak V̇o2 of ≤10 mL·min−1·kg−1 (n=15) tended to have a lower survival (82% at 1 year, P=NS compared with patients with peak V̇o2 of >10 mL·min−1·kg−1) (Fig 4⇓).
Until recently, the selection criteria for prospective candidates for cardiac transplantation have focused on resting hemodynamic measurements, ejection fraction, and NYHA functional classification.1 5 8 17 Although these variables have been shown to predict survival in heterogeneous populations of heart failure patients, they are inadequate predictors in potential transplantation candidates.1 6 7 8 12 18 19 20 21 22 23
In an effort to identify a more reliable prognostic tool, several groups have examined the relation between survival in heart failure and peak exercise V̇o2, a noninvasive index of maximal CO.7 24 25 26 Their studies uniformly found that peak V̇o2 correlates with survival in these patients. Subsequently, Mancini et al8 prospectively used peak exercise V̇o2 to help guide the timing of heart transplantation in ambulatory patients. One-year survival rates in patients with peak V̇o2 levels of >14 mL·min−1·kg−1 (in whom transplantation was deferred) was 94%, whereas survival in patients with peak V̇o2 of ≤14 mL·min−1·kg−1 was only 70%. After confirmation of these findings by others,27 several consensus and expert groups strongly recommended adopting a peak exercise V̇o2 level of ≤14 mL·min−1·kg−1 as a major criterion for heart transplantation.1 5
However, the use of a specific peak exercise V̇o2 level as a selection criterion for heart transplantation has several key limitations. Most important, there is growing evidence that peak exercise V̇o2 is influenced by noncardiac factors, such as deconditioning, patient motivation, and body composition.9 10 Indeed, exercise training has been shown to improve the peak exercise V̇o2 of patients with heart failure without concomitant changes in central hemodynamics.28 29 Second, the use of a peak V̇o2 cutoff of 14 mL·min−1·kg−1 is somewhat arbitrary because the study of Mancini et al8 did not demonstrate any significant difference in survival between patients with peak V̇o2 of 14 to 18 mL·min−1·kg−1 versus those with peak V̇o2 levels of 10 to 14 mL·min−1·kg−1. Third, it is unclear how best to normalize peak exercise V̇o2 for individual variations in body composition, sex, and age.9 11
One potential method of avoiding these limitations is to directly measure the CO response to exercise. The relation between peak exercise V̇o2 and prognosis is presumably due to a link between exercise V̇o2 and the CO response. Therefore, one might anticipate that the CO response to exercise would provide a more accurate prognostic tool than peak exercise V̇o2.
In the present study, we examined the relative value of standard prognostic variables and compared their predictive accuracy with that of the CO response to exercise. The overall 1-year survival rate (83%) was comparable to previous studies in similar patient populations.8 30 Sixty-six percent of patients died from sudden death, whereas only 34% died from progressive heart failure. The disproportionately low frequency of death due to progressive heart failure has been seen in other studies involving potential transplantation candidates.8 This is likely a reflection of the close monitoring and management strategies of heart failure/transplantation centers in which patients with worsening heart failure are often hospitalized until a donor heart becomes available. Of note, there were no deaths related to progressive heart failure in patients with normal CO responses to exercise despite the markedly reduced LVEF (mean, 24%) and exercise capacity (mean peak V̇o2, 13.8 mL·min−1·kg−1) in this group.
Forty-five percent of patients had a CO response to exercise that was within the normal range, whereas 55% had a reduced CO response. This is consistent with prior observations from our laboratory.10 Patients with a normal CO response had lower resting pulmonary wedge pressures and higher peak V̇o2 levels, but the degree of difference in peak exercise V̇o2 between patients with normal and reduced CO responses was small, with the average peak exercise V̇o2 of the patients with normal flows still being markedly depressed at 13.8 mL·min−1·kg−1.
