Timing of Cardiac Transplantation in Patients With Heart Failure ...

3 downloads 0 Views 163KB Size Report
Linda R. Peterson, MD, FACC,a,b Kenneth B. Schechtman, PhD,a,c. Gregory A. Ewald, MD, FACC,a Edward M. Geltman, MD, FACC,a. Lisa de las Fuentes, MD ...
Timing of Cardiac Transplantation in Patients With Heart Failure Receiving ␤-adrenergic Blockers Linda R. Peterson, MD, FACC,a,b Kenneth B. Schechtman, PhD,a,c Gregory A. Ewald, MD, FACC,a Edward M. Geltman, MD, FACC,a Lisa de las Fuentes, MD,a Timothy Meyer, MS,b Pamela Krekeler, MA,b Martha L. Moore, MD,a and Joseph G. Rogers, MD, FACCa Background: Previous work shows that patients with heart failure patients who have peak oxygen consumption (VO2 peak) ⬎14 ml/kg/min do not derive a survival benefit from cardiac transplantation. However, this was shown before ␤-blocker therapy for patients with systolic heart failure became common, and ␤-blockers improve survival in patients with heart failure without changing VO2 peak. Our purpose was to re-evaluate the utility of VO2 peak ⬎14 ml/kg/min as an indicator of the need for cardiac transplantation in patients with heart failure who are taking ␤-blockers. Methods: Actuarial, hemodynamic, and exercise ventilatory data were collected from 540 patients with heart failure, 256 of whom were taking ␤-blockers. We tracked death and cardiac transplantation. We stratified the percentage of patients event-free 1 and 3 years after VO2 peak study by their VO2 peak and ␤-blocker status, and compared 1and 3-year post-transplant survival (United Network of Organ Sharing [UNOS] data). We also compared total mortality for the patients with heart failure as stratified by ␤blocker stats and VO2 peak (excluding the 42 who underwent transplantation) with UNOS post-transplant survival. Results: Patients with heart failure who were receiving ␤-blockers and whose VO2 peak was ⱖ12 ml/kg/min had greater 1- and 3-year event-free survival rates (95% confidence intervals, 92.6%–96.6% and 85.8%–96.0%) than did post-transplant patients (83.9%– 86.3% and 75.4%–76.6%). However, in patients with heart failure not taking ␤blockers, VO2 peak ⬍14 ml/kg/min was associated with worse 3-year survival (38.9 62.1%) than that for post-transplant patients. Excluding the 42 patients with heart failure in our study who underwent transplantation and then evaluating survival of the remaining patients with heart failure (not event-free survival) did not substantially change these results. Conclusions: Patients with heart failure who are receiving ␤-blockers do not derive a survival advantage at 1 and 3 years after cardiac transplantation if VO2 peak is ⱖ12 ml/ kg/min. Patients not taking ␤-blockers whose VO2 peak is ⬍14 ml/kg/min have superior survival with cardiac transplantation. J Heart Lung Transplant 2003;22:1141–1148.

From the aCardiovascular Division and bDivision of Geriatrics and Gerontology, Department of Medicine, and cDivision of Biostatistics, Washington University School of Medicine, St. Louis, Missouri. Submitted June 10, 2002; revised October 8, 2002; accepted November 11, 2002. Supported in part by the Washington University Claude D. Pepper Older Americans Independence Center, AG 13629. Reprint requests: Linda R. Peterson, MD, FACC, Washington

University School of Medicine, Cardiovascular Division, Campus Box 8086, 660 S. Euclid Avenue, St. Louis, Missouri 63110. Telephone: 314-362-4577. Fax: 314-362-9982. E-mail: [email protected] Copyright © 2003 by the International Society for Heart and Lung Transplantation. 1053-2498/03/$–see front matter doi:10.1016/S1053-2498(02)01225-1

1141

1142

D

Peterson et al.

