New priorities: Analysis of the New Kidney Allocation System on UCLA patients transplanted from the deceased donor waitlist

NephroCAGE-German-Canadian Consortium on AI for Improved Kidney Transplantation Outcome: Protocol for an Algorithm Development and Validation Study

BACKGROUND

Recent advances in hardware and software enabled the use of artificial intelligence (AI) algorithms for analysis of complex data in a wide range of daily-life use cases. We aim to explore the benefits of applying AI to a specific use case in transplant nephrology: risk prediction for severe posttransplant events. For the first time, we combine multinational real-world transplant data, which require specific legal and technical protection measures.

OBJECTIVE

The German-Canadian NephroCAGE consortium aims to develop and evaluate specific processes, software tools, and methods to (1) combine transplant data of more than 8000 cases over the past decades from leading transplant centers in Germany and Canada, (2) implement specific measures to protect sensitive transplant data, and (3) use multinational data as a foundation for developing high-quality prognostic AI models.

METHODS

To protect sensitive transplant data addressing the first and second objectives, we aim to implement a decentralized NephroCAGE federated learning infrastructure upon a private blockchain. Our NephroCAGE federated learning infrastructure enables a switch of paradigms: instead of pooling sensitive data into a central database for analysis, it enables the transfer of clinical prediction models (CPMs) to clinical sites for local data analyses. Thus, sensitive transplant data reside protected in their original sites while the comparable small algorithms are exchanged instead. For our third objective, we will compare the performance of selected AI algorithms, for example, random forest and extreme gradient boosting, as foundation for CPMs to predict severe short- and long-term posttransplant risks, for example, graft failure or mortality. The CPMs will be trained on donor and recipient data from retrospective cohorts of kidney transplant patients.

RESULTS

We have received initial funding for NephroCAGE in February 2021. All clinical partners have applied for and received ethics approval as of 2022. The process of exploration of clinical transplant database for variable extraction has started at all the centers in 2022. In total, 8120 patient records have been retrieved as of August 2023. The development and validation of CPMs is ongoing as of 2023.

CONCLUSIONS

For the first time, we will (1) combine kidney transplant data from nephrology centers in Germany and Canada, (2) implement federated learning as a foundation to use such real-world transplant data as a basis for the training of CPMs in a privacy-preserving way, and (3) develop a learning software system to investigate population specifics, for example, to understand population heterogeneity, treatment specificities, and individual impact on selected posttransplant outcomes.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

DERR1-10.2196/48892.

Race, Age, and Kidney Transplant Waitlisting Among Patients Receiving Incident Dialysis in the United States

Rationale & Objective

Patients with kidney failure from racial and ethnic minority groups and older patients have reduced access to the transplant waitlist relative to White and younger patients. Although racial disparities in the waitlisting group have declined after the 2014 kidney allocation system change, whether there is intersectionality of race and age in waitlisting access is unknown.

Study Design

Retrospective cohort study.

Setting & Participants

439,455 non-Hispanic White and non-Hispanic Black US adults initiating dialysis between 2015 and 2019 were identified from the United States Renal Data System, and followed through 2020.

Exposures

Patient race and ethnicity (non-Hispanic White and non-Hispanic Black) and age group (18-29, 30-49, 50-64, and 65-80 years).

Outcomes

Placement on the United Network for Organ Sharing deceased donor waitlist.

Analytical Approach

Age- and race-stratified waitlisting rates were compared. Multivariable Cox proportional hazards models, censored for death, examined the association between race and waitlisting, and included interaction term for race and age.