The CO response to exercise was the most powerful predictor of survival in this study population according to both univariate and multivariate analyses. In patients with a reduced CO response, the overall 1-year survival rate was poor (72%). The additional finding of a peak V̇o2 of ≤10 mL·min−1·kg−1 in such patients identified a group with a particularly unfavorable prognosis (38% were alive at 1 year). In contrast, the overall 1-year survival rate of patients with a normal CO response was 95%, which is far superior to that expected after transplantation.1 5 8 The favorable prognosis of this group was evident regardless of the peak V̇o2. This was particularly noteworthy in patients with a peak V̇o2 of ≤14 mL·min−1·kg−1 (1-year survival rate, 94%), a group that would be recommended for transplantation based on the most recent American Heart Association consensus statement.5 Although it was not statistically significant (P=.09), survival rates did tend to be lower in patients with a normal CO response and peak V̇o2 of ≤10 mL·min−1·kg−1 (82% at 1 year). However, this finding was the result of only two deaths in this relatively small group (n=15) and therefore must be viewed with circumspection.
The relative importance of hemodynamic performance during exercise has been previously suggested by the results of several small studies. In a study of 49 patients, Griffin et al30 demonstrated that exercise stroke work index was the strongest independent predictor of prognosis, whereas exercise duration and peak V̇o2 could not be used to discriminate survivors from nonsurvivors. Roul et al31 studied 50 NYHA class II and III patients and similarly found that hemodynamic data at maximum exercise provided the best predictors of 1-year outcomes by both univariate and multivariate analyses. They did, however, identify a correlation between peak V̇o2 and peak exercise hemodynamics and concluded that the less-invasive cardiopulmonary testing provided adequate prognostic information. These studies support the intuitive assumption that the direct assessment of circulatory adaptations to a failing heart will provide important prognostic information. However, these and other similar studies are generally limited by small patient numbers, heterogeneous populations, and varying prognostic end points.
Peak exercise V̇o2, when treated as a continuous variable or split at a level of 14 mL·min−1·kg−1, was not found to correlate with survival in our study. At first glance, this finding appears to contradict prior studies by Mancini et al8 and Pilote et al.26 However, in the present study, patients with peak exercise V̇o2 levels of >20 mL·min−1·kg−1 were excluded by prior evidence that these patients have an excellent prognosis. For instance, in the study of Mancini et al,8 much of the prognostic power of peak V̇o2 was due to its ability to define this patient group. Indeed, as alluded to previously, survival of patients with peak V̇o2 levels of 14 to 18 mL·min−1·kg−1 was similar to those with peak V̇o2 levels of 10 to 14 mL·min−1·kg−1.
As in prior studies,8 12 a peak V̇o2 of ≤10 mL·min−1·kg−1 did identify a relatively high-risk population. However, even in this group, a number of patients were identified as having normal CO responses and a relatively favorable prognosis.
There are several limitations to this study. First, this study focused on a relatively well-defined group of patients: ambulatory patients with heart failure being considered for heart transplantation and having peak exercise V̇o2 levels of ≤20 mL·min−1·kg−1. Whether the results of this trial have practical implications for all patients with heart failure remains to be determined. For example, the average age of the patients in this study was relatively young, with few patients over the age of 65. In contrast, the majority of patients with heart failure are over the age of 65.
A second relative limitation is the influence of management approaches used in this patient population. Many of these ambulatory patients were listed for heart transplantation, based in part on peak V̇o2 levels ≤14 mL·min−1·kg−1, and they subsequently received a transplant. As Stevenson et al12 noted, it can be misleading in survival analyses to treat transplanted patients merely as censored data because many of them would not have survived without this therapy. It is similarly misleading to assume that all transplanted patients would have died in the absence of this treatment.
To address this issue, we examined the effect of prognostic variables on a combined end point of death and/or hospitalization for urgent transplantation (status 1). As expected, peak V̇o2 dichotomized at a level of 14 mL·min−1·kg−1 was a significant predictor of outcome using this approach. However, the CO response to exercise continued to be a more significant predictor.
