Abstract
Background:
With rates of potential donor heart discard as high as 66% nationally, quality improvement efforts must seek to optimize donor utilization. Whether the timing of donor brain death declaration (BDD) influences organ acceptance is understudied. The authors sought to characterize the impacts of time between donor hospital admission and BDD on heart utilization and posttransplant outcomes.
Methods:
All potential heart donors and recipients in the United Network for Organ Sharing database were identified (2006–2021). Admission-to-BDD cohorts were: 1–2d-(n=52 469), 3–4d-(n=44 033), 5–7d-(n=24 509), 8–10d-(n=8576). Donor clinical characteristics were compared between cohorts, and donor acceptance was assessed using multivariable binary logistic regression. Recipient posttransplant survival was assessed with Kaplan-Meier.
Results:
Donor demographics and comorbidity profiles (diabetes, hypertension) were comparable across cohorts. Anoxia/overdose deaths more common (10%>21%>24%>18%, respectively) and CPR requirements were higher (37%>52%>58%>47%) when BDD occurred longer after admission. Renal dysfunction (44%>44%>35%>29%) and inotrope-requirements (52%>25%>36%>29%) were lower in the later BDD cohorts. Proportions of hepatic dysfunction (18–21%) and LVEF<50% (13–16%) were clinically equivalent. Donor acceptance differed by admission-to-BDD cohort [36%–(1–2d), 34%–(3–4d), 30%–(5–7d), 28%–(8–10d)]. Admission-to-BDD >4d was independently associated with lower odds of acceptance on multivariable analysis (OR 0.79, p<0.001). Recipients experienced equivalent posttransplant survival for all donor admission-to-BDD cohorts (p=0.999–adults & p=0.260–pediatrics).
Conclusions:
Heart donors with later BDD were disproportionately discarded despite similar-to-favorable overall clinical profiles, resulting in nearly 3000 fewer transplants during the study. Increased utilization of donors with later BDD and “high-risk” characteristics (eg, anoxia/overdose, CPR requirement), can improve rates of transplantation without compromising outcomes.
Introduction
Demand for donor organs continues to outpace supply for heart transplantation in the United States, yet approximately two-thirds of potential hearts go unused.1–4 While rates of donation after circulatory death (DCD) are increasing, the majority of heart transplants continue to be from brain-dead donors, and this remains standard practice even at adult programs at this point. Brain death results in eventual irreversible multisystem organ dysfunction, in large part due to alterations in the hormonal milieu and signaling pathways from brainstem dysfunction, thereby making organ procurement for transplantation a time-sensitive matter.5–7 While the onset of brain-death occurs prior to or shortly after admission in a large proportion of potential heart donors, the actual declaration of brain death may be influenced by external factors such as physician availability and hospital practices. Whether donors with later brain death declaration (BDD; defined as the duration between donor hospital admission and clinical declaration of brain-death) is associated with less favorable donor features, lower donor utilization rates, or inferior posttransplant outcomes is unclear.
The authors hypothesized that donors for whom BDD occurred later after admission were less likely to be accepted for heart transplantation, but posited that accepted donors, regardless of time from admission to BDD, yielded similar posttransplant outcomes. The primary aim of the current study was to characterize whether later BDD was associated with lesser acceptance of donor hearts for transplantation, with secondary aims focused on characterizing differences in donor characteristics and posttransplant outcomes depending upon the timing of donor BDD.
Materials and Methods
Institutional review board approval was obtained from the authors’ institution via waiver of informed consent for nonhuman subject research.
All data were obtained from the United Network for Organ Sharing (UNOS) registry. Brain death declaration date and time were available in UNOS beginning on January 1, 2006. Data were available through June 30, 2022, thereby defining the study period. Potential heart donors (those consented for heart donation, per UNOS) with BDD occurring within 10 days of admission (95% of all potential donors in UNOS) were included in analysis. Restricting the study to potential donors with BDD within 10 days of admission also served to eliminate from analysis those donors with prolonged or chronic illnesses, thereby generally limiting the study cohort to donors with acute mortality.
Only the date of donor admission is recorded in UNOS, rather than date and time as for BDD, so 48- and 72-hour intervals were used to define donor cohorts to account for the margin of error accorded by the inability to precisely calculate the number of hours between admission and BDD. Donor cohorts included admission-to-BDD within 1–2 days (n=52 469; 40% of donor population), 3–4 days (n=44 033; 34%), 5–7 days (n=24 509; 19%), and 8–10 days (n=8576; 7%), with day 1 representing BDD on the same day as admission. Median time between BDD and heart offering was 20 (IQR 11–29) hours. For donors with an organ accepted for transplant, the median time from BDD to aortic cross-clamp (synonymous with termination of brain death and organ procurement time) was 34 (IQR 22–48) hours.
Recipients were identified in UNOS by encrypted donor identification code and were studied by age at transplant: adult (≥18 years) and pediatric. Demographic and clinical data were obtained from UNOS. Renal dysfunction was defined as estimated glomerular filtration rate of less than 60 mL/min/1.73 m2, calculated from serum creatinine using the Schwartz formula, and hepatic dysfunction was defined as serum bilirubin of 1.2 mg/dL or greater.8,9 Left ventricular ejection fraction (LVEF) was analyzed as a categorical variable, with LVEF<50% considered marginal. As a dataset predicated primarily upon donor utilization, UNOS contains limited data on posttransplant recipient outcomes. Posttransplant outcomes available included hospital length of stay, incidence of acute rejection during the index hospitalization or within the first year posttransplant (defined by UNOS as hospitalization requirement for treatment of rejection), and overall survival (median 4.5 years of follow-up available). Outcomes were analyzed separately for adult and pediatric recipients.
