Objectives: There are conflicting reports in the literature regarding outcomes after kidney transplant for patients of black ethnicity. To investigate further, we compared outcomes for black versus white kidney transplant recipients in a single UK transplant center.
Materials and Methods: We analyzed 1066 kidney transplant recipients (80 black patients, 986 white patients) within a single-center cohort (2007-2017) in the United Kingdom, with cumulative 4446 patient-year follow-up. Data were electronically extracted from the Department of Health Informatics database for every study recruit, with manual data linkage to the UK Transplant Registry (for graft survival, delayed graft function, and rejection data) and Office for National Statistics (for mortality data). Primary outcomes of interest were graft/patient survival.
Results: Black recipients have increased baseline risk profiles with longer wait times, difficulty in matching, worse HLA matching, more socioeconomic deprivation, and lower rates of living kidney donors. Postoperatively, black versus white recipients had increased risk for delayed graft function (34.3% vs 10.2%; P < .001), increased 1-year rejection (16.7% vs 7.3%; P = .012), higher 1-year creatinine levels (166 vs 138 mmol/L; P = .003), and longer posttransplant length of stay (14.5 vs 9.5 days; P = .020). Although black recipients did not have increased risk of death versus white recipients (10.0% vs 11.0%, respectively; P = .486), they did have increased risk for death-censored graft loss (23.8% vs 11.1%; P = .002). However, in an adjusted Cox regression model, black ethnicity was not associated with increased risk for death-censored graft loss (hazard ratio of 1.209, 95% confidence interval, 0.660-2.216; P = .539).
Conclusions: Black kidney transplant recipients in the United Kingdom have increased risk of adverse graft-related outcomes due to high-risk baseline variables rather than their black ethnicity per se.
Key words : Outcomes, Renal transplant, Survival
Black ethnicity comprised 3% of the general population in the 2011 census report for the United Kingdom.1 Black individuals are at significantly greater risk of developing end-stage kidney failure, with national registry data highlighting a higher incidence of end-stage kidney failure, requiring renal replacement therapy commencement, in areas with greater black (and south Asian) ethnicity.2 Although survival outcomes after commencement of renal replacement therapy/dialysis are improved among black patients,3 outcomes after kidney transplant in black versus other ethnic groups are conflicting.
In the United States, kidney transplant recipients of black ethnicity have consistently shown inferior outcomes compared with white recipients,4 although contemporary data from the United States now suggest that the disparity between black and white kidney allograft outcomes is narrowing (but still not equivalent).5 This contrasts starkly with Canadian data, which suggest that black kidney transplant recipients have similar graft survival but improved patient survival compared with white kidney transplant recipients.6 From a European perspective, no differences in outcomes have been observed between black and white kidney transplant recipients in France,7 but historical data from the United Kingdom have suggested that black recipients may have inferior graft survival.8,9 Data between countries are not transferable, with comparative analyses confirming that outcomes for black kidney transplant recipients in Canada10 and England11 are inferior to those for black recipients in the United States.
These conflicting results point to a need for contemporary, country-specific data to ascertain whether risks after kidney transplant are different for end-stage kidney failure patients of black ethnicity. This is important as black patients comprise 8% and 3% of all deceased-donor and living-donor kidney transplants in the United Kingdom, respectively,12 and may require targeted counseling. Although our previous work had shown disparate outcomes for black kidney transplant recipients in England versus the United States, we do not have any evidence to suggest outcomes are different for black recipients versus other ethnic groups within the United Kingdom. Therefore, the aim of this study was to investigate this in a large, single-center analysis.
Materials and Methods
We undertook a retrospective cohort analysis of all consecutive kidney-alone transplants performed at a single center between January 2007 and January 2017 (with survival analysis censored to event or September 2017, whichever occurred first). Our cohort only included kidney-alone allograft recipients aged 18 and over; all other kidney allograft recipients were included for analysis. Data for every study participant were electronically extracted from the Department of Health Informatics database, with manual data linkage to additional electronic patient records. Graft survival, delayed graft function, and rejection data were acquired and linked from the UK Transplant Registry held by NHS Blood and Transplant. Patient survival data were obtained by record linkage to the Office for National Statistics.
Definition of ethnicity
We utilized existing predetermined ethnicity classifications as obtained from electronic patient records, which were cross checked against data from the UK Transplant Registry. Ethnicity was classified into the following categories: white, black, south Asian (also referred to as Indo-Asian), Chinese, mixed, or other. For this analysis, we used data for black versus white ethnicity only.