uring 1997 and 1998, more than 4,300 orthotopic heart transplantations (OHTx) were performed in the United States, with an average survival rate of 85.1% at 1 year.1 Thus, OHTx has become a standard therapeutic option for patients with severe heart failure (HF). However, the widespread use of cardiac transplantation is limited principally by a lack of suitable donor organs. The imbalance between the number of patients who may benefit from OHTx and the limited number of donors requires the judicious distribution of hearts to those most likely to benefit from transplantation. The risk– benefit considerations of transplantation must be considered, particularly as they pertain to the morbidity and mortality associated with this treatment. It is critical that patients with HF undergo OHTx only if there is an expected survival advantage. Thus, the timing of cardiac transplantation has become an important issue for physicians and their patients with HF. A landmark article by Mancini et al2 in 1991 established peak oxygen consumption (VO2 peak) testing as a useful tool for predicting survival in patients with HF and for optimal timing of cardiac transplantation. Analyzing VO2 peak as a dichotomous variable, they demonstrated that OHTx could be deferred safely in patients with severe HF who had a VO2 peak ⬎14 ml/kg/min, because their event-free 1- and 2-year survival rates of 94% and 84% compared favorably with 1- and 2-year survival rates after OHTx. However, recent advances in the medical treatment of HF have resulted in marked decreases in mortality. The most profound change has been the widespread use of ␤-blocker therapy. Several large, prospective, randomized, doubleblind studies evaluating the effect of bisoprolol, metoprolol, and carvedilol have demonstrated a clear decrease in morbidity and mortality with the addition of ␤-blocker therapy.3–9 Beta-blockers are now considered part of the standard of care for HF.10 However, ␤-blockers blunt the chronotropic response to exercise, whereas they have little or no effect on the VO2 peak achieved by patients with HF.11 Based on these observations, one would expect improved survival rates in patients with HF who are taking ␤-blockers (␤-blocker⫹) without a marked, concomitant change in the VO2 peak. Thus, we hypothesized that ␤-blocker⫹ patients who had VO2 peak levels ⱖ13 or ⱖ12 ml/kg/min would have improved transplant-free survival and, consequently, the level of VO2 peak above which OHTx could be safely deferred would be lower than the previous standard of ⬎14 ml/kg/min set by Mancini

The Journal of Heart and Lung Transplantation October 2003

et al.2 To test this hypothesis, we evaluated the 1and 3-year event-free (OHTx-free) survival rates in 540 patients with HF (256 of whom were ␤-blocker⫹) as a function of their VO2 peak and compared this with 1- and 3-year post-OHTx survival rates, provided by the United Network for Organ Sharing (UNOS).

METHODS Subjects We reviewed the data from 577 consecutive patients with HF from the Division of Applied Physiology of Washington University School of Medicine to determine VO2 peak between December 1994 and September 7, 2001. Clinical data and medications were recorded on the day of the VO2 peak testing, and all patients underwent symptom-limited exercise treadmill testing, as described previously.12 If a patient had multiple tests while taking and not taking ␤-blockers, he or she was excluded. We also excluded patients if they had incomplete actuarial data, if they were lost to follow-up, or if they were taking ␤-blockers other than carvedilol or metoprolol CR/XL, because these 2 agents are used widely for HF treatment in our community and because controlled clinical trials demonstrate their salutary effects on survival in patients with HF.3– 8 A patient was not excluded based on VO2 peak or respiratory exchange ratio achieved. Of the 577 patients referred for VO2 testing during the time period listed above, 540 patients met these inclusion/exclusion criteria. We considered OHTx and death in this analysis, and OHTx was considered a censored event. Dates of death were obtained from the clinical records or the Social Security Death Index (www.ancestry.com). Dates of OHTx were obtained from the Washington University School of Medicine transplant database. The Washington University School of Medicine Human Studies Committee approved the study protocol. We obtained the 1- and 3-year survival rates after OHTx in the analyses of this study from the UNOS database because the number of patients who underwent OHTx at our institution during the study period was relatively small (n ⫽ 42).

Determining Peak VO2 Peak oxygen consumption exercise ventilatory and hemodynamic data were obtained during exercise treadmill testing in all patients, as described in detail previously.12 Briefly, VO2 peak was determined by standard open-circuit spirometry. Inspiratory volume or ventilation (VO2) was measured using a

The Journal of Heart and Lung Transplantation Volume 22, Number 10

Parkinson-Cowan CD-4 dry-gas meter (Parvomedics; Salt Lake City, UT). Fractional concentrations of expired oxygen and carbon dioxide were measured from a mixing chamber using electronic gas analyzers. Ventilatory equivalent was calculated as the ratio of ventilation (V, liter/min) to VO2 peak (liter/min). The respiratory exchange ratio, a standard measure of exercise effort, was calculated as the ratio of the peak production of carbon dioxide (liter/min) to VO2 peak (liter/min).

