The United Network for Organ Sharing (UNOS) adopted new criteria for the heart allocation score on October 18, 2018 to reflect the changing trends of candidates’ mortality while awaiting transplant. We examined the impact of these policy changes on rates of left ventricular assist device (LVAD) implantation and outcomes after transplant from a relatively newer UNOS database. The UNOS registry was used to identify first-time adult heart recipients with LVAD at listing or transplant who underwent transplantation between January 1, 2016 and March 10, 2020. Survival data were collected through March 30, 2023. Those listed before October 18, 2018 but transplanted after were excluded. Patients were divided into before or after change groups. Demographics and clinical parameters were compared. Survival was analyzed with Kaplan–Meier curves and log-rank tests. A p <0.05 was considered significant. We identified 4,387 heart recipients with LVAD in the before (n = 3,606) and after (n = 781) score change eras. The after group had a lower rate of LVAD implantation while listed than the before group (20.4% vs 34.9%, p <0.0001), and were more likely to be female (25.1% vs 20.2%, p = 0.002); in both groups, most recipients (62.8%) were white. There was significantly farther distance from the donor hospital to transplant center in the after group (264.4 NM vs 144.2 NM, p <0.0001) and decreased waitlist days (84.9 ± 105.1 vs 369.2 ± 459.5, p <0.0001). Recipients in the after group were more likely to use extracorporeal membrane oxygenation (3.7% vs 0.5%, p <0.0001) and intravenous inotropes (19.1% vs 7.5%, p <0.0001) and receive a Centers for Disease Control and Prevention increased risk donor organ (37.9% vs 30.5%, p <0.0001). Survival at 3 years was comparable between the 2 groups.
The allocation score change in 2018 yielded considerable changes in mechanical circulatory support device implantation strategy and outcomes. The rate of LVAD implantation decreased with increased utilization of temporary mechanical circulatory support devices.
Heart transplant allocation scoring criteria were implemented to better stratify candidates based on waitlist mortality and improve access to transplantation in the most medically urgent cases. As mechanical circulatory support devices (MCSDs) continue to improve in terms of durability, reduced adverse events, and overall effect on survival, there was a need to rethink the prioritization of donor hearts. The United Network for Organ Sharing (UNOS) adopted new criteria for the heart allocation score in October 2018, in part, to reflect changing mortality trends of heart transplant candidates implanted with MCSDs. The scoring system changed in several ways but notably by further stratifying the existing 3 categories into 6 and deprioritizing stable candidates with left ventricular assist devices (LVADs). , We examined the impact of these policy changes on rates of LVAD implantation and post-transplant outcomes.
Before this score change, candidates living with durable LVADs received a discretional 30-day period during which they were assigned equal priority to critical patients on venoarterial extracorporeal membrane oxygenation (ECMO). LVAD technology, especially with the advent of the magnetically levitated centrifugal-flow HeartMate III (HMIII), has improved so successfully that a majority of patients are alive after 5 years, with significantly reduced complication rates. Several studies have described various outcomes since the allocation score change including decreased deterioration of candidates on the waitlist yet significantly decreased utilization of LVADs and post-transplant survival. , However, survival data in particular have been limited thus far given the relative recency of the UNOS policy change. At present, to the best of our knowledge, this is the first study to consider 4-year survival data available for patients who received the most advanced and life-prolonging LVAD technology and heart transplant recipients after the score change. It is valuable to investigate the impact of policy change on the utilization and outcomes of these MCSDs and expand on previous studies showing some alarming trends.
Our primary objective in this study was to compare the survival outcomes and implantation strategy between patients with LVAD who received a heart transplant before and after the change in allocation policy. We also compared patient demographics, transplant parameters, and LVAD implantation between the groups.
Methods
This is a retrospective review of the UNOS registry to evaluate the LVAD utilization and survival outcomes in patients with LVAD who received a heart transplant before and after the allocation score change. Data were obtained through the UNOS registry and entered into an online Health Insurance Portability and Accountability Act–compliant database and deidentified for analysis. The institutional review board at Temple University waived approval for this study because the UNOS data set contains deidentified information.
A total of 4,387 patients were identified through the UNOS registry who had an LVAD at the time of listing and/or at the time of transplant. Those transplanted between January 1, 2016 and March 10, 2020 were included. Recipients then were stratified into 2 groups based on transplantation before or after the allocation score change on October 18, 2018. Those transplanted from January 1, 2016 to October 17, 2018 were in the “before” group; those transplanted from October 18, 2018 to March 10, 2020 were in the “after” group. Survival and follow-up information extended through March 30, 2023.