Statistical Analysis
Statistical analyses were performed using R Studio (v2023.06.1+524, R Foundation for Statistical Computing, Vienna, Austria, https://www.R-project.org). Temporal trends in donor organ acceptance (utilization) from time (days) between donor admission and BDD were assessed using simple linear regression. Comparative analyses between admission-to-BDD cohorts (1–2 vs 3–4 vs 5–7 vs 8–10 days) utilized Mood’s median test for continuous variables and Chi-squared test with continuity correction for categorical variables. Multivariable binary logistic regression was used to identify factors independently associated with heart acceptance or rejection. Given collinearity between admission-to-BDD cohorts, separate models using the same covariates were used to compare odds of organ acceptance for each donor cohort compared to all others (models A-D corresponding to each respective donor cohort). Covariates were selected using an a priori method based upon the authors’ clinical experience and the results of univariate analyses. Model diagnostics included goodness-of-fit (model Chi-square p>0.05) and noncollinearity (variance inflation factor >5 for all covariates) testing. Recipient posttransplant survival was estimated using Kaplan-Meier with log-rank test, with Cox proportional hazards regression with stepwise backward selection of covariates with p>0.10 performed to identify donor and recipient factors independently associated with overall posttransplant survival. Separate models were used to study adult and pediatric recipient survival, with potential covariates including those variables included in univariate analysis.
Results
Donor utilization and characteristics
Overall, 34% (43 819/129 587) of potential donor hearts were accepted for transplantation. A strong negative temporal trend existed in donor heart acceptance by duration between donor hospital admission and BDD, with significantly lower rates of heart acceptance for donors in whom BDD occurred longer after admission (slope: −0.011, R2 0.889, p<0.001; Figure 1). To enable comparisons of donor-recipient pair characteristics and transplant outcomes, donors were grouped into cohorts based upon the number of days between donor hospital admission and BDD (see Methods). By admission-to-BDD cohort, organ acceptance was: 36% (19 012/52 469) for 1–2d; 34% (15 019/44 033) for 3–4d, 30% (7350/24 509) for 5–7d, and 28% (2438/8576) for 8–10d (p<0.001 between all).
Figure 1.
Proportion of potential heart donors accepted for transplantation by days between admission and brain death declaration. Day 1 corresponds to the brain death declaration occurring on the same day as hospital admission.
The characteristics of all potential donors are shown in Table 1. Donor age, weight, and sex were clinically similar across admission-to-BDD cohorts, while the cohorts with later BDD were comprised of a greater proportion of non-White donors (68%-[1–2d] > 66%-[3–4d] > 65%-[5–7d] - 61%-[8–10d]; p<0.001). The comorbidity profiles of potential donors were also clinically similar across admission-to-BDD cohorts (diabetes: 10–13%, hypertension: 34–35%, cigarette smoking: 21–24%, heavy alcohol use: 18–20%; all p<0.001). Donor mechanisms of death differed between cohorts, with anoxia/overdose accounting for a greater proportion of donors with later BDD (10>21>24>18%; p<0.001) and penetrating trauma accounting for a greater proportion of donors with earlier BDD (16>6>3>3%; p<0.001). The proportion of potential donors for whom CPR was performed mirrored the proportion of anoxia/overdose deaths and was greater in the later BDD cohorts (37>52>58>47%; p<0.001). At time of transplant, the proportions of potential donors with renal dysfunction (44>44>35>29%; p<0.001) and requiring inotropes (52>45>36> 34%, respectively; p<0.001) were lower in the later admission-to-BDD groups, while hepatic dysfunction (21>19>18>19%; p<0.001) and LVEF<50% (16>16>16>13%; p<0.001) were clinically comparable. Similar trends existed for accepted donors from each cohort (Table S1). Across all admission-to-BDD cohorts, the donors accepted for heart transplantation expectedly had overall more favorable characteristics (eg, younger age, fewer comorbidities, less end-organ dysfunction), more traumatic deaths, and fewer anoxia/overdose-associated deaths than donors not accepted for transplantation (Tables 2–5).
Table 1.
Characteristics of all potential heart donors by time from admission to brain death declaration.
