OPTN/SRTR 2015 Annual Data Report: Early effects of the new kidney allocation system

Abstract

In December 2014, a new kidney allocation system (KAS) was implemented in the United States in an attempt to improve access to transplant for historically underrepresented groups, and to incorporate longevity matching such that donor kidneys with the longest projected graft survival are given to recipients with the longest projected patient survival. The development of organ allocation policies is often guided by simulated allocation models, computer programs that simulate the arrival of donated organs and new candidates on the waiting list over a 1-year period to project outcomes under a new allocation method. We examined the early outcomes under the new KAS using quarterly data beginning in 2013, revealing whether trends were already underway before implementation. Quarterly data also serve to reveal any bolus effect, or a rapid rise or fall in the proportion of transplants in a given group due to reordering of the list, followed by tapering toward a new steady state. Post-KAS changes were notable for an increase in the proportion of transplants among younger candidates, black and Hispanic candidates, highly sensitized candidates, and those on dialysis for at least 5 years. Transplants among blood type B candidates increased slightly but these candidates remain underrepresented relative to their prevalence on the waiting list. Regional and national sharing increased under the new KAS, but transplants of kidneys with a kidney donor profile index above 85% decreased. Early graft survival appears unchanged, but given the increases in regional sharing, cold ischemia time, and transplants among highly sensitized candidates and candidates with long pretransplant dialysis time, long-term graft survival will need to monitored.

Introduction

After more than a decade of debate and analytic modeling, the Organ Procurement and Transplantation Network (OPTN) in the United States approved a new kidney allocation system (KAS), implemented in December 2014. The new KAS replaced a largely first-come, first-served system essentially unchanged for three decades despite dramatic shifts in the demographics of waitlisted patients. As the supply-demand gap steadily widened, waiting time became the disproportionate driver of access to transplant over other patient-level characteristics. The system also unintentionally perpetuated racial disparities in access to transplant given a greater prevalence of blood type B, higher sensitization, and delayed referral to the transplant waiting list among racial and ethnic minorities. Charged with upholding the tenets of the OPTN Final Rule, the new KAS was intended to balance equitable distribution of deceased donor kidneys with maximal utility of this increasingly scarce resource, goals that are unfortunately at times in direct opposition.

Under the new KAS, the intended improvement in utility occurs through risk stratification using the kidney donor risk index (KDRI), in which donor kidneys in the highest 20% of expected posttransplant allograft survival based on 10 donor characteristics are offered first to recipients in the highest 20% of estimated posttransplant survival (EPTS). The KDRI score includes donor age, height, weight, race/ethnicity, history of hypertension, history of diabetes, cause of death, serum creatinine, hepatitis C status, and donation after circulatory death status. This score is converted to a kidney donor profile index (KDPI) score of 0% to 100%, such that kidneys with KDPI above 85% are somewhat comparable to previously designated expanded criteria donor kidneys. The EPTS score includes the following patient characteristics: age, duration on dialysis, prior transplant, and diabetes status.

The new KAS addresses issues of equity through the allocation of priority points and increased regional and national sharing of organs. Priority is given to candidates awaiting multiple organs, candidates with calculated panel-reactive antibodies (CPRA) 98% and above, zero-HLA mismatch kidneys, pediatric candidates, and prior living donors. Additional priority points are given on a sliding scale to candidates with CPRA greater than 19%, and the highest priority is given to the most highly sensitized candidates, with local, regional, and national priority for organ offers to those with CPRA 98%, 99%, and 100%, respectively. Blood type A2 and A2B kidneys are now offered to medically suitable type B candidates at transplant programs willing to accept such organs. In addition, children receive priority for kidneys with KDPI below 35%. Finally, candidates who are listed after initiating dialysis are given credit for time on dialysis before listing.

Trends in kidney transplants before and after the new allocation system

Distributions of transplants are shown by age, race, PRA, blood type, dialysis time, cold ischemia time, geographic sharing, and KDPI. Data are shown by quarter beginning in 2013, 2 years before implementation of the new KAS, revealing whether trends were already underway before implementation, and revealing any bolus effect, or a rapid rise or fall in the proportion of transplants in a given group due to reordering of the list, followed by tapering toward a new steady state.