Statistics We used SAS software version 8 (SAS Institute Inc.; Cary, NC) for statistical analyses. Student’s t-tests were used to determine the differences between the continuous variables that described the 2 groups of patients: those taking ␤-blockers (␤-blocker⫹) and those patients not taking ␤-blockers (␤-blocker⫺). Chi-squared tests were used to analyze the differences between the discontinuous variables. We examined VO2 peak data as dichotomous variables (i.e., ⱖ14 vs ⬍14 ml/kg/min; ⱖ13 vs ⬍13 ml/kg/min; and ⱖ12 vs ⬍12 ml/kg/min). We generated KaplanMeier curves of event-free survival using the date of the VO2 peak test as follow-up Day 1 and with the date of an event or, if a patient was event-free, 09/07/01 as the last follow-up day. Log-rank tests were used to determine the differences among the Kaplan-Meier curves. A stepwise Cox regression analysis was performed to determine the independent predictors of event-free survival. For dichotomous predictors, risk ratios measured the magnitude of the increased risk for an event that was associated with the variable. For the purposes of analysis, we presented the ventilatory equivalent in units of 5 and presented age in units of 10. Thus, all reported risk ratios for these variables reflect the magnitude of change as these 2 variables change by 5 and 10 units, respectively. A p value ⬍0.05 indicated a statistically significant difference. To compare event-free survival rates of the ␤-blocker⫹ group and survival rates of post-OHTx patients, we calculated the 95% confidence intervals from the standard errors of the mean event-free survival rates for each of the groups at 1 and 3 years. If there was no overlap of the 95% confidence intervals for the patients with HF (either taking or not taking ␤-blockers) and those of the post-OHTx patients, we were 95% confident of a statistical difference between the 2 groups. To compare the overall survival rates of patients with HF as stratified by ␤-blocker status and by VO2 peak with that of post-OHTx patients, we also performed the afore-

Peterson et al.

1143

mentioned analyses excluding the 42 patients with HF in our database who underwent transplantation.

RESULTS Of the 540 patients who underwent VO2 peak testing and who were observed between December 1994 and September 2001, 256 (47%) were receiving ␤-blockers and 284 (53%) were not. One-hundred thirty-four patients (25%) had an event during the follow-up period. Ninety-two (17%) died, and 42 (8%) underwent transplantation. Of the patients who died, 21 (23%) were receiving ␤-blockers, and 71 (77%) were not (p ⫽ 0.001). Moreover, we observed a statistically significant difference (p ⫽ 0.0008) by log-rank test between the event-free survival curves of the ␤-blocker⫹ and the ␤-blocker⫺ patients, indicating a better event-free survival in the ␤-blocker⫹ patients. Of the patients taking ␤-blockers, 60% were taking carvedilol and 40% were taking metoprolol. The median event-free survival follow-up time for ␤-blocker⫹ patients was 20 months, and for the ␤-blocker⫺ patients it was 30.5 months. The longest event-free survival was 78 months by the end of the follow-up period. Table I lists the baseline characteristics of the patients in the study: those treated with ␤-blockers were more likely to be white and to be taking spironolactone. However, we found no difference between the 2 groups of patients in terms of age, sex, body mass index, cause of HF, the percentage taking medications for diabetes mellitus, or the percentage taking other medications for HF or arrhythmias. Table II lists the resting and peak exercise hemodynamic data, peak exercise VO2, and ventilatory data. A total of 174 (32%) patients had VO2 peak ⬍14 ml/kg/min. The majority (86%) of patients achieved a respiratory exchange ratio ⬎1.0, indicating an adequate exercise effort. We found no statistical differences in VO2 peak between the ␤-blocker⫺ and the ␤-blocker⫹ patients. The only difference in hemodynamics between the 2 groups of patients was a greater heart rate reserve (peak resting heart rate) in the ␤-blocker⫺ group (p ⬍ 0.05). The ventilatory equivalent, which was calculated as the ratio of V ⫼ VO2 peak, was greater in the ␤-blocker⫺ group (p ⬍ 0.0005). Table III shows the mean and standard error of the event-free survival of the 2 groups stratified by VO2 peak. Table IV shows the 95% confidence intervals of the percentage of patients who were event-free at 1 and 3 years, stratified by their VO2 peak. A higher VO2 peak was associated with better event-free survival. Patients with VO2 peak ⱖ14

1144

Peterson et al.

The Journal of Heart and Lung Transplantation October 2003

TABLE I Patient characteristics Age (years, mean ⫾ SD) Body mass index (kg/m2) % Male % Non-hispanic white % Taking spironolactone % Taking amiodarone % Taking angiotensin-converting enzyme inhibitors % Taking digoxin % Taking medications for diabetes mellitus % Non-ischemic cardiomyopathy

␤-blockerⴙ

␤-blockerⴚ

49 ⫾ 11 29 ⫾ 5 67% 85* 27† 18% 80% 76% 13% 66%

48 ⫾ 9 28 ⫾ 5 74% 77% 12% 23% 84% 83% 15% 69%

By chi-square testing: *p ⬍ 0.05, †p ⬍ 0.0001, all other p ⫽ not significant.