Results

Over a median follow-up period of 1 year, the proportion of non-Hispanic White and non-Hispanic Black patients waitlisted was 20.7% and 20.5%, respectively. In multivariable models, non-Hispanic Black patients were 14% less likely to be waitlisted (aHR, 0.86, 95% CI, 0.77-0.95). Relative differences between non-Hispanic Black and non-Hispanic White patients were different by age group. Non-Hispanic Black patients were 27%, 12%, and 20% less likely to be waitlisted than non-Hispanic White patients for ages 18-29 years (aHR, 0.73; 95% CI, 0.61-0.86), 50-64 (aHR, 0.88; 95% CI, 0.80-0.98), and 65-80 years (aHR, 0.80; 95% CI, 0.71-0.90), respectively, but differences were attenuated among patients aged 30-49 years (aHR, 0.89; 95% CI, 0.77-1.02).

Limitations

Race and ethnicity data is physician reported, residual confounding, and analysis is limited to non-Hispanic White and non-Hispanic Black patients.

Conclusions

Racial disparities in waitlisting exist between non-Hispanic Black and non-Hispanic White individuals and are most pronounced among younger patients with kidney failure. Results suggest that interventions to address inequalities in waitlisting may need to be targeted to younger patients with kidney failure.

Plain-Language Summary

Research has shown that patients from racial and ethnic minority groups and older patients have reduced access to transplant waitlisting relative to White and younger patients; nevertheless, how age impacts racial disparities in waitlisting is unknown. We compared waitlisting between non-Hispanic Black and non-Hispanic White patients with incident kidney failure, within age strata, using registry data for 439,455 US adults starting dialysis (18-80 years) during 2015-2019. Overall, non-Hispanic Black patients were less likely to be waitlisted and relative differences between the two racial groups differed by age. After adjusting for patient-level factors, the largest disparity in waitlisting was observed among adults aged 18-29 years. These results suggest that interventions should target younger adults to reduce disparities in access to kidney transplant waitlisting.

Evaluating the outcomes of donor-recipient age differences in young adults undergoing liver transplantation

The current liver allocation system may be disadvantaging younger adult recipients as it does not incorporate the donor-recipient age difference. Given the longer life expectancy of younger recipients, the influences of older donor grafts on their long-term prognosis should be elucidated. This study sought to reveal the long-term prognostic influence of the donor-recipient age difference in young adult recipients. Adult patients who received initial liver transplants from deceased donors between 2002 and 2021 were identified from the UNOS database. Young recipients (patients 45 years old or below) were categorized into 4 groups: donor age younger than the recipient, 0-9 years older, 10-19 years older, or 20 years older or above. Older recipients were defined as patients 65 years old or above. To examine the influence of the age difference in long-term survivors, conditional graft survival analysis was conducted on both younger and older recipients. Among 91,952 transplant recipients, 15,170 patients were 45 years old or below (16.5%); these were categorized into 6,114 (40.3%), 3,315 (21.9%), 2,970 (19.6%), and 2,771 (18.3%) for groups 1-4, respectively. Group 1 demonstrated the highest probability of survival, followed by groups 2, 3, and 4 for the actual graft survival and conditional graft survival analyses. In younger recipients who survived at least 5 years post-transplant, inferior long-term survival was observed when there was an age difference of 10 years or above (86.9% vs. 80.6%, log-rank p <0.01), whereas there was no difference in older recipients (72.6% vs. 74.2%, log-rank p =0.89). In younger patients who are not in emergent need of a transplant, preferential allocation of younger aged donor offers would optimize organ utility by increasing postoperative graft survival time.

Continuation of immunosuppression vs. immunosuppression weaning in potential repeat kidney transplant candidates: a care management perspective

Management of immunosuppression in patients with a failing or failed kidney transplant requires a complete assessment of their clinical condition. One of the major considerations in determining immunosuppression is whether or not such an individual is considered a candidate for re-transplantation. Withdrawal of immunosuppression in a re-transplant candidate can result in allosensitization and markedly reduce the chances of a repeat transplant. In this review, we summarize the effects of immunosuppression reduction on HLA sensitization, discuss the impacts of allosensitization in these patients, and explore reduction protocols and future directions. Risks of chronic immunosuppression, medical management of the failing allograft, and the effect of nephrectomy are covered elsewhere in this issue.