| Time from Admission to Brain Death Declaration | |||||
|---|---|---|---|---|---|
| 1–2 days (n=52 469) |
3–4 days (n=44 033) |
5–7 days (n=24 509) |
8–10 days (n=8576) |
p-value | |
| Donor heart accepted | 19 012 (36%) | 15 019 (34%) | 7350 (30%) | 2438 (28%) | <0.001 |
| Demographics | |||||
| Age (yr) | 41 (26–54) | 41 (27–54) | 41 (28–53) | 42 (28–53) | 0.013 |
| Weight (kg) | 79 (67–93) | 79 (66–95) | 80 (66–96) | 80 (67–96) | <0.001 |
| Male sex (n=78 167) | 31 960 (61%) | 26 210 (60%) | 14 763 (60%) | 5234 (61%) | <0.001 |
| Race/ethnicitya | |||||
| Black (n=20 434) | 8103 (15%) | 6906 (16%) | 3998 (16%) | 1427 (17%) | <0.001 |
| Hispanic (n=18 708) | 6917 (13%) | 6470 (15%) | 3783 (15%) | 1538 (18%) | |
| White (n=85 970) | 35 870 (68%) | 29 020 (66%) | 15 828 (65%) | 5252 (61%) | |
| Comorbidities | |||||
| Diabetes (n=15 286) | 5242 (10%) | 5887 (13%) | 3148 (13%) | 1009 (12%) | <0.001 |
| Hypertension (n=44 285) | 17 782 (34%) | 15 300 (35%) | 8271 (34%) | 2932 (35%) | 0.013 |
| Cig. smoker (n=29 382) | 12 201 (24%) | 10 057 (23%) | 5392 (23%) | 1732 (21%) | <0.001 |
| Heavy alcohol (n=23 887) | 9445 (18%) | 8044 (19%) | 4745 (20%) | 1653 (20%) | <0.001 |
| Mechanism of death* | |||||
| Anoxia/OD (n=21 733) | 5022 (10%) | 9277 (21%) | 5911 (24%) | 1524 (18%) | <0.001 |
| CVA (n=40 328) | 19 602 (37%) | 12 393 (28%) | 5878 (24%) | 2456 (29%) | |
| Trauma–blunt (n=26 567) | 12 867 (25%) | 8026 (18%) | 3925 (16%) | 1749 (20%) | |
| Trauma–penetr. (n=12 211) | 8597 (16%) | 2609 (6%) | 769 (3%) | 236 (3%) | |
| Clinical status | |||||
| CPR performed (n=58 252) | 18 447 (37%) | 22 156 (52%) | 13 743 (58%) | 3906 (47%) | <0.001 |
| Renal dysfxn (n=53 326) | 22 769 (44%) | 19 509 (44%) | 8557 (35%) | 2491 (29%) | <0.001 |
| Hepatic dysfxn (n=25 552) | 10 981 (21%) | 8515 (19%) | 4411 (18%) | 1645 (19%) | <0.001 |
| Inotropes (n=58 065) | 26 920 (52%) | 19 595 (45%) | 8682 (36%) | 2868 (34%) | <0.001 |
| LVEF <50% (n=14 561) | 6395 (16%) | 4849 (16%) | 2592 (16%) | 725 (13%) | <0.001 |
Values expressed as median (IQR) or n (%) as appropriate.
Percentage by column.
Renal dysfunction, eGFR<60ml/min/1.73m2; hepatic dysfunction, total serum bilirubin >1.2 mg/dL. CPR, cardiopulmonary resuscitation; CVA, cerebrovascular accident; LVEF, left ventricular ejection fraction; OD, drug overdose.
Table 2.
Characteristics of accepted vs nonaccepted heart donors in whom brain death declaration occurred on admission day 1–2.
| Accepted heart donor (n=19 012) |
Nonaccepted heart donor (n=33 457) |
p-value | |
|---|---|---|---|
| Demographics | |||
| Age (yr) | 28 (21–39) | 50 (35–59) | <0.001 |
| Weight (kg) | 79 (68–91) | 79 (66–93) | 0.014 |
| Male sex | 13 922 (73%) | 18 038 (54%) | <0.001 |
| Race/ethnicity | |||
| Black | 3184 (17%) | 4919 (15%) | <0.001 |
| Hispanic | 3016 (16%) | 3901 (12%) | |
| White | 12 343 (65%) | 23 527 (70%) | |
| Comorbidities | |||
| Diabetes | 506 (3%) | 4736 (14%) | <0.001 |
| Hypertension | 2546 (13%) | 15 236 (46%) | <0.001 |
| Cigarette smoker | 2193 (12%) | 10 008 (31%) | <0.001 |
| Heavy alcohol | 2871 (15%) | 6574 (20%) | <0.001 |
| Mechanism of death | |||
| Anoxia/overdosea | 1971 (39%) | 3051 (61%) | <0.001 |
| CVAa | 3551 (18%) | 16 051 (82%) | |
| Trauma–blunta | 6500 (51%) | 6367 (49%) | |
| Trauma–penetratinga | 5550 (65%) | 3047 (35%) | |
| Clinical status | |||
| Donor CPR | 6206 (34%) | 12 241 (38%) | <0.001 |
| Renal dysfunction | 6636 (35%) | 16 133 (48%) | <0.001 |
| Hepatic dysfunction | 4388 (23%) | 6593 (20%) | <0.001 |
| Inotropes | 8447 (45%) | 18 473 (55%) | <0.001 |
| LVEF <50% | 371 (2%) | 6024 (34%) | <0.001 |
Values expressed as median (IQR) or n (%) as appropriate.
Percentage by row.
Renal dysfunction, eGFR<60ml/min/1.73m2; hepatic dysfunction, total serum bilirubin >1.2 mg/dL. CPR, cardiopulmonary resuscitation; CVA, cerebrovascular accident; LVEF, left ventricular ejection fraction.
Table 5.
