Under the Kidney Allocation System (KAS) of 2014, highly sensitized (HS) candidates receive substantial allocation priority.1-5 This has led to a significant increase in deceased donor kidney transplantation (DDKT) rates for candidates with a rounded calculated panel-reactive antibodies (cPRA) of 100% (24.3-fold increase for those with a cPRA 99.5%–99.9%, and an 11.6-fold increase for those with a cPRA ≥99.9%).6 However, HS candidates have been reported to have worse posttransplant outcomes than non-HS candidates, with higher rates of DGF and lower graft survival.7-10 As such, the changing donor pool and increased DDKT rates for these challenging candidates who previously might have died on the waitlist (and who might now die posttransplant) have led to concerns that KAS might amplify their poorer posttransplant outcomes, especially for cPRA-100% candidates who have seen the largest increase in DDKT rates under KAS.11
One study shortly after KAS examined outcomes of 18 cPRA-100% recipients, and reported 100% patient and graft survival over a median 13.9 months of follow-up.12 However, this study was performed at a single center, with only 18 patients, and their recipient composition and management strategies may not be broadly reflective of national practices. Moreover, no studies have compared posttransplant outcomes for cPRA-100% candidates before and after KAS, which might have changed in light of KAS-related changes in donor/recipient case-mix.13 For example, increased national sharing might lead to cPRA-100% candidates receiving more kidneys with prolonged cold ischemia time, leading to worse posttransplant outcomes.
In light of concerns that these recipients may have worse posttransplant outcomes after KAS, and concerns that KAS has led to unusual priority for patients who ultimately have bad outcomes, we hypothesized that outcomes might have worsened for cPRA-100% recipients under KAS. Therefore, we used national registry data to compare posttransplant outcomes for cPRA-100% recipients before and after KAS. We compared posttransplant mortality, graft loss, DGF, acute rejection (AR), and index hospitalization length of stay (LOS) among cPRA-100% recipients before and after KAS.
MATERIALS AND METHODS
This study used data from the Scientific Registry of Transplant Recipients (SRTR). The SRTR data system includes data on all donor, wait-listed candidates, and transplant recipients in the United States, submitted by the members of the Organ Procurement and Transplantation Network (OPTN), and has been described elsewhere.14 The Health Resources and Services Administration of the US Department of Health and Human Services provides oversight to the activities of the OPTN and SRTR contractors.
We compared adult (≥18 y old) DDKT recipients with a cPRA of 100% who were transplanted in the 5 years pre-KAS (December 4, 2009, to December 3, 2014) to those transplanted in the 3 years post-KAS (December 4, 2014, to December 3, 2017). This study was approved by the Johns Hopkins University Institutional Review Board. This study complies with the ethical principles outlined by the Declaration of Helsinki and the Declaration of Istanbul.
Mortality and Graft Loss
Our primary outcomes were mortality and death-censored graft loss. For death-censored graft loss, deaths with a functioning graft were treated as censored events. We analyzed mortality and death-censored graft loss using Cox proportional hazards regression with a sandwich estimator, which accounted for clustering of outcomes by center. We included the year of transplant in our regression models to account for secular trends in mortality and graft loss. We first used a univariable model, and then created a multivariable model, adjusting for the following recipient factors: age, gender, race (as reported to SRTR), ABO blood type, years on dialysis, cause of end-stage renal disease, and history of prior transplant. We considered donor and transplant characteristics (such as CIT, number of HLA mismatches, and the Kidney Donor Profile Index [KDPI]) to be mediators of any potential effect of KAS (in other words, along the mechanistic pathway between any changes that occurred through KAS and any differences observed in outcomes), rather than confounders, and so we did not adjust for them in our multivariable models.
Delayed Graft Function, Acute Rejection, and Length of Stay
We also examined changes in the incidence of DGF, AR, and LOS following KAS implementation. DGF was defined as the need for dialysis within the first week after transplantation as reported to SRTR and was analyzed using logistic regression. AR was defined as a clinically suspected or biopsy-proven episode of acute rejection within the first year after transplant, as reported by individual centers to SRTR, and was also analyzed using logistic regression. LOS was defined as the length of the index hospitalization in which the recipient was transplanted, and was analyzed as over dispersed count data using negative binomial regression. For each outcome, we first used a univariable model, and then a multivariable model adjusted for the same recipient characteristics as for mortality and graft loss.
We compared baseline characteristics between pre-KAS and post-KAS cPRA-100% recipients using the χ2 test for categorical variables, Student’s ttest for normally distributed continuous variables, and the Kruskal-Wallis test for nonnormally distributed continuous variables. We also compared posttransplant outcomes between first time and retransplant cPRA-100% recipients using the same regression model as our primary analysis. We then compared posttransplant outcomes between cPRA-100% recipients and cPRA-0% recipients under KAS, but included KDPI, number of HLA mismatches, and cold ischemia time (CIT) in our regression model to account for donor and transplant characteristics because we limited this comparison to recipients in the KAS era. All regression models accounted for center-level clustering of outcomes by using sandwich estimation for standard errors.15 Confidence intervals are reported as per the method of Louis and Zeger.16 All analyses were performed using Stata 15.0/IC for Windows (College Station, TX).
We identified 3026 cPRA-100% recipients in the 3 years post-KAS (1009/y) and 525 cPRA-100% recipients in the 5 years pre-KAS (105/y). Compared to pre-KAS recipients, post-KAS recipients were younger (33.4 versus 35.1 y, P = 0.01) and more likely to receive a kidney with a lower KDPI (median 30% versus 35%, P < 0.001) (Figure 1; Table 1). Additionally, post-KAS recipients were more likely to be African American (34.1% versus 26.1%, P < 0.001), have more HLA mismatches with their donor (58.2% with 1–3 mismatches versus 23.1%, P < 0.001), receive a kidney with a longer CIT (mean 21.0 versus 18.6 h, P < 0.001), and receive a kidney via national sharing (76.1% versus 46.1%, P < 0.001). The median waitlist time was 3.8 years (1.7 to 6.3) for pre-KAS recipients versus 4.8 years (2.7 to 8) for post-KAS recipients (P < 0.001). There was no difference in the percentage of kidneys that were pumped for pre-KAS and post-KAS recipients (20.2% versus 17.3%, P = 0.1).
Mortality and Graft Loss
Recipients were followed for a mean of 2.0 years. One-year survival for post-KAS recipients was 97.3% versus 95.4% for pre-KAS recipients, and 3-year survival was 93.6% versus 91.4% for pre-KAS recipients (P = 0.04) (Figure 2). This translated to a 30% decreased mortality risk for KAS in an unadjusted model (hazard ratio [HR]: 0.490.700.99, P = 0.04). However, after adjusting for secular trends (to disambiguate the effect of KAS versus general improvement in mortality over time), there was no statistically significant difference in mortality risk for post-KAS recipients compared to pre-KAS recipients (adjusted HR [aHR]: 0.911.743.30, P = 0.09). And after further adjusting for recipient factors, and accounting for secular trends, post-KAS recipients were at not at a statistically significant higher risk of mortality than pre-KAS recipients (aHR: 0.861.623.06, P = 0.1) (Table 2). That said, it should be noted that the HRs were somewhat far from 1.0 and the P values were somewhat close to 0.05, so these findings must be viewed in that light.
One-year death-censored graft survival for post-KAS recipients was 97.6% versus 94.8% for pre-KAS recipients, and 3-year death-censored graft survival was 93.7% versus 90.6% (P = 0.005) (Figure 3). Although post-KAS recipients had a lower unadjusted HR for death-censored graft failure (HR: 0.440.620.87, P = 0.006), this became nonsignificant after account for secular trends (aHR: 0.540.991.79, P > 0.9) and recipient factors (aHR: 0.521.001.91, P > 0.9) (Table 2).