ml/kg/min had a better event-free survival than did those with VO2 peak ⬍14 ml/kg/min (p ⫽ 0.0001, log-rank test), regardless of their ␤-blocker status. Similarly, patients with VO2 peak ⱖ13 had better event-free survival than did those with ⬍13 ml/kg/ min (p ⫽ 0.0001), and those with VO2 peak ⱖ12 had better event-free survival than did those with ⬍12ml/kg/min (p ⫽ 0.0001), regardless of their ␤-blocker status. Among the sub-set of patients taking ␤-blockers, those with VO2 peak ⱖ14 had better event-free survival than did those with ⬍14 ml/kg/min (p ⫽ 0.0004, Figure 1); those with VO2 peak ⱖ13 had better event-free survival than did those with ⬍13 ml/kg/min (p ⬍ 0.0001, Figure 1); those with VO2 peakⱖ12 had better event-free survival than did those with ⬍12 ml/kg/min (p ⫽ 0.0007, Figure 1). To compare the event-free survival and the survival excluding patients who underwent transplantation in

TABLE II VO2 peak testing results VO2 peak (ml/kg/min) RER V (liter/min) VE Heart rate reserve (bpm) Rest SBP (mm Hg) Rest DBP (mm Hg) Max. SBP (mm Hg) Max. DBP (mm Hg)

␤-blockerⴙ

␤-blockerⴚ

16.3 ⫾ 4.8 1.1 ⫾ .1 49.4 ⫾ 17.7 35.9 ⫾ 9.2† 50 ⫾ 21* 116 ⫾ 19 77 ⫾ 11 142 ⫾ 30 77 ⫾ 13

16.1 ⫾ 5.5 1.2 ⫾ 0.5 51.5 ⫾ 17.7 39.1 ⫾ 10.5 55 ⫾ 26 114 ⫾ 17 76 ⫾ 10 140 ⫾ 30 75 ⫾ 13

Data are expressed as mean ⫾ 1 SD. By unpaired t-test: *p ⬍ 0.05, †p ⬍ 0.0005, all other p ⫽ not significant. DBP, diastolic blood pressure (mm Hg); RER, respiratory exchange ratio; SBP, systolic blood pressure (mm Hg); V, ventilation (liter/min); VE, ventilatory equivalent ⫽ V/(VO2) (liters).

this database with survival rates of post-OHTx patients, we consulted the UNOS website. The 1- and 3-year mean ⫾ standard errors of percent survival are 85.1% ⫾ 0.6% and 76.0% ⫾ 0.3%, respectively. The 95% confidence intervals of the 1- and 3-year survival of transplant recipients in the UNOS database are 83.9% to 86.3% and 75.4% to 76.7%, respectively. Patients who were taking ␤-blockers and whose VO2 peak was either ⱖ14 ml/kg/min, ⱖ13 ml/kg/min, or ⱖ12 ml/kg/min had event-free survival rates at 1 and 3 years that were better than those of patients who underwent OHTx (Figure 1 and Table III). The 95% confidence intervals of event-free survival in these 3 groups of patients with HF did not overlap with and were better than the 1- and 3-year survival rates in post-OHTx patients (see Table IV). Among the sub-set of patients not taking ␤-blockers, those with VO2 peak ⱖ14, ⱖ13, and ⱖ12 had better event-free survival compared with those with VO2 peak ⬍14, ⬍13, and ⬍12 ml/kg/min, respectively (all p values are ⬍0.0001, Figure 2). Patients who did not take ␤-blockers but had VO2 peak ⱖ14 ml/kg/min had 1- and 3-year event-free survival rates superior to those of patients who had undergone OHTx, and we found no overlap between their 95% confidence intervals. However, ␤-blocker⫺ patients with VO2 peak ⬍14 ml/kg/min had a 3-year eventfree survival markedly worse than that of postOHTx patients (no overlap between their 95% confidence intervals; see Figure 2 and Table IV). To compare survival only with post-transplant survival (UNOS), we re-analyzed the data excluding the patients from our institution with HF who underwent transplantation. Table V shows that overall survival in patients with HF taking ␤-blockers who have VO2 peak ⱖ12 ml/kg/min had better 1-

The Journal of Heart and Lung Transplantation Volume 22, Number 10

Peterson et al.

