Abstract
Background:
Donation after circulatory death (DCD) heart transplantation has promising early survival, but effects on rejection remain unclear.
Methods:
The United Network for Organ Sharing (UNOS) database was queried for adult heart transplants from 12/01/2019 to 12/31/2021. Multiorgan transplants and loss to follow-up were excluded. Primary outcome was acute rejection, comparing DCD and donation after brain death (DBD) transplants.
Results:
A total of 292 DCD and 5582 DBD transplants met study criteria. Most DCD transplants were transplanted at status 3–4 (61.0%) compared to 58.6% of DBD recipients at status 1–2. DCD recipients were less likely to be hospitalized at transplant (26.7% vs 58.3%, p<0.001) and to require IABP (9.6% vs 28.9%, p<0.001), ECMO (0.3% vs 5.9%, p<0.001) or temporary LVAD (1.0% vs 2.7%, p<0.001).
DCD recipients were more likely to have acute rejection prior to discharge (23.3% vs 18.4%, p=0.044) and to be hospitalized for rejection (23.4% vs 11.4%, p=0.003) at a median follow-up of 15 months; the latter remained significant after propensity-matching. On multivariable logistic regression, DCD donation was an independent predictor of acute rejection (OR 1.47, 95% CI 1.00–2.15, p=0.048) and hospitalization for rejection (OR 2.03, 95% CI 1.06–3.70, p=0.026). On center-specific subgroup analysis, DCD recipients continued to have higher rates of hospitalization for rejection (23.4% vs 13.8%, p=0.043).
Conclusion:
DCD recipients are more likely to experience acute rejection. Early survival is similar between DCD and DBD recipients but long term implications of increased early rejection in DCD recipients requires further investigation.
Keywords: donation after circulatory death, DCD, Acute rejection
Introduction
Donation after circulatory death (DCD) heart transplantation has historically been avoided due to concerns about ischemia and reperfusion injury, but with recent advances to organ perfusion techniques, DCD heart transplants are becoming a global reality. In 2014, Australia implemented the Sydney DCD heart program, pioneering a method of direct recovery and ex-vivo reanimation during transport [1]. This method allows the donor heart to be reperfused to a beating physiologic state during transport, optimizing oxygen and substrate delivery, and allowing for the assessment of the heart prior to transplantation. Transplant centers in the U.K. [2] and Belgium [3] have followed suit, with the development of alternative procurement techniques utilizing normothermic regional perfusion, which reperfuses the heart and assesses its function in situ. In 2019, the first adult DCD heart transplant in the modern era was performed in the U.S. Since then, DCD heart transplants have increased to 186 transplants in 2021 and continue to rise, accompanied by research efforts to understand the post-transplant consequences of this new strategy.
Early studies demonstrate comparable perioperative and early survival, with 1-year survival rates of 91–97% compared to 89–92% in DBD recipients [1, 2, 4, 5]. Medium term follow-up is available from the Sydney group with 5-year survival rates of 94% compared to 82% in DBD transplants [1] but has yet to be corroborated by other studies. Important risk factors that affect long-term survival have also yet to be investigated, including the effects of DCD donation on acute and chronic rejection. While the latter is limited by the relative recency of this technique, in this study, we aim to examine acute rejection in the early post-transplant years, which has been shown to significantly impact graft dysfunction and late mortality [6]. Our study investigates the effect of DCD heart donation on acute rejection, utilizing the United Network for Organ Sharing (UNOS) database to summarize the U.S. experience.
Methods
Study Design
This retrospective cohort study was approved by the Institutional Review Board of the Massachusetts General Hospital (protocol 2017P001969, approved on September 28, 2017) and complies with the International Society for Heart and Lung Transplantation ethics statement. We queried the UNOS database for adult heart transplant recipients from 12/1/2019 to 12/31/2021, ensuring >6 months follow-up for all patients, with a mean follow-up time of 15 months. We excluded multi-organ transplants, prior heart transplants, and those lost to follow-up. Loss to follow-up was determined by missing information for patient survival in the UNOS Thoracic Dataset or lack of inclusion in the UNOS Thoracic Follow-up Dataset. We compared donation after brain death (DBD) and donation after circulatory death (DCD) transplants including all procurement and reperfusion methods. Primary outcome was acute rejection measured by the following metrics: acute rejection prior to discharge (as defined by reported rejection to UNOS which includes both treated and untreated episodes, and assumes both humoral and cellular rejection), treatment for rejection within 1 year (including those treated for rejection prior to discharge), and hospitalization for rejection after discharge. Secondary outcomes were overall survival measured by Kaplan Meier analysis.
Statistical Analysis
R software was used for statistical analyses [7]. Missing data were excluded from calculations and percent missingness is included in each table. Normality was tested with the Anderson-Darling test, with null hypothesis (normal distribution) rejected if p<0.05 (Supplementary Table 1). Normal continuous variables were compared with the two-way t-test, nonnormal continuous variables with the Mann Whitney U test, and categorical with the Fisher exact test. For all analyses, p-value of <0.05 was considered significant. Propensity-matching was performed with nearest-neighbor matching as previously described in cardiovascular research [8], using a 3:1 ratio with successful matching defined as SMD <0.150 (Supplementary Table 1). Kaplan Meier analysis was performed for survival comparison, and multivariable logistic regression for acute rejection metrics. Multivariable factors were chosen based on significance in univariable regression or clinical significance (Supplementary Table 2, 3).