Definitions of variables
Baseline and posttransplant data were extracted and classified as described here. HLA mismatch levels were defined and graded in accordance to NHS Blood and Transplant classification levels. Matchability scores were calculated from a standardized pool of 10 000 recent donors, from which numbers of blood group-identical donors who were well or favorably HLA-mismatched with recipients were counted. This number was converted to a standardized score between 1 and 10, which was used to categorize recipients into 1 of 3 matchability groups: easy (1-3), moderate (4-6), or difficult (7-10) to match. Determination of socioeconomic deprivation was based on the Index of Multiple Deprivation, a multiple deprivation model calculated at the local level. The model is a composite construct of multiple domains reflective of areas of socioeconomic deprivation, including (1) income deprivation, (2) employment deprivation, (3) health deprivation and disability, (4) education skills and training deprivation, (5) barriers to housing and services, (6) living environment deprivation, and (7) crime. Individual domains are measured in isolation and subsequently combined (utilizing appropriately judged weighting) into a single composite termed the Index of Multiple Deprivation. On the Index of Multiple Deprivation quintile scale, 1 represents the most deprived and 5 represents the least deprived area.
All patients received the same immunosuppression with minimization of tacrolimus exposure, in line with the SYMPHONY protocol13 over the study period. Induction therapy was with basiliximab (10 mg twice/day) and methylprednisolone (500 mg). Maintenance therapy included tacrolimus (target 12-hour trough level of 5-8 ng/L), mycophenolate mofetil (2 g/day with tapering to 1 g/day after 6 months), and maintenance corticosteroids. Biopsies were indication based in the context of transplant dysfunction (categorized as ≥ 20% creatinine rise or new-onset proteinuria). Biopsy data were classified in accordance with the latest Banff criteria.14
Management protocols remained consistent over the study period. Briefly, episodes of acute cellular rejection were treated with a bolus of corticosteroids, with T-cell depletion therapy used for cases of steroid-resistant rejection. Antibody-mediated rejection was treated with antibody removal by plasmapheresis with or without intravenous immunoglobulin. Viral serology (eg, polyoma virus) and/or anti-HLA antibodies were checked on an indication-basis based on transplant dysfunction. Standard antibiotic prophylaxis after kidney transplant consisted of nystatin (3 months), cotrimoxazole (12 months), valganciclovir (3 months if deemed high risk [donor cytomegalovirus positive/recipient cytomegalovirus negative]), and isoniazid/pyridoxine (12 months if high risk for tuberculosis [previous tuberculosis or ethnic minority]).
Differences in baseline demographic and transplant characteristics between ethnicity groups were initially assessed with univariate analyses, using chi-square tests for categorical data, t tests for parametric continuous data, and Wilcoxon tests for nonparametric continuous data. Data comparisons were made between patients of black versus nonblack ethnicity.
Our primary outcomes of interest were graft and patient survival. Graft failure was taken as the time from transplant to return to dialysis, graft nephrectomy, or repeat kidney transplant (whichever occurred first, with death data censored). Patient survival was defined as the time from transplant until death. Follow-up analyses of the entire transplant study cohort included all data up to September 2017, which was used as the date of censoring. For these outcomes, Kaplan-Meier analyses were run with Mantel-Cox (log-rank) tests used to compare patient groups and to test the proportionality hazards assumption. Time to date of last follow-up or an event (death or graft loss) was calculated in days. Mortality outcome data were available for all 1066 kidney transplant recipients (100.0% completeness), but graft survival data were only available for 995 kidney transplant recipients (93.3%).
Cox proportional hazards regression models were then used for time-to-event outcome analyses for graft loss. Covariates included in the multivariable analysis were (1) any with values of P < .10 in the unadjusted analyses and (2) automatic inclusion of age, ethnicity (black versus white), recipient diabetes, matchability, socioeconomic deprivation, HLA mismatch, living donation, ABO incompatibility, first transplant, cold ischemic time, wait list time, and donor age based on documented literature of an association with patient or graft survival.
This study received institutional approval and was registered as an audit (audit identifier; CARMS-12578). The corresponding author had full access to all data.
Data were extracted for all consecutive kidney transplant recipients at our center, with median follow-up of 858 days (interquartile range, 192-1887 days) and cumulative 4446 patient-years of follow-up. Breakdown of kidney transplant recipients by ethnicity was as follows: 986 white (62.6%), 80 black (6.1%), 271 south Asian (17.2%), 9 mixed (0.6%), and 75 other (4.8%). For this analyses, we only included data for black and white recipients (n = 1066).
Table 1 highlights important baseline demographic differences between kidney transplant recipients of black ethnicity compared with those of white ethnicity. Black kidney transplant recipients had longer wait times, more difficult matchability scores, greater residence in socioeconomically deprived areas, and greater HLA-level mismatches compared with white kidney transplant recipients. With regard to donors, black kidney transplant recipients were less likely to receive kidneys from living donors and received kidneys from older donors with longer cold ischemic times versus white kidney transplant recipients.
