|
|
||||||||
J Thorac Cardiovasc Surg 2008;135:1159-1166
© 2008 The American Association for Thoracic Surgery
Cardiothoracic Transplantation |
a Department of Thoracic and Cardiovascular Surgery, Cleveland Clinic, Cleveland, Ohio
b Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, Ohio
c Department of Cardiovascular Medicine, Cleveland Clinic, Cleveland, Ohio
d Kaufman Center for Heart Failure, Cleveland Clinic, Cleveland, Ohio
Received for publication November 13, 2007; accepted for publication January 25, 2008. * Address for reprints: Katherine J. Hoercher, RN, Kaufman Center for Heart Failure, Cleveland Clinic, 9500 Euclid Ave/JJ-40, Cleveland, OH 44195. (Email: hoerchk{at}ccf.org).
| Abstract |
|---|
|
|
|---|
Methods: From January 1985 to February 2004, 100 status 2 patients were delisted for medical improvement (median on-list duration, 314 days). Return of heart failure, sudden cardiac death, and all-cause mortality were determined from follow-up (mean, 7.7 ± 3.9 years among survivors; 10% followed >12 years). Hazard function modeling, competing-risks analyses, simulation, and propensity matching to equivalent patients undergoing transplantation were used to analyze and compare outcomes and predict benefit of interventions.
Results: Freedom from return of heart failure was 77% at 5 years. The most common mode of death was sudden cardiac death, with risk peaking at 2.5 years after delisting but remaining at 3.5% per year thereafter. Event-free survival at 1, 5, and 10 years was 94%, 55%, and 28%, respectively; simulation demonstrated that implantable cardioverter–defibrillators could have improved this to 45% at 10 years. Overall survival after delisting was better than that of matched status 2 patients who underwent transplantation, but was demonstrably worse after 30 months.
Conclusions: Status 2 patients, including those delisted, require vigilant surveillance and optimal medical management, implantable cardioverter–defibrillators, and a revised approach to transplantation timing, such that overall salvage is maximized while allocation of scarce organs is optimized.
| Introduction |
|---|
|
|
|---|
In the absence of such a trial, we sought to augment short-term data with long-term information by studying the fate of status 2 patients removed over the years from our heart transplant waiting list (delisted) because of medical improvement. We considered such patients as surrogates for medical therapy versus transplantation in stable outpatients with chronic heart failure. Specifically assessed were return of heart failure and occurrence of sudden cardiac death (SCD), interventions to improve outcomes, and survival compared with that of status 2 patients undergoing transplantation.
| Materials and Methods |
|---|
|
|
|---|
Cases were identified through the Cleveland Clinic Unified Transplant Database, which is used for UNOS reporting and has been approved for use in research by the institutional review board, with patient consent waived. These data were confirmed and supplemented by review of each patient's clinical records to determine the reason for delisting, time on the waiting list, UNOS status and New York Heart Association (NYHA) functional class at delisting, and, when available, peak oxygen consumption (
) during exercise at listing and delisting.
Clinical characteristics of patients at listing and delisting are presented in
Table 1. Of this group, 91% were aggressively treated with optimal medical therapy for a minimum of 9 months before listing. Despite optimal medical therapy, persistent NYHA class III or IV symptoms and
of less than 14 mL · kg–1 · min–1 guided the decision to list potential transplant candidates. The remaining patients were admitted with cardiogenic shock requiring mechanical circulatory support and listed for transplantation within a week of referral.
|
Definitions and Indications
Medically improved patients were those who were in UNOS status 2 for more than 6 months with objective improvement in NYHA class (all but 1 known to be in NYHA class I or II at delisting, Table 1). Indications for delisting were improved
during exercise from listing to delisting, subjective improvement in heart failure symptoms, and a heart failure team consensus that cardiac transplantation would not provide a survival benefit over medical management. No formal surveillance program was implemented; rather, patients were remanded to the care of their local health care providers.
Outcomes
Patients or their surviving families were followed cross-sectionally by using an institutional review board–approved telephone script, with a common closing date of September 15, 2005. Follow-up was 100% complete, with mean follow-up of 7.7 ± 3.9 years among survivors; 10% were followed more than 12 years. Outcomes elicited included (1) return of heart failure, defined as return of NYHA class IV heart failure symptoms, relisting for transplantation, institution of mechanical circulatory support, or reported death in heart failure; (2) mode of death (heart failure, SCD, or noncardiac death); (3) insertion of an implantable cardioverter–defibrillator (ICD) among the 99 patients not having such a device inserted before delisting; and (4) all-cause mortality at any time during follow-up.