Subgroup Analysis
Two different subgroup analyses were performed, focusing on 1) waitlist status and 2) center-specific data. For waitlist status subgroups, we separated patients based on priority: Status 1–2, Status 3–4, and Status 5–6 as a proxy for acuity. Acute rejection metrics were compared between DCD and DBD recipients in each of these 3 subgroups using the comparison tests detailed above. For center-specific data, we included a smaller subset of patients, focusing only on centers performing both DCD and DBD heart transplants. Donor and recipient characteristics were again compared (Supplementary Table 3), and acute rejection metrics reported.
Results
Recipient and Donor Characteristics
A total of 292 patients received DCD heart donations compared to 5582 DBD heart transplants within the study period. DCD recipients were more likely to be white (70.9% vs 60.7%, p=0.008), ABO type O (57.9% vs 40.0%, p<0.001), with higher BMI (median 29.5 vs 27.7, p<0.001), and history of smoking (48.3% vs 41.8%, p=0.033). DCD donors were younger (8.2% over age 40 years, compared to 22.7%, p<0.001), with a higher percentage of male (88.0% vs 71.9%, p<0.001) and white (78.1% vs 61.3%, p<0.001) donors. There were no significant differences in donor comorbidities. DCD transplants traveled further distances (median: 360 vs 223 miles, p<0.001), with longer “out of body” times – time from procurement to implantation (median: 5.13 vs 3.43 hours, p<0.001). Additional characteristics are summarized in Table 1.
Table 1 –
Recipient and Donor Characteristics
| DBD | DCD | p | Missing | |
|---|---|---|---|---|
|
| ||||
| n | 5582 | 292 | ||
| Age (Recipient) (median [IQR]) | 57.0 [46.0, 64.0] | 57.5 [45.0, 64.0] | 0.593 | 0.0 |
| Male (Recipient) (%) | 4107 (73.6) | 225 (77.1) | 0.212 | 0.0 |
| Ethnicity (Recipient) (%) | 0.008 | 0.0 | ||
| White | 3391 (60.7) | 207 (70.9) | ||
| Black | 1347 (24.1) | 59 (20.2) | ||
| Hispanic | 565 (10.1) | 18 (6.2) | ||
| Asian | 214 (3.8) | 6 (2.1) | ||
| Other | 65 (1.2) | 2 (0.7) | ||
| Blood type (Recipient) (%) | <0.001 | 0.0 | ||
| A | 2185 (39.1) | 83 (28.4) | ||
| AB | 285 (5.1) | 7 (2.4) | ||
| B | 879 (15.7) | 33 (11.3) | ||
| O | 2233 (40.0) | 169 (57.9) | ||
| BMI (Recipient) (median [IQR]) | 27.7 [24.3, 31.5] | 29.5 [26.6, 32.9] | <0.001 | 0.1 |
| Diabetes (Recipient) (%) | 1553 (27.8) | 91 (31.2) | 0.242 | 0.0 |
| Steroids (Recipient) (%) | 268 (4.8) | 11 (3.8) | 0.484 | 0.9 |
| History of smoking (Recipient) (%) | 2333 (41.8) | 141 (48.3) | 0.033 | 0.0 |
| Transfused prior to transplant (Recipient) (%) | 737 (13.3) | 16 (5.5) | <0.001 | 0.7 |
| Ischemic cardiomyopathy (Recipient) (%) | 1489 (26.7) | 88 (30.1) | 0.217 | 0.0 |
| IABP at transplant (%) | 1611 (28.9) | 28 (9.6) | <0.001 | 0.0 |
| ECMO at transplant (%) | 329 (5.9) | 1 (0.3) | <0.001 | 0.0 |
| LVAD at transplant (%) | <0.001 | 0.0 | ||
| None | 4009 (71.8) | 175 (59.9) | ||
| Durable | 1422 (25.5) | 114 (39.0) | ||
| Temporary | 151 (2.7) | 3 (1.0) | ||
| Functional status at the time of transplant (Recipient) (%) | <0.001 | 5.9 | ||
| 10–30% | 3053 (58.3) | 77 (26.7) | ||
| 40–60% | 1361 (26.0) | 110 (38.2) | ||
| 70–90% | 804 (15.3) | 101 (35.1) | ||
| 100% | 22 (0.4) | 0 (0.0) | ||
| PVR >3 Wood units at the time of transplant (Recipient) (%) | 1219 (24.4) | 45 (16.0) | 0.002 | 10.2 |
| Transpulmonary gradient >12 mmHg at the time of transplant (Recipient) (%) | 1376 (27.2) | 63 (22.3) | 0.084 | 9.1 |