Delayed Graft Function, Acute Rejection, and Length of Stay
The incidence of DGF in pre-KAS recipients was 29.3%, compared with 29.2% in post-KAS recipients (P = 0.9). Post-KAS cPRA-100% recipients were at no higher risk than pre-KAS cPRA-100% for DGF in an unadjusted model (odds ratio [OR]: 0.780.991.27, P = 0.9), or after accounting for secular trends (adjusted OR [aOR]: 0.610.891.31, P = 0.6) and recipient factors (aOR: 0.580.861.27, P = 0.4) (Table 2).
The incidence of acute rejection in pre-KAS recipients in the first year posttransplant was 11.2% compared with 11.7% for post-KAS recipients (P = 0.8). Post-KAS recipients were at no higher risk in an unadjusted model (OR: 0.711.041.52, P = 0.8), or after accounting for secular trends (aOR: 0.841.322.09, P = 0.2) and recipient factors (aOR: 0.610.941.43, P = 0.8) (Table 2).
The median index hospitalization LOS was 5 days for pre-KAS recipients, and 5 days for post-KAS recipients (P = 0.2). Post-KAS recipients did not have a higher LOS in an unadjusted model (LOS ratio: 0.911.001.09, P = 0.9), or after accounting for secular trends (LOS ratio: 0.981.171.40, P = 0.09) and recipient factors (adjusted LOS ratio: 0.981.161.36, P = 0.08) (Table 2).
cPRA-100% Retransplant Recipients
The majority of recipients pre-KAS (70.1%) and post-KAS (66.4%) were retransplant recipients (P = 0.1). Compared with the first time transplant recipients, retransplant recipients had an equivalent risk of mortality (aHR: 0.651.001.55, P > 0.9) and death-censored graft failure (aHR: 0.701.011.46, P > 0.9), but a 69% higher risk of delayed graft function (aOR: 1.371.692.08, P < 0.001), a 48% higher risk of acute rejection (aOR: 1.061.482.07, P = 0.02), and a 14% longer LOS (adjusted LOS ratio: 1.021.141.29, P = 0.03).
cPRA-100% Recipients Compared to cPRA-0% Recipients Under KAS
Under KAS, there were 21 306 cPRA-0% recipients and 3026 cPRA-100% recipients, who were followed for a median 1.4 years posttransplant. Compared with cPRA-0% recipients, cPRA-100% recipients were at equivalent risk of mortality (aHR: 0.750.961.24, P = 0.8), death-censored graft failure (aHR: 0.811.041.33, P = 0.8), DGF (aOR: 0.750.911.09, P = 0.3), and LOS (adjusted LOS ratio: 0.880.971.06, P = 0.5). However, cPRA-100% recipients had a 61% higher risk of acute rejection (aOR: 1.261.612.04, P < 0.001).
In this national study of 3551 cPRA-100% DDKT recipients cPRA-100%, we found no statistically significant differences in posttransplant outcomes post-KAS versus pre-KAS. In particular, after adjusting for recipient characteristics and accounting for secular trends (general improvement in outcomes over time, irrespective of KAS), we did not find any evidence for post-KAS changes in mortality, graft loss, DGF, AR, and LOS. The median KDPI for cPRA-100% recipients improved from 35% pre-KAS to 30% post-KAS, although the median CIT increased from 18.6 to 21.0 hours, and the percentage of recipients with 1-3 HLA mismatches increased from 23.1% to 58.2%. Despite these changes, short-term patient and graft survival were excellent for post-KAS recipients, who had a 3-year patient survival of 93.6% and a 3-year death-censored graft survival of 93.7%. Although DGF rates were high (29.2% for post-KAS recipients), the incidence of acute rejection (11.7% for post-KAS recipients) and median LOS (5 days) were more modest, and none were statistically significantly higher post-KAS compared with pre-KAS. Moreover, posttransplant mortality and death-censored graft failure were similar between cPRA-100% and cPRA-0% recipients under KAS, although cPRA-100% recipients had a 61% higher risk of acute rejection. Our results do not support the notion that cPRA-100% candidates are doing poorly under KAS, although careful monitoring of posttransplant mortality is warranted given our somewhat equivocal findings.