European Guideline for the Management of Kidney Transplant Patients With HLA Antibodies: By the European Society for Organ Transplantation Working Group

This guideline, from a European Society of Organ Transplantation (ESOT) working group, concerns the management of kidney transplant patients with HLA antibodies. Sensitization should be defined using a virtual parameter such as calculated Reaction Frequency (cRF), which assesses HLA antibodies derived from the actual organ donor population. Highly sensitized patients should be prioritized in kidney allocation schemes and linking allocation schemes may increase opportunities. The use of the ENGAGE 5 ((Bestard et al., Transpl Int, 2021, 34: 1005–1018) system and online calculators for assessing risk is recommended. The Eurotransplant Acceptable Mismatch program should be extended. If strategies for finding a compatible kidney are very unlikely to yield a transplant, desensitization may be considered and should be performed with plasma exchange or immunoadsorption, supplemented with IViG and/or anti-CD20 antibody. Newer therapies, such as imlifidase, may offer alternatives. Few studies compare HLA incompatible transplantation with remaining on the waiting list, and comparisons of morbidity or quality of life do not exist. Kidney paired exchange programs (KEP) should be more widely used and should include unspecified and deceased donors, as well as compatible living donor pairs. The use of a KEP is preferred to desensitization, but highly sensitized patients should not be left on a KEP list indefinitely if the option of a direct incompatible transplant exists.

Posttransplant Outcomes for cPRA-100% Recipients Under the New Kidney Allocation System

BACKGROUND

There is concern in the transplant community that outcomes for the most highly sensitized recipients might be poor under Kidney Allocation System (KAS) high prioritization.

METHODS

To study this, we compared posttransplant outcomes of 525 pre-KAS (December 4, 2009, to December 3, 2014) calculated panel-reactive antibodies (cPRA)-100% recipients to 3026 post-KAS (December 4, 2014, to December 3, 2017) cPRA-100% recipients using SRTR data. We compared mortality and death-censored graft survival using Cox regression, acute rejection, and delayed graft function (DGF) using logistic regression, and length of stay (LOS) using negative binomial regression.

RESULTS

Compared with pre-KAS recipients, post-KAS recipients were allocated kidneys with lower Kidney Donor Profile Index (median 30% versus 35%, P < 0.001) but longer cold ischemic time (CIT) (median 21.0 h versus 18.6 h, P < 0.001). Compared with pre-KAS cPRA-100% recipients, those post-KAS had higher 3-year patient survival (93.6% versus 91.4%, P = 0.04) and 3-year death-censored graft survival (93.7% versus 90.6%, P = 0.005). The incidence of DGF (29.3% versus 29.2%, P = 0.9), acute rejection (11.2% versus 11.7%, P = 0.8), and median LOS (5 d versus 5d, P = 0.2) were similar between pre-KAS and post-KAS recipients. After accounting for secular trends and adjusting for recipient characteristics, post-KAS recipients had no difference in mortality (adjusted hazard ratio [aHR]: 0.861.623.06, P = 0.1), death-censored graft failure (aHR: 0.521.001.91, P > 0.9), DGF (adjusted odds ratio [aOR]: 0.580.861.27, P = 0.4), acute rejection (aOR: 0.610.941.43, P = 0.8), and LOS (adjusted LOS ratio: 0.981.161.36, P = 0.08).

CONCLUSIONS

We did not find any statistically significant worsening of outcomes for cPRA-100% recipients under KAS, although longer-term monitoring of posttransplant mortality is warranted.