The proportion of transplants among candidates aged 18 to 34 years sharply increased from 10.0% in the third quarter of 2014 to 14.6% in the first quarter of 2015 (a 45.5% increase), then declined back to 12.6% at the end of 2015, an increase of 26.2% over the pre-KAS distribution in this age group (Figure KAS 1). A smaller but similar trend was noted for candidates aged 35 to 49 years, with a 19.0% increase initially in the first quarter of 2015 and an increase of 12.0% over pre-KAS rates by the end of 2015. Conversely, the proportion of transplants among patients aged 65 years or older decreased sharply, from 22.0% in the third quarter of 2014 to 16.0% in the first quarter of 2015 (a 27.0% decrease). By the end of 2015, 19.9% of recipients were aged 65 years or older, a 9.4% decrease over the pre-KAS proportion.

Reducing disparities in access to transplant for racial and ethnic minority groups was a proposed goal of the new allocation system. The number and proportion of black transplant recipients increased after implementation of the new KAS, from 31.1% (n = 872) pre-KAS to 37.8% (n = 1024) in the first quarter of 2015 (21.6% increase) (Figure KAS 2). Most of this increase was sustained through the end of 2015. We observed a corresponding decrease in the proportion of white recipients, from 40.8% to 33.9% (16.9% decrease); by the end of 2015, 35.4% of recipients were white, a 13.3% decrease compared with the third quarter of 2014. Proportions of Hispanic recipients also increased in 2015, by 9.1% from 17.7% to 19.4% by the end of 2015; this increase had not clearly leveled off to a new steady state by the end of 2015.

Perhaps one of the greatest surprises after implementation of the new KAS was the magnitude of the increase in transplant recipients with the highest CPRAs. Historically, these candidates have had very low rates of transplant per 100 waitlist days and have been consistently underrepresented among transplant recipients relative to their prevalence on the waiting list. Priority point allocation for CPRA changed dramatically from a flat 4-point allocation for CPRA 80% or higher to a sliding scale that progressively increases priority point allocation as CPRA rises. The previous 2-tiered system did not account for the large biological differences as CPRA rises from 1% to 80% and from 80% to 100%, and resulted in transplant rates that did not reflect the prevalence of high CPRA groups on the list, with notable overrepresentation of recipients with CPRA 80% to 94%. The new sliding scale system results in a point allocation difference of about 2 points between CPRA 80% and 89%, but more than 195 points between a CPRA of 90% and 100%. This sliding scale was supported by simulations that predicted an 18.6% decrease in the proportion of transplants among candidates with CPRA 80% to 89%, from 7.0% to 5.7%, reaching levels more consistent with (but still above) their prevalence on the waiting list. In fact, transplants in this group declined from 6.6% in the third quarter of 2014 to 3.2% by the end of 2015, a 52.3% decrease, but to a level commensurate with prevalence on the waiting list (Figure KAS 5). Transplants among patients with CPRA 99% to 100% increased from 2.2% pre-KAS, peaked at 13.3% in the first quarter of 2015, and tapered off to 10.1% by the end of 2015, ultimately representing a 366% increase, slightly above these candidates' waitlist prevalence of 8.1%. Transplants among candidates with CPRA 98% also doubled from 1.2% pre-KAS to 2.4% post-KAS, with no appreciable bolus effect. The greater than anticipated effect of the KAS on high-CPRA transplants may be explained by an underestimate of the acceptance of high-quality regional and national offers for these highly sensitized candidates.

The new KAS sought to decrease disparity in access to transplant for patients with blood type B (more prevalent in ethnic minority populations) by allocating non-A1 (aka, A2) and non-A1B (aka, A2B) kidneys to type B recipients with low anti-A antibody titers. While the total number of A2/A2B to B transplants increased 4-fold after implementation of the new KAS, the number as a proportion of all waitlisted type B candidates remained small. Consequently, the intended increase in transplants among type B candidates more proportional to their prevalence on the waiting list did not occur (Figure KAS 3). The lack of substantial increase is likely due to transplant programs reporting very small numbers of eligible candidates (blood type B with low anti-A antibody titers). Given a published prevalence of low anti-A antibody titers ranging from 80% to 90% of type B candidates, the small number listed raises concern that programs may not be identifying all eligible candidates.

Transplants among patients on dialysis for longer than 10 years increased dramatically after the new KAS was implemented, from 14.3% to 25.8% of transplants in the first quarter of 2015 with a post-KAS bolus effect, and decreased to 17.7% by the end of 2015, ultimately a 23.9% increase from the third quarter of 2014 (Figure KAS 4). Similarly, transplants among patients on dialysis for 5 to 10 years increased from 28.5% to 33.8% in the first quarter of 2015, an 18.7% increase. However, no bolus effect was evident, as the percentage of transplants in this group did not subsequently decline toward pre-KAS levels and continued to rise through 2015 to 34.9% of transplants.