The results of the present study suggest that the CO response to exercise is an important predictor of survival in ambulatory patients being considered for heart transplantation, particularly when used in conjunction with cardiopulmonary exercise testing. Fig 5⇓ demonstrates a proposed algorithm for the evaluation of ambulatory transplant candidates based on our findings. It should be noted that the approach outlined does not take into account certain individual circumstances, such as the patient with recurrent ventricular tachycardia, which can have a significant impact on the decision to proceed with transplantation.
It seems most appropriate to use cardiopulmonary exercise testing as a screening technique in all ambulatory patients being considered for heart transplantation. If the peak exercise V̇o2 is >20 mL·min−1·kg−1, transplantation can be safely deferred. If the peak exercise V̇o2 is <20 mL·min−1·kg−1, exercise hemodynamic measurements should be obtained to confirm that the patient's exercise intolerance is due to circulatory factors. If the CO response is normal, transplantation should be deferred with close monitoring, and serial cardiopulmonary testing should be performed to identify any further deterioration in the patient's status. In addition, noncardiac factors that may contribute to exercise intolerance should be addressed, such as deconditioning. Data from our center suggest that cardiac rehabilitation is particularly beneficial in patients with exercise intolerance despite normal exercise hemodynamics.32 In patients with a normal CO response despite a peak V̇o2 of ≤10 mL·min−1·kg−1, we believe that the prognosis remains uncertain due to the small number of such individuals in this study. We suggest that exercise rehabilitation along with repeat hemodynamic exercise testing be done to follow their response. In addition, consideration for outpatient listing for transplantation while exercise training is conducted seems reasonable.
If the CO response to exercise is reduced, our findings suggest that a complete transplant evaluation is advisable. Within this population, however, there appears to be two subgroups. Patients with a reduced CO response to exercise but a peak V̇o2 of >10 mL·min−1·kg−1 had a 1-year survival rate of 81%. Transplantation is likely to improve survival in this group. Outpatient evaluation and listing are probably safe, although the patients clearly require careful follow-up and serial testing to monitor any further deterioration. If the peak V̇o2 is ≤10 mL·min−1·kg−1 and the CO response is reduced, the 1-year prognosis is so poor (38%) that monitoring of such patients at home as a status 2 candidate carries an extremely high risk. It may be advisable to list such patients for urgent transplantation (status 1). If such patients are relatively stable, current United Network of Organ Sharing criteria would prohibit them from being admitted and considered an urgent transplantation candidate. Based on the results of the present study, revision of these criteria may be indicated. However, these results need to be confirmed by others before exercise hemodynamic testing is widely implemented at transplantation centers.
This work was supported in part by an RO-1 grant from the National Institutes of Health, Bethesda, Md, and a Grant-in-Aid from the American Heart Association, National Center, Dallas, Tex.
- Received March 19, 1996.
- Revision received July 24, 1996.
- Accepted July 31, 1996.
- Copyright © 1996 by American Heart Association
O'Connell JB, Bourge RC, Costanzo-Nordin MR, Driscoll DJ, Morgan JP, Rose EA, Uretsky BF. Cardiac transplantation: recipient selection, donor procurement, and medical follow-up: a statement for health professionals from the Committee on Cardiac Transplantation of the Council on Clinical Cardiology, American Heart Association. Circulation.. 1992;86:1061-1079.
Stevenson LW, Warner SL, Steimle AE, Fonarow GC, Hamilton MA, Moriguchi JD, Kobashigawa JA, Tillisch JH, Drinkwater DC, Laks H. The impending crisis awaiting cardiac transplantation: modeling a solution based on selection. Circulation.. 1994;89:450-457.
Benenson E, ed. United Network of Organ Sharing Bulletin, Vol. 8. Richmond, Va: UNOS; 1993.
Miller LW. Working Group of Transplant Cardiologists First Annual Report. Fort Lauderdale, Fla; Ninth Scientific Sessions of the International Society for Heart Transplantation; 1990.
Costanzo MR, Augustine S, Bourge R, Bristow M, O'Connell JB, Driscoll D, Rose E. Selection and treatment of candidates for heart transplantation: a statement for health professionals from the Committee on Heart Failure and Cardiac Transplantation of the Council on Clinical Cardiology, American Heart Association. Circulation.. 1995;92:3593-3612.