Design of a state of the art reporting system and process improvement for reporting of high complexity single antigen bead data for transplant patients to the electronic medical record

The UCLA Immunogenetics Center is an Immunogenetics and Histocompatibility laboratory that performs testing for multiple transplant programmes within and outside of UCLA. The single antigen bead (SAB) test is a high complexity luminex bead test used to assess pretransplant and post-transplant patients for the presence of pathogenic human leucocyte antigen donor-specific antibody associated with allograft rejection. Efficient reporting of the SAB test has been difficult as data analysis and reports are generated in the laboratory information system (LIS) and uploaded to the electronic medical record (EMR) as PDFs. To solve this, we recently developed a state of the art reporting workflow allowing discrete reporting of SAB data (antibody specificity, mean fluorescent intensity and interpretative comments) from the LIS HistoTrac to UCLA Health System’s EMR EPIC:CareConnect. However, a proportion of tests did not report to the EMR appropriately. Baseline system performance data evaluated over a 10-week period showed that ~4.5/100 tests resulted in EPIC as ‘preliminary result’ or ‘in process’ instead of ‘final result’ with only common cause variation. Quality improvement methods were employed to improve the process with the SMART Aim of reporting 100% of tests as ‘final result’. Pareto analysis identified two errors accounting for 79% of common system-level failures—status errors and interface errors. We hypothesised that addressing the status error would reduce or eliminate the interface errors. We used the Model For Improvement to test a reprogramming intervention. Status and interface errors were completely resolved through the process improvement. Continuous monitoring revealed a system-level shift with only ~1.9/100 tests resulting inappropriately. Through the audit process, the remaining common system-level failures were identified and resolved. Therefore, 100% of tests result to EPIC as ‘final result’. The study demonstrates that high complexity SAB bead data can be efficiently reported EPIC:CareConnect from HistoTrac as discrete data.

Outcomes of cPRA 100% deceased donor kidney transplant recipients under the new Kidney Allocation System: A single-center cohort study

In light of changes in donor/recipient case-mix and increased cold ischemia times under the Kidney Allocation System (KAS), there is some concern that cPRA 100% recipients might be doing poorly under KAS. We used granular, single-center data on 109 cPRA 100% deceased donor kidney transplant (DDKT) recipients to study post-KAS posttransplant outcomes not readily available in national registry data. We found that 3-year patient (96.4%) and death-censored graft survival (96.8%) was excellent. We also found that cPRA 100% recipients had a relatively low incidence of T cell-mediated rejection (9.2%) and antibody-mediated rejection (AMR) (13.8%). T cell-mediated rejection episodes tended to be relatively mild-50% (5 episodes) were grade 1, 50% (5 episodes) were grade 2, and none were grade 3. Only 1 episode was associated with graft loss, but this was in the context of a mixed rejection. Although only 15 recipients (13.8%) developed an AMR episode, 2 of these were associated with a graft loss. Despite the rejection episodes, the vast majority of recipients had excellent graft function 3 years posttransplant (median serum creatinine 1.5 mg/dL). In conclusion, cPRA 100% DDKT recipients are doing well under KAS, although every effort should be made to prevent AMR to ensure long-term outcomes remain excellent.

How do highly sensitized patients get kidney transplants in the United States? Trends over the last decade

Prioritization of highly sensitized (HS) candidates under the kidney allocation system (KAS) and growth of large, multicenter kidney-paired donation (KPD) clearinghouses have broadened the transplant modalities available to HS candidates. To quantify temporal trends in utilization of these modalities, we used SRTR data from 2009 to 2017 to study 39 907 adult HS (cPRA ≥ 80%) waitlisted candidates and 19 003 recipients. We used competing risks regression to quantify temporal trends in likelihood of DDKT, KPD, and non-KPD LDKT for HS candidates (Era 1: January 1, 2009-December 31, 2011; Era 2: January 1, 2012-December 3, 2014; Era 3: December 4, 2014-December 31, 2017). Although the likelihood of DDKT and KPD increased over time for all HS candidates (adjusted subhazard ratio [aSHR] Era 3 vs 1 for DDKT: 1.74 1.851.97 , P < .001 and for KPD: 1.70 2.202.84 , P < .001), the likelihood of non-KPD LDKT decreased (aSHR: 0.69 0.820.97 , P = .02). However, these changes affected HS recipients differently based on cPRA. Among recipients, more cPRA 98%-99.9% and 99.9%+ recipients underwent DDKT (96.2% in Era 3% vs 59.1% in Era 1 for cPRA 99.9%+), whereas fewer underwent non-KPD LDKT (1.9% vs 30.9%) or KPD (2.0% vs 10.0%). Although KAS increased DDKT likelihood for the most HS candidates, it also decreased the use of non-KPD LDKT to transplant cPRA 98%+ candidates.