Data Analysis
All time-related post-delisting events were analyzed nonparametrically by using the method of Kaplan and Meier and parametrically by multiphase hazard modeling.5
(For additional details, see http://www.clevelandclinic.org/heartcenter/hazard.) Multivariable analysis of outcomes was performed in the hazard function domain by using bagging as the primary variable selection method.6,7
For this, 1000 bootstrap data sets were analyzed with a selection criterion of P = .05, followed by aggregation of models by the median rule. Variables used in these analyses were age at delisting (and its potential linearizing transformations of scale), sex, race, indication for listing (ischemic vs dilated cardiomyopathy vs other), dysrhythmia during listing (atrial fibrillation or flutter, sustained ventricular tachycardia, and SCD), NYHA class at listing and delisting, UNOS status at listing, time on the waiting list, and date of delisting, expressed as the interval between 1985 and delisting.
Event-free survival was defined as freedom from 3 end states: return of heart failure, SCD, and all other modes of death, each censored at any subsequent transplantation. Nonparametric simultaneous estimation of cumulative incidence for these events used the multiple decrement method of Andersen and colleagues.8
For parametric estimation, rate of transition (hazard function) from event-free survival to each end state was estimated, and event-free survival and cumulative incidence of end states were calculated by using numeric integration. To estimate the potential benefit of ICD insertion, we simulated event-free survival were SCD completely eliminated.
Survival after delisting for medical improvement was compared with survival after transplantation for similar patients. For this, patients who underwent transplantation as UNOS status 2 were matched with delisted patients on age, sex, race, UNOS status at listing, diagnosis at listing, transplantation date, and waiting time to transplantation or delisting by using propensity score methodology.7,9
This yielded 85 matched pairs.
| Results |
|---|
|
|
|---|
Risk of return of heart failure increased rapidly to 3.5% per year and then rose more slowly thereafter—but never leveled off or decreased—such that by 5 years it was nearly double that at 1 year (7.2% per year). This resulted in freedom from return of heart failure at 1, 5, and 10 years of 97%, 77%, and 51%, respectively (
Figure 1). The only incremental risk factor for return of heart failure was older age at delisting (
Table 2).
|
|
|
|
|
|
|
| Discussion |
|---|
|
|
|---|
Temporal Changes in Heart Failure Severity
An increase in peak
levels that occurred during listing, which might have been due to the intensity of therapy in such patients, conferred only a short-term benefit but was not predictive of late outcomes. Favorable short-term outcomes (<2 years) in listed patients who increased their
levels has also been reported in some,10,11
but not all,12
series. The 9 patients listed for cardiogenic shock and requiring temporary mechanical circulatory support all had dramatic return of cardiac function and were subsequently delisted.
Risk of return of heart failure symptoms increased rapidly within a year of delisting and then slowly but steadily continued to increase. Late return of heart failure symptoms is not surprising because heart failure is a chronic and progressive disease. Importantly, follow-up care, especially with regard to contemporary disease management, was neither systematic nor intensive. Only 5% of patients were followed by a heart failure specialist after delisting; thus, the levels of care and compliance with heart failure therapies were unknown, and critically, the majority of patients who returned to NYHA class IV heart failure were not referred back to our transplantation center. Previous studies, as well as recent data, suggest that outcomes are improved by the attentive care provided within a heart failure disease management program.13-15
Sudden Cardiac Death
SCD was the most common mode of death for these patients, accounting for 45% of deaths. By means of simulation, we estimated that an ICD would have substantially improved intermediate and long-term event-free survival in this population after delisting because it clearly affects a mechanism of death not captured by improved functional capacity or exercise tolerance. Results of this simulation are similar to those observed in the Multicenter Automatic Defibrillator Implantation Trial II study of ICD therapy in patients with prior myocardial infarction and reduced left ventricular function, as well as in clinical studies of stable heart transplant candidates.16-18
Beyond optimal medical therapy, the Sudden Cardiac Death in Heart Failure Trial demonstrated a 46% relative reduction in the risk of SCD in NYHA class II patients with an ICD and further explains the high risk of SCD in our stable heart failure group, patients who, for the most part, did not have an ICD.19
In a recent analysis of 50,000 UNOS registrants awaiting heart transplantation between 1990 and 2005, the presence of an ICD in status 1 candidates on the day of listing correlated with improved outcomes.20
This same study demonstrated that 1-year survival of non-transplanted status 2 candidates in the most recent era (2000–2005) is approaching that of survival after transplantation. Although the authors were unable to determine the specific reasons for these improved outcomes, it is important to note that in the most recent era, 47% of listed status 2 patients had an ICD compared with 18% between 1995 and 1999 and 0% between 1990 and 1994.