| Creatinine at the time of transplant (Recipient) (median [IQR]) | 1.1 [0.9, 1.4] | 1.2 [1.0, 1.4] | 0.449 | 0.0 |
| Inotropes at the time of transplant (Recipient) (%) | 2209 (39.6) | 66 (22.6) | <0.001 | 0.0 |
| Age >40 (Donor) (%) | 1267 (22.7) | 24 (8.2) | <0.001 | 0.0 |
| Male (Donor) (%) | 4016 (71.9) | 257 (88.0) | <0.001 | 0.0 |
| Ethnicity (Donor) (%) | <0.001 | 0.0 | ||
| White | 3424 (61.3) | 228 (78.1) | ||
| Black | 960 (17.2) | 29 (9.9) | ||
| Hispanic | 1019 (18.3) | 29 (9.9) | ||
| Asian | 93 (1.7) | 3 (1.0) | ||
| Other | 86 (1.5) | 3 (1.0) | ||
| BMI (Donor) (median [IQR]) | 27.1 [23.7, 31.5] | 26.6 [24.1, 30.7] | 0.649 | 0.0 |
| History of smoking (Donor) (%) | 668 (12.3) | 25 (8.7) | 0.079 | 2.6 |
| History of cocaine use (Donor) (%) | 1476 (27.1) | 70 (24.4) | 0.346 | 2.4 |
| History of heavy alcohol (Donor) (%) | 1008 (18.7) | 77 (26.9) | 0.001 | 3.5 |
| CDC high risk donor (%) | 1721 (30.8) | 67 (22.9) | 0.005 | 0.0 |
| Diabetes (Donor) (%) | 240 (4.4) | 6 (2.1) | 0.079 | 1.3 |
| Hypertension (Donor) (%) | 870 (15.8) | 35 (12.0) | 0.094 | 1.4 |
| LVEF <45% (Donor) (%) | 43 (0.8) | 0 (0.0) | 0.249 | 0.1 |
| Last Creatinine (Donor) (median [IQR]) | 1.0 [0.8, 1.7] | 0.8 [0.7, 1.0] | <0.001 | 0.0 |
| Gender mismatch (F->M) (%) | 630 (11.3) | 12 (4.1) | <0.001 | 0.0 |
| % Predicted heart mass mismatch (median [IQR]) | −0.7 [−12.9, 8.2] | −2.2 [−13.6, 8.2] | 0.369 | 0.0 |
| ABO match - Identical (%) | 4785 (85.7) | 245 (83.9) | 0.432 | 0.0 |
| HLA mismatch (# alleles) (median [IQR]) | 5.0 [4.0, 5.0] | 5.0 [4.0, 5.0] | 0.740 | 8.4 |
| Thymoglobulin induction (%) | 1201 (21.5) | 19 (6.5) | <0.001 | 0.0 |
| Basiliximab induction (%) | 1362 (24.4) | 114 (39.0) | <0.001 | 0.0 |
| Transplant year (%) | <0.001 | 0.0 | ||
| 2019 | 206 (3.7) | 5 (1.7) | ||
| 2020 | 2701 (48.4) | 101 (34.6) | ||
| 2021 | 2675 (47.9) | 186 (63.7) | ||
| Region (%) | <0.001 | 0.0 | ||
| 1 | 262 (4.7) | 53 (18.2) | ||
| 2 | 566 (10.1) | 0 (0.0) | ||
| 3 | 648 (11.6) | 20 (6.8) | ||
| 4 | 501 (9.0) | 1 (0.3) | ||
| 5 | 909 (16.3) | 34 (11.6) | ||
| 6 | 163 (2.9) | 0 (0.0) | ||
| 7 | 497 (8.9) | 32 (11.0) | ||
| 8 | 396 (7.1) | 13 (4.5) | ||
| 9 | 432 (7.7) | 16 (5.5) | ||
| 10 | 470 (8.4) | 0 (0.0) | ||
| 11 | 738 (13.2) | 123 (42.1) | ||
| Distance (miles) (median [IQR]) | 223.0 [99.0, 396.0] | 360.5 [136.0, 574.5] | <0.001 | 0.0 |
| Out of body time* (hours) (median [IQR]) | 3.4 [2.9, 4.0] | 5.1 [3.5, 6.4] | <0.001 | 0.2 |
| Waitlist status at transplant (%) | <0.001 | 0.4 | ||
| 1 | 501 (9.0) | 3 (1.0) | ||
| 2 | 2755 (49.6) | 53 (18.2) | ||
| 3 | 902 (16.2) | 62 (21.2) | ||
| 4 | 1116 (20.1) | 116 (39.7) | ||
| 5 | 2 (0.0) | 0 (0.0) | ||
| 6 | 280 (5.0) | 58 (19.9) | ||
| Total waitlist days (median [IQR]) | 31.0 [9.0, 158.8] | 46.0 [14.0, 175.8] | 0.004 | 0.0 |
Time from procurement to implantation
Overall, DCD recipients were transplanted in a less acute state. They were less likely to be hospitalized at the time of transplant (26.7% vs 58.3%, p<0.001), and less likely to require IABP (9.6% vs 28.9%, p<0.001), ECMO (0.3% vs 5.9%, p<0.001), or temporary LVAD (1.0% vs 2.7%, p<0.001) at transplant. The majority of DCD recipients were transplanted at status 3–4 (61.0%) while DBD recipients were transplanted at higher acuity status 1–2 (58.6%), p<0.001 (Figure 2A).
Figure 2. Waitlist Subgroup Analysis.