Prolonged Cold Ischemia Time Offsets the Benefit of Human Leukocyte Antigen Matching in Deceased Donor Kidney Transplant

BACKGROUND

The consequences of prolonging cold ischemia time (CIT) to facilitate HLA matching in kidney transplantation are not known.

METHODS

Patients with a history of kidney transplant in the United States (2000-2016) with 0 HLA mismatch (MM) were categorized based on CIT (< 10; 10 to < 15; 15 to < 20; 20 to < 25; 25 to < 30; and ≥ 30 hours). Time to graft loss was compared for each CIT category to a reference group of individuals with > 0 HLA MM and short CIT (< 10 hours) using a multivariable Cox proportional hazards model.

RESULTS

The adjusted risk of graft failure was significantly lower for 0 HLA MM with the shortest CIT compared to the reference group (hazard ratio, 0.82; 95% confidence interval, 0.72-0.94), and this survival advantage persisted to a threshold of < 20 hours of CIT. No survival advantage was observed for the 0 HLA MM group once CIT was > 20 hours. This trend persisted after excluding highly sensitized recipients (panel reactive antibody > 98%) where shipping of organs occurs to achieve more equitable access to organs rather than optimize HLA match.

CONCLUSIONS

CIT > 20 hours offsets the benefit of 0 HLA MM in kidney transplantation. This may have implications in organ shipping to facilitate immunologic match.

Pediatric deceased donor kidney transplant outcomes under the Kidney Allocation System

The Kidney Allocation System (KAS) has resulted in fewer pediatric kidneys being allocated to pediatric deceased donor kidney transplant (pDDKT) recipients. This had prompted concerns that post-pDDKT outcomes may worsen. To study this, we used SRTR data to compare the outcomes of 953 pre-KAS pDDKT (age <18 years) recipients (December 4, 2012-December 3, 2014) with the outcomes of 934 post-KAS pDDKT recipients (December 4, 2014-December 3, 2016). We analyzed mortality and graft loss by using Cox regression, delayed graft function (DGF) by using logistic regression, and length of stay (LOS) by using negative binomial regression. Post-KAS recipients had longer pretransplant dialysis times (median 1.26 vs 1.07 years, P = .02) and were more often cPRA 100% (2.0% vs 0.1%, P = .001). Post-KAS recipients had less graft loss than pre-KAS recipients (hazard ratio [HR]: _0.35 0.54_0.83 , P = .005) but no statistically significant differences in mortality (HR: _0.29 0.72_1.83 , P = .5), DGF (odds ratio: _0.93 1.32_1.93 , P = .2), and LOS (LOS ratio: _0.96 1.06_1.19 , P = .4). After adjusting for donor-recipient characteristics, there were no statistically significant post-KAS differences in mortality (adjusted HR: _0.37 1.04_2.92 , P = .9), DGF (adjusted odds ratio: _0.94 1.41_2.13 , P = .1), or LOS (adjusted LOS ratio: _0.93 1.04_1.16 , P = .5). However, post-KAS pDDKT recipients still had less graft loss (adjusted HR: _0.38 0.59_0.91 , P = .02). KAS has had a mixed effect on short-term posttransplant outcomes for pDDKT recipients, although our results are limited by only 2 years of posttransplant follow-up.