Our findings of good short-term outcomes in cPRA-100% recipients are consistent with a single-center study that reported a 100% patient and graft survival for 18 cPRA-100% recipients post-KAS, over a median follow-up time of 13 months.12 We have extended this work by using national registry data on 3026 post-KAS cPRA-100% recipients (compared with 18), including all transplant centers and accounting for differences in management of these recipients across the country, and comparing to pre-KAS outcomes. However, posttransplant mortality risk was equivocally higher post-KAS compared with pre-KAS (aHR: 0.861.623.06, P = 0.1). Although this did not reach statistical significance, it warrants ongoing monitoring as experience under KAS grows. Moreover, we found that cPRA-100% recipients had a similar risk of mortality, death-censored graft failure, DGF, and LOS compared with cPRA-0% recipients under KAS, but a 61% higher risk of acute rejection. Although these short-term findings are encouraging, monitoring of long-term outcomes is important because acute rejection might be associated with an increased risk of longer-term mortality and graft loss.17
We also found no post-KAS changes in the incidence of DGF, acute rejection, or LOS. Our relatively high incidence of DGF (29.2% for post-KAS recipients) is consistent with previous reports that HS candidates are at higher risk of DGF than non-HS candidates.8,9 Interestingly, although KAS increased the CIT of kidneys allocated to cPRA-100% candidates, and increased CIT is associated with higher rates of DGF,18,19 we found that post-KAS cPRA-100% candidates had an equivalent incidence of DGF compared with pre-KAS recipients. It might be that the improved KDPI of kidneys allocated to post-KAS recipients (median 30% post-KAS versus 35% pre-KAS) partially offset the effect of a slight increase in CIT (median 21.0 h post-KAS versus 18.6 h pre-KAS). Finally, in the absence of an increase in immediate posttransplant complications, such as DGF, our finding of equivalent LOS after KAS compared with before KAS is not surprising, and is consistent with a report using University HealthSystem Consortium data that showed no change in overall LOS across all post-KAS recipients20 but extends this report specifically to the cPRA-100% population.
The limitations of our study deserve consideration. First, our use of national registry data precludes assessment of certain aspects of our outcomes. For example, we do not have granular information on the type of acute rejection (acute cellular rejection versus antibody-mediated rejection) or more nuanced data on graft function (serial creatinine measurements), which might be important when considering outcomes of cPRA-100% candidates. Additionally, the OPTN does not collect data on crossmatch specifics or the presence of donor-specific antibody, which has been associated with decreased graft survival. Despite this limitation, our use of national registry data facilitates broadly generalizable inferences, such that our findings are likely to be applicable at most centers across the country. Additionally, we are limited to a median 2 years of follow-up data. Although our data are consistent with no worsening of outcomes after KAS, this might change with long-term follow-up. For example, post-KAS cPRA-100% recipients are more likely to have 1–3 HLA mismatches than pre-KAS recipients, who more commonly had 0 HLA mismatches. This difference may not manifest itself as worsened short-term graft outcomes, and may only become apparent with longer-term follow-up. Despite this, we have also examined short-term outcomes that could have been plausibly affected by KAS (acute rejection, delayed graft function, and LOS) and would have likely shown a difference within our median 2 years of follow-up.
In conclusion, in this national study, we found that posttransplant outcomes have not significantly worsened for cPRA-100% DDKT recipients post-KAS. Despite significant changes to characteristics of organs allocated to cPRA-100% candidates after KAS (improved KDPI, decreased HLA matching, increased CIT), we found no statistically significant changes in mortality, graft loss, DGF, AR, or LOS. Our results are reassuring given the substantial allocation priority these candidates now receive under KAS, and support continued prioritization of these candidates, although posttransplant mortality should be closely monitored as experience under KAS grows.
The analyses described here are the responsibility of the authors alone and do not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products or organizations imply endorsement by the US Government.
The data reported here have been supplied by the Hennepin Healthcare Research Institute (HHRI) as the contractor for the SRTR. The interpretation and reporting of these data are the responsibility of the author(s) and in no way should be seen as an official policy of or interpretation by the SRTR or the US Government.