1145

TABLE III 1- and 3-year event-free survival (mean ⫾ SE) ␤-blockerⴚ

284

91.7 ⫾ 2% 73.8 ⫾ 3% (ⱖ14 ml/kg/min) 97.0 ⫾ 3% 86.2 ⫾ 3% (⬍14 ml/kg/min) 82.0 ⫾ 4% 50.5 ⫾ 6% (ⱖ13 ml/kg/min) 95.7 ⫾ 1% 83.9 ⫾ 3% (⬍13 ml/kg/min) 81.4 ⫾ 5% 47.0 ⫾ 7% (ⱖ12 ml/kg/min) 95.1 ⫾ 2% 82.0 ⫾ 3% (⬍12 ml/kg/min) 79.8 ⫾ 5% 42.5 ⫾ 8%

1 year 3 years 1 year 3 years 1 year 3 years 1 year 3 years 1 year 3 years 1 year 3 years 1 year 3 years

␤-blockerⴙ

n

182 153 91 102 70 28 202 167 98 82 55 21 220 180 105 64 42 14

94.0 ⫾ 2% 88.6 ⫾ 3% (ⱖ14 ml/kg/min) 96.8 ⫾ 1% 93.0 ⫾ 3% (⬍14 ml/kg/min) 87.0 ⫾ 4% 78.0 ⫾ 6% (ⱖ13 ml/kg/min) 97.1 ⫾ 1% 92.6 ⫾ 3% (⬍13 ml/kg/min) 82.9 ⫾ 6% 74.4 ⫾ 7% (ⱖ12 ml/kg/min) 95.6 ⫾ 2% 90.9 ⫾ 3% (⬍12 ml/kg/min) 84.8 ⫾ 6% 75.9 ⫾ 8%

n 256 174 137 47 82 47 16 190 144 52 66 33 11 209 155 57 47 22 6

n ⫽ number of event-free patients at the beginning of the follow-up, and at 1 year and 3 years.

and 3-year survival than did those who underwent transplantation. In contrast, those not receiving ␤-blockers who had VO2 peak ⬍14 ml/kg/min had a worse 3-year survival than that of post-OHTx patients.

Table VI shows the results of a multiple, stepwise regression analysis. The independent predictors of mortality in our study were decreased VO2 peak,

TABLE IV 1- and 3-year event-free survival (95% confidence intervals)

1 year 3 years 1 year 3 years 1 year 3 years 1 year 3 years 1 year 3 years 1 year 3 years

␤-blockerⴚ

␤-blockerⴙ

(ⱖ14 ml/kg/min) (94.4–99.5) (80.6–92.0) (⬍14 ml/kg/min) (74.3–89.8) (38.9–62.1) (ⱖ13 ml/kg/min) (92.8–98.6) (78.1–89.8) (⬍13 ml/kg/min) (72.5–90.2) (33.9–60.1) (ⱖ12 ml/kg/min) (92.1–98.1) (76.1–87.8) (⬍12 ml/kg/min) (69.5–90.1) (27.3–57.7)

(ⱖ14 ml/kg/min) (94.0–99.6) (87.8–98.1) (⬍14 ml/kg/min) (78.4–95.5) (66.2–89.3) (ⱖ13 ml/kg/min) (94.5–99.6) (87.6–97.6) (⬍13 ml/kg/min) (71.8–93.9) (60.9–87.8) (ⱖ12 ml/kg/min) (92.6–98.6) (85.8–96.0) (⬍12 ml/kg/min) (72.2–97.5) (59.6–92.2)

FIGURE 1 A Kaplan-Meier curve of the event-free

survival of ␤-blocker (⫹) patients with heart failure as stratified by VO2 peak (ⱖ12 vs ⬍12 ml/kg/min, p ⬍ 0.0001) compared with the mean survival after heart transplantation at 1 and 3 years. The error bars at 1 and 3 years for the patients with heart failure depict the SEM. (The SEMs of the transplant recipients at 1 and 3 years are encompassed within the symbol depicting the mean survival). ␤-blocker (⫹) patients, those taking ␤blockers; UNOS, United Network for Organ Sharing.

1146

Peterson et al.

The Journal of Heart and Lung Transplantation October 2003

TABLE VI Independent predictors of mortality VO2 ⬍14 ml/kg/min VO2 ⬍13 ml/kg/min VO2 ⬍12 ml/kg/min Ischemic cause Men Non-Hispanic white race No ␤-blockers Age Increased VE

p p p p p p p p p

⬍ ⬍ ⬍ ⬍ ⫽ ⫽ ⬍ ⫽ ⬍

.0001 .0001 .0001 .0001 .01 NS .0001 .009 .0001

NS, not significant; VE, ventilatory equivalent.