The national landscape of deceased donor kidney transplantation for the highly sensitized: Transplant rates, waitlist mortality, and posttransplant survival under KAS

Deceased donor kidney transplantation (DDKT) rates for highly sensitized (HS) candidates increased early after implementation of the Kidney Allocation System (KAS) in 2014. However, this may represent a bolus effect, and a granular investigation of the current state of DDKT for HS candidates remains lacking. We studied 270 722 DDKT candidates from the SRTR from 12/4/2011 to 12/3/2014 ("pre-KAS") and 12/4/2014 to 12/3/2017 ("post-KAS"), analyzing DDKT rates for HS candidates using adjusted negative binomial regression. Post-KAS, candidates with the highest levels of sensitization had an increased DDKT rate compared with pre-KAS (cPRA 98% adjusted incidence rate ratio [aIRR]:1.27 1.772.46 P = .001, cPRA 99% aIRR:3.18 4.365.98 P < .001, cPRA 99.5-99.9% aIRR:16.91 24.2934.89 P < .001, and cPRA 99.9%+ aIRR:8.79 11.5815.26 P < .001). To determine whether these changes produced more equitable access to DDKT, we compared DDKT rates of HS to non-HS candidates (cPRA 0-79%). Post-KAS, cPRA, 98% candidates had an equivalent DDKT rate (aIRR:0.65 0.941.36 , P = .8) to non-HS candidates, whereas 99% candidates had a higher DDKT rate (aIRR:1.19 1.682.38 , P = .02). Although cPRA 99.5-99.9% candidates had an increased DDKT rate (aIRR:2.46 3.504.98 , P < .001) compared to non-HS candidates, cPRA 99.9%+ candidates had a significantly lower DDKT rate (aIRR:0.29 0.400.56 , P < .001). KAS has improved access to DDKT for HS candidates, although substantial imbalance exists between cPRA 99.5-99.9% and 99.9%+ candidates.

Transplantation in highly sensitized patients: challenges and recommendations

INTRODUCTION

Highly sensitized patients awaiting a kidney transplant accrue on the transplant waiting list. The breadth of HLA antibodies in this subpopulation of patients precludes receiving a compatible organ offer through regular allocation within an acceptable time-frame. Areas covered: Several alternative options to receive a transplant exist for these patients, including additional priority in regular allocation, special programs based on allocation through acceptable antigens, kidney paired donation programs, desensitization protocols, or a combination of the latter two. In this review, these options and their outcomes are discussed as well as some possibilities to further enhance transplantation of this disadvantaged group of patients. Expert commentary: For highly sensitized patients, special attention is required, for which several strategies may apply. A step-wise approach may be the optimal strategy to facilitate successful transplantation of highly sensitized patients.

Donor Specificity but Not Broadness of Sensitization Is Associated With Antibody-Mediated Rejection and Graft Loss in Renal Allograft Recipients

Panel-reactive antibodies are widely regarded as an important immunological risk factor for rejection and graft loss. The broadness of sensitization against HLA is most appropriately measured by the "calculated population-reactive antibodies" (cPRA) value. In this study, we investigated whether cPRA represent an immunological risk in times of sensitive and accurate determination of pretransplantation donor-specific HLA antibodies (DSA). Five hundred twenty-seven consecutive transplantations were divided into four groups: cPRA 0% (n = 250), cPRA 1-50% (n = 129), cPRA 51-100% (n = 43), and DSA (n = 105). Patients without DSA were considered as normal risk and received standard immunosuppression without T cell-depleting induction. Patients with DSA received an enhanced induction therapy and maintenance immunosuppression. Surveillance biopsies were performed at 3 and 6 months. Median follow-up was 5.7 years. Among the three cPRA groups, there were no differences regarding the 1-year incidence of ABMR (p = 0.16) and TCMR (p = 0.75). The 5-year allograft survival rates were similar and around 87% (p = 0.28). The estimated glomerular filtration rate at last follow-up was 50-53 mL/min (p = 0.45). On multivariable Cox proportional hazard analysis, the strongest independent predictor for ABMR and (death-censored) graft survival was pretransplantation DSA. cPRA were not predictive for ABMR, TCMR, or (death-censored) graft survival. We conclude that with current DSA assignment, the broadness of sensitization measured by cPRA does not imply an immunological risk.