Eligibility criteria included patients with LVAD aged ≥18 years who received a heart transplant. Those listed before the score change date but transplanted after were excluded. Other exclusion criteria were multiorgan transplant, retransplant recipients, and those missing follow-up or survival information.
We collected recipient and donor demographic variables including age, gender, ethnicity, body mass index (BMI), height, hepatitis C virus serostatus, cause of death, and Centers for Disease Control and Prevention risk status. Clinical variables included days on the waitlist, distance from donor hospital to transplant center, length of status, graft status, cause of graft failure, renal and hepatic function at time of transplant, ECMO at listing and transplant, LVAD implantation at and after listing, inotrope utilization, ventilatory support, allocation type, and postoperative complications. Baseline recipient characteristics, donor characteristics, and clinical parameters were collected from UNOS.
Continuous variables were compared using 2-sample Student’s t tests and were reported as means and SDs; categorical variables were compared using chi-square tests and were reported as counts and percentages.
A multivariate Cox proportional hazards model was used to identify variables that were significant predictors of mortality. Covariates were selected based on clinical relevance and included the following: era of transplantation (before vs after the score change), recipient age, gender, BMI, ethnicity, donor age, ischemic time, ECMO status, creatinine, and bilirubin.
Given differences in baseline characteristics between groups, a propensity score–matched analysis was performed. We matched those transplanted before and after the score change (1:1) based on the following variables: recipient age, gender, BMI, ethnicity, donor age, ischemic time, ECMO status, creatinine, and bilirubin.
In the matched and unmatched groups, the primary outcome of interest was survival. Survival time was calculated from the date of transplant to the last date of follow-up and was assessed up to 3 years after transplant. Survival was analyzed with Kaplan–Meier curves and log-rank tests. The secondary outcomes included length of stay, graft status, and post-transplant complications. A value of p <0.05 was considered significant. Statistical analyses were conducted with SAS 9.4 (SAS Institute Inc., Cary, North Carolina).
Results
Of 4,387 patients with LVAD included in the study, 3,606 patients (82.2%) were transplanted before the allocation score change compared with 781 patients (17.8%) after the score change. Table 1 lists the demographics and baseline characteristics for heart transplant recipients and donors, stratified by transplantation before versus after the score change. In heart transplant recipients, those transplanted before the score change were more likely to be men (79.8% vs 74.9%, p = 0.002) and taller in height (175.1 vs 174.0 cm, p = 0.004) than those transplanted after the score change.
| Variable | Before (n=3606) | After (n=781) | P-value |
|---|---|---|---|
| Recipient age | 54.0 (12.0) | 53.0 (12.2) | 0.05 |
| Recipient sex | 0.002 | ||
| – Male | 2878 (79.8%) | 585 (74.9%) | |
| – Female | 728 (20.2%) | 196 (25.1%) | |
| Recipient ethnicity | 0.85 | ||
| – White | 2263 (62.8%) | 491 (62.8%) | |
| – Black | 892 (24.7%) | 192 (24.6%) | |
| – Asian | 108 (3.0%) | 24 (3.1%) | |
| – Hispanic | 304 (8.4%) | 69 (8.8%) | |
| – Other/unknown | 39 (1.1%) | 5 (0.6%) | |
| Recipient BMI (kg/m 2 ) | 28.9 (4.8) | 28.7 (5.1) | 0.30 |
| Recipient height (cm) | 175.1 (9.6) | 174.0 (9.7) | 0.004 |
| Recipient HCV serostatus | 0.39 | ||
| – Positive | 84 (3.2%) | 20 (3.2%) | |
| – Negative | 2521 (95.1%) | 595 (95.8%) | |
| – Not done | 46 (1.7%) | 6 (1.0%) | |
| Donor age | 31.7 (10.5) | 32.8 (10.7) | 0.01 |
| Donor sex | 0.09 | ||
| – Male | 2728 (75.7%) | 568 (72.7%) | |
| – Female | 878 (24.3%) | 213 (27.3%) | |
| Donor ethnicity | 0.17 | ||
| – White | 2403 (66.6%) | 519 (66.5%) | |
| – Black | 584 (16.2%) | 108 (13.8%) | |
| – Asian | 51 (1.4%) | 17 (2.2%) | |
| – Hispanic | 527 (14.6%) | 125 (16.0%) | |
| – Other/unknown | 41 (1.1%) | 12 (1.5%) | |
| Donor BMI (kg/m 2 ) | 28.0 (6.0) | 28.5 (6.4) | 0.06 |
| Donor height (cm) | 175.3 (9.2) | 174.2 (9.0) | 0.003 |
| Donor cause of death | <0.0001 | ||
| – Anoxia | 1305 (36.2%) | 365 (46.7%) | |
| – CNS tumor | 17 (0.5%) | 1 (0.1%) | |
| – Cerebrovascular/stroke | 507 (14.1%) | 107 (13.7%) | |
| – Head trauma | 1710 (47.4%) | 281 (36%) | |
| – Other/unknown | 67 (1.9%) | 27 (3.5%) | |
| Donor CDC risk status | <0.0001 | ||
| – Increased risk | 1098 (30.5%) | 296 (37.9%) | |
| – Standard risk | 2506 (69.5%) | 485 (62.1%) |
In heart transplant donors, those before the score change were more likely to be younger (31.7 vs 32.8 years, p = 0.01), taller in height (175.3 vs 174.2 cm, p = 0.003), and standard risk (as opposed to increased risk) Centers for Disease Control and Prevention status (69.5% vs 62.1%, p ≤0.0001) than those transplanted after the score change. The donor cause of death was significantly different between the 2 groups (p ≤0.0001). Donors before the score change were less likely to have anoxia (36.2% vs 46.7%) and more likely to have head trauma (47.4% vs 35.9%) as a cause of death than donors after the change.