Characteristics of accepted vs nonaccepted heart donors in whom brain death declaration occurred on admission day 8–10.
| Accepted heart donor (n=2438) |
Nonaccepted heart donor (n=6138) |
p-value | |
|---|---|---|---|
| Demographics | |||
| Age (yr) | 28 (20–38) | 47 (34–56) | <0.001 |
| Weight (kg) | 76 (63–90) | 82 (69–99) | <0.001 |
| Male sex | 1609 (66%) | 3625 (59%) | <0.001 |
| Race/ethnicity | |||
| Black | 459 (19%) | 968 (16%) | <0.001 |
| Hispanic | 555 (23%) | 983 (16%) | |
| White | 1338 (55%) | 3914 (64%) | |
| Comorbidities | |||
| Diabetes | 84 (3%) | 925 (15%) | <0.001 |
| Hypertension | 301 (12%) | 2631 (43%) | <0.001 |
| Cigarette smoker | 208 (9%) | 1524 (25%) | <0.001 |
| Heavy alcohol | 332 (14%) | 1321 (22%) | <0.001 |
| Mechanism of death | |||
| Anoxia/overdosea | 582 (38%) | 941 (62%) | <0.001 |
| CVAa | 477 (19%) | 1979 (81%) | |
| Trauma–blunta | 790 (45%) | 959 (55%) | |
| Trauma–penetratinga | 111 (47%) | 125 (53%) | |
| Clinical status | |||
| Donor CPR | 1080 (46%) | 2826 (48%) | 0.134 |
| Renal dysfunction | 568 (23%) | 1923 (32%) | <0.001 |
| Hepatic dysfunction | 524 (18%) | 1121 (22%) | 0.001 |
| Inotropes | 1011 (42%) | 1857 (31%) | <0.001 |
| LVEF <50% | 34 (1%) | 691 (21%) | <0.001 |
Values expressed as median (IQR) or n (%) as appropriate.
Percentage by row.
Renal dysfunction, eGFR<60ml/min/1.73m2; hepatic dysfunction, total serum bilirubin >1.2 mg/dL. CPR, cardiopulmonary resuscitation; CVA, cerebrovascular accident; LVEF, left ventricular ejection fraction.
Upon multivariable binary logistic regression analysis, BDD occurring within 1–2 days of admission was independently associated with greater odds of acceptance compared to other admission-to-BDD cohorts (OR 1.26 [95% CI 1.21–1.31]), whereas BDD occurring 5–7 days (OR 0.79 [95% CI 0.76–0.83]) and 8–10 days (OR 0.79 [95% CI 0.76–0.83]) after admission were independently associated with lower odds of acceptance (Table S2). Other factors positively associated with donor acceptance included death due to penetrating trauma, male sex, and white race (all p≤0.01; Table S2). Factors independently associated with lower odds of acceptance included greater age, CPR requirement, renal dysfunction, hepatic dysfunction, hypertension, diabetes, and marginal LVEF (all p≤0.01; Table S2).
Recipient characteristics
Adult recipient characteristics were clinically similar for all donor admission-to-BDD cohorts with regards to recipient age and weight, race and ethnicity, cardiac diagnosis, UNOS status at transplant, renal and hepatic dysfunction at transplant, and requirement for inotropes, mechanical ventilation, VAD, and ECMO (Table 6). For pediatric recipients, more variability existed between admission-to-BDD cohorts (Table 7). Recipient ages across donor cohorts mirrored the mechanisms of death within the cohort, with older recipients corresponding to greater proportions of penetrating trauma in the earlier admission-to-BDD cohort. Clinical differences also existed in the proportions of pediatric recipients with renal dysfunction, hepatic dysfunction, ventilator-dependence, and inotrope requirements by donor admission-to-BDD cohort, although no clear trends existed by cohort (Table 7).
Table 6.
Characteristics of adult heart transplant recipients by time from donor admission to brain death declaration.
| Time from Admission to Brain Death Declaration | |||||
|---|---|---|---|---|---|
| 1–2 days (n=16 784) |
3–4 days (n=11 885) |
5–7 days (n=5888) |
8–10 days (n=1975) |
p-value | |
| Demographics | |||||
| Age | 56 (46–63) | 56 (46–63) | 56 (46–63) | 56 (45–63) | 0.020 |
| Weight | 83 (71–95) | 82 (70–95) | 82 (70–95) | 81 (70–95) | 0.005 |
| Male sex | 12 720 (76%) | 8649 (73%) | 4249 (72%) | 1383 (70%) | <0.001 |
| Race/ethnicitya | |||||
| Black | 3745 (22%) | 2676 (23%) | 1285 (22%) | 453 (23%) | <0.001 |
| Hispanic | 1362 (8%) | 1057 (9%) | 569 (10%) | 228 (12%) | |
| White | 10 968 (65%) | 7615 (64%) | 3737 (63%) | 1190 (60%) | |
| Diagnosisa | |||||
| Cardiomyopathy | 14 578 (89%) | 10294 (88%) | 5100 (88%) | 1712 (89%) | <0.001 |
| CHD | 472 (3%%) | 407 (3%) | 196 (3%%) | 86 (4%) | |
| Retransplant | 481 (3%) | 382 (3%) | 194 (3%) | 60 (3%) | |
| Clinical status | |||||
| UNOS status 1A/1–3 | 5894 (35%) | 3995 (34%) | 2019 (34%) | 711 (36%) | 0.029 |
| Renal dysfunction | 8025 (48%) | 5756 (49%) | 2863 (49%) | 920 (47%) | 0.213 |
| Hepatic dysfunction | 3438 (21%) | 2430 (21%) | 1137 (20%) | 380 (19%) | 0.145 |
| Ventilator | 303 (2%) | 220 (2%) | 94 (2%) | 32 (2%) | 0.609 |
| Inotropes | 6472 (39%) | 4580 (39%) | 2239 (38%) | 758 (38%) | 0.903 |
| VAD | 6537 (39%) | 4537 (38%) | 2184 (37%) | 759 (38%) | 0.085 |
| ECMO | 339 (2%) | 308 (3%) | 142 (2%) | 51 (3%) | 0.010 |
Values expressed as median (IQR) or n (%) as appropriate.