Mancini DM, Eisen H, Kussmaul W, Mull R, Edmunds LH Jr, Wilson JR. Value of peak exercise oxygen consumption for optimal timing of cardiac transplantation in ambulatory patients with heart failure. Circulation.. 1991;83:778-786.
Fleg JL, Lakata EG. Role of muscle loss in the age-associated reduction in V̇o2 max. J Appl Physiol.. 1988;65:1147-1151.
Becklake MR, Frank H, Dagenais GR, Ostiguy GL, Guzman CA. Influence of age and sex on exercise cardiac output. J Appl Physiol.. 1965;20:938-947.
Stevenson LW, Couper G, Natterson B, Fonarow G, Hamilton MA, Woo M, Creaser JW. Target heart failure populations for newer therapies. Circulation. 1995;92(suppl II):II-174-II-181.
Higginbotham MB, Morris KG, Williams RS, McHale PA, Coleman RE, Cobb FR. Regulation of stroke volume during submaximal and maximal upright exercise in normal man. Circ Res.. 1986;58:281-291.
Damato AN, Galante JG, Smith WM. Hemodynamic response to treadmill exercise in normal subjects. J Appl Physiol.. 1966;21:959-966.
Julius S, Amery A, Whitlock LS, Conway J. Influence of age on the hemodynamic response to exercise. Circulation.. 1967;36:222-226.
Copeland J, Emery R, Levinson M, Icenogle T, Carrier M, Ott R, Copeland J, McAleer-Rhenman M, Nicholson S. Selection of patients for cardiac transplantation. Circulation.. 1987;75:2-9.
Schwarz F, Mall G, Zebe H, Schmitzer E, Manthey J, Scheurlen H, Kubler W. Determinants of survival in congestive cardiomyopathy: quantitative morphologic findings and left ventricular hemodynamics. Circulation.. 1984;70:923-928.
Stevenson LW, Tillisch JH, Hamilton M, Luu M, Chelimsky-Fallick C, Moriguchi J, Kobashigawa J, Walden J. Importance of hemodynamic response to therapy in predicting survival with ejection fraction less than or equal to 20% secondary to ischemic or nonischemic dilated cardiomyopathy. Am J Cardiol.. 1990;66:1348-1354.
Packer M, Pitt B. Which objective measurements parallel the course of disease progression in patients with congestive heart failure? J Am Coll Cardiol.. 1988;11:202-A. Abstract.
Pilote L, Silberberg J, Lisbona R, Sniderman A. Prognosis in patients with low left ventricular ejection fraction after myocardial infarction: importance of exercise capacity. Circulation.. 1989;80:1636-1641.
Cohn JN, Johnson G, Shabetai R, Loeb H, Tristani F, Rector T, Smith R, Fletcher R, for the V-HeFT VA Cooperative Studies Group. Ejection fraction, peak exercise oxygen consumption, cardiothoracic ratio, ventricular arrhythmias, and plasma norepinephrine as determinants of prognosis in heart failure. Circulation. 1993;87(suppl VI):VI-5-V-16.
Coats AJS, Adamopoulos S, Radaelli A, McCance A, Meyer TE, Bernardi L, Solda PL, Davey P, Ormerod O, Forfar C, Conway J, Sleight P. Controlled trial of physical training in chronic heart failure. Circulation.. 1992;85:2119-2131.
Sullivan MJ, Higginbotham MB, Cobb FR. Exercise training in patients with chronic heart failure delays ventilatory anaerobic threshold and improves submaximal exercise performance. Circulation.. 1989;79:324-329.
Roul G, Moulichon ME, Bareiss P, Gries P, Koegler A, Sacrez J, Germain PH, Mossard JM, Sacrez A. Prognostic factors of chronic heart failure in NYHA class II or III: value of invasive exercise hemodynamic data. Eur Heart J.. 1995;16:1387-1398.
Wilson JR, Groves J, Rayos GH. Circulatory status and response to cardiac rehabilitation in patients with heart failure. Circulation.. 1996;94:1567-1572.