1. Friedewald JJ, Samana CJ, Kasiske BL, et al. The kidney allocation system. Surg Clin North Am. 2013; 93:1395–1406
2. Stegall MD, Stock PG, Andreoni K, et al. Why do we have the kidney allocation system we have today? A history of the 2014 kidney allocation system. Hum Immunol. 2017; 78:4–8
3. Colovai AI, Ajaimy M, Kamal LG, et al. Increased access to transplantation of highly sensitized patients under the new kidney allocation system. A single center experience. Hum Immunol. 2017; 78:257–262
4. Hickey MJ, Zheng Y, Valenzuela N, et al. New priorities: analysis of the new kidney allocation system on UCLA patients transplanted from the deceased donor waitlist. Hum Immunol. 2017; 78:41–48
5. Parsons RF, Locke JE, Redfield RR III, et al. Kidney transplantation of highly sensitized recipients under the new kidney allocation system: A reflection from five different transplant centers across the United States. Hum Immunol. 2017; 78:30–36
6. Jackson KR, Covarrubias K, Holscher CM, et al. The national landscape of deceased donor kidney transplantation for the highly sensitized: transplant rates, waitlist mortality, and posttransplant survival under KAS. Am J Transplant. 2019; 19:1129–1138
7. Süsal C, Döhler B, Opelz G. Presensitized kidney graft recipients with HLA class I and II antibodies are at increased risk for graft failure: a collaborative transplant study report. Hum Immunol. 2009; 70:569–573
8. Redfield RR, Scalea JR, Zens TJ, et al. The mode of sensitization and its influence on allograft outcomes in highly sensitized kidney transplant recipients. Nephrol Dial Transplant. 2016; 31:1746–1753
9. Peräsaari JP, Kyllönen LE, Salmela KT, et al. Pre-transplant donor-specific anti-human leukocyte antigen antibodies are associated with high risk of delayed graft function after renal transplantation. Nephrol Dial Transplant. 2016; 31:672–678
10. Barama A, Oza U, Panek R, et al. Effect of recipient sensitization (peak PRA) on graft outcome in haploidentical living related kidney transplants. Clin Transplant. 2000; 14:212–217
11. Stewart DE, Kucheryavaya AY, Klassen DK, et al. Changes in deceased donor kidney transplantation one year after KAS implementation. Am J Transplant. 2016; 16:1834–1847
12. Parajuli S, Redfield RR, Astor BC, et al. Outcomes in the highest panel reactive antibody recipients of deceased donor kidneys under the new kidney allocation system. Clin Transplant. 2017; 31
13. Massie AB, Luo X, Lonze BE, et al. Early changes in kidney distribution under the new allocation system. J Am Soc Nephrol. 2016; 27:2495–2501
14. Massie AB, Kucirka LM, Kuricka LM, et al. Big data in organ transplantation: registries and administrative claims. Am J Transplant. 2014; 14:1723–1730
15. Williams RL. A note on robust variance estimation for cluster-correlated data. Biometrics. 2000; 56:645–646
16. Louis TA, Zeger SL. Effective communication of standard errors and confidence intervals. Biostatistics. 2009; 10:1–2
17. Clayton PA, McDonald SP, Russ GR, et al. Long-term outcomes after acute rejection in kidney transplant recipients: an ANZDATA analysis. J Am Soc Nephrol. 2019; 30:1697–1707
18. Serrano OK, Vock DM, Chinnakotla S, et al. The relationships between cold ischemia time, kidney transplant length of stay, and transplant-related costs. Transplantation. 2019; 103:401–411
19. Kayler LK, Srinivas TR, Schold JD. Influence of CIT-induced DGF on kidney transplant outcomes. Am J Transplant. 2011; 11:2657–2664
20. Taber DJ, DuBay D, McGillicuddy JW, et al. Impact of the new kidney allocation system on perioperative outcomes and costs in kidney transplantation. J Am Coll Surg. 2017; 224:585–592