Percentage by column.
Renal dysfunction, eGFR<60ml/min/1.73m2; hepatic dysfunction, total serum bilirubin >1.2 mg/dL. CHD, congenital heart disease; ECMO, extracorporeal membrane oxygenation; VAD, ventricular assist device.
Table 7.
Characteristics of pediatric heart transplant recipients by time from donor admission to brain death declaration.
| Time from Admission to Brain Death Declaration | |||||
|---|---|---|---|---|---|
| 1–2 days (n=1775) |
3–4 days (n=2767) |
5–7 days (n=1281) |
8–10 days (n=408) |
p-value | |
| Demographics | |||||
| Age | 11 (3–15) | 3 (0–11) | 3 (0–12) | 7 (1–13) | <0.001 |
| Weight | 38 (13–58) | 14 (7–35) | 14 (7–39) | 21 (11–42) | <0.001 |
| Male sex | 1037 (58%) | 1507 (54%) | 698 (54%) | 211 (52%) | 0.017 |
| Race/ethnicitya | |||||
| Black | 399 (22%) | 527 (19%) | 248 (19%) | 80 (20%) | 0.283 |
| Hispanic | 329 (19%) | 585 (21%) | 270 (21%) | 84 (21%) | |
| White | 927 (52%) | 1462 (53%) | 673 (53%) | 214 (52%) | |
| Diagnosisa | |||||
| CHD | 938 (54%) | 1203 (45%) | 555 (45%) | 205 (51%) | <0.001 |
| Cardiomyopathy | 641 (37%) | 1349 (50%) | 622 (50%) | 168 (42%) | |
| Retransplant | 125 (7%) | 114 (4%) | 53 (4%) | 23 (6%) | |
| Clinical status | |||||
| UNOS status 1A/1 | 1494 (84%) | 2297 (83%) | 1078 (84%) | 341 (84%) | 0.706 |
| Renal dysfunction | 238 (13%) | 321 (12%) | 162 (13%) | 34 (8%) | 0.029 |
| Hepatic dysfunction | 375 (22%) | 477 (18%) | 253 (20%) | 61 (15%) | 0.001 |
| Ventilator | 194 (11%) | 427 (15%) | 218 (17%) | 58 (14%) | <0.001 |
| Inotropes | 804 (45%) | 1304 (47%) | 656 (51%) | 192 (47%) | 0.013 |
| VAD | 467 (26%) | 703 (25%) | 307 (24%) | 108 (26%) | 0.493 |
Values expressed as median (IQR) or n (%) as appropriate.
Percentage by column.
Renal dysfunction, eGFR<60ml/min/1.73m2; hepatic dysfunction, total serum bilirubin >1.2 mg/dL. CHD, congenital heart disease; UNOS, United Network for Organ Sharing; VAD, ventricular assist device.
Posttransplant outcomes
Posttransplant outcomes were similar for both adult and pediatric recipients of heart donors from each admission-to-BDD cohort (Table 8).
For adult heart transplant recipients, posttransplant hospital lengths of stay by donor admission-to-BDD cohort were: median 15 (IQR 11–24) days, 16 (11–24) days, 16 (11–23) days, and 15 (11–23) days, respectively (p=0.062). The incidence of acute in-hospital rejection was 17–18% (p=0.785) and rejection requiring treatment within the first year posttransplant was 28–29% (p=0.947) for recipients from each donor admission-to-BDD cohort. Posttransplant survival for adult recipients was comparable for donors from each cohort (p=0.999; Figure 2), with neither donor admission-to-BDD cohort nor mechanism of death independently associated with differences in posttransplant survival upon multivariable analysis (Table S3).
Figure 2.
Posttransplant survival for adult heart transplant recipients based upon time from donor admission to brain death declaration (BDD). 95% confidence intervals shown.
For pediatric heart transplant recipients, posttransplant hospital lengths of stay differed slightly depending upon time to donor BDD, but no clear clinical trends emerged. By donor admission-to-BDD cohort, median lengths of stay were: 18 (IQR 12–30) days, 21 (IQR 13–38) days, 22 (IQR 14–39) days, and 20 (IQR 13–30) days, respectively (p<0.001). The incidence of in-hospital acute rejection (13–15%; p=0.288) and rejection requiring treatment within the first year posttransplant (23–25%; p=0.730) were equivalent. As in adults, posttransplant survival was similar for pediatric recipients of all donor admission-to-BDD cohorts (p=0.230; Figure 3), with neither donor admission-to-BDD cohort nor mechanism of death independently associated with differences in posttransplant survival upon multivariable analysis (Table S4).
Figure 3.
Posttransplant survival for pediatric heart transplant recipients based upon time from donor admission to brain death declaration (BDD). 95% confidence intervals shown.
Discussion
Heart donor utilization remains suboptimal across the United States. Despite a near-annual increase in the number of potential donors consented, the proportion with hearts accepted for transplantation has remained stable over the past 15 years (~33%).1,2,10 Accordingly, rates of waitlist mortality are 8–12% for adults and twice as high for certain pediatric candidates.4,11–14 An understanding of the ways in which donor utilization can be optimized is therefore imperative to meet the ongoing demand for donor hearts, thereby ensuring that the greatest possible number of candidates receive transplants that yield optimal long-term outcomes.