FIGURE 2 A Kaplan-Meier curve of the event-free survival of ␤-blocker (-) patients with heart failure as stratified by VO2 peak (ⱖ14 vs ⬍14 ml/kg/min, p ⬍ 0.0001). The error bars at 1 and 3 years for the patients with heart failure depict the standard error of the mean. The mean and standard errors of the transplant recipients at 1 and 3 years also are shown. (The standard errors are encompassed within the symbols depicting the mean survival). ␤-blocker (-) patients, patients not taking ␤-blockers; UNOS, United Network for Organ Sharing. ischemic cause, male sex, no ␤-blocker treatment, increased age, and increased ventilatory equivalent.

DISCUSSION The selection of a population of patients with HF most likely to benefit from OHTx is important for several reasons. First, the shortage of donor hearts has created a mismatched supply/demand balance.13,14 Thus, apportioning donor hearts only to those most likely to benefit from OHTx would improve survival for the entire population. Second, OHTx is not without attendant morbidity and mortality. Heart transplantation is associated with operative mortality and may be complicated by right ventricular failure; acute rejection; toxicity of immunosuppressive drugs; and increased incidence of

TABLE V 1- and 3-year survival

␤-blocker⫹ ␤-blocker⫺

VO2 peak (ml/kg/min)

1-year % survival

3-year % survival

ⱖ12 ⬍12 ⱖ14 ⬍14

93–99 72–97 94–100 74–90

86–96 60–92 80–92 39–62

␤-blocker⫹, patients taking ␤-blockers; ␤-blocker⫺, patients not taking ␤-blockers.

infection, coronary artery disease, and immunosuppression-related neoplasms.14 –22 Exposing a patient with HF to these risks without a clear predicted survival benefit may be inappropriate. Therefore, it is important periodically to review and revise risk stratification methods for determining the optimal timing of transplantation in the context of advances in medical and surgical HF treatments. Mancini et al2,23 established VO2 peak as a simple test that would predict mortality in patients with HF and contribute important information about the timing of OHTx. The VO2 peak is a better predictor of survival and a better tool for timing OHTx than is New York Heart Association (NYCA) class, ejection fraction, 6-minute walk tests, and other markers of HF severity.23,24 The VO2 peak also predicts (and has a trend toward being an independent predictor [p ⫽ 0.055]) of event-free survival in patients with HF even in the current era of HF treated with ␤-blockers.25 However, the question of how ␤-blocker status may change the VO2 peak value at which transplantation no longer confers a survival advantage had not been addressed previously. Metra et al11 demonstrated that ␤-blockers do not significantly change VO2 peak. In a prospective, randomized study of 150 patients with HF that compared the effects of carvedilol with those of metoprolol, neither drug had a marked effect on VO2 peak (metoprolol, 13.7 ⫾ 4.5–15.0 ⫾ 5.1 ml/kg/ min, p ⬍ 0.01; carvedilol, 14.2 ⫾ 3.9 –14.0 ⫾ 4.6 ml/kg/min, p ⫽ not significant) despite significant improvements in ejection fraction.11 Because ␤-blocker therapy did not markedly change VO2 peak, despite a marked benefit in survival for patients with HF, we hypothesized that ␤-blocker⫹ patients would have a better survival rate even if their VO2 peak level was ⱖ12 ml/kg/min compared with patients who underwent OHTx. If this were the

The Journal of Heart and Lung Transplantation Volume 22, Number 10

case, ␤-blocker status should be considered when evaluating patients with HF for possible OHTx. The findings of our study support this hypothesis. The patients receiving ␤-blockers who had VO2 peak ⱖ12 ml/kg/min had 1- and 3-year event-free survival and 1- and 3-year survival (excluding those who were transplanted) rates that were superior to those of post-OHTx patients when the 95% confidence intervals of these 2 groups were compared. In contrast, the ␤-blocker⫺ group with VO2 peak ⬍14 ml/kg/min had a poorer 3-year event-free survival rate and a poorer survival rate (excluding those who were transplanted) than did post-OHTx patients. These findings suggest that the historical cutoff level for deferral of OHTx should be changed to VO2 peak ⱖ12 ml/kg/min for patients with HF who are taking ␤-blockers. Unfortunately, the number of patients receiving ␤-blocker with VO2 peak ⬍11 (n ⫽ 32) or ⬍10 ml/kg/min (n ⫽ 22) in this study was relatively small. Thus, the standard errors and standard deviations of the mean survival in these groups were so large that the calculated 95% survivals of these 2 groups were spread widely. Thus, these 95% confidence intervals were not very helpful in distinguishing the level of VO2 peak at which ␤-blocker⫹ patients would derive a survival benefit from OHTx. We found a few differences in the baseline characteristics of patients taking and those not taking ␤-blockers. The ␤-blocker⫹ subjects were more likely to be taking spironolactone and were more likely to be non-Hispanic, white patients (Table I). However, even though spironolactone has been shown to improve survival in patients with HF, this discrepancy is unlikely to account completely for the improvement in survival for those receiving ␤-blockers for the following reasons: 1) spironolactone conferred less of a survival benefit than did ␤-blockers in large clinical trials of each drug; 2) a relatively low percentage of ␤-blocker⫹ patients were receiving spironolactone; and 3) in a multivariate analysis of our findings, spironolactone status was not an independent predictor of mortality (Table VI).26 In addition, it is unlikely that race was a confounding factor in the prediction of survival based on VO2 peak and ␤-blocker status because race was not an independent predictor of events in the multivariate analysis. Recent studies that suggest that AfricanAmerican race does not have a marked effect on survival in patients with HF support this conclusion.27,28 The median event-free follow-up period was longer for the group of patients who were not taking ␤-blockers. This is most likely because ␤-blocker therapy for HF has become more com-

