- Email: [email protected]
Outcomes in Patients Older Than 60 Years of Age Undergoing Orthotopic Heart Transplantation: An Analysis of the UNOS Database
Outcomes in Patients Older Than 60 Years of Age Undergoing Orthotopic Heart Transplantation: An Analysis of the UNOS Database Eric S. Weiss, MD, Lois U. Nwakanma, MD, Nishant D. Patel, BA, and David D. Yuh, MD Background: Patients 60 years and older have traditionally not been considered candidates for orthotopic heart transplantation (OHT). Recent studies have shown equivalent survival between older and younger patients, leading many to question this traditional ethos. As these studies may lack significant power to draw meaningful conclusions, the United Network for Organ Sharing (UNOS) database provides a unique opportunity to examine the effects of age on OHT. Methods: We retrospectively reviewed the UNOS dataset to identify 14,401 first-time OHT recipients between the years 1999 and 2006. Stratification was by age into those ⱖ60 years and younger patients aged 18 to 59 years. Baseline demographic and clinical factors were recorded. The primary end-point was all-cause mortality during the study period. Secondary outcomes included length of hospital stay (LOS), post-operative stroke, post-operative infections, acute renal failure (ARF) and rejection within 1 year of transplant. Post-transplant survival was modeled using the Kaplan–Meier method and compared between groups using Cox proportional hazard regression. Results: Of the 14,401 patients who met the inclusion criteria, 30% (n ⫽ 4,273) were ⱖ60 years of age. The elderly group had higher serum creatinine levels (1.5 vs 1.3, p ⬍ 0.001), longer waitlist times (255 vs 212 days, p ⬍ 0.001), and were more likely to have hypertension (HTN; 46% vs 37%, p ⬍ 0.001) or diabetes mellitus (DM; 25% vs 20%, p ⬍ 0.001). Survival at 30 days, 1 year and 5 years was 94%, 87% and 75% for the young group, and 93%, 84% and 69% for the older group (p ⬍ 0.001). Multivariate analysis revealed age ⱖ60 years, donor age, ischemic time, creatinine, HTN and DM to be independent predictors of mortality. Older patients had more infections (26% vs 23%, p ⬍ 0.001), ARF (9% vs 7%, p ⬍ 0.001) and longer LOS (21 vs 19 days, p ⬍ 0.001), but had lower rates of rejection (34% vs 43%, p ⬍ 0.001) as compared with younger recipients. Conclusions: The UNOS database has provided a large multi-institutional sample examining OHT in the elderly. Although our analysis shows lower survival in patients ⱖ60 years of age, the cumulative 5-year survival in these patients of close to 70% is acceptable. OHT should not be restricted based on age, as encouraging long-term results exist. J Heart Lung Transplant 2008;27:184 –91. Copyright © 2008 by the International Society for Heart and Lung Transplantation.
With life expectancy on the rise and older individuals comprising the fastest growing segment of the U.S. population, the demographics of orthotopic heart transplantation (OHT) are in flux. Although it is known that OHT represents the best treatment for many forms of From the Division of Cardiac Surgery, The Johns Hopkins University School of Medicine, Baltimore, Maryland. Submitted July 24, 2007; revised November 12, 2007; accepted November 15, 2007. Presented at the 27th annual meeting and scientific sessions of the International Society of Heart and Lung Transplantation, San Francisco, California, April 2007. Supported in part by the Health Resources and Services Administration, Contract No. 234-2005-370011C. Reprint requests: Eric S. Weiss, MD, Division of Cardiac Surgery, The Johns Hopkins University School of Medicine, 1820 Bank Street, Baltimore, MD 21231. Telephone: 410-539-7677. Fax: 410-522-5048. E-mail: [email protected] Copyright © 2008 by the International Society for Heart and Lung Transplantation. 1053-2498/08/$–see front matter. doi:10.1016/ j.healun.2007.11.566
184
advanced heart failure, the utility of this therapeutic modality in elderly patients continues to be a subject of debate. With advances in the medical and surgical management of heart failure and the paucity of available organs, it is clear that a consensus must be reached regarding the appropriate utilization of OHT for elderly patients. Despite this lack of uniform agreement, it has become clear in recent years that elderly patients are undergoing more frequent OHT nationwide. The International Society of Heart and Lung Transplantation reports that, since 1995, the percentage of transplant patients over the age of 65 has increased steadily, while the percentage of those between 40 and 49 years of age has decreased steadily.1 This trend is echoed by UNOS data, which show a steady increase in the proportion of recipients ⬎60 years of age undergoing OHT (Figure 1). Notwithstanding this increase, many centers still view advanced age as a contraindication to OHT. The literature examining this issue has been mixed, with
The Journal of Heart and Lung Transplantation Volume 27, Number 2
some series reporting worse outcomes among elderly patients2–5 and others reporting equivalent rates of mortality, rejection and transplant coronary artery disease.6 –9 Examination of age has extended to patients as old as 70 years of age, showing that OHT can be performed safely in selected candidates.10 A central problem surrounding the literature examining OHT in the elderly is the fact that most studies directly examining outcomes in elderly patients have been conducted in single centers with small patient cohorts. The United Network for Organ Sharing (UNOS) dataset is a nationwide, physician-regulated registry that provides a unique opportunity to examine OHT in a large population of elderly patients free of institutional bias. We conducted a retrospective examination of the multiinstitutional UNOS dataset to examine both short- and long-term outcomes of elderly patients undergoing OHT. We hypothesize that elderly patients have equivalent outcomes to younger patients when examined on a large scale with a modern cohort. METHODS Data Source UNOS provided de-identified patient data (Standard Transplant Analysis and Research [STAR] files with follow-up files) from the thoracic organ transplant registry for the years 1987 to 2006, with follow-up through January 2007. The data include all U.S. patients undergoing thoracic organ transplantation and who were reported to the Organ Procurement Network during the time period. No patient or center identifiers are included. Study Design and Patient Population We conducted a retrospective review of the UNOS registry from January 1999 to December 2006, with follow-up through January 2007. The time-points were chosen to identify a modern cohort of patients who underwent OHT free of bias related to advances in heart transplantation techniques and immunosuppressive regimens. All first-time OHT patients ⬎18 years of age were included. Patients were primarily stratified by age (⬍60 years vs ⱖ60 years) and were secondarily stratified into additional age groups for subsequent analysis (ⱕ49, 50 to 54, 55 to 59, 60 to 64, 65 to 69 and ⱖ70 years). Outcome Measures Baseline demographic and clinical factors were examined. The primary end-point was all-cause mortality during the study period. Secondary outcomes included length of hospital stay (LOS), post-operative infections (defined by the need for antibiotics), post-operative stroke, development of transplant coronary artery disease (TCAD), new-onset dialysis requirement, and rejec-
Weiss et al.
185
tion within 1 year (defined as any treatment of rejection for all causes). Statistical Analysis Comparisons of baseline characteristics between study groups were performed using Student’s t-test for continuous variables and the chi-square test for categorical variables. Cumulative survival was modeled using the Kaplan–Meier method with statistical differences between survival curves assessed using the Mantel–Cox log-rank test. In addition, modeling based on 30-day and 1-year conditional survival was employed to assess long-term survival independent of early mortality. Mortality was assessed for all risk factors using a univariate model. Multivariate analysis was conducted with a Cox proportional hazard regression model for cumulative survival and with multivariate logistic regression for short-term mortality. Significant univariate predictors of mortality were incorporated into the multivariate models in a stepwise fashion to assess the effect of technique on mortality. Only well-represented variables (⬍33% missing in the registry) were included in the multivariate analysis. In addition, univariate and multivariate logistic and linear regression analyses, when appropriate, were performed for all additional outcome variables. Two-tailed p ⬍ 0.05 was considered statistically significant. All odds ratios and regression coefficients are presented with along with the 95% confidence interval (CI). All statistical analyses were performed using STATA software, version 9.0 (StataCorp LP, College Station, TX). RESULTS A total of 14,401 adult patients underwent primary OHT during the study period. Of these, 4,273 (30%) were ⱖ60 years of age. In addition, 212 (1.4%) were ⱖ70 years of age. Although the mean age of OHT recipients has remained relatively constant since 1987, the percentage of patients ⬎60 years of age has increased steadily. This trend continued during the study period of 1999 to 2006 (Figure 1). As expected, older patients had a higher incidence of baseline co-morbidities, such as diabetes and hypertension and higher baseline creatinine levels, although, perhaps surprisingly, their body mass indices (BMIs) were comparable to those of younger patients (Table 1). Patient age groups were well matched for other markers of clinical acuity such as hospitalization and ICU admission rates, as well as the prevalence of pre-operative intra-aortic balloon counterpulsation (IABP) and mechanical ventilation. In addition, patients were well matched for pre-operative (baseline) hemodynamic variables. Older patients did have longer waitlist times (255 vs 212 days, p ⬍ 0.001), which in part
186
Weiss et al.
The Journal of Heart and Lung Transplantation February 2008
Figure 1. Trends in age of OHT recipients over time. Bars: transplants performed in patients ⱖ60 years of age; solid line: mean age for adult OHTs for reference. Although the mean age for adult OHTs has remained stable, the mean age for elderly patients has increased since 1987 (based on OPTN data, September 2006).
may have been related to a greater percentage being listed as UNOS Status 2 (Table 1). Older patients also had a greater history of malignancy (5.3% vs 3.7%, p ⬍ 0.001). Both females and ethnic minorities comprised a smaller percentage of the total number of older patients in the study population (Table 1). Older recipients did receive hearts from older donors and donors with a greater prevalence of diabetes and infections, indicating that older patients may be receiving organs from higher risk donors. Cumulatively, 3,027 patients died during the study period, for an all-cause mortality of 21%. Older patients (ⱖ60 years old) had higher 5-year overall mortality rates (24.5% vs 19.9%, p ⬍ 0.001) as well as increased 30-day and 1-year mortality rates (30-day: 6.7% vs 5.6%, p ⫽ 0.002; 1-year: 14.7% vs 11.5%, p ⬍ 0.001) (Table 2). With multivariate Cox proportional hazards regression analysis, age as a continuous variable was not significantly associated with mortality, but ages ⱖ55, ⱖ60 and ⱖ65 years were all significant predictors of mortality independent of other known risk factors for death (Table 3). When examining different age-range cut-offs, patients ⱖ60 years possessed the highest 5-year cumulative hazard ratio (HR) for death (HR ⫽ 1.28 [CI 1.15 to 1.42], p ⬍ 0.001) as compared with the other age cut-offs (Table 3). When examining both 30-day and 1-year mortality rates, controlling for confounders in a mutivariate logistic regression model, older patients only showed a significantly greater risk of mortality at 1 year posttransplant (odds ratio [OR] for death at 30 days ⫽ 1.17 [CI 0.95 to 1.46], p ⫽ 0.12; OR for death at 1 year ⫽ 1.29 [CI 1.10 to 1.50], p ⫽ 0.001). Additional predictors of cumulative mortality on multivariate analysis included: female gender; ischemic time; age of donor; and markers of acuity such as history of diabetes, history of
hypertension, mechanical ventilation prior to transplant, and ICU admission prior to transplant. On subset analysis, patients ⱖ70 years of age (n ⫽ 212) also had significantly worse cumulative mortality, although 30-day and 1-year mortality did not reach statistical significance. On multivariate analysis (likely related to small sample size), age ⱖ70 years did not emerge as an independent predictor of mortality. When survival was modeled using the Kaplan–Meier method, the older group only showed a 6% lower rate of survival at 5 years. Although this reached statistical significance, the absolute difference was small (69% vs 75%) (Figure 2). The difference in survival narrowed when cumulative survival conditional on 30-day and 1-year survival was modeled. When censoring those patients who died at 1 year, cumulative survival differed by only 3% between older and younger patients (83% vs 86%, respectively, p ⫽ 0.01), indicating that a substantial risk of mortality for older patients occurs in the first year (Figure 3). When patients were stratified by age range, there was also a progressive stepwise decrease in cumulative survival for all progressive age groups after the age of 60 years (Figure 4). Other outcome variables examined included the development of transplant coronary artery disease (TCAD) at follow-up, development of post-operative infections (requiring antibiotic treatment), new-onset dialysis, post-operative stroke, rejection requiring treatment (for all causes) at 1 year, and LOS. Older patients did have increased new-onset dialysis needs, longer LOS, more post-operative infections and a higher risk of post-operative stroke, although only differences in LOS and new-onset dialysis were confirmed on multivariate analysis (Table 4). Interestingly, TCAD did not differ between older and younger patients on either univari-
The Journal of Heart and Lung Transplantation Volume 27, Number 2
Weiss et al.
187
Table 1. Baseline Characteristics of Elderly (ⱖ60 years) Versus Young Patients (⬍60 years) Young (⬍60 years) N ⫽ 10,128a
Number with data available (%)
Elderly (ⱖ60 years) N ⫽ 4,273a
Number with data available (%)
p-valueb
Demographics Mean age Female Black, Hispanic, Native American, Asian/Pacific Islanderc
47 (⫾10.7) 2,743 (27)
10,128 (100) 10,128 (100)
64 (⫾3.1) 688 (16)
4,273 (100) 4,273 (100)
⬍0.001 ⬍0.001
2,896 (29)
10,128 (100)
621 (14)
4,273 (100)
⬍0.001
Acuity Diabetes HTN Creatinine BMI Hospitalizedd ICU prior to transplant Ventilator prior to transplant IABP prior to transplant Donor age (years) Donor diabetes Donor Hypertension Donor creatinine Donor infection ⫾ ischemic time (hours) Days on waitlist History of malignancy UNOS Status 1e
1,936 (20) 3,562 (37) 1.33 (⫾0.9) 26.5 (⫾4.8) 5,230 (52) 3,120 (31) 287 (2.8) 527 (5.2) 30.8 (12) 181 (1.8) 1,626 (16.2) 1.20 (0.01) 2,605 (27.3) 3.12 (⫾1.0) 212 (⫾352) 373 (5.3) 7,642 (76)
9,930 (98) 9,730 (96) 9,640 (95) 9,355 (92) 9,988 (99) 9,988 (99) 9,841 (97) 10,128 (100) 10,123 (99) 10,080 (99) 10,029 (99) 10,043 (99) 9,524 (94) 8,996 (88) 9,413 (93) 10,081 (99) 10,121 (99)
1,064 (25) 1,896 (46) 1.45 (⫾0.7) 26.1 (⫾4.0) 2,049 (49) 1,251 (29) 131 (3.0) 226 (5.2) 33.3 (13) 100 (2.3) 718 (16.9) 1.21 (0.01) 1,168 (29.2) 3.16 (⫾1.0) 255 (⫾419) 225 (3.7) 3,025 (71)
4,184 (98) 4,102 (96) 4,053 (97) 3,940 (95) 4,220 (99) 4,220 (99) 4,142 (97) 4,273 (100) 4,269 (99) 4,242 (99) 4,236 (99) 4,237 (99) 3,999 (94) 3,796 (89) 3,953 (93) 4,257 (99) 4,272 (99)
⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.06 0.45 0.88 ⬍0.001 0.04 0.2
3,523 (82) 3,234 (76) 3,493 (82) 3,395 (79)
⬍0.001 0.87 0.40 0.91
Hemodynamic variables Mean PA pressure PVR Cardiac index TPG
28.5 (⫾10) 2.4 (⫾1.9) 1.5 (⫾0.55) 9.5 (⫾5.7)
8,339 (82) 7,516 (74) 8,217 (81) 7,924 (78)
27.9 (⫾10) 2.3 (⫾2.2) 1.5 (⫾0.49) 9.5 (⫾5.6)
0.03 0.10 ⬍0.001 ⬍0.001 ⬍0.001
PA, pulmonary artery; HTN, hypertension; BMI, body mass index; ICU, intensive care unit; IABP, intra-aortic balloon pump; PVR, pulmonary vascular resistance (defined by [mean PA pressure (mPAP) ⫺ pulmonary capillary wedge pressure (PCWP)]/cardiac output); TPG, transpulmonary gradient (defined by MPAP ⫺ PCWP); Donor infection ⫾ represents any type of infection present in the donor at time of transplant. a Data presented as number of patients (%) or mean (⫾SD). b p-value is based on comparison between two groups by either chi-square or Student’s t-test (with p ⬍ 0.05 considered statistically significant). c Race or ethnic group was a variable present in the dataset. d Hospitalized refers to hospitalization prior to transplantation. e UNOS Status 1 refers to patients listed as Status 1a, 1b or old UNOS Status 1.
ate or multivariate analyses. Importantly, univariate and multivariate analyses showed that older age was associated with significantly less treatment for rejection within 1 year of transplant (34 vs 43%, OR ⫽ 0.67, p ⬍ 0.001). DISCUSSION From national registry reports it is clear that the number of orthotopic heart transplantations is increasing in the elderly.1 Because the elderly population is growing, it is likely that the demand for OHT in this population will continue to increase. Due to the organ donor supply and improvements in the medical and surgical treatments of heart failure, examination of the utility of OHT in this population is warranted.
Studies examining the effect of age on outcomes in OHT have reported mixed results.2– 4,6 –9,11 Noteworthy is the fact that many of these studies have been conducted at single institutions with small patient numbers. In reviewing the literature, it is more meaningful to examine those studies conducted in the past 5 to 10 years as advances in the field of transplantation have rendered older studies somewhat obsolete. Borkon and colleagues reported on 153 consecutive patients undergoing OHT at their institution. Recipients ⱖ55 years of age experienced significantly worse 1- and 5-year survival rates, with the majority of early deaths due to significantly higher rates of severe infections among older patients.5 More recent series, however, have
188
Weiss et al.
The Journal of Heart and Lung Transplantation February 2008
Table 2. Unadjusted Univariate Mortality Rates for Elderly Versus Young Patients
Overall mortality 30-day mortality 1-year mortality
Overall mortality 30-day mortality 1-year mortality
Age ⬍60 (N ⫽ 10,128) 1,991 (19.9%) 567 (5.6%) 1,168 (11.5)
Age ⱖ60 (N ⫽ 4,273) 1,036 (24.5%) 286 (6.7%) 632 (14.7%)
p-Value ⬍0.001 0.002 ⬍0.001
Age ⬍70 (N ⫽ 14,189)
Age ⱖ70 (N ⫽ 212)
p-Value
2,963 (20.8%) 842 (5.9%) 1,762 (12.4)
64 (30%) 11 (⬍5.1%) 38 (17.9%)
0.001 0.53 0.25
Chi-square analysis was used, with p ⬍ 0.05 considered statistically significant. Data presented as number of patients with percentage in parentheses.
refuted their conclusions, demonstrating acceptable outcomes in patients ⬎60 years of age. Demers and colleagues examined 81 patients between 60 and 70 years of age who underwent OHT at their institution Table 3. Multivariate Cox Proportional Hazards Regression Model Risk factor Elderly subsets Age (continuous) Elderly, primary stratification (ⱖ60 y vs ⬍60 y) Additional age stratifications Age ⱖ50 y vs ⬍50 y Age ⱖ55 y vs ⬍55 y Age ⱖ65 y vs ⬍65 y Age ⱖ70 y vs ⬍70 y Additional variables Gender (M vs F) Hypertension Diabetes Donor diabetes Ischemic time Creatinine Donor age HLA mismatch (HLA ⱕ4 vs 5 or 6) ICU prior to transplant Mechanical ventilation prior to transplant IABP PA mean pressure Cardiac index TPG PVR Transplant year
HR (95% CI)
p-valuea
1.0 (1.00–1.00)
0.9
1.28 (1.15–1.42)
⬍0.001
0.95 (0.85–1.06) 1.15 (1.04–1.28) 1.26 (1.09–1.46) 1.30 (0.87–1.93)
0.34 0.01 0.002 0.2
0.83 (0.74–0.94) 1.13 (1.02–1.26) 1.18 (1.05–1.33)
0.002 0.02 0.005
1.06 (1.01–1.12) 1.09 (1.05–1.13) 1.01 (1.01–1.02)
0.014 ⬍0.001 ⬍0.001
1.08 (0.98–1.2) 1.16 (1.03–1.30)
0.12 0.02
2.41 (1.88–3.11) 1.07 (0.86–1.34) 1.01 (1.00–1.01) 0.93 (0.82–1.05) 1.00 (0.98–1.02) 0.99 (0.94–1.05) 1.03 (1.00–1.05)
⬍0.001 0.52 0.01 0.26 0.91 0.74 0.1
HR, hazard ratio; CI, confidence interval; F, female; M, male; y, years; BMI, body mass index; HLA, human leukocyte antigen; IABP, intra-aortic balloon pump; PA, pulmonary artery; ICU, intensive care unit; TPG, transpulmonary gradient; PVR, pulmonary vascular resistance. a p-value based on multivariate Cox proportional hazard regression analysis, using factors significant on univariate analysis.
Figure 2. Results of Kaplan–Meier models of survival. Solid gray line: elderly patients; dashed black line: young patients. A survival table is given with number of patients at risk in parentheses. The p-value corresponds to Mantel–Cox log-rank test results (based on OPTN data, September 2006).
with up to 10 years of follow-up and found survival rates comparable to those of younger patients (50% vs 51%).8 Nagendran examined 50 patients ⬎60 years of age who underwent OHT at the University of Alberta between 1990 and 2000. Again, similar rates of survival were seen between older and younger patient age groups and no statistical differences were noted in TCAD, rejection episodes or quality of life.9 Further studies have been conducted focusing on patients ⬎65 and ⬎70 years of age.6,7,10,12 Finally, a series by Radovancevic and colleagues focusing on long-term mortality after OHT did not find recipient age to be a significant predictor of mortality.13 Although the majority of recent studies examining this issue have shown encouraging results, the data are far from conclusive and this issue remains highly controversial. The relative lack of patients of advanced age and the short follow-up time intervals within each series make it difficult to draw meaningful statistical comparisons. Furthermore, the single-center focus provides the opportunity for institutional bias and limits broad applicability. Although multi-institutional databases exist (e.g., the ISHLT registry), to our knowledge there have been no previous studies utilizing registry data to focus solely on the elderly. The annual report published with ISHLT data gives a broad overview of statistics on ⬎70,000 OHTs performed worldwide, but does not provide a focused examination of outcomes stratified by age.1 We attempted to mitigate these concerns and provide a more comprehensive assessment of outcomes in elderly patients by performing a retrospective review using data from the Organ Procurement and Transplantation Network. The UNOS dataset provides a large multi-institutional sample of elderly patients undergoing OHT. Our analysis has focused on the years be-
The Journal of Heart and Lung Transplantation Volume 27, Number 2
Figure 3. Kaplan–Meier estimates of 5-year survival conditional on both 30-day (A) and 1-year (B) survival. Solid gray line: elderly patients; dashed black line: young patients. A survival table is given with number of patients at risk in parentheses. The p-value corresponds to Mantel–Cox log-rank test results (based on OPTN data, September 2006).
tween 1999 and 2006, comprising a modern cohort free of bias relating to differences in technique or immunosuppression regimens. By using the UNOS registry, we were able to examine all centers, not just specialized centers of expertise, in order to ascertain whether a true difference in survival exists for elderly patients. We chose the age of 60 years or older as a cut-off point for several reasons. First, the literature is mixed with regard to what is considered advanced age with OHT, but many studies have defined 60 years of age or older as “elderly.” In addition, by using age 60 years as a cut-off, we have a robust amount of data in the elderly subset. In this way, we can be sure that our results are accurate and not compromised by an underpowered analysis due to small sample size. It is noteworthy that the most significant differences with multivariate analysis occurred at the 60-year cut-off, confirming that our stratification was appropriate. Our results show that older patients enjoy excellent survival rates with OHT. Although statistically signifi-
Weiss et al.
189
cant differences in survival exist between older and younger patients, the cumulative 5-year estimates of survival differed by only 6%. Even patients ⱖ70 years of age, comprising 212 patients, achieved a 5-year survival rate of 61% with OHT. Furthermore, long-term survival conditional on surviving the first year differed by only 3% between groups. This indicates that if an older patient survives the first year of transplantation, that patient has an excellent chance of achieving long-term survival. When we examined secondary outcomes, we observed no difference in allograft vasculopathy between older and younger patients. This suggests that TCAD is not associated with age but rather immune-mediated donor factors, as described elsewhere.14 As expected, older patients did have higher rates of post-operative stroke, infection and new-onset dialysis. As reported previously, it is probable that these sequelae (particularly infection) may contribute to the significantly lower 30-day and 1-year survival associated with advanced age.5 Interestingly, older patients underwent fewer treatments for rejection within the first year after transplantation.8,11 This is consistent with what has been reported previously. It is possible that older patients enjoy decreased immune reactivity, which is consistent with our observation of increased infection rates among older recipients. It is important to mention that, although we are encouraged by what seems to be acceptable survival in the older population, the age of 60 years proved to be an independent predictor of mortality on multivariate analysis. This significant risk of death is likely related to the large patient population that we examined and may explain why previous studies have not been able to show this risk independent of other predictors. In this regard it is noteworthy that older patients appeared to receive organs from higher-risk donors. Although the UNOS database does not provide specific
Figure 4. Kaplan–Meier estimates of 5-year survival stratified by age for patients ⬍60 (solid gray line), 60 to 64 (solid black line), 65 to 69 (dashed black line) and ⱖ70 (dotted gray line) years of age (based on OPTN data, September 2006).
190
Weiss et al.
The Journal of Heart and Lung Transplantation February 2008
Table 4. Univariate and Multivariate Analysis of Secondary Outcome Variables for Elderly (ⱖ60 years) Versus Young (⬍60 years) Patients Outcome Bivariate TCAD Post-operative infections New-onset dialysis Post-operative stroke Rejection (first year) Length of hospital stay (SEM)
Young (⬍60 years), N ⫽ 10,128 2,719 (11.2) 2,233 (23.2) 685 (7.2) 218 (2.3) 2,796 (43) 19.6 (0.28)
Number with data (%)
Elderly (ⱖ60 years) N ⫽ 4,273
N/A ⫾ 1,077 (11.0) 9,595 (95) 1,043 (25.9) 9,464 (93) 349 (8.8) 9,602 (95) 116 (2.9) 6,528 (65) 897 (34) 8,762 (87) 21.3 (0.41) HR, OR or regression coefficient (95% CI)
Multivariate Time to TCAD Post-operative infections New-onset dialysis Post-operative stroke Acute rejection (1 year) Length of hospital stay
1.15 (0.68–1.96) 1.11 (0.98–1.25) 1.2 (1.01–1.44) 1.16 (0.84–1.16) 0.67 (0.60–0.77) 1.45 (0.15–2.7)
Number with data (%)
p-valuea
N/A 4,027 (94) 3,960 (93) 4,019 (94) 2,650 (62) 3,676 (86)
0.52 0.001 0.002 0.03 ⬍0.001 ⬍0.001
0.58 0.07 0.04 0.34 ⬍0.001 0.03
Transplant coronary artery disease (⫾) based on UNOS follow-up files. There were 31,314 follow-up points for young patients and 13,082 follow-up points for elderly patients. Approximately 70% of patients had data on allograft vasculopathy available. HR, hazard ratio; OR, odds ratio for logistic regression; CI, confidence interval; CI, confidence interval; SEM, standard error of mean; TCAD, transplant coronary artery disease; N/A, not applicable. a p-value for univariate analysis based on either the chi-square (pacemaker) or Student’s t-test (LOS). p-values for multivariate analysis based on either logistic (pacemaker) or linear (LOS) regression analysis. p ⬍ 0.05 considered statistically significant.
data on high-risk donors, we did investigate donor factors that may indicate marginal status, including donor age, infection, diabetes, hypertension and creatinine. Older patients received organs from older donors and donors with a higher percentage of infection and diabetes. Although this is far from conclusive, the results may indicate a willingness by some centers to allocate organs from high-risk donors to elderly patients. This fact may in part have led to the increased mortality in older patients. Given the fact that the number of donor hearts has decreased steadily since 1994 (approximately 35%),1 the increased mortality in elderly patients does raise the question of whether society as a whole benefits from placing elderly patients on the transplant list, thus reducing the availability of organs for younger candidates. It is also true that elderly patients might be candidates for mechanical assistance, further lessening the burden on an already shortened donor supply. This issue is more ethical than scientific and decisions along this line are beyond the scope of this study. However, it is clear that, for a patient with advanced heart failure, OHT is the best therapeutic option—a statement that holds true no matter what age group is examined. The preceding discussion, however, does reinforce the notion that transplantation in older patients must be undertaken with careful consideration, in appropriate candidates, and at centers familiar with the specialized care required for older patients undergoing OHT.
Limitations We recognize that there are several limitations to our study. First, our study is retrospective and cannot account for inherent undocumented differences in patient characteristics. Although we attempted to control for selection bias with multivariate statistical methods, we concede that a fundamental limitation of any retrospective study examining outcomes is a lack of control of all potential confounders. Second, nationwide administrative clinical registries, including the one we used, and studies based on them are reliant on accurate coding of information. We acknowledge that the data presented herein were not necessarily entered by individuals with clinical expertise. However, although errors and variance in the data undoubtedly exist, we have assumed that these are randomly distributed and should not lead to significant bias in our conclusions. Third, by using a nationwide sample, we are dependent on the variables and definitions present in the dataset. In many circumstances, we would have chosen to include additional variables under our control. Specifically, the UNOS dataset offers limited information on immunosuppressive regimens and rejection types. In addition, follow-up is limited with the dataset. Although we acknowledge the weaknesses inherent in our design, we are encouraged by the strengths of our study. Specifically, we have been able to provide an unbiased multi-center examination of outcomes among
The Journal of Heart and Lung Transplantation Volume 27, Number 2
elderly heart transplant recipients. Without a randomized, controlled clinical trial, we believe that a retrospective study of a nationwide sample with compensatory statistical methodology constitutes a reasonable approach for providing a thorough analysis of this issue. Our Approach At our institution, we do not use age cut-offs as criteria for OHT. We believe that each patient must be considered irrespective of age in favor of what medical co-morbidities they possess. We do, however, advocate careful patient selection as it is clear that the older recipient population carries higher risks for several contraindications for transplantation, including previous malignancy. Incidentally, this was seen in our UNOS sample as well (Table 1).15,16 In conclusion, the current study has examined OHT in the elderly. Although our analysis revealed significantly lower survival (independent of other predictors of mortality) in patients ⱖ60 years of age, cumulative 5-year survival rates, even up to the age of 70 years, were acceptable. Given the dismal survival and poor quality of life associated with advanced heart failure, our results should encourage surgeons to consider OHT in carefully selected older patients when indicated and appropriate. The content of this article is the responsibility of the authors alone and does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products or organizations imply endorsement by the U.S. Government. REFERENCES 1. Taylor DO, Edwards LB, Boucek MM, et al. Registry of the International Society for Heart and Lung Transplantation: twentysecond official adult heart transplant report—2005. J Heart Lung Transplant 2005;24:945–55. 2. Bull DA, Karwande SV, Hawkins JA, et al. Long-term results of cardiac transplantation in patients older than sixty years. UTAH Cardiac Transplant Program. J Thorac Cardiovasc Surg 1996;111:423–7.
Weiss et al.
191
3. Favaloro R, Diez M, Bertolotti A, et al. Orthotopic heart transplantation in elderly patients: a 10-year experience at a single center. Transplant Proc 2004;36:1692– 4. 4. Peraira JR, Segovia J, Fuentes R, et al. Differential characteristics of heart transplantation in patients older than 60 years. Transplant Proc 2003;35:1959 – 61. 5. Borkon AM, Muehlebach GF, Jones PG, et al. An analysis of the effect of age on survival after heart transplant. J Heart Lung Transplant 1999;18:668 –74. 6. Blanche C, Takkenberg JJ, Nessim S, et al. Heart transplantation in patients 65 years of age and older: a comparative analysis of 40 patients. Ann Thorac Surg 1996;62:1442– 6. 7. Crespo-Leiro MG, Paniagua-Martin MJ, Muniz J, et al. Long-term results of heart transplant in recipients older and younger than 65 years: a comparative study of mortality, rejections, and neoplasia in a cohort of 445 patients. Transplant Proc 2005; 37:4031–2. 8. Demers P, Moffatt S, Oyer PE, et al. Long-term results of heart transplantation in patients older than 60 years. J Thorac Cardiovasc Surg 2003;126:224 –31. 9. Nagendran J, Wildhirt SM, Modry D, et al. A comparative analysis of outcome after heart transplantation in patients aged 60 years and older: the University of Alberta experience. J Card Surg 2004;19:559 – 62. 10. Blanche C, Blanche DA, Kearney B, et al. Heart transplantation in patients seventy years of age and older: a comparative analysis of outcome. J Thorac Cardiovasc Surg 2001;121:532– 41. 11. Frazier OH, Macris MP, Duncan JM, Van Buren CT, Cooley DA. Cardiac transplantation in patients over 60 years of age. Ann Thorac Surg 1988;45:129 –32. 12. Morgan JA, John R, Weinberg AD, et al. Long-term results of cardiac transplantation in patients 65 years of age and older: a comparative analysis. Ann Thorac Surg 2003;76:1982–7. 13. Radovancevic B, Konuralp C, Vrtovec B, et al. Factors predicting 10-year survival after heart transplantation. J Heart Lung Transplant 2005;24:156 –9. 14. Caforio AL, Tona F, Fortina AB, et al. Immune and nonimmune predictors of cardiac allograft vasculopathy onset and severity: multivariate risk factor analysis and role of immunosuppression. Am J Transplant 2004;4:962–70. 15. Caforio AL, Gambino A, Piaserico S, et al. De novo noncutaneous malignancies in heart transplantation: a single-centre 15-year experience and risk factor analysis. Transplant Proc 2001;33: 3658 –9. 16. Conraads VM, Denollet J, Vorlat A, Moulijn AC, Vrints CJ. Screening for solid organ malignancies prior to heart transplantation. Transplantation 2001;71:1481–3.
With life expectancy on the rise and older individuals comprising the fastest growing segment of the U.S. population, the demographics of orthotopic heart transplantation (OHT) are in flux. Although it is known that OHT represents the best treatment for many forms of From the Division of Cardiac Surgery, The Johns Hopkins University School of Medicine, Baltimore, Maryland. Submitted July 24, 2007; revised November 12, 2007; accepted November 15, 2007. Presented at the 27th annual meeting and scientific sessions of the International Society of Heart and Lung Transplantation, San Francisco, California, April 2007. Supported in part by the Health Resources and Services Administration, Contract No. 234-2005-370011C. Reprint requests: Eric S. Weiss, MD, Division of Cardiac Surgery, The Johns Hopkins University School of Medicine, 1820 Bank Street, Baltimore, MD 21231. Telephone: 410-539-7677. Fax: 410-522-5048. E-mail: [email protected] Copyright © 2008 by the International Society for Heart and Lung Transplantation. 1053-2498/08/$–see front matter. doi:10.1016/ j.healun.2007.11.566
184
advanced heart failure, the utility of this therapeutic modality in elderly patients continues to be a subject of debate. With advances in the medical and surgical management of heart failure and the paucity of available organs, it is clear that a consensus must be reached regarding the appropriate utilization of OHT for elderly patients. Despite this lack of uniform agreement, it has become clear in recent years that elderly patients are undergoing more frequent OHT nationwide. The International Society of Heart and Lung Transplantation reports that, since 1995, the percentage of transplant patients over the age of 65 has increased steadily, while the percentage of those between 40 and 49 years of age has decreased steadily.1 This trend is echoed by UNOS data, which show a steady increase in the proportion of recipients ⬎60 years of age undergoing OHT (Figure 1). Notwithstanding this increase, many centers still view advanced age as a contraindication to OHT. The literature examining this issue has been mixed, with
The Journal of Heart and Lung Transplantation Volume 27, Number 2
some series reporting worse outcomes among elderly patients2–5 and others reporting equivalent rates of mortality, rejection and transplant coronary artery disease.6 –9 Examination of age has extended to patients as old as 70 years of age, showing that OHT can be performed safely in selected candidates.10 A central problem surrounding the literature examining OHT in the elderly is the fact that most studies directly examining outcomes in elderly patients have been conducted in single centers with small patient cohorts. The United Network for Organ Sharing (UNOS) dataset is a nationwide, physician-regulated registry that provides a unique opportunity to examine OHT in a large population of elderly patients free of institutional bias. We conducted a retrospective examination of the multiinstitutional UNOS dataset to examine both short- and long-term outcomes of elderly patients undergoing OHT. We hypothesize that elderly patients have equivalent outcomes to younger patients when examined on a large scale with a modern cohort. METHODS Data Source UNOS provided de-identified patient data (Standard Transplant Analysis and Research [STAR] files with follow-up files) from the thoracic organ transplant registry for the years 1987 to 2006, with follow-up through January 2007. The data include all U.S. patients undergoing thoracic organ transplantation and who were reported to the Organ Procurement Network during the time period. No patient or center identifiers are included. Study Design and Patient Population We conducted a retrospective review of the UNOS registry from January 1999 to December 2006, with follow-up through January 2007. The time-points were chosen to identify a modern cohort of patients who underwent OHT free of bias related to advances in heart transplantation techniques and immunosuppressive regimens. All first-time OHT patients ⬎18 years of age were included. Patients were primarily stratified by age (⬍60 years vs ⱖ60 years) and were secondarily stratified into additional age groups for subsequent analysis (ⱕ49, 50 to 54, 55 to 59, 60 to 64, 65 to 69 and ⱖ70 years). Outcome Measures Baseline demographic and clinical factors were examined. The primary end-point was all-cause mortality during the study period. Secondary outcomes included length of hospital stay (LOS), post-operative infections (defined by the need for antibiotics), post-operative stroke, development of transplant coronary artery disease (TCAD), new-onset dialysis requirement, and rejec-
Weiss et al.
185
tion within 1 year (defined as any treatment of rejection for all causes). Statistical Analysis Comparisons of baseline characteristics between study groups were performed using Student’s t-test for continuous variables and the chi-square test for categorical variables. Cumulative survival was modeled using the Kaplan–Meier method with statistical differences between survival curves assessed using the Mantel–Cox log-rank test. In addition, modeling based on 30-day and 1-year conditional survival was employed to assess long-term survival independent of early mortality. Mortality was assessed for all risk factors using a univariate model. Multivariate analysis was conducted with a Cox proportional hazard regression model for cumulative survival and with multivariate logistic regression for short-term mortality. Significant univariate predictors of mortality were incorporated into the multivariate models in a stepwise fashion to assess the effect of technique on mortality. Only well-represented variables (⬍33% missing in the registry) were included in the multivariate analysis. In addition, univariate and multivariate logistic and linear regression analyses, when appropriate, were performed for all additional outcome variables. Two-tailed p ⬍ 0.05 was considered statistically significant. All odds ratios and regression coefficients are presented with along with the 95% confidence interval (CI). All statistical analyses were performed using STATA software, version 9.0 (StataCorp LP, College Station, TX). RESULTS A total of 14,401 adult patients underwent primary OHT during the study period. Of these, 4,273 (30%) were ⱖ60 years of age. In addition, 212 (1.4%) were ⱖ70 years of age. Although the mean age of OHT recipients has remained relatively constant since 1987, the percentage of patients ⬎60 years of age has increased steadily. This trend continued during the study period of 1999 to 2006 (Figure 1). As expected, older patients had a higher incidence of baseline co-morbidities, such as diabetes and hypertension and higher baseline creatinine levels, although, perhaps surprisingly, their body mass indices (BMIs) were comparable to those of younger patients (Table 1). Patient age groups were well matched for other markers of clinical acuity such as hospitalization and ICU admission rates, as well as the prevalence of pre-operative intra-aortic balloon counterpulsation (IABP) and mechanical ventilation. In addition, patients were well matched for pre-operative (baseline) hemodynamic variables. Older patients did have longer waitlist times (255 vs 212 days, p ⬍ 0.001), which in part
186
Weiss et al.
The Journal of Heart and Lung Transplantation February 2008
Figure 1. Trends in age of OHT recipients over time. Bars: transplants performed in patients ⱖ60 years of age; solid line: mean age for adult OHTs for reference. Although the mean age for adult OHTs has remained stable, the mean age for elderly patients has increased since 1987 (based on OPTN data, September 2006).
may have been related to a greater percentage being listed as UNOS Status 2 (Table 1). Older patients also had a greater history of malignancy (5.3% vs 3.7%, p ⬍ 0.001). Both females and ethnic minorities comprised a smaller percentage of the total number of older patients in the study population (Table 1). Older recipients did receive hearts from older donors and donors with a greater prevalence of diabetes and infections, indicating that older patients may be receiving organs from higher risk donors. Cumulatively, 3,027 patients died during the study period, for an all-cause mortality of 21%. Older patients (ⱖ60 years old) had higher 5-year overall mortality rates (24.5% vs 19.9%, p ⬍ 0.001) as well as increased 30-day and 1-year mortality rates (30-day: 6.7% vs 5.6%, p ⫽ 0.002; 1-year: 14.7% vs 11.5%, p ⬍ 0.001) (Table 2). With multivariate Cox proportional hazards regression analysis, age as a continuous variable was not significantly associated with mortality, but ages ⱖ55, ⱖ60 and ⱖ65 years were all significant predictors of mortality independent of other known risk factors for death (Table 3). When examining different age-range cut-offs, patients ⱖ60 years possessed the highest 5-year cumulative hazard ratio (HR) for death (HR ⫽ 1.28 [CI 1.15 to 1.42], p ⬍ 0.001) as compared with the other age cut-offs (Table 3). When examining both 30-day and 1-year mortality rates, controlling for confounders in a mutivariate logistic regression model, older patients only showed a significantly greater risk of mortality at 1 year posttransplant (odds ratio [OR] for death at 30 days ⫽ 1.17 [CI 0.95 to 1.46], p ⫽ 0.12; OR for death at 1 year ⫽ 1.29 [CI 1.10 to 1.50], p ⫽ 0.001). Additional predictors of cumulative mortality on multivariate analysis included: female gender; ischemic time; age of donor; and markers of acuity such as history of diabetes, history of
hypertension, mechanical ventilation prior to transplant, and ICU admission prior to transplant. On subset analysis, patients ⱖ70 years of age (n ⫽ 212) also had significantly worse cumulative mortality, although 30-day and 1-year mortality did not reach statistical significance. On multivariate analysis (likely related to small sample size), age ⱖ70 years did not emerge as an independent predictor of mortality. When survival was modeled using the Kaplan–Meier method, the older group only showed a 6% lower rate of survival at 5 years. Although this reached statistical significance, the absolute difference was small (69% vs 75%) (Figure 2). The difference in survival narrowed when cumulative survival conditional on 30-day and 1-year survival was modeled. When censoring those patients who died at 1 year, cumulative survival differed by only 3% between older and younger patients (83% vs 86%, respectively, p ⫽ 0.01), indicating that a substantial risk of mortality for older patients occurs in the first year (Figure 3). When patients were stratified by age range, there was also a progressive stepwise decrease in cumulative survival for all progressive age groups after the age of 60 years (Figure 4). Other outcome variables examined included the development of transplant coronary artery disease (TCAD) at follow-up, development of post-operative infections (requiring antibiotic treatment), new-onset dialysis, post-operative stroke, rejection requiring treatment (for all causes) at 1 year, and LOS. Older patients did have increased new-onset dialysis needs, longer LOS, more post-operative infections and a higher risk of post-operative stroke, although only differences in LOS and new-onset dialysis were confirmed on multivariate analysis (Table 4). Interestingly, TCAD did not differ between older and younger patients on either univari-
The Journal of Heart and Lung Transplantation Volume 27, Number 2
Weiss et al.
187
Table 1. Baseline Characteristics of Elderly (ⱖ60 years) Versus Young Patients (⬍60 years) Young (⬍60 years) N ⫽ 10,128a
Number with data available (%)
Elderly (ⱖ60 years) N ⫽ 4,273a
Number with data available (%)
p-valueb
Demographics Mean age Female Black, Hispanic, Native American, Asian/Pacific Islanderc
47 (⫾10.7) 2,743 (27)
10,128 (100) 10,128 (100)
64 (⫾3.1) 688 (16)
4,273 (100) 4,273 (100)
⬍0.001 ⬍0.001
2,896 (29)
10,128 (100)
621 (14)
4,273 (100)
⬍0.001
Acuity Diabetes HTN Creatinine BMI Hospitalizedd ICU prior to transplant Ventilator prior to transplant IABP prior to transplant Donor age (years) Donor diabetes Donor Hypertension Donor creatinine Donor infection ⫾ ischemic time (hours) Days on waitlist History of malignancy UNOS Status 1e
1,936 (20) 3,562 (37) 1.33 (⫾0.9) 26.5 (⫾4.8) 5,230 (52) 3,120 (31) 287 (2.8) 527 (5.2) 30.8 (12) 181 (1.8) 1,626 (16.2) 1.20 (0.01) 2,605 (27.3) 3.12 (⫾1.0) 212 (⫾352) 373 (5.3) 7,642 (76)
9,930 (98) 9,730 (96) 9,640 (95) 9,355 (92) 9,988 (99) 9,988 (99) 9,841 (97) 10,128 (100) 10,123 (99) 10,080 (99) 10,029 (99) 10,043 (99) 9,524 (94) 8,996 (88) 9,413 (93) 10,081 (99) 10,121 (99)
1,064 (25) 1,896 (46) 1.45 (⫾0.7) 26.1 (⫾4.0) 2,049 (49) 1,251 (29) 131 (3.0) 226 (5.2) 33.3 (13) 100 (2.3) 718 (16.9) 1.21 (0.01) 1,168 (29.2) 3.16 (⫾1.0) 255 (⫾419) 225 (3.7) 3,025 (71)
4,184 (98) 4,102 (96) 4,053 (97) 3,940 (95) 4,220 (99) 4,220 (99) 4,142 (97) 4,273 (100) 4,269 (99) 4,242 (99) 4,236 (99) 4,237 (99) 3,999 (94) 3,796 (89) 3,953 (93) 4,257 (99) 4,272 (99)
⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.06 0.45 0.88 ⬍0.001 0.04 0.2
3,523 (82) 3,234 (76) 3,493 (82) 3,395 (79)
⬍0.001 0.87 0.40 0.91
Hemodynamic variables Mean PA pressure PVR Cardiac index TPG
28.5 (⫾10) 2.4 (⫾1.9) 1.5 (⫾0.55) 9.5 (⫾5.7)
8,339 (82) 7,516 (74) 8,217 (81) 7,924 (78)
27.9 (⫾10) 2.3 (⫾2.2) 1.5 (⫾0.49) 9.5 (⫾5.6)
0.03 0.10 ⬍0.001 ⬍0.001 ⬍0.001
PA, pulmonary artery; HTN, hypertension; BMI, body mass index; ICU, intensive care unit; IABP, intra-aortic balloon pump; PVR, pulmonary vascular resistance (defined by [mean PA pressure (mPAP) ⫺ pulmonary capillary wedge pressure (PCWP)]/cardiac output); TPG, transpulmonary gradient (defined by MPAP ⫺ PCWP); Donor infection ⫾ represents any type of infection present in the donor at time of transplant. a Data presented as number of patients (%) or mean (⫾SD). b p-value is based on comparison between two groups by either chi-square or Student’s t-test (with p ⬍ 0.05 considered statistically significant). c Race or ethnic group was a variable present in the dataset. d Hospitalized refers to hospitalization prior to transplantation. e UNOS Status 1 refers to patients listed as Status 1a, 1b or old UNOS Status 1.
ate or multivariate analyses. Importantly, univariate and multivariate analyses showed that older age was associated with significantly less treatment for rejection within 1 year of transplant (34 vs 43%, OR ⫽ 0.67, p ⬍ 0.001). DISCUSSION From national registry reports it is clear that the number of orthotopic heart transplantations is increasing in the elderly.1 Because the elderly population is growing, it is likely that the demand for OHT in this population will continue to increase. Due to the organ donor supply and improvements in the medical and surgical treatments of heart failure, examination of the utility of OHT in this population is warranted.
Studies examining the effect of age on outcomes in OHT have reported mixed results.2– 4,6 –9,11 Noteworthy is the fact that many of these studies have been conducted at single institutions with small patient numbers. In reviewing the literature, it is more meaningful to examine those studies conducted in the past 5 to 10 years as advances in the field of transplantation have rendered older studies somewhat obsolete. Borkon and colleagues reported on 153 consecutive patients undergoing OHT at their institution. Recipients ⱖ55 years of age experienced significantly worse 1- and 5-year survival rates, with the majority of early deaths due to significantly higher rates of severe infections among older patients.5 More recent series, however, have
188
Weiss et al.
The Journal of Heart and Lung Transplantation February 2008
Table 2. Unadjusted Univariate Mortality Rates for Elderly Versus Young Patients
Overall mortality 30-day mortality 1-year mortality
Overall mortality 30-day mortality 1-year mortality
Age ⬍60 (N ⫽ 10,128) 1,991 (19.9%) 567 (5.6%) 1,168 (11.5)
Age ⱖ60 (N ⫽ 4,273) 1,036 (24.5%) 286 (6.7%) 632 (14.7%)
p-Value ⬍0.001 0.002 ⬍0.001
Age ⬍70 (N ⫽ 14,189)
Age ⱖ70 (N ⫽ 212)
p-Value
2,963 (20.8%) 842 (5.9%) 1,762 (12.4)
64 (30%) 11 (⬍5.1%) 38 (17.9%)
0.001 0.53 0.25
Chi-square analysis was used, with p ⬍ 0.05 considered statistically significant. Data presented as number of patients with percentage in parentheses.
refuted their conclusions, demonstrating acceptable outcomes in patients ⬎60 years of age. Demers and colleagues examined 81 patients between 60 and 70 years of age who underwent OHT at their institution Table 3. Multivariate Cox Proportional Hazards Regression Model Risk factor Elderly subsets Age (continuous) Elderly, primary stratification (ⱖ60 y vs ⬍60 y) Additional age stratifications Age ⱖ50 y vs ⬍50 y Age ⱖ55 y vs ⬍55 y Age ⱖ65 y vs ⬍65 y Age ⱖ70 y vs ⬍70 y Additional variables Gender (M vs F) Hypertension Diabetes Donor diabetes Ischemic time Creatinine Donor age HLA mismatch (HLA ⱕ4 vs 5 or 6) ICU prior to transplant Mechanical ventilation prior to transplant IABP PA mean pressure Cardiac index TPG PVR Transplant year
HR (95% CI)
p-valuea
1.0 (1.00–1.00)
0.9
1.28 (1.15–1.42)
⬍0.001
0.95 (0.85–1.06) 1.15 (1.04–1.28) 1.26 (1.09–1.46) 1.30 (0.87–1.93)
0.34 0.01 0.002 0.2
0.83 (0.74–0.94) 1.13 (1.02–1.26) 1.18 (1.05–1.33)
0.002 0.02 0.005
1.06 (1.01–1.12) 1.09 (1.05–1.13) 1.01 (1.01–1.02)
0.014 ⬍0.001 ⬍0.001
1.08 (0.98–1.2) 1.16 (1.03–1.30)
0.12 0.02
2.41 (1.88–3.11) 1.07 (0.86–1.34) 1.01 (1.00–1.01) 0.93 (0.82–1.05) 1.00 (0.98–1.02) 0.99 (0.94–1.05) 1.03 (1.00–1.05)
⬍0.001 0.52 0.01 0.26 0.91 0.74 0.1
HR, hazard ratio; CI, confidence interval; F, female; M, male; y, years; BMI, body mass index; HLA, human leukocyte antigen; IABP, intra-aortic balloon pump; PA, pulmonary artery; ICU, intensive care unit; TPG, transpulmonary gradient; PVR, pulmonary vascular resistance. a p-value based on multivariate Cox proportional hazard regression analysis, using factors significant on univariate analysis.
Figure 2. Results of Kaplan–Meier models of survival. Solid gray line: elderly patients; dashed black line: young patients. A survival table is given with number of patients at risk in parentheses. The p-value corresponds to Mantel–Cox log-rank test results (based on OPTN data, September 2006).
with up to 10 years of follow-up and found survival rates comparable to those of younger patients (50% vs 51%).8 Nagendran examined 50 patients ⬎60 years of age who underwent OHT at the University of Alberta between 1990 and 2000. Again, similar rates of survival were seen between older and younger patient age groups and no statistical differences were noted in TCAD, rejection episodes or quality of life.9 Further studies have been conducted focusing on patients ⬎65 and ⬎70 years of age.6,7,10,12 Finally, a series by Radovancevic and colleagues focusing on long-term mortality after OHT did not find recipient age to be a significant predictor of mortality.13 Although the majority of recent studies examining this issue have shown encouraging results, the data are far from conclusive and this issue remains highly controversial. The relative lack of patients of advanced age and the short follow-up time intervals within each series make it difficult to draw meaningful statistical comparisons. Furthermore, the single-center focus provides the opportunity for institutional bias and limits broad applicability. Although multi-institutional databases exist (e.g., the ISHLT registry), to our knowledge there have been no previous studies utilizing registry data to focus solely on the elderly. The annual report published with ISHLT data gives a broad overview of statistics on ⬎70,000 OHTs performed worldwide, but does not provide a focused examination of outcomes stratified by age.1 We attempted to mitigate these concerns and provide a more comprehensive assessment of outcomes in elderly patients by performing a retrospective review using data from the Organ Procurement and Transplantation Network. The UNOS dataset provides a large multi-institutional sample of elderly patients undergoing OHT. Our analysis has focused on the years be-
The Journal of Heart and Lung Transplantation Volume 27, Number 2
Figure 3. Kaplan–Meier estimates of 5-year survival conditional on both 30-day (A) and 1-year (B) survival. Solid gray line: elderly patients; dashed black line: young patients. A survival table is given with number of patients at risk in parentheses. The p-value corresponds to Mantel–Cox log-rank test results (based on OPTN data, September 2006).
tween 1999 and 2006, comprising a modern cohort free of bias relating to differences in technique or immunosuppression regimens. By using the UNOS registry, we were able to examine all centers, not just specialized centers of expertise, in order to ascertain whether a true difference in survival exists for elderly patients. We chose the age of 60 years or older as a cut-off point for several reasons. First, the literature is mixed with regard to what is considered advanced age with OHT, but many studies have defined 60 years of age or older as “elderly.” In addition, by using age 60 years as a cut-off, we have a robust amount of data in the elderly subset. In this way, we can be sure that our results are accurate and not compromised by an underpowered analysis due to small sample size. It is noteworthy that the most significant differences with multivariate analysis occurred at the 60-year cut-off, confirming that our stratification was appropriate. Our results show that older patients enjoy excellent survival rates with OHT. Although statistically signifi-
Weiss et al.
189
cant differences in survival exist between older and younger patients, the cumulative 5-year estimates of survival differed by only 6%. Even patients ⱖ70 years of age, comprising 212 patients, achieved a 5-year survival rate of 61% with OHT. Furthermore, long-term survival conditional on surviving the first year differed by only 3% between groups. This indicates that if an older patient survives the first year of transplantation, that patient has an excellent chance of achieving long-term survival. When we examined secondary outcomes, we observed no difference in allograft vasculopathy between older and younger patients. This suggests that TCAD is not associated with age but rather immune-mediated donor factors, as described elsewhere.14 As expected, older patients did have higher rates of post-operative stroke, infection and new-onset dialysis. As reported previously, it is probable that these sequelae (particularly infection) may contribute to the significantly lower 30-day and 1-year survival associated with advanced age.5 Interestingly, older patients underwent fewer treatments for rejection within the first year after transplantation.8,11 This is consistent with what has been reported previously. It is possible that older patients enjoy decreased immune reactivity, which is consistent with our observation of increased infection rates among older recipients. It is important to mention that, although we are encouraged by what seems to be acceptable survival in the older population, the age of 60 years proved to be an independent predictor of mortality on multivariate analysis. This significant risk of death is likely related to the large patient population that we examined and may explain why previous studies have not been able to show this risk independent of other predictors. In this regard it is noteworthy that older patients appeared to receive organs from higher-risk donors. Although the UNOS database does not provide specific
Figure 4. Kaplan–Meier estimates of 5-year survival stratified by age for patients ⬍60 (solid gray line), 60 to 64 (solid black line), 65 to 69 (dashed black line) and ⱖ70 (dotted gray line) years of age (based on OPTN data, September 2006).
190
Weiss et al.
The Journal of Heart and Lung Transplantation February 2008
Table 4. Univariate and Multivariate Analysis of Secondary Outcome Variables for Elderly (ⱖ60 years) Versus Young (⬍60 years) Patients Outcome Bivariate TCAD Post-operative infections New-onset dialysis Post-operative stroke Rejection (first year) Length of hospital stay (SEM)
Young (⬍60 years), N ⫽ 10,128 2,719 (11.2) 2,233 (23.2) 685 (7.2) 218 (2.3) 2,796 (43) 19.6 (0.28)
Number with data (%)
Elderly (ⱖ60 years) N ⫽ 4,273
N/A ⫾ 1,077 (11.0) 9,595 (95) 1,043 (25.9) 9,464 (93) 349 (8.8) 9,602 (95) 116 (2.9) 6,528 (65) 897 (34) 8,762 (87) 21.3 (0.41) HR, OR or regression coefficient (95% CI)
Multivariate Time to TCAD Post-operative infections New-onset dialysis Post-operative stroke Acute rejection (1 year) Length of hospital stay
1.15 (0.68–1.96) 1.11 (0.98–1.25) 1.2 (1.01–1.44) 1.16 (0.84–1.16) 0.67 (0.60–0.77) 1.45 (0.15–2.7)
Number with data (%)
p-valuea
N/A 4,027 (94) 3,960 (93) 4,019 (94) 2,650 (62) 3,676 (86)
0.52 0.001 0.002 0.03 ⬍0.001 ⬍0.001
0.58 0.07 0.04 0.34 ⬍0.001 0.03
Transplant coronary artery disease (⫾) based on UNOS follow-up files. There were 31,314 follow-up points for young patients and 13,082 follow-up points for elderly patients. Approximately 70% of patients had data on allograft vasculopathy available. HR, hazard ratio; OR, odds ratio for logistic regression; CI, confidence interval; CI, confidence interval; SEM, standard error of mean; TCAD, transplant coronary artery disease; N/A, not applicable. a p-value for univariate analysis based on either the chi-square (pacemaker) or Student’s t-test (LOS). p-values for multivariate analysis based on either logistic (pacemaker) or linear (LOS) regression analysis. p ⬍ 0.05 considered statistically significant.
data on high-risk donors, we did investigate donor factors that may indicate marginal status, including donor age, infection, diabetes, hypertension and creatinine. Older patients received organs from older donors and donors with a higher percentage of infection and diabetes. Although this is far from conclusive, the results may indicate a willingness by some centers to allocate organs from high-risk donors to elderly patients. This fact may in part have led to the increased mortality in older patients. Given the fact that the number of donor hearts has decreased steadily since 1994 (approximately 35%),1 the increased mortality in elderly patients does raise the question of whether society as a whole benefits from placing elderly patients on the transplant list, thus reducing the availability of organs for younger candidates. It is also true that elderly patients might be candidates for mechanical assistance, further lessening the burden on an already shortened donor supply. This issue is more ethical than scientific and decisions along this line are beyond the scope of this study. However, it is clear that, for a patient with advanced heart failure, OHT is the best therapeutic option—a statement that holds true no matter what age group is examined. The preceding discussion, however, does reinforce the notion that transplantation in older patients must be undertaken with careful consideration, in appropriate candidates, and at centers familiar with the specialized care required for older patients undergoing OHT.
Limitations We recognize that there are several limitations to our study. First, our study is retrospective and cannot account for inherent undocumented differences in patient characteristics. Although we attempted to control for selection bias with multivariate statistical methods, we concede that a fundamental limitation of any retrospective study examining outcomes is a lack of control of all potential confounders. Second, nationwide administrative clinical registries, including the one we used, and studies based on them are reliant on accurate coding of information. We acknowledge that the data presented herein were not necessarily entered by individuals with clinical expertise. However, although errors and variance in the data undoubtedly exist, we have assumed that these are randomly distributed and should not lead to significant bias in our conclusions. Third, by using a nationwide sample, we are dependent on the variables and definitions present in the dataset. In many circumstances, we would have chosen to include additional variables under our control. Specifically, the UNOS dataset offers limited information on immunosuppressive regimens and rejection types. In addition, follow-up is limited with the dataset. Although we acknowledge the weaknesses inherent in our design, we are encouraged by the strengths of our study. Specifically, we have been able to provide an unbiased multi-center examination of outcomes among
The Journal of Heart and Lung Transplantation Volume 27, Number 2
elderly heart transplant recipients. Without a randomized, controlled clinical trial, we believe that a retrospective study of a nationwide sample with compensatory statistical methodology constitutes a reasonable approach for providing a thorough analysis of this issue. Our Approach At our institution, we do not use age cut-offs as criteria for OHT. We believe that each patient must be considered irrespective of age in favor of what medical co-morbidities they possess. We do, however, advocate careful patient selection as it is clear that the older recipient population carries higher risks for several contraindications for transplantation, including previous malignancy. Incidentally, this was seen in our UNOS sample as well (Table 1).15,16 In conclusion, the current study has examined OHT in the elderly. Although our analysis revealed significantly lower survival (independent of other predictors of mortality) in patients ⱖ60 years of age, cumulative 5-year survival rates, even up to the age of 70 years, were acceptable. Given the dismal survival and poor quality of life associated with advanced heart failure, our results should encourage surgeons to consider OHT in carefully selected older patients when indicated and appropriate. The content of this article is the responsibility of the authors alone and does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products or organizations imply endorsement by the U.S. Government. REFERENCES 1. Taylor DO, Edwards LB, Boucek MM, et al. Registry of the International Society for Heart and Lung Transplantation: twentysecond official adult heart transplant report—2005. J Heart Lung Transplant 2005;24:945–55. 2. Bull DA, Karwande SV, Hawkins JA, et al. Long-term results of cardiac transplantation in patients older than sixty years. UTAH Cardiac Transplant Program. J Thorac Cardiovasc Surg 1996;111:423–7.
Weiss et al.
191
3. Favaloro R, Diez M, Bertolotti A, et al. Orthotopic heart transplantation in elderly patients: a 10-year experience at a single center. Transplant Proc 2004;36:1692– 4. 4. Peraira JR, Segovia J, Fuentes R, et al. Differential characteristics of heart transplantation in patients older than 60 years. Transplant Proc 2003;35:1959 – 61. 5. Borkon AM, Muehlebach GF, Jones PG, et al. An analysis of the effect of age on survival after heart transplant. J Heart Lung Transplant 1999;18:668 –74. 6. Blanche C, Takkenberg JJ, Nessim S, et al. Heart transplantation in patients 65 years of age and older: a comparative analysis of 40 patients. Ann Thorac Surg 1996;62:1442– 6. 7. Crespo-Leiro MG, Paniagua-Martin MJ, Muniz J, et al. Long-term results of heart transplant in recipients older and younger than 65 years: a comparative study of mortality, rejections, and neoplasia in a cohort of 445 patients. Transplant Proc 2005; 37:4031–2. 8. Demers P, Moffatt S, Oyer PE, et al. Long-term results of heart transplantation in patients older than 60 years. J Thorac Cardiovasc Surg 2003;126:224 –31. 9. Nagendran J, Wildhirt SM, Modry D, et al. A comparative analysis of outcome after heart transplantation in patients aged 60 years and older: the University of Alberta experience. J Card Surg 2004;19:559 – 62. 10. Blanche C, Blanche DA, Kearney B, et al. Heart transplantation in patients seventy years of age and older: a comparative analysis of outcome. J Thorac Cardiovasc Surg 2001;121:532– 41. 11. Frazier OH, Macris MP, Duncan JM, Van Buren CT, Cooley DA. Cardiac transplantation in patients over 60 years of age. Ann Thorac Surg 1988;45:129 –32. 12. Morgan JA, John R, Weinberg AD, et al. Long-term results of cardiac transplantation in patients 65 years of age and older: a comparative analysis. Ann Thorac Surg 2003;76:1982–7. 13. Radovancevic B, Konuralp C, Vrtovec B, et al. Factors predicting 10-year survival after heart transplantation. J Heart Lung Transplant 2005;24:156 –9. 14. Caforio AL, Tona F, Fortina AB, et al. Immune and nonimmune predictors of cardiac allograft vasculopathy onset and severity: multivariate risk factor analysis and role of immunosuppression. Am J Transplant 2004;4:962–70. 15. Caforio AL, Gambino A, Piaserico S, et al. De novo noncutaneous malignancies in heart transplantation: a single-centre 15-year experience and risk factor analysis. Transplant Proc 2001;33: 3658 –9. 16. Conraads VM, Denollet J, Vorlat A, Moulijn AC, Vrints CJ. Screening for solid organ malignancies prior to heart transplantation. Transplantation 2001;71:1481–3.