The present analysis reveals that potential heart donors for whom brain death is declared longer after admission are less likely to be accepted for transplantation. Independent of donor demographic or clinical features, BDD occurring after 4 days of admission was associated with lower odds of acceptance, with donors with BDD occurring 8–10 days after admission being 30% less likely to be used for transplantation than donors with BDD occurring on the first or second day as admission. However, donors with later BDD were not of inferior quality and, in some regards, had favorable clinical characteristics. With the exception of mechanism of death, the characteristics of nonaccepted donors within each admission-to-BDD cohort were generally similar, suggesting against the notion that donors with later BDD were rightfully discarded in greater numbers compared to donors with earlier BDD. For the donors accepted for transplantation, posttransplant outcomes were equivalent for both adult and pediatric recipients irrespective of the duration between admission and BDD, and admission-to-BDD cohort was not associated with posttransplant survival differences upon multivariable analysis. The impact of this finding is profound. Had donor utilization remained consistent for all admission-to-BDD durations at 36% (the rate of acceptance for donors with BDD on admission day 1–2), an additional 2,865 additional patients would have received heart transplants during the study period.
Brain death results in a massive hormonal and hemodynamic changes from widespread systemic inflammation and loss of biofeedback mechanisms.5–7 Clinical management of the brain-dead donor therefore becomes both resource-intensive and time-sensitive, and the goal is to procure organs prior to the onset of irreversible injury. Thus, the notion of a “transplant window” – that is, the ideal timeframe for organ procurement – exists, but remains undefined. Several prior single-institutional studies have suggested that earlier procurement is better, reporting inferior posttransplant survival among adult transplant recipients who received hearts from donors with prolonged brain death (≤66 hours of brain death duration in one study, ≤14 hours in another).15,16 While the admission-to-BDD interval represents only a general surrogate for the total duration of brain death, the current study calls this notion into question, revealing that posttransplant hospital and long-term survival outcomes were comparable for both adult and pediatric recipients regardless of donor admission-to-BDD interval.
Nevertheless, the fact remains that donors with later BDD after hospital admission are less frequently accepted for heart transplantation, directly resulting in fewer recipients receiving transplants and the discard of a number of potentially suitable organs. It is likely that surgeons do not explicitly consider the admission-to-BDD interval as part of their selection process, but the study results demonstrate that a persistent bias seems to exist suggesting that donors with longer durations of brain death (ie, later BDD) are somehow less favorable. This presumption may stem from the presumed inferiority of donors expiring from overdose or anoxia and requiring CPR, the proportion of which were greater among potential donors with later BDD after admission. It is known that donors requiring CPR and those with death from anoxia or overdose are less likely to be utilized for heart transplantation.10,17 In the current study, donors with anoxic/overdose-associated deaths were less frequently accepted for transplantation in every donor admission-to-BDD cohort, suggesting this bias to be real. However, beliefs regarding the inferiority of donors expiring from anoxia or drug overdose and requiring CPR have not been substantiated.18–20 Several analyses have shown survival to be equivalent following heart transplantation with overdose-death donors compared to donors of other mechanisms of death, and a recent study from the authors’ group showing that up to an hour of total donor CPR yields comparable posttransplant survival compared to no CPR.17,18,20,21 The former finding was echoed by the present study, in which anoxic or overdose deaths were not associated with differences in posttransplant survival upon univariate or multivariable analysis. When deciding to accept organs, transplant teams may therefore be weighing the supposed negative impacts of anoxic/overdose death and CPR requirements to a greater extent than other clinical factors that have been demonstrated to negatively impact donor quality, and thus posttransplant outcomes, such as renal dysfunction and inotrope requirements – both less common in donors with later BDD.
Certainly, donor BDD and heart transplant offer acceptance practices are multifaceted and complex and cannot therefore be comprehensively examined in a single study. The present study, though, confirms that room exists for quality improvement in heart donor utilization in the United States and reveals that a large proportion of potential donors with favorable clinical characteristics go unused. As the authors show, increased utilization of donors with later BDD can help bridge the current discrepancy between the number of patients in need of heart transplantation and the number of donors considered without compromising posttransplant outcomes.
Limitations and Future Directions
Like all retrospective analyses, the possibility of biases and statistical errors exist in the current study. While UNOS is subject to reporting and selection biases, mandatory reporting decreases the impact of biases of omission. One major benefit of using national datasets like UNOS is the vast amount of data present, although this can lend itself to statistical errors. Type 1 statistical errors likely influenced the statistical significance seen in comparisons between admission-to-BDD groups since few clinically meaningful differences existed. Of note, donors were not studied by age group (ie, adult versus pediatric) since many potential donors are offered both to adult and pediatric candidates and since UNOS does not differentiate between “adult” and “pediatric” donors. The unexpected finding that rates of donor noncardiac organ dysfunction were lower as the interval from admission to BDD increased deserves further attention, especially since the donor pool differed with regards to mechanisms of death between admission-to-BDD groups. Additional study is needed to more fully characterize the progression of clinical deterioration following brain death.
Additionally, as a dataset predicated primarily upon donor utilization, UNOS unfortunately contains limited information on posttransplant outcomes beyond hospital length of stay, 1-year rejection episodes, and survival (a reasonable surrogate for severe coronary vasculopathy and late graft failure in the case of heart transplantation). Future studies should therefore seek to include more granular and expanded data on posttransplant outcomes not readily available in UNOS (eg, posttransplant ventilation, incidence and management of delayed graft function, mechanical circulatory support requirement, chronic vasculopathy, infection), which would add clinically meaningful information to help clinicians improve donor organ utilization and, therefore, optimize the proportion of recipients who receive long-term successful transplants.