Peterson et al.

1147

mon. However, the median length of follow-up was relatively long in both groups, and the difference between the survival curves of the ␤-blocker⫹ and ⫺ patients demonstrating an improved survival in ␤-blocker⫹ patients (p ⫽ 0.0008) suggests that the difference in survival between these 2 groups is not solely because of a longer follow-up time in the ␤-blocker⫺ group. Table VI lists the independent predictors of events (and by extension, event-free survival). These are consistent with multiple published findings of others.25,29,30

Limitations This was an analysis of an observational database, not a randomized, controlled trial. Observed similarities or differences in outcomes may relate to differences in patients’ baseline characteristics or patient selection rather than to treatment effects. Pill counts were not performed during this study, so compliance with ␤-blockers therapy could not be assessed in the ␤-blocker⫹ patient group. Other predictors of survival and severity of HF, such as NYHA class and ejection fraction, were not available in this database. The VO2 peak levels and the ␤-blocker status before transplantation are not available in the UNOS database. The data from this study cannot necessarily be used to predict outcomes in patients whose profiles do not fit the entry criteria used in our study. This study was not designed to compare the effects of the different ␤-blocking agents on VO2 peak as it pertains to the timing of OHTx, nor was it designed to evaluate the effect of ␤-blocker therapy on the cause of mortality in the patients who died.

CONCLUSIONS The results of this study suggest that patients with HF treated with ␤-blockers have a prognosis superior to that of patients who undergo OHTx, as long as the former have VO2 peak ⱖ12 ml/kg/min. Patients with HF who are taking ␤-blockers with VO2 peak ⬍12 ml/kg/min and those unable to tolerate ␤-blockers who have VO2 peak ⬍14 ml/kg/min derive a survival advantage from OHTx. Thus, ␤-blocker status should be considered when evaluating patients with HF for possible OHTx. The authors thank Dr. Ali Ehsani for his mentorship and gratefully acknowledge the secretarial assistance of Ava Ysaguirre; the editorial assistance of Beth Engeszer; and the data collection efforts of Dr. Omar Shaher, Allison Barrett, Tamara Donahue, Rhonda Barrett-Avery, Matt

1148

Peterson et al.

The Journal of Heart and Lung Transplantation October 2003

Kueper, and Elizabeth Ballard in the Division of Applied Physiology.