(A) DCD heart transplants were more likely to be transplanted at lower priority status, with the majority transplanted at status 3–4, while the majority of DBD transplants were transplanted at status 1–2. (B)
Patients were separated by waitlist status into subgroups: Status 1–2, Status 3–4, and Status 5–6. DCD and DBD transplants were compared within each subgroup, and DCD patients were found to have higher rates of hospitalization for rejection, with statistical significance in the Status 3–4 subgroup.
There was no significant difference in ABO match (p=0.432) or HLA mismatch (p=0.740). DCD recipients were less likely to have PRA >10% (15.3% vs 21.3%, p=0.031) and peak PRA >10% (21.0% vs 27.6%, p=0.027). Basiliximab induction was more common in DCD transplants (39.0% vs 24.4%, p<0.001), and anti-thymocyte globulin (ATG) induction less common (6.5% vs 21.5%, p<0.001).
Outcomes – Rejection and Survival
In the total cohort, DCD recipients were more likely to have an acute rejection episode prior to discharge (23.3% vs 18.4%, p=0.044), and after discharge, were more likely to be hospitalized for rejection (23.4% vs 11.4%, p=0.003) at a mean follow-up of 15 months (Table 2). There was a trend towards increased need for treatment for rejection within 1 year, but this did not reach statistical significance (13.0% vs 10.1%, p=0.134).
Table 2 –
Rejection Outcomes
| Total Cohort | Propensity-Matched Cohort | |||||
|---|---|---|---|---|---|---|
|
| ||||||
| DBD | DCD | p | DBD | DCD | p | |
| n | 5582 | 292 | 358 | 152 | ||
| PRA >10% (%) | 869 (21.3) | 38 (15.3) | 0.031 | 41 (15.7) | 19 (13.1) | 0.574 |
| Peak PRA >10% (%) | 1127 (27.6) | 52 (21.0) | 0.027 | 58 (22.3) | 28 (19.3) | 0.562 |
| Acute rejection prior to discharge (%) | 1027 (18.4) | 68 (23.3) | 0.044 | 48 (13.4) | 26 (17.1) | 0.344 |
| Treated for acute rejection prior to discharge (%) | 543 (9.7) | 38 (13.0) | 0.083 | 23 (6.4) | 17 (11.2) | 0.099 |
| Treated for rejection within 1 year (%) | 564 (10.1) | 38 (13.0) | 0.134 | 22 (6.1) | 19 (12.5) | 0.025 |
| Hospitalized for rejection (%) | 230 (11.4) | 18 (23.4) | 0.003 | 9 (8.7) | 12 (27.3) | 0.007 |
| Follow-up time (months) for hospitalization for rejection (mean (SD)) | 14.56 (5.76) | 14.98 (6.04) | 0.529 | 14.25 (5.90) | 14.91 (5.66) | 0.531 |
| Hospitalized for infection (%) | 754 (37.5) | 33 (42.9) | 0.401 | 44 (42.3) | 18 (40.9) | 1.000 |
| Follow-up time (months) for hospitalization for infection (mean (SD)) | 14.56 (5.76) | 14.98 (6.04) | 0.522 | 14.25 (5.90) | 14.91 (5.66) | 0.531 |
| Coronary vasculopathy (%) | 257 (6.8) | 11 (6.3) | 0.926 | 11 (5.4) | 8 (9.1) | 0.365 |
| Follow-up time (months) for coronary vasculopathy (mean (SD)) | 15.41 (5.88) | 14.15 (5.38) | 0.006 | 14.06 (5.58) | 13.98 (5.41) | 0.913 |
Propensity-matching was performed, accounting for recipient, donor, and transplant characteristics (Supplementary Table 1). On propensity-matched analysis, DCD recipients were more likely to be hospitalized for rejection (27.3% vs 8.7%, p=0.007) within mean follow-up times of 14.3 and 14.9 months; and they were more likely to be treated for rejection within 1 year (12.5% vs 6.1%, p=0.025). There was no difference in rates of hospitalization for infection (Table 2). Kaplan Meier analysis showed no difference in longitudinal survival in the total cohort or propensity-matched analysis (Figure 1).
Figure 1. Kaplan Meier Analysis.
Kaplan Meier survival analysis demonstrated comparable survival in patients after heart transplants from donation after circulatory death (DCD) donors vs. donation after brain death (DBD) donors in the total cohort (A) and 3:1 propensity-matched analysis (B). 95% confidence intervals are displayed.
Multivariable Analysis
On univariable and multivariable analysis, DCD heart donation was an independent predictor of acute rejection prior to discharge, and hospitalization for rejection after discharge. DCD donation increased the risk of acute rejection prior to discharge with an odds ratio of 1.47 (95% CI 1.00–2.15, p=0.048), accounting for age, gender, comorbidities, region, and risk factors for rejection (Table 3). Additional negative predictors included older age (OR 0.98, 95% CI 0.98–0.99, p<0.001) and ATG induction (OR 0.34, 95% CI 0.26–0.45, p<0.001); positive predictors for rejection included higher BMI (OR 1.03, 95% CI 1.01–1.05, p<0.001) and UNOS regions summarized in Table 3.