Conclusions
Heart donors for whom brain death is declared 4 days or more after admission are disproportionately discarded despite similar-to-favorable clinical characteristics, resulting in nearly 3000 fewer transplants during the study period. While donors with later brain death declaration do have several traits widely considered to suggest inferior donor and graft quality (eg, anoxic or overdose death, and CPR requirements), these traits have not been shown in the literature to yield inferior posttransplant outcomes. Indeed, the current study reveals that posttransplant hospital outcomes and long-term survival are equivalent regardless of the duration between donor admission and brain death declaration. More active consideration of donors with later brain death declaration, including those with CPR requirements or anoxic brain death, may therefore help increase the number of patients who receive heart transplants, and also further the discussions of the role for standardized criteria for organ acceptance and donor-recipient matching algorithms rather than relying on individual provider or programmatic practices.
Supplementary Material
Table 3.
Characteristics of accepted vs nonaccepted heart donors in whom brain death declaration occurred on admission day 3–4.
| Accepted heart donor (n=15 019) |
Nonaccepted heart donor (n=29 014) |
p-value | |
|---|---|---|---|
| Demographics | |||
| Age (yr) | 28 (19–38) | 49 (36–58) | <0.001 |
| Weight (kg) | 77 (62–91) | 81 (67–97) | <0.001 |
| Male sex | 10 057 (67%) | 16 153 (56%) | <0.001 |
| Race/ethnicity | |||
| Black | 2504 (17%) | 4402 (15%) | <0.001 |
| Hispanic | 2722 (18%) | 3748 (13%) | |
| White | 9323 (62%) | 19 697 (68%) | |
| Comorbidities | |||
| Diabetes | 550 (4%) | 5337 (19%) | <0.001 |
| Hypertension | 1952 (13%) | 13 348 (46%) | <0.001 |
| Cigarette smoker | 1653 (11%) | 8404 (30%) | <0.001 |
| Heavy alcohol | 2102 (14%) | 5942 (21%) | <0.001 |
| Mechanism of death | |||
| Anoxia/overdosea | 4196 (45%) | 5081 (55%) | <0.001 |
| CVAa | 2342 (19%) | 10 051 (81%) | |
| Trauma–blunta | 4145 (52%) | 3881 (48%) | |
| Trauma–penetratinga | 1610 (62%) | 999 (38%) | |
| Clinical status | |||
| Donor CPR | 7680 (53%) | 14 476 (52%) | 0.028 |
| Renal dysfunction | 5267 (35%) | 14 242 (49%) | <0.001 |
| Hepatic dysfunction | 2875 (19%) | 5640 (20%) | 0.411 |
| Inotropes | 5903 (40%) | 13 692 (47%) | <0.001 |
| LVEF <50% | 246 (2%) | 4603 (29%) | <0.001 |
Values expressed as median (IQR) or n (%) as appropriate.
Percentage by row.
Renal dysfunction, eGFR<60ml/min/1.73m2; hepatic dysfunction, total serum bilirubin >1.2 mg/dL. CPR, cardiopulmonary resuscitation; CVA, cerebrovascular accident; LVEF, left ventricular ejection fraction.
Table 4.
Characteristics of accepted vs nonaccepted heart donors in whom brain death declaration occurred on admission day 5–7.
| Accepted heart donor (n=7350) |
Nonaccepted heart donor (n=17 159) |
p-value | |
|---|---|---|---|
| Demographics | |||
| Age (yr) | 28 (19–38) | 47 (34–56) | <0.001 |
| Weight (kg) | 76 (62–91) | 82 (68–99) | <0.001 |
| Male sex | 4810 (65%) | 9953 (58%) | <0.001 |
| Race/ethnicity | |||
| Black | 1290 (18%) | 2708 (16%) | <0.001 |
| Hispanic | 1469 (20%) | 2314 (13%) | |
| White | 4339 (59%) | 11 489 (67%) | |
| Comorbidities | |||
| Diabetes | 253 (3%) | 2895 (17%) | <0.001 |
| Hypertension | 886 (12%) | 7385 (44%) | <0.001 |
| Cigarette smoker | 722 (10%) | 4670 (28%) | <0.001 |
| Heavy alcohol | 1067 (15%) | 3678 (22%) | <0.001 |
| Mechanism of death | |||
| Anoxia/overdosea | 2486 (42%) | 3425 (58%) | <0.001 |
| CVAa | 1113 (19%) | 4765 (81%) | |
| Trauma–blunta | 1737 (44%) | 2188 (56%) | |
| Trauma–penetratinga | 426 (55%) | 343 (45%) | |
| Clinical status | |||
| Donor CPR | 4117 (58%) | 9626 (58%) | 0.812 |
| Renal dysfunction | 2085 (28%) | 6472 (38%) | <0.001 |
| Hepatic dysfunction | 1372 (19%) | 3039 (18%) | 0.097 |
| Inotropes | 2811 (38%) | 5871 (34%) | <0.001 |
| LVEF <50% | 113 (2%) | 2479 (27%) | <0.001 |
Values expressed as median (IQR) or n (%) as appropriate.
Percentage by row.
Renal dysfunction, eGFR<60ml/min/1.73m2; hepatic dysfunction, total serum bilirubin >1.2 mg/dL. CPR, cardiopulmonary resuscitation; CVA, cerebrovascular accident; LVEF, left ventricular ejection fraction.
Financial Disclosure
Funding was received from the National Institutes of Health: R01HL147957 - “Novel Methods to Grow the Impact of Pediatric Thoracic Transplantation” - Principal Investigators: David L. S. Morales, MD, Farhan Zafar, MD, MS.
Abbreviations
- BDD
brain-death declaration
- CDC
Centers for Disease Control and Prevention
- CPR
cardiopulmonary resuscitation
- DCD
donation after circulatory death
- LVEF
left ventricular ejection fraction
- UNOS
United Network for Organ Sharing database
Footnotes
Meeting Presentation: Presented at the 59th Annual Meeting of the Society of Thoracic Surgeons, January 22, 2023, San Diego, California
Disclaimer
Dr. Zafar is a transplant procurement surgeon for TransMedics, Inc. The remaining authors have no relevant financial disclosures and no conflicts of interest.
References
- 1.Israni AK, Zaun D, Hadley N, et al. OPTN/SRTR 2018 annual data report: deceased organ donation. Am J Transplant. 2020;20(s1):509–541. [DOI] [PubMed] [Google Scholar]
- 2.Giafaglione J, Morrison A, Gowda C, et al. Pediatric donor heart allocation in the United States, 2006‐2017: Current patterns and potential for improvement. Pediatr Transplant. 2020;24(6):e13743. [DOI] [PubMed] [Google Scholar]
- 3.Colvin M, Smith JM, Ahn Y, et al. OPTN/SRTR 2020 annual data report: heart. Am J Transplant. 2022;22(S2):350–437. [DOI] [PubMed] [Google Scholar]
- 4.Hsich EM. Matching the market for heart transplantation. Circ Heart Fail. 2018;9(4):e002679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Amado JoséA, López-Espadas F, Vázquez-Barquero A, et al. Blood levels of cytokines in brain-dead patients: Relationship with circulating hormones and acute-phase reactants. Metabolism. 1995;44(6):812–816. [DOI] [PubMed] [Google Scholar]
- 6.Smith M Physiologic changes during brain stem death—lessons for management of the organ donor. J Heart Lung Transplant. 2004;23(9):S217–S222. [DOI] [PubMed] [Google Scholar]
- 7.Barklin A Systemic inflammation in the brain‐dead organ donor. Acta Anaesth Scand. 2009;53(4):425–435. [DOI] [PubMed] [Google Scholar]
- 8.Staples A, LeBlond R, Watkins S, et al. Validation of the revised Schwartz estimating equation in a predominantly non-CKD population. Pediatr Nephrol. 2010;25(11):2321–2326. [DOI] [PubMed] [Google Scholar]
- 9.Chokshi A, Cheema FH, Schaefle KJ, et al. Hepatic dysfunction and survival after orthotopic heart transplantation: application of the MELD scoring system for outcome prediction. J Heart Lung Transplant. 2012;31(6):591–600. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Dharmavaram N, Hess T, Jaeger H, et al. National trends in heart donor usage rates: are we efficiently transplanting more hearts? J Am Heart Assoc. 2020;10(15):e019655. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Colvin M, Smith JM, Ahn Y, et al. OPTN/SRTR 2019 annual data report: heart. Am J Transplant. 2021;21(S2):356–440. [DOI] [PubMed] [Google Scholar]
- 12.Goldstein BA, Thomas L, Zaroff JG, et al. Assessment of heart transplant waitlist time and pre- and post-transplant failure. Epidemiology. 2016;27(4):469–476. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Rana A, Fraser CD, Scully BB, et al. Inferior outcomes on the waiting list in low‐volume pediatric heart transplant centers. Am J Transplant. 2017;17(6):1515–1524. [DOI] [PubMed] [Google Scholar]
- 14.Denfield SW, Azeka E, Das B, et al. Pediatric cardiac waitlist mortality—still too high. Pediatr Transplant. 2020;24(3):e13671. [DOI] [PubMed] [Google Scholar]
- 15.Cantin B, Kwok BWK, Chan MCY, et al. The impact of brain death on survival after heart transplantation: time is of the essence. Transplantation. 2003;76(9):1275–1279. [DOI] [PubMed] [Google Scholar]
- 16.Ramjug S, Hussain N, Yonan N. Prolonged time between donor brain death and organ retrieval results in an increased risk of mortality in cardiac transplant recipients. Interact Cardiovasc Thorac Surg. 2011;12(6):938–942. [DOI] [PubMed] [Google Scholar]
- 17.Durand CM, Bowring MG, Thomas AG, et al. The drug overdose epidemic and deceased-donor transplantation in the United States: a national registry study. Ann Intern Med. 2018;168(10):702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Cohen O, Zerda DJDL, Beygui R, et al. Donor brain death mechanisms and outcomes after heart transplantation. Transplant Proc. 2007;39(10):2964–2969. [DOI] [PubMed] [Google Scholar]
- 19.Hsich EM, Blackstone EH, Thuita LW, et al. Heart transplantation an in-depth survival analysis. JACC Heart Fail. 2020;8(7):557–568. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Khush KK, Menza R, Nguyen J, et al. Donor predictors of allograft use and recipient outcomes after heart transplantation. Circ Heart Fail. 2018;6(2):300–309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Shudo Y, Cohen JE, Lingala B, et al. Impact of “increased-risk” donor hearts on transplant outcomes: a propensity-matched analysis. J Thorac Cardiovasc Surg. 2019;157(2):603–610. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.