REFERENCES 1. UNOS Scientific Registry data as of September 5, 2000. OPTN/SR 2000 Annual Report. 2. Mancini DM, Eisen H, Kussmaul W, Mull R, Edmunds LH, Wilson JR. Value of peak exercise consumption for optimal timing of cardiac transplantation in ambulatory patients with heart failure. Circulation 1991;83:778 –6. 3. Packer M, Bristow MR, Cohn JN, et al, for the US Carvedilol Heart Failure Study Group. The effect of carvedilol on morbidity and mortality in patients with chronic heart failure. N Engl J Med 1996;334:1349 –55. 4. MERIT-HF Study Group. Effect of metoprolol CR/XL in chronic heart failure: Metoprolol CR/XL Randomised Intervention Trial in Congestive Heart Failure (MERIT-HF). Lancet 1999;353:2001–7. 5. Packer M, Coats AJS, Fowler MB, et al. Effect of carvedilol on survival in severe chronic heart failure. N Engl J Med 2001;344:1651–8. 6. Di Lenarda A, Maria RD, Gavassa A, et al, on behalf of the Italian Multicenter Cardiomyopathy Study Group. Long term survival effect of metoprolol in dilated cardiomyopathy. Heart 1998;79:337–44. 7. Waagstein F, Bristow MR, Swedberg K, et al, for the Metoprolol in Dilated Cardiomyopathy Trial Study Group. Beneficial effects of metoprolol in idiopathic dilated cardiomyopathy. Lancet 1993;342:1441–6. 8. The CAPRICORN Investigators. Effect of carvedilol on outcome after myocardial infarction in patients with left ventricular dysfunction: the CAPRICORN randomized trail. Lancet 2001;357:1385–90. 9. CIBIS II Investigators and Committees. The Cardiac Insufficiency Bisoprolol Study II: a randomized trial (CIBIS II). Lancet 1999;353:9 –13. 10. Metra M, Nodari S, D’Aloia A, et al. A rationale for the use of ␤-blockers as standard treatment for heart failure. Am Heart J 2000;139:511–21. 11. Metra M, Giubbini R, Nodari S, et al. Differential effects of beta-blockers in patients with heart failure. A prospective, randomized, double-blind comparison of the long-term effects of metoprolol vs. carvedilol. Circulation 2000;102:546 – 51. 12. Kohrt WM, Malley MT, Coggan AR, et al. Effects of gender, age, and fitness level on the responses of VO2max to training in 60 –71 yr-olds. J Appl Physiol 1991;71:2004 –11. 13. Rayburn BK, Burton TM, Wannenburg T, et al. Are efforts at expanding the donor pool misdirected? J Heart Lung Transplant 1998;17:998 –1003. 14. Miniati DN, Robbins RC, Reitz BA. Heart and heart-lung

15.

16.

17. 18. 19.

20. 21.

22. 23. 24. 25.

26.

27.

28.

29.

30.

transplantation. In: Braunwald E, Zipes DP, Libby P, eds. Heart Disease: A Textbook of Cardiovascular Medicine, 6th ed. Philadelphia: WB Saunders, 2001, pp. 615–34. Puig M, Ballester M, Matias-Guiu X, et al. Burden of myocardial damage in cardiac allograft rejection. Scintigraphic evidence of myocardial injury and histologic evidence of myocytes necrosis and apoptosis. J Nuclear Cardiol 2000; 7:132–9. Normann SJ, Salomon DR, Leelachaikul P, et al. Acute vascular rejection of the coronary arteries in human heart transplantation. Pathology and correlations with immunosuppression and cytomegalovirus infection. J Heart Lung Transplant 1991;10:674 –87. Andoh TF, Bennett WM. Chronic cyclosporine nephrotoxicity. Curr Opin Nephrol Hypertens 1998;7:265–70. Luke RG. Mechanism of cyclosporine-induced hypertension. Am J Hypertens 1991;4:468 –71. Horn JE, Bartlett JG. Infectious complications following heart transplantation. In: Baumgartner WA, Reitz BA, Achuff SA, eds. Heart and Heart-Lung Transplantation. Philadelphia: WB Saunders, 1990, p. 220. Penn I. Tumors after renal and cardiac transplantation. Hematol Oncol Clin North Am 1993;7:431–45. Miller LW, Schlant RC, Kobashigawa J, et al. 24th Bethesda conference: cardiac transplantation. Task Force 5: complications. J Am Coll Cardiol 1993;22:41–54. Thomson JG. Production of severe atherosclerosis in a transplant human heart. Lancet 1969;2:1088 –92. Mancini D, LeJemtel T, Aaronson K. Peak VO2: a simple yet enduring standard. Circulation 2000;101:1080 –2. Rector TS, Cohn JN. Prognosis in congestive heart failure. Ann Rev Med 1994;45:341. Peterson LR, Shectman KB, Ewald GA, et al. The effect of beta-adrenergic blockers on the prognostic value of peak exercise oxygen uptake in patients with heart failure. J Heart Lung Transplant 2003;22:70 –7. Pitt B, Zannad F, Remme WJ, et al. The effect of spironolactone on morbidity and mortality in patients with severe heart failure. N Engl J Med 1999;341:709 –17. Alexander M, Grumbach K, Remy L, et al. Congestive heart failure hospitalizations and survival in California: patterns according to race/ethnicity. Am Heart J 1999;137:919 –27. Chin M, Goldman L. Gender differences in 1-year survival and quality of life among patients admitted with congestive heart failure. Med Care 1998;36:1033–46. Ho KK, Anderson KM, Kannel WB, et al. Survival after the onset of congestive heart failure in Framingham Heart Study subjects. Circulation 1993;88:107. Adams KF, Dunlap SH, Suteta CA, et al. Relation between gender, etiology and survival in patients with symptomatic heart failure. J Am Coll Cardiol 1997;28:1781.