Table 3 –
Multivariable Regression: Acute Rejection Prior to Discharge
| Characteristic | OR1 | 95% CI1 | p-value |
|---|---|---|---|
|
| |||
| Donor type - DCD | 1.47 | 1.00, 2.15 | 0.048 |
| Age (Recipient) | 0.98 | 0.98, 0.99 | <0.001 |
| Male (Recipient) | 0.80 | 0.64, 1.00 | 0.051 |
| Ethnicity (Recipient) | |||
| White | — | — | |
| Black | 1.02 | 0.83, 1.25 | 0.841 |
| Hispanic | 0.73 | 0.53, 1.00 | 0.056 |
| Asian | 0.67 | 0.37, 1.13 | 0.151 |
| Other | 1.40 | 0.63, 2.83 | 0.378 |
| BMI (Recipient) | 1.03 | 1.01, 1.05 | <0.001 |
| Ischemic cardiomyopathy (Recipient) | 1.14 | 0.92, 1.41 | 0.218 |
| Creatinine at the time of transplant (Recipient) | 0.86 | 0.70, 1.04 | 0.133 |
| Male (Donor) | 0.84 | 0.68, 1.05 | 0.127 |
| PRA >10% | 1.06 | 0.85, 1.32 | 0.588 |
| HLA mismatch (# alleles) | 1.14 | 1.05, 1.24 | 0.001 |
| Thymoglobulin induction | 0.34 | 0.26, 0.45 | <0.001 |
| Region | |||
| 1 | — | — | |
| 2 | 1.12 | 0.69, 1.88 | 0.652 |
| 3 | 2.54 | 1.58, 4.21 | <0.001 |
| 4 | 1.93 | 1.17, 3.28 | 0.012 |
| 5 | 0.58 | 0.34, 0.99 | 0.040 |
| 6 | 0.99 | 0.46, 2.04 | 0.973 |
| 7 | 1.76 | 1.08, 2.96 | 0.028 |
| 8 | 0.57 | 0.29, 1.11 | 0.100 |
| 9 | 1.32 | 0.78, 2.27 | 0.310 |
| 10 | 1.16 | 0.67, 2.03 | 0.611 |
| 11 | 0.64 | 0.39, 1.08 | 0.088 |
| Distance (miles) | 1.00 | 1.00, 1.00 | 0.076 |
| Out of body time* (hours) | 0.96 | 0.87, 1.05 | 0.361 |
OR = Odds Ratio, CI = Confidence Interval
Time from procurement to implantation
DCD donation also increased the risk of hospitalization for rejection after discharge (OR 2.03, 95% CI 1.06–3.70, p=0.026) on multivariable analysis (Table 4). Additional risk factors included Black and Hispanic ethnicity (OR 1.48, 95% CI 1.03–2.12, p=0.031; and OR 1.76, 95% CI 1.06–2.84, p=0.025 respectively), while older recipient age was a negative predictor (OR 0.97, 95% CI 0.96–0.98, p<0.001).
Table 4 –
Multivariable Regression: Hospitalization for Rejection
| Characteristic | OR1 | 95% CI1 | p-value |
|---|---|---|---|
|
| |||
| Donor type - DCD | 2.03 | 1.06, 3.70 | 0.026 |
| Age (Recipient) | 0.97 | 0.96, 0.98 | <0.001 |
| Male (Recipient) | 0.89 | 0.63, 1.27 | 0.513 |
| Ethnicity (Recipient) | |||
| White | — | — | |
| Black | 1.48 | 1.03, 2.12 | 0.031 |
| Hispanic | 1.76 | 1.06, 2.84 | 0.025 |
| Asian | 0.54 | 0.09, 1.84 | 0.406 |
| Other | 1.89 | 0.42, 6.20 | 0.339 |
| Diabetes (Recipient) | 0.96 | 0.66, 1.38 | 0.821 |
| Creatinine at the time of transplant (Recipient) | 0.98 | 0.72, 1.17 | 0.860 |
| PRA >10% | 1.13 | 0.76, 1.65 | 0.528 |
| HLA mismatch (# alleles) | 1.01 | 0.87, 1.18 | 0.895 |
| Basiliximab induction | 1.35 | 0.96, 1.89 | 0.084 |
OR = Odds Ratio, CI = Confidence Interval
Subgroup Analysis – Waitlist Status
ohorts were subdivided into waitlist status as a measure of acuity: Status 1–2 (high priority), Status 3–4 (moderate priority), and Status 5–6 (low priority). Differences in acute rejection were observed in the Status 3–4 group. DCD recipients were more likely to be hospitalized for rejection (25.5% vs 9.8%, p=0.001) at a mean follow-up time of 15.4 and 14.6 months respectively (Figure 2B), despite lower rates of PRA >10% (13.3% vs 23.3%, p=0.007). There was a trend towards increased rates of acute rejection prior to discharge but this did not reach statistical significance (24.2% vs 18.4%, p=0.074, Supplementary Table 4). There were no significant differences in acute rejection prior to discharge, treatment for rejection within 1 year, or hospitalizations for rejection in the Status 1–2 and Status 5–6 groups (Supplementary Table 4).
Subgroup Analysis – Center-specific Data
The 292 DCD recipients were compared with 1604 DBD recipients from the same transplant centers (n=22 centers). Similar recipient and donor characteristic trends were observed including higher rates of IABP, ECMO and temporary LVAD in DBD recipients, who were more likely to be hospitalized at the time of transplant (Supplementary Table 5). There were no differences in ABO match (p=0.221) or HLA mismatch (p=0.564). PRA >10% was noted in 15.3% of DCD recipients compared to 21.6% of DBD (p=0.031). ATG induction was used in 6.5% of DCD recipients compared to 15.9% of DBD (p<0.001), and basiliximab used in 39.0% of DCD patients compared to 30.6% of DBD (p=0.006). Higher rates of hospitalization for rejection were again observed in DCD recipients (23.4% vs 13.8%, p=0.043), with no difference in hospitalizations for infection (Table 5).
Table 5 –
Subgroup (Center-specific) Rejection Outcomes
| DBD | DCD | p | |
|---|---|---|---|
|
| |||
| n | 1604 | 292 | |
| PRA >10% (%) | 315 (21.6) | 38 (15.3) | 0.031 |
| Peak PRA >10% (%) | 402 (27.6) | 52 (21.0) | 0.035 |
| Acute rejection prior to discharge (%) | 336 (20.9) | 68 (23.3) | 0.412 |
| Treated for acute rejection prior to discharge (%) | 178 (11.1) | 38 (13.0) | 0.396 |
| Treated for rejection within 1 year (%) | 190 (11.8) | 38 (13.0) | 0.641 |
| Hospitalized for rejection (%) | 77 (13.8) | 18 (23.4) | 0.043 |
| Follow-up time (months) for hospitalization for rejection (mean (SD)) | 14.82 (5.64) | 14.98 (6.04) | 0.815 |
| Hospitalized for infection (%) | 221 (39.8) | 33 (42.9) | 0.700 |
| Follow-up time (months) for hospitalization for infection (mean (SD)) | 14.80 (5.63) | 14.98 (6.04) | 0.794 |
| Coronary vasculopathy (%) | 104 (9.6) | 11 (6.3) | 0.209 |
| Follow-up time (months) for coronary vasculopathy (mean (SD)) | 15.54 (5.74) | 14.15 (5.38) | 0.003 |
Discussion
Donation after circulatory death (DCD) heart transplantation has emerged as a promising option to expand the donor pool, estimating an increase heart transplant rates by 17–48% and portending a future in which DCD hearts will serve as a substantial portion of new transplants [2, 9]. While multiple studies have shown comparable early survival, the effects on long-term survival and important risk factors such as rejection remain unknown.
Given the recency of DCD heart transplantation, prior literature comparing rejection rates is sparse. The Papworth experience described by Messer et al. included 79 DCD recipients and reported no difference in treatment for acute rejection within 1 year (DCD: 25% vs DBD: 23%, p=0.85) [2]; however, additional metrics were not investigated. In the 23 DCD recipients of the Sydney report, there was no difference in antibody-mediated rejection or acute cellular rejection within the first year, defined by endomyocardial biopsy, but does not compare clinical measures such as treatment or hospitalization for rejection [4]. Case series from the U.S. including NYU Langone [10], Vanderbilt [11], and Mayo [12] show promising survival but do not report rejection outcomes. Our institution reported comparable survival and early rejection on endomyocardial biopsy at 4 weeks (DCD: 31.9% vs DBD: 25.9%, p=0.46) but do not report rejection beyond 4 weeks[5].
Our study examined acute rejection after DCD heart transplantation in the U.S., including 292 DCD transplants and their DBD contemporaries. We observed increased risk of acute rejection, noted on multiple metrics - acute rejection prior to discharge, treatment for rejection within 1 year, and hospitalization for rejection at a mean follow-up of 15 months. This was examined in the total cohort, propensity-matched analysis, and subgroups investigating waitlist status and center-specific data. In each of these tests, all three measures of acute rejection trended towards higher rates in DCD recipients, with hospitalization for rejection reaching statistical significance on all analyses (and acute rejection prior to discharge and treatment for rejection within 1 year intermittently significant). On multivariable regression, DCD donation was an independent risk factor for acute rejection prior to discharge and hospitalization for rejection. We continued to observe no difference in overall survival or 1 year mortality, similar to prior studies, and no difference in hospitalization rates for infection.
The mechanism by which DCD transplants may increase acute rejection risk is unclear. Key differences exist in recipient selection, procurement techniques, and early post-operative courses, any of which could affect the inflammatory response. In our study, DCD recipients were healthier and less likely to be hospitalized or require mechanical support pre-transplant compared to DBD recipients. Critical illness is understood to cause immune system dysregulation, with chronic critical illness leading to a state of immunosuppression [13–15]. With over half of DBD recipients hospitalized at the time of transplant (compared to 26.7% of DCD patients), it is possible that more of these hospitalized DBD recipients have immune dysregulation, whereas outpatient DCD recipients are more likely to mount a stronger immune response, leading to higher rates of acute rejection. In subgroup analysis, we observed that in status 1–2 (higher acuity patients and more likely to be hospitalized), there was no difference in acute rejection, which may be explained by this hypothesis. Differences were observed only in status 3–4 and status 5–6 groups, the latter of which did not meet statistical significance likely due to low power. However, there are likely other components that affect rejection risk, as in our multivariable regression analyses, we include recipient comorbidities, functional status, induction therapy, and waitlist status in our multivariable analyses and propensity-match, and continued to find DCD donation an independent risk factor for rejection.