Regional and national sharing of kidneys increased post-KAS from 9.8% to 12.7% and 12.2% to 19.3% of transplants, respectively. National sharing decreased by the end of 2015 but remained above pre-KAS levels at 16.6%, while regional sharing remained stable at 13.1% of transplants (Figure KAS 8). The increase in non-local transplants not surprisingly resulted in longer cold ischemia time (CIT), with a 10.6% increase in the proportion of kidneys transplanted with 24 to 36 hours of CIT and a corresponding 18.7% decrease in those transplanted with less than 12 hours of CIT (Figure KAS 6).

Despite implementation of regional sharing of high KDPI kidneys, transplants with KDPI above 85% decreased, from between 9% and 11% in the 2 years before the new KAS to between 7.5% and 8.9% in 2015 (Figure KAS 7). This change appears to be mirrored by an increase in discards of these high KDPI kidneys from 55.1% pre-KAS to 59.8% in the first quarter of 2015, a proportion that appeared relatively stable for the remainder of 2015 (Figure KAS 12). However, the decrease in transplants with KDPI above 85% appears to have begun as far back as the first quarter of 2013, 2 years before the implementation of KAS. Should this trend continue, it could further reduce access to transplant for older candidates, who make up a smaller group of recipients under the new KAS, and it should therefore be closely monitored.

Early graft survival under the new allocation system

Simulations of the new allocation policy projected improvements in both patient and graft survival, with an average 7.0% increase in median patient life-years per transplant and a 2.8% increase in median allograft years of life compared with the previous allocation policy. It is too early to evaluate whether these predictions will hold. However, higher than expected transplant rates among candidates with CPRA 99% to100%, increased regional and national sharing with longer CIT, and more transplants in candidates with longer dialysis time have raised concerns that this intended effect will not be realized. Encouragingly, 6- and 12-month graft survival pre- and post-KAS appears unchanged, even when examined by age, race, KDPI, and CPRA (Figure KAS 13, Figure KAS 14, Figure KAS 15, Figure KAS 16, Figure KAS 17, Figure KAS 23, Figure KAS 24, Figure KAS 25, Figure KAS 26, Figure KAS 27). Twelve-month graft survival curves by age and era may indicate lower post-KAS graft survival for recipients aged 50 to 64 and 65 years or older, groups that will continue to be monitored closely in years to come.

Figure List

Kidney allocation system

Figure KAS 1. Distribution of kidney transplants by age
Figure KAS 2. Distribution of kidney transplants by race
Figure KAS 3. Distribution of kidney transplants by blood type
Figure KAS 4. Distribution of kidney transplants by time on dialysis
Figure KAS 5. Distribution of kidney transplants by CPRA
Figure KAS 6. Distribution of kidney transplants by cold ischemia time
Figure KAS 7. Distribution of kidney transplants by KDPI
Figure KAS 8. Distribution of kidney transplants by geographic sharing
Figure KAS 9. Kidneys recovered for transplant and not transplanted by age
Figure KAS 10. Kidneys recovered for transplant and not transplanted by race
Figure KAS 11. Kidneys recovered for transplant and not transplanted by sex
Figure KAS 12. Kidneys recovered for transplant and not transplanted by KDPI
Figure KAS 13. Six-month graft survival by era
Figure KAS 14. Six-month graft survival by age and era
Figure KAS 15. Six-month graft survival by race and era
Figure KAS 16. Six-month graft survival by KDPI and era
Figure KAS 17. Six-month graft survival by CPRA and era
Figure KAS 18. Six-month graft survival by coldi ischemia time and era
Figure KAS 19. Six-month graft survival by time on dialysis and era
Figure KAS 20. Six-month graft survival by blood type and era
Figure KAS 21. Six-month graft survival by adult vs. pediatric and era
Figure KAS 22. Six-month graft survival by EPTS group and era
Figure KAS 23. Twelve-month graft survival by era
Figure KAS 24. Twelve-month graft survival by age and era
Figure KAS 25. Twelve-month graft survival by race and era
Figure KAS 26. Twelve-month graft survival by KDPI and era
Figure KAS 27. Twelve-month graft survival by CPRA and era
Figure KAS 28. Twelve-month graft survival by coldi ischemia time and era
Figure KAS 29. Twelve-month graft survival by time on dialysis and era
Figure KAS 30. Twelve-month graft survival by blood type and era
Figure KAS 31. Twelve-month graft survival by adult vs. pediatric and era
Figure KAS 32. Twelve-month graft survival by EPTS group and era