Table 2 lists the clinical parameters for heart transplant recipients before and after the score change. Recipients before the score change spent more days on the waitlist (369.2 vs 84.9 days, p <0.0001), received organs from shorter distances to the transplant center (144.2 vs 264.4 NM, p <0.0001), and had decreased ischemic time (3.0 vs 3.5 hours, p <0.0001) compared with recipients after the score change. Allocation type was different between the 2 groups (p <0.0001), with recipients before the score change more likely to have a local allocation than recipients after (68.1% vs 34.9%). Recipients after the score change were more likely to require intravenous inotropes (19.1% vs 7.5%, p <0.0001) and ventilatory support (1.8% vs 0.5%, p <0.0001) at the time of transplant than recipients before. There was no significant difference in the serum creatinine and total bilirubin in recipients in both eras.
| Variable | Before (n=3606) | After (n=782) | P-value |
|---|---|---|---|
| Days on the waitlist | 369.2 (459.5) | 84.9 (105.1) | <0.0001 |
| Distance from donor hospital to transplant center (nautical miles) | 144.2 (182.6) | 264.4 (249.7) | <0.0001 |
| Length of stay (days) | 22.7 (22.9) | 24.4 (25.1) | 0.08 |
| Graft status at follow-up | 0.024 | ||
| – Functioning | 3357 (93.8%) | 746 (95.9%) | |
| – Failed | 222 (6.2%) | 32 (4.1%) | |
| ECMO at registration | 34 (0.9%) | 25 (3.2%) | <0.0001 |
| ECMO at transplant | 17 (0.5%) | 29 (3.7%) | <0.0001 |
| Ischemic time (hours) | 3.0 (1.1) | 3.5 (1.1) | <0.0001 |
| LVAD implantation during listing | 1260 (34.9%) | 159 (20.4%) | <0.0001 |
| Allocation type | <0.0001 | ||
| – Local | 2454 (68.1%) | 273 (35%) | |
| – Regional | 710 (19.7%) | 328 (42.0%) | |
| – National | 437 (12.1%) | 179 (22.9%) | |
| – Foreign | 5 (0.1%) | 1 (0.1%) | |
| Dialysis prior to discharge? | 0.06 | ||
| – Yes | 476 (13.2%) | 123 (15.7%) | |
| – No | 3127 (86.8%) | 658 (84.3%) | |
| Permanent pacemaker implanted prior to discharge? | 0.70 | ||
| – Yes | 84 (2.3%) | 20 (2.6%) | |
| – No | 3515 (97.7%) | 760 (97.4%) | |
| Stroke prior to discharge? | <0.0001 | ||
| – Yes | 128 (3.6%) | 56 (7.2%) | |
| – No | 3469 (96.4%) | 722 (92.8%) | |
| LVAD brand at time of transplant | <0.0001 | ||
| – HeartMate II | 1320 (36.6%) | 121 (15.5%) | |
| – HeartMate III | 11 (0.3%) | 203 (26.0%) | |
| – HeartWare HVAD | 674 (18.7%) | 194 (24.8%) | |
| – Other/unknown | 478 (13.3%) | 170 (21.7) | |
| Recipient creatinine | 1.2 (0.4) | 1.2 (0.5) | 0.13 |
| Recipient total bilirubin | 0.8 (1.3) | 1.0 (1.5) | 0.05 |
| Ventilatory support at transplant | 17 (0.5%) | 14 (1.8%) | <0.0001 |
| IV inotropes at transplant | 272 (7.5%) | 149 (19.1%) | <0.0001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