Differences in procurement techniques lead to downstream consequences that may affect rejection risk – including warm ischemic time and exposure of the heart to machine perfusion. Total ischemic time is known upregulate inflammatory cytokines and affect acute and chronic rejection across all solid organ transplants [16–18]. The advent of normothermic perfusion devices reduces cold ischemic time during transport and as a result, focuses the discussion on warm ischemia preceding procurement. In lung transplantation, longer warm ischemia times have been shown to increase acute rejection in mouse studies [19] but data in heart transplant and larger animals is lacking. Unfortunately, warm ischemia times for DCD heart transplants are not available in the UNOS database, limiting our assessment of this important variable, and will be important to assess in more granular detail in institutional studies. The procurement technique itself is another significant factor to consider- the majority of DCD heart transplants utilize the OCS machine perfusion device for transport compared to static cold preservation. This difference in reperfusion may cause histopathologic changes that may be interpreted as rejection on endomyocardial biopsy, which may explain the early rates of acute rejection prior to discharge; however, our other metrics such as treatment for rejection within 1 year, and hospitalization for rejection after discharge, continue to show higher rates in DCD patients. The early prospective, multicenter, randomized trial (PROCEED II) demonstrated no difference in high grade rejection within 30 days using the Organ Care System compared to standard cold preservation [20], but this study does not report rejection episodes beyond the perioperative period or readmissions for rejection. The trial compares DBD transplants only, and does not factor in the period of warm ischemia prior to transport on OCS.
Finally, the perioperative course of DCD transplants appears to differ as well. In each of the larger institutional studies (Messer et al. from the U.K. and Chew et al. from Australia), DCD recipients were noted to have high rates of initial graft dysfunction, with 31–35% of patients requiring mechanical support (ECMO or IABP) in the perioperative period [1, 4, 5]. D’Alessandro et al. demonstrated a similar trend of higher MCS rates within the DCD cohort (DCD: 10.6% vs DBD: 5.4%, p=0.20) [5]. The use of ECMO itself is associated with an inflammatory response caused by exposure of to the nonepithelialized ECMO circuit, leading to activation of the innate immune system[21]. The transient dysfunction may also alter postoperative management of immunosuppression. Prior studies have suggested the failing myocardium to be a source of proinflammatory cytokines [22, 23], and while the early graft dysfunction appears to be transient, this initial episode of myocardial edema, dysfunction, and mechanical support may contribute to a state of heightened inflammation with downstream consequences on rejection. In these studies, there remains little information on endothelial cell preservation, which is an important factor to consider in the inflammatory response, release of cytokines, and long-term vasculopathy. Assessment of this requires more granular detail through controlled prospective studies.
Limitations
As a retrospective analysis, our study is limited by the potential for selection bias, particularly in a small DCD cohort. For patients lost to follow-up who were excluded from the study, we do not have information on reasons for loss to follow-up, which may also influence selection bias. However, the percentage of DCD and DBD recipients excluded for this reason are similar (3.9% vs 4.2% excluded). There are also significant differences in recipient baseline characteristics that may affect rejection risk. We aimed to address this by our use of propensity-matching, multivariable analyses, and subgroup analysis, but additional confounding may persist. Using the UNOS database, we are restricted to variables that are reported to the network, which unfortunately does not include important factors such as warm ischemic time, post-operative mechanical support, and details of rejection treatment. We emphasize the need for further prospective studies to assess details of rejection (cellular vs humoral, and grade) and how these rejection episodes were treated. As a national database study, we face a level of heterogeneity in key variables such as the definition of acute rejection prior to discharge (with different methods of diagnosis and lack of granularity in grades or type of rejection), hospitalization for infection (without details of types of infection and relation to immunosuppression), and DCD procurement techniques (normothermic regional perfusion vs direct recovery). With these limitations, we report a trend of increased acute rejection in DCD transplants but cannot address causality or mechanism, which require further investigation. Our study does not proport to answer all of these questions but to underlie the need and provide evidence for a future controlled prospective study.
Conclusion
Donation after circulatory death (DCD) transplantation is already a clinical reality and will continue to develop as a strategy to address donor availability. As this technique becomes more common, it is important to investigate beyond the promising early survival rates and understand the effects on predictors of long-term survival. We report increased rates of acute rejection after DCD transplantation, that persist after accounting for differences in baseline recipient, donor, and transplant characteristics. Future prospective investigations at both the institutional and national level are needed to corroborate and understand the cause of this increased rejection risk, which may highlight opportunities to optimize DCD protocols and improve long-term outcomes.
Supplementary Material
Funding statement:
We have no external sources of funding to report for this study.