Although the results of the Multicenter Automatic Defibrillator Implantation Trial II and Sudden Cardiac Death in Heart Failure Trial in NYHA class II patients are well known, and furthermore, ICD implantation is a class IA recommendation from the 2006 American College of Cardiology/American Heart Association/European Society of Cardiology guidelines,21
Lietz and Miller20
have shown that even in the most recent era, fewer than half of status 2 patients on the UNOS waiting list had ICDs. The ominous findings of SCD in our own study clearly demonstrate the need for ICDs in these "low-risk" patients.
Survival
Recent studies question whether cardiac transplantation provides a survival benefit over optimal medical therapy in status 2 patients.1,2
To date, all published survival data comparing medical management of transplant candidates with transplanted patients are limited by follow-up of only 1 to 5 years.10,11,22,23
The strength of our investigation is the 10-year follow-up. If one assumes that our delisted patients were the healthiest or lowest-risk patients, their survival should be better than that of the average status 2 patient after transplantation, and in our experience this is in fact the case out to at least 18 months. However, heart failure and SCD escalate thereafter, and data from both our study and prior investigations are insufficient to predict who will deteriorate over time.
Limitations
This is a nonrandomized observational study encompassing 20 years of experience from a single high-volume transplantation center but involving a relatively small number of patients. Numerous changes in clinical management occurred during this time, and multiple surgeons and cardiologists were involved in caring for these patients. Unfortunately, we were unable to quantify the intensity and quality of care after delisting to determine how these might have affected outcome. Major advances in medical and device therapies occurred as well, including use of β-blockers, ICDs, cardiac resynchronization therapy, and dissemination of evidence-based care. Importantly, because of improved survival in patients treated with β-blockers, the traditional cutoff point of less than 14 mg · kg–1 · min–1 for transplantation is being reassessed; a value of less than 12 mg · kg–1 · min–1 might be more appropriate.
Clinical Implications and Recommendations for Practice Guidelines
Although the outcomes of this study cannot be extrapolated to all status 2 transplant candidates, and we do not believe it to be a direct substitute for a trial of tailored medical therapy versus transplantation, we believe our findings have several implications for both status 2 patients and those being considered for delisting.
First, all patients meeting guideline criteria should have an ICD implanted.
Second, follow-up must be intense and is best done by heart failure specialists as if patients remained on the transplant waiting list.
Third, close periodic surveillance (every 3–6 months) is required to identify patients whose cardiac function is deteriorating.
Fourth, additional methods for monitoring both listed and delisted patients should be investigated, including implantable hemodynamic devices, microvolt T-wave alternans, and biomarkers, to anticipate deterioration earlier and prompt renewed consideration for relisting or status upgrade.
Fifth, timing of transplantation in this population remains a dilemma. More detailed and longer-term investigation into the prediction of survival for status 2 patients is needed, as has already been done for status 1 patients. Reliable predictive models for status 2 patients might help discriminate low- from high-risk patients, helping to guide time of listing.
Sixth, UNOS modifications to the current allocation system might be required when accurate predictive models become available. If the number of patients listed continues to decrease and efforts to increase organ donation are successful, more status 2 patients might be offered organs. In that case, one would need to balance long-term survival of the graft and morbidity associated with long-term immunosuppression against good short-term survival of low-risk patients who do not undergo transplantation. Consequently, it might be that in this situation, the lowest-risk patients should have transplantation delayed as long as possible and undergo transplantation only when matched to a perfect heart.
| Acknowledgments |
|---|
| Footnotes |
|---|
Randall C. Starling reports consulting fees from Medtronic, James B. Young reports consulting fees from Guidant and Medtronic and grant support from Medtronic.
* Supported in part by the Kenneth Gee and Paula Shaw, PhD, Chair in Heart Research. ![]()
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
S. Westaby, M. Siegenthaler, F. Beyersdorf, M. Massetti, J. Pepper, A. Khayat, R. Hetzer, and O. H. Frazier Destination therapy with a rotary blood pump and novel power delivery Eur J Cardiothorac Surg, February 1, 2010; 37(2): 350 - 356. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |