If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Reprint requests: Bartley P. Griffith, MD, University of Maryland School of Medicine, Division of Cardiac Surgery, 110 S Paca St, 7th Flr, Baltimore, MD 21201. Tel: +410-328-5842. Fax: +410-328-2750
Current lung transplantation guidelines stipulate that the ideal donor is aged younger than 55 years, but several institutions have reported that outcomes using donors aged 55 years and older are comparable with those of younger donors.
Methods
We retrospectively reviewed the United Network for Organ Sharing (UNOS) database to identify all adult lung transplants between 2000 and 2010 in the United States. Patients were stratified by donor age 18 to 34 (reference), 35 to 54, 55 to 64, and ≥ 65 years. Primary outcomes included survival at 30 days and at 1, 3, and 5 years and rates of bronchiolitis obliterans syndrome (BOS). Survival was assessed using the Kaplan-Meier method. Risk factors for mortality were identified by multivariable Cox and logistic regression.
Results
We identified 10,666 recipients with median follow-up of 3 years (range, 0–10 years). Older donors were more likely to have died of cardiovascular or cerebrovascular causes, but there were no differences in recipient diagnosis, lung allocation score, or incidence of BOS as a function of donor age. The use of donors aged 55 to 64 years was not a risk factor for mortality at 1 year (odds ratio, 1.1; p = 0.304) or 3 years (odds ratio, 0.923; p = 0.571) compared with the reference group; however, use of donors aged > 65 years was associated with increased mortality at both time points (odds ratio, 2.8 and 2.4, p < 0.02).
Conclusions
Outcomes after lung transplantation using donors aged 55 to 64 years were similar to those observed with donors meeting conventional age criteria. Donors aged ≥ 65 years, however, were associated with decreased intermediate-term survival, although there was no increased risk of BOS for this group.
Lung transplantation could improve quality of life and survival for many of these individuals but is constrained by an inadequate number of suitable donors (15% of donors offered),
and relatively rapid progression to chronic rejection, also referred to as bronchiolitis obliterans syndrome (BOS). Considering that donor factors are known be important in determining graft function after transplant, guidelines have been established defining acceptable donor parameters.
As waiting list mortality continues to rise, there has been a growing trend toward accepting lungs that do not meet these criteria. The use of these “marginal” donors has the potential to afford life-saving therapy to many more patients but demands thorough and ongoing assessment of the effect of more liberal acceptance criteria on clinical outcomes.
One of the more controversial guidelines stipulates that the ideal donor is aged younger than 55 years. As the population ages, and as medicine facilitates improved health at advanced ages, several centers have published outcome data suggesting that organs from donors aged older than 55 years may be comparable to those from younger donors.
Some centers have reported no association between donor age and outcomes, whereas others report decreased long-term survival and decreased time to the development of BOS. Another concern is that these older organs are only being allocated to older or sicker recipients and that recipient characteristics may, in fact, be determining outcomes. We examined the United Network for Organ Sharing (UNOS) database to evaluate the effect of donor age on outcomes after lung transplantation in a large, contemporary patient cohort in the United States.
Methods
The UNOS database was queried for all patients who underwent lung transplantation in the United States between January 2000 and December 2009. In all, 12,633 patients were identified. We excluded 503 pediatric recipients and 1,919 donors aged < 18 years. The remaining 10,666 adults were stratified by donor age 18 to 34 (reference group), 35 to 54, 55 to 64, and ≥ 65 years. Donor information included age, sex, smoking history, arterial partial pressure of oxygen, mechanism of death, and lung ischemic time. Recipients were characterized by age, sex, diagnosis, lung allocation score (LAS), and type of transplant. Because the LAS was first routinely recorded in May 2005, analyses involving this variable were limited to transplants after this date (n = 5,986). Primary outcomes included survival at 30 days and at 1, 3, and 5 years, as well as incidence of BOS and time to its development.
Data are expressed as mean ± standard error of the mean or number (%), unless otherwise specified. The Student’s t-test was used to test continuous variables with normal distributions, and the Mann-Whitney U test was used for non-normal continuous variables. Categoric variables were analyzed using the chi-square test or Fisher’s exact test. Survival data were analyzed using the Kaplan-Meier method, and survival curves were compared using the log-rank test. A multivariable Cox regression was performed to identify risk factors for death and then repeated on data segregated into pre-LAS and post-LAS eras, using as the dividing point the official inclusion of the LAS in the UNOS database on May 3, 2005. Further comparison of older donors and their recipients was conducted, followed by Cox regression assessing the effect of transplantation era on outcomes while controlling for known risk factors and other differences between donor and recipient groups. Finally, multivariable logistic regression was used to assess the effect of donor age at specific time points. All variables included in these calculations were found to be significant in Cox regression or univariate analyses (LAS, not shown). The results are presented as odds ratio (OR) ± standard error of the mean. Stata 9.2 software (StataCorp LP, College Station, TX) was used for all analyses.
Results
The number of transplants performed in each donor age group is shown in Figure 1. Almost all lung transplants continue to use lungs from donors who meet ideal age criteria, but an increasing number of organs are being procured from donors of advanced age. In 2000, donors aged ≥ 55 years represented only 5.8% of the total, whereas in 2009, this increased to 10.0%.
Figure 1Number of lung transplants in the United States by donor age.
Donor characteristics are presented in Table 1. Older donors were more likely to be female and less likely to have had a significant tobacco smoking history (≥ 20 pack-years, p < 0.004). The most striking difference between younger and older donors is in the mechanism of death. Younger donors were more likely to die from gunshots and blunt trauma. As donor age increases, donors are more likely to have died from stroke and cardiovascular disease. Very few non-heart-beating donors were found in the database (n = 80), and the partial pressure of arterial oxygen (Pao2) did not vary significantly among groups.
Recipient characteristics are reported in Table 2. Patients who received lungs from older donors were more likely to be female and more likely to be older. Recipient diagnosis varied among groups, with a peak in cystic fibrosis in the 34- to 55-year-olds (37.4%) and a progressive increase in the prevalence of idiopathic pulmonary fibrosis with increasing age (41.3% in the group aged ≥65 years). Medical comorbidities are not well documented in the database, but patients aged ≥ 65 were more likely to be diabetic (22.1% vs 14.2%, p = 0.022) and had similar serum creatinine levels as younger donors (0.80 vs 0.83 mg/dl, p = 0.720).
Interestingly, there was no difference in median LAS among the donor age groups nor was there a difference in the proportion of patients with LAS in the top quartile for their age. There was, however, an increase in ischemic time of about 40 minutes in the oldest donors relative to the youngest. There was also a statistical difference in transplant type (double vs single) between the groups aged 34 to 55 years and those aged 54 to 65 years and the reference, but the magnitude of the difference (0.1%, p = 0.003) was of no clinical significance.
Recipient outcomes are summarized in Table 3. Patients who received lungs from older donors were characterized by a 2-day increase in the median post-operative hospital stay, but were no more likely to experience airway dehiscence, new-onset kidney failure requiring dialysis, or stroke. Short-term 30-day survival was similar among the groups, but age-related differences began to appear at 1 and 5 years. Donors aged 35 to 54 were characterized by slightly decreased 5-year survival relative to the 18- to 34-year-old reference group (49% vs 51%, p = 0.025). This decrease in survival became considerably more pronounced in the group aged ≥65 years (30% vs 51%, p < 0.001). These survival data are presented graphically in Figure 2. Log-rank comparisons of all age groups with the reference group are significant (p < 0.016) except for the group aged 55 to 64 years (p = 0.080).
The results of a multivariable Cox regression for risk factors for all-cause mortality are presented in Table 4. There is a slightly increased risk of morality for the 35- to 54-year-old donor group relative to controls (hazard ratio [HR], 1.080; p = 0.016), something not seen with the 55- to 64-year-old donor group. The most striking association with mortality is seen in patients who received lungs from the oldest donors, who had a nearly 2-fold increased risk of death (HR, 1.853; p < 0.001) that persisted even when accounting for a number of common risk factors, as well as all of the previously presented differences between donors and recipients.
Table 4Risk Factors for Mortality by Donor Age Group
Because this analysis included patients from multiple transplant eras, the association between advanced donor age and death was re-evaluated using the official inclusion of the LAS in the UNOS database (May 3, 2005) as the dividing point between significant time periods. This event is clinically relevant and happens to fall near the chronologic midpoint of this study. The number of patients in each era and follow-up available are summarized in Table 5. Median follow-up was just over 4 years (up to 9 years) for the pre-LAS group, and 1 to 2 years (up to 5) for the post-LAS group. This relatively short follow-up may explain the observation in Table 6 that donor age 35 to 54 years is associated with increased mortality in the pre-LAS era (HR, 1.109; p = 0.001) but not in the post-LAS (HR, 1.036; p = 0.487). As reported in Table 3, the difference in survival for this donor age group is not apparent until 5 years after transplant and, therefore, may not have been captured in this sub-population.
Table 5Distribution of Donors by Transplantation Era
Pre-LAS includes transplants performed between January 1, 2000, and May 2, 2005. Post-LAS includes transplants performed between May 3, 2005, and December 31, 2009.
Number of patients
Follow-up, days
Variable
In group
In group + reference
Median (range)
Donor age 35–54 years
Overall
4,281
9,648
767 (0–3,270)
Pre-LAS
1,947
4,319
1,664 (0–3,270)
Post-LAS
2,334
5,329
540 (0–1,877)
Donor age 55–64 years
Overall
914
6,281
728 (0–3,663)
Pre-LAS
332
2,704
1,611 (0–3,663)
Post-LAS
582
3,577
579 (0–1,850)
Donor age ≥ 65 years
Overall
104
5,471
531 (1–3,601)
Pre-LAS
33
2,405
1,240 (1–3,601)
Post-LAS
71
3,066
371 (1–1,835)
LAS, lung allocation score.
a Pre-LAS includes transplants performed between January 1, 2000, and May 2, 2005. Post-LAS includes transplants performed between May 3, 2005, and December 31, 2009.
Univariate and multivariable Cox regression for risk factors for mortality, stratified by donor age and transplantation era. Pre-LAS includes transplants performed between January 1, 2000, and May 2, 2005. Post-LAS includes transplants performed between May 3, 2005 and December 31, 2009.
CI, confidence interval; HR, hazard ratio; LAS, lung allocation score; Pao2, arterial partial pressure of oxygen at an inspired oxygen concentration of 100%.
a Univariate and multivariable Cox regression for risk factors for mortality, stratified by donor age and transplantation era. Pre-LAS includes transplants performed between January 1, 2000, and May 2, 2005. Post-LAS includes transplants performed between May 3, 2005 and December 31, 2009.
The opposite is seen with donors aged ≥65 years; namely, the effect on mortality appears to be greater in the more modern era (2.333 vs 1.417). These donors were specifically re-examined by transplantation era, yielding the data reported in Table 7. The more modern cohort contained a larger proportion of women, but donor age, smoking history, mechanism of death, and Pao2 were similar between the groups. In the same way, the recipients of these older lungs were re-evaluated with a focus on transplantation era (Table 8). Age, sex, transplant type (double vs single), and ischemic time were no different among the groups. The modern recipients, however, spent considerably less time on the waiting list (41 vs 141 days, p = 0.044), were more likely to be diagnosed with idiopathic pulmonary fibrosis (49.3% vs 24.2%), and were more likely to have diabetes (22.5% vs 3.0%), 2 recipient factors known to be associated with poorer outcomes.
Table 7Characteristics of Donors Aged 65 Years and Older by Transplant Era
Pre-LAS includes transplants performed between January 1, 2000, and May 2, 2005. Post-LAS includes transplants performed between May 3, 2005, and December 31, 2009.
Pre-LAS includes transplants performed between January 1, 2000, and May 2, 2005. Post-LAS includes transplants performed between May 3, 2005, and December 31, 2009.
Categoric data are shown as number (%) and continuous variables as mean ± standard error of the mean.
(n = 33)
(n = 71)
Donor age, years
66.7 ± 2.7
67.1 ± 1.7
0.286
Female
12 (36.4)
43 (60.6)
0.018
Smoking history
2 (6.1)
2 (2.8)
0.289
Mechanism of death
0.302
Stroke
24 (72.7)
62 (87.3)
Blunt trauma
4 (12.1)
4 (5.6)
Gunshot
0 (0.0)
0 (0.0)
Asphyxiation
0 (0.0)
0 (0.0)
Other cardiovascular
4 (12.1)
3 (4.2)
Pao2, mm Hg
424.6 ± 106.8
387.0 ± 154.6
0.220
LAS, lung allocation score; Pao2, partial pressure of arterial oxygen at an inspired oxygen concentration of 100%.
a Categoric data are shown as number (%) and continuous variables as mean ± standard error of the mean.
b Pre-LAS includes transplants performed between January 1, 2000, and May 2, 2005. Post-LAS includes transplants performed between May 3, 2005, and December 31, 2009.
Pre-LAS includes transplants performed between January 1, 2000, and May 2, 2005. Post-LAS includes transplants performed between May 3, 2005, and December 31, 2009.
Pre-LAS includes transplants performed between January 1, 2000, and May 2, 2005. Post-LAS includes transplants performed between May 3, 2005, and December 31, 2009.
a Categoric data are shown as number (%) and continuous variables as mean ± standard error of the mean or median (interquartile range).
b Pre-LAS includes transplants performed between January 1, 2000, and May 2, 2005. Post-LAS includes transplants performed between May 3, 2005, and December 31, 2009.
Further analysis using multivariable Cox regression was conducted to better assess the effect of transplantation era within this older donor group. As reported in Table 9, pre-LAS or post-LAS status was not significantly associated with mortality in the unadjusted analysis or when adjusting for the previously used risk factors, with the addition of diagnosis of idiopathic pulmonary fibrosis or diabetes. Log-rank comparison of the pre-LAS and post-LAS survival function also failed to demonstrate a significant difference (p = 0.128).
Univariate(crude) and multivariable Cox regression examining the association between transplantation era and mortality. Post-LAS is a binary variable equal to 0 for transplants performed between January 1, 2000, and May 2, 2005, and equal to 1 for transplants performed between May 3, 2005 and December 31, 2009.
CI, confidence interval; HR, hazard ratio; IPF, idiopathic pulmonary fibrosis; LAS, lung allocation score; Pao2, partial pressure of arterial oxygen at an inspired oxygen concentration of 100%.
a Univariate(crude) and multivariable Cox regression examining the association between transplantation era and mortality. Post-LAS is a binary variable equal to 0 for transplants performed between January 1, 2000, and May 2, 2005, and equal to 1 for transplants performed between May 3, 2005 and December 31, 2009.
To better define the post-transplant time points at which donor age had the strongest effect, a multivariable logistic regression using the previously identified risk factors for sub-acute and chronic graft dysfunction (Table 10) revealed that when controlling for recipient sex, age, and ischemic time, donor age ≥ 65 years was associated with a 2-fold to 3-fold increased risk for 1-year and 3-year mortality relative to donors aged 18 to 35 years. No such increase was observed in the other donor age groups.
Table 10Risk Factors for Recipient Mortality Over Time
Although the number of patients awaiting lung transplantation continues to increase, the number of transplants performed has remained relatively stable. In an effort to provide life-saving treatment for those on the waiting list, the transplant community continues to re-evaluate organ selection criteria with the aim of maximizing our use of this scarce resource, with much attention given to the “ideal” donor age younger than 55 years. In this review of nearly a decade of lung transplantation in the United States, we have concluded that outcomes after transplantation using older lungs (ages 55–64) have been similar to those observed using much younger, ideal donors, despite the observation that these lungs tended to go to slightly older recipients (with similar LASs). Moreover, there was no difference in cause of death, rate of BOS, or time to the development of BOS. The use of donors aged older than 65 years, however, was associated with significantly decreased long-term survival in a similar recipient population. When controlling for several known risk factors for organ dysfunction and transplantation era, donor age older than 65 remained an important risk factor for death.
These results are consistent with the published literature regarding the effect of donor age on transplantation outcomes. Dahlman et al
reported their experience with 212 lung transplants, 41 of which involved donors aged between 55 and 64 years. These donors were largely female (80% vs our 61%) and were much less likely to have a significant smoking history (14% vs our 36%). Survival was similar to the present study, with evidence that recipient age had a stronger effect on mortality than donor age. Horai et al
present similar support for the predominance of recipient age over donor age in their review of 584 patients, 86 aged older than 55 years, as do Pizanis et al
with a series of 186 patients, 19 of whom were older than 55 years (range, 55–69 years). In both studies, male gender and increasing recipient age were strongly predictive of death, whereas survival and time to BOS were independent of donor age. Fischer et al
presented 49 transplants involving donors aged older than 50 years (range, 50–64 years) and found no difference in survival up to 5 years. In addition, there were no differences in duration of post-transplant mechanical ventilation, intensive care unit length of stay, or rates of acute graft failure.
A recent article from De Perrot et al10 out of Toronto presented a similar analysis in an older patient population. They reported the outcomes of 467 lung transplant recipients, 60 of whom received lungs from donors aged between 60 and 77 years, with a median follow-up of 25 months (range, 0–136 months). Survival at 5 and 10 years was significantly decreased in recipients of these older organs, and the rates of BOS were nearly twice as high as in the reference population, representing the most common cause of death in the group aged >60 years.
We observed relatively few cases of BOS in a similar population. This is likely due to a significant difference in recipient age—the mean age of patients receiving lungs from a donor aged >65 years in our study was 57 years, whereas the mean age of the patients receiving older lungs in the Toronto study was 49 years. Increasing recipient age, as demonstrated by the reports cited above as well as our analysis, is associated with decreased survival and increases the likelihood that the patient will suffer cerebrovascular, respiratory, and infectious complications. Because BOS typically takes several years to develop, the baseline decreased survival and death due to other causes may result in few cases of BOS being observed in our older recipient population.
This study has the limitations inherent in the retrospective review of a large, voluntary database. In particular, it is impossible to discern the consideration given to donor age or other donor factors in each transplant scenario, and therefore, some physician bias may be present with respect to any of these donor characteristics that is not accounted for in the present analysis. In addition, certain post-transplant variables of interest, such as primary graft dysfunction and episodes of rejection, are not uniformly documented and could not be evaluated. These concerns are better addressed through institutional reports similar to those cited above and may provide particular insight into the decision making regarding donors aged 65 and older, as only 104 such individuals were represented in the current data set.
In conclusion, although donor age has long been considered an important predictor of graft failure, the level at which this risk becomes unacceptable (age 55 years) has come into question, and many centers have reported the routine use of lungs from older donors. The results of this large investigation of the effect of donor age on outcomes after lung transplantation in the last decade in the United States show that organs from donors aged 55 to 64 years are associated with outcomes similar to those observed with much younger donors, with donors aged 18 to 34 years as the reference for comparison. The same non-inferiority cannot be claimed for donors aged 65 and older, whose organs are associated increased mortality when controlling for recipient age and acuity, especially at 1 and 3 years after transplant. Ongoing surveillance, supplemented by better reporting of short-term outcomes, is warranted to better understand any differences in the rates of primary graft dysfunction and acute rejection. Our findings, however, support the expansion of lung acceptance criteria to include donors aged younger than 65 years.
Disclosure Statement
None of the authors has a financial relationship with a commercial entity that has an interest in the subject of the presented manuscript or other conflicts of interest to disclose.
References
Centers for Disease Control and Prevention
Deaths from chronic obstructive pulmonary disease—United States, 2000-2005.
01216-3&id=gr1.jpg){kind=link}
01216-3&id=gr2.jpg){kind=link}