Glossary of Abbreviations
- CDC
Center for Disease Control
- CI
confidence interval
- DCD
donation after circulatory death
- DBD
donation after brain death
- ECMO
extracorporeal membrane oxygenation
- HR
hazard ratio
- LVAD
left ventricular assist device
- PVR
pulmonary vascular resistance
- SMD
standardized mean difference
- UNOS
United Network for Organ Sharing
Footnotes
Financial Disclosure Statement
Disclosure statement: There are no disclosures.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Dhital K, et al. , DCD donations and outcomes of heart transplantation: the Australian experience. Indian J Thorac Cardiovasc Surg, 2020. 36(Suppl 2): p. 224–232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Messer S, et al. , A 5-year single-center early experience of heart transplantation from donation after circulatory-determined death donors. J Heart Lung Transplant, 2020. 39(12): p. 1463–1475. [DOI] [PubMed] [Google Scholar]
- 3.Tchana-Sato V, et al. , Successful clinical transplantation of hearts donated after circulatory death using normothermic regional perfusion. J Heart Lung Transplant, 2019. 38(6): p. 593–598. [DOI] [PubMed] [Google Scholar]
- 4.Chew HC, Macdonald PS, and Dhital KK, The donor heart and organ perfusion technology. J Thorac Dis, 2019. 11(Suppl 6): p. S938–S945. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.D’Alessandro DA, et al. , Hemodynamic and Clinical Performance of Hearts Donated After Circulatory Death. J Am Coll Cardiol, 2022. 80(14): p. 1314–1326. [DOI] [PubMed] [Google Scholar]
- 6.Alyaydin E, et al. , Predisposing factors for late mortality in heart transplant patients. Cardiol J, 2021. 28(5): p. 746–757. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.R Core Team. R: A Language and environment for statistical computing. 2020. [cited 2020; Available from: https://www.R-project.org. [Google Scholar]
- 8.Deb S, et al. , A Review of Propensity-Score Methods and Their Use in Cardiovascular Research. Can J Cardiol, 2016. 32(2): p. 259–65. [DOI] [PubMed] [Google Scholar]
- 9.Scheuer SE, Jansz PC, and Macdonald PS, Heart transplantation following donation after circulatory death: Expanding the donor pool. J Heart Lung Transplant, 2021. 40(9): p. 882–889. [DOI] [PubMed] [Google Scholar]
- 10.Ngai J, Masuno K, and Moazami N Early US experience with cardiac donation after circulatory death (DCD) using normothermic regional perfusion, Anesthetic Considerations During Heart Transplantation Using Donation After Circulatory Death. J Cardiothorac Vasc Anesth, 2020. 34(11): p. 3073–3077. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hoffman JRH, et al. ,. J Heart Lung Transplant, 2021. 40(11): p. 1408–1418. [DOI] [PubMed] [Google Scholar]
- 12.Alomari M, et al. , Is the Organ Care System (OCS) Still the First Choice With Emerging New Strategies for Donation After Circulatory Death (DCD) in Heart Transplant? Cureus, 2022. 14(6): p. e26281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Hawkins RB, et al. , Chronic Critical Illness and the Persistent Inflammation, Immunosuppression, and Catabolism Syndrome. Front Immunol, 2018. 9: p. 1511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rabaan AA, et al. , Role of Inflammatory Cytokines in COVID-19 Patients: A Review on Molecular Mechanisms, Immune Functions, Immunopathology and Immunomodulatory Drugs to Counter Cytokine Storm . Vaccines (Basel), 2021. 9(5). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mira JC, et al. , Persistent Inflammation, Immunosuppression and Catabolism Syndrome. Crit Care Clin, 2017. 33(2): p. 245–258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Cai S, Ichimaru N, and Takahara S, How Do Dendritic Cells Play the Role in Ischemia/Reperfusion Triggered Kidney Allograft Rejection. Curr Gene Ther, 2017. 17(6): p. 400–404. [DOI] [PubMed] [Google Scholar]
- 17.Zhai Y, et al. , Ischaemia-reperfusion injury in liver transplantation--from bench to bedside. Nat Rev Gastroenterol Hepatol, 2013. 10(2): p. 79–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Jernryd V, et al. , The influence of ischemia and reperfusion time on outcome in heart transplantation. Clin Transplant, 2020. 34(5): p. e13840. [DOI] [PubMed] [Google Scholar]
- 19.Hirano Y, et al. , Prolonged warm ischemia exacerbated acute rejection after lung transplantation from donation after cardiac death in a mouse. Gen Thorac Cardiovasc Surg, 2020. 68(1): p. 57–62. [DOI] [PubMed] [Google Scholar]
- 20.Ardehali A, et al. , Ex-vivo perfusion of donor hearts for human heart transplantation (PROCEED II): a prospective, open-label, multicentre, randomised non-inferiority trial. Lancet, 2015. 385(9987): p. 2577–84. [DOI] [PubMed] [Google Scholar]
- 21.Millar JE, et al. , The inflammatory response to extracorporeal membrane oxygenation (ECMO): a review of the pathophysiology. Crit Care, 2016. 20(1): p. 387. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Torre-Amione G, et al. , Tumor necrosis factor-alpha and tumor necrosis factor receptors in the failing human heart. Circulation, 1996. 93(4): p. 704–11. [DOI] [PubMed] [Google Scholar]
- 23.Habib FM, et al. , Tumour necrosis factor and inducible nitric oxide synthase in dilated cardiomyopathy. Lancet, 1996. 347(9009): p. 1151–5. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.