Black recipients had worse posttransplant outcomes than white recipients (Table 2). Black recipients (vs white recipients) had longer postoperative length of stay after transplant surgery (14.5 vs 9.5 days; P = .020), higher risk for delayed graft function (34.3% vs 10.2%; P < .001), increased risk for any episode of rejection within the first year (16.7% vs 7.3%; P = .012), and higher creatinine level at the 1-year time point among surviving kidneys (166 vs 138 mmol/L; P = .003). However, there were no differences shown with other histologic findings or development of donor-specific anti-HLA antibody at any time point after kidney transplant.
Patient and graft survival
Although black recipients did not have an increased risk of death versus white recipients (10.0% versus 11.0%, respectively, P = .486), black recipients did have a significantly increased risk for death-censored graft loss (23.8% vs 11.1%; P = .002). Figure 1 shows the unadjusted Kaplan-Meier plots of patient and graft survival stratified by black versus white ethnicity. However, in a Cox regression model adjusted for baseline variables, black ethnicity was not associated with increased risk for death-censored graft loss (hazard ratio [HR] = 1.209; 95% confidence interval [CI], 0.660-2.216; P = .539).
We undertook a subanalyses to probe the data further. First, we compared black versus all other minority ethnic kidney transplant recipients, and the results mirrored the comparison between black versus white ethnicity groups. Black versus all other kidney transplant recipients had increased risk for delayed graft function (34.3% vs 16.2%; P = .001), increased risk for rejection within the first year (16.7% vs 7.1%; P = .009), increased creatinine level at 1 year (166 vs 140 mmol/L; P = .046), and increased risk for death-censored graft loss (23.8% vs 11.7%; P = .003), but no difference in risk for death (10.0% vs 9.8%; P = .532). When we replicated the Cox regression model, after adjustment with baseline covariates, we did not find black versus nonblack ethnicity to be a risk factor for death-censored graft loss.
Second, we stratified the analyses into living- versus deceased-donor kidney transplant procedures. In the setting of living-donor kidney transplant, we found no significant differences in any variable comparing black versus white ethnicity regarding risk for delayed graft function (10.0% vs 4.0%, respectively; P = .348), risk for rejection within the first year (0.0% vs 8.5%, respectively; P = .382), 1-year creatinine (162 vs 128 mmol/L, respectively; P = .192), risk for death (7.1% vs 7.0%, respectively; P = .644), or death-censored graft loss (7.1% vs 6.3%, respectively; P = .605). However, we must be cautious with this subanalysis due to the small numbers (only 14 black recipients underwent living-donor kidney transplant). Findings in the setting of deceased-donor kidney transplant were similar to our main findings.
In our single-center retrospective analysis, we observed significant baseline differences between black and white kidney transplant recipients at time of transplant that rendered them at higher risk for adverse graft-related outcomes. In our unadjusted analyses, we have shown that black versus white kidney transplant recipients have increased risk for delayed graft function (deceased-donor kidney transplants), rejection within the first year, worse creatinine levels at 1-year posttransplant, and increased risk for death-censored graft loss. However, after adjustment for these baseline factors, black ethnicity in itself was not a risk factor for death-censored graft loss. Our data suggest that black kidney transplant recipients should be counseled about their increased risk for adverse graft-related outcomes posttransplant, but this risk is related to baseline covariates rather than their black ethnicity per se.
Although the examination of outcomes in kidney transplant recipients with black ethnicity is not novel, we believe our results are important as they highlight the lack of generalizability between countries for posttransplant outcomes stratified by ethnicity and the need for country-specific data. Most data on black ethnicity outcomes after kidney transplant come from the United States, but our study and others have shown that those findings cannot be directly translated to black recipients elsewhere. Published reports have consistently shown that outcomes after kidney transplant are inferior for black recipients in the United States compared with Canada10 and the United Kingdom.11 In the United States, black recipients have shown inferior outcomes compared with white recipients,4 although this gap has narrowed with contemporary immunosuppression.5 Similar results have been reported in the United Kingdom, but each study has had specific limitations that curb our interpretation. For example, Rudge and colleagues compared kidney transplant outcomes for different ethnicity groups in the United Kingdom between 1998 and 2003 in a population-cohort study.8 They observed worse 3-year graft survival for black patients versus white or south Asian patients (76% vs 82% vs 82%, respectively; P = .06), with the effect more apparent after risk adjustment for age (recipient and donor) and HLA match (P = .03). However, this analysis had limited risk adjustments, ignoring many important baseline factors that could influence graft survival, and did not reflect outcomes with contemporary immunosuppression. In a smaller multiple-center cohort of 2650 kidney transplant recipients, Medcalf and colleagues observed worse graft survival for both black (HR = 1.57; 95% CI, 1.10-2.24) and south Asian (HR = 1.39; 95% CI, 1.03-1.85) versus white recipients.9 Again, risk adjustment was limited and significant center variation was identified. National registry analyses have also shown that black versus white kidney transplant recipients have worse 1-year (HR = 1.7; 95% CI, 1.2-2.3; P < .001) and 5-year (HR = 1.5; 95% CI, 1.1-2.1; P = .02) death-censored graft survival in adjusted Cox regression models.15 Risk adjustment in the Cox regression model did not include donor age, cold ischemic times, or HLA matching, all important variables that affect graft survival. Finally, in a contemporary population cohort analyses by Pisavadia and colleagues of 27 790 kidney transplant recipients in the United Kingdom from 2003 and 2015, investigators demonstrated that recipient black ethnicity was an independent risk factor for deceased-donor graft loss (HR = 1.34; 95% CI, 1.04-1.74; P = .025), although the group did not have access to socioeconomic variables for the regression analysis.16 No increased risk was identified with living donors, similar to our subanalysis, but our small numbers suggest cautious interpretation. Importantly, ethnicity matching between the donor and recipient was not shown to improve kidney transplant outcomes regardless of donor type.16
Our work has the advantage of a single-center study, including access to raw patient-level data, but has the disadvantage of having a smaller study cohort, which may be underpowered to determine true differences. However, our results are important as they demonstrate that kidney transplant recipients of black ethnicity have significantly high-risk pretransplant profiles that contribute to their posttransplant risk profile. All of these factors render black recipients at increased risk for adverse graft-related outcomes after kidney transplant and many are interlinked in their relationship. Stephens and colleagues observed that residence in deprived areas was associated with increased risk for rejection and graft survival among 621 kidney transplant recipients from a single transplant center,17 although there were no specific data with regard to black ethnicity (6% of cohort was nonwhite). Socioeconomic factors and poor HLA matching have been linked to poor survival with deceased donor kidneys among black recipients in the United States,18 although the effect is diminished in the setting of living-donor kidney transplant.19 However, living donor rates for kidneys among patients of black ethnicity are suboptimal,20 and there is growing awareness of the link between ethnicity, socioeconomic deprivation, and poor rates for living kidney donors.21 Understanding socioeconomic confounders for patient and graft survival after kidney transplant is a significant unmet need in the context of clinical research and merits further work.
Our results are important because disparate outcomes for black kidney transplant recipients in the literature have been explained at both the macro-level (eg, healthcare system, universal health coverage, population) and micro-level (eg, biological or genetic determinants). Our significant finding is that, after adjustment for baseline differences, black ethnicity in itself is not a risk factor for graft loss and implies micro-level factors have negligible effects on graft survival. The concept that black ethnicity portends an increased risk for graft loss, independent of socioeconomic factors, stems from published data. For example, Chakkera and colleagues demonstrated similar increased risk for graft loss among African Americans in US-based healthcare systems, in which those with universal health access (through the US Department of Veteran Affairs) were compared with those who did not have universal health access.22 However, limitations included whether population cohorts within the US Department of Veterans Affairs reflect the general population and the confounding effects of variations in care. In addition, association does not reflect causality, and the underlying biological rationale for inferior outcomes for black recipients remains unclear. Black recipients have increased expression of CYP3A5,23 with variations in drug metabolism possibly leading to subtherapeutic immunosuppression dosing. However, Ng and colleagues observed black transplant recipients had worse long-term graft survival than other CYP3A5 expressers in their single-center study, with the latter not significant in a multivariate analysis.24 However, socioeconomic variables were not included in this analysis, which limits its interpretation. Kidneys donated by black individuals, both deceased and living, have been shown to be predictors of graft loss independent of recipient ethnicity in a population-cohort analysis,16 lending some support to a putative biologic risk associated with black ethnicity such as predisposed genetic risk.25 However, further work is warranted in this area before drawing definitive conclusions.
Regardless of any inherent risk, our data confirm that black kidney transplant recipients have more adverse graft outcomes (due to multifactorial factors) that warrant attenuation strategies. Living kidney donor rates remain low among black individuals, with some suggestion that black individuals are more likely to withdraw from the living-donor work-up than other ethnic groups.26 This consideration may be related to a delayed progress from referral to donation27 and needs attention. Improved HLA matching, by encouraging more organ donation from the minority ethnic community, will help improve graft outcomes for black recipients, and this remains a topical organ procurement issue concerning how we overcome such disengagement.28 Better access to kidney transplantation is acknowledged as key to helping black individuals gain the benefit of a kidney transplant procedure versus dialysis, with geographical variations in access to transplant not fully explained by patient-related factors in the United Kingdom.29 However, the French analyses suggested that socioeconomic inequalities did not influence access to kidney transplant.30 Finally, perhaps the biggest obstacle is an array of socioeconomic variables that are poorly understood but that affect kidney transplant outcomes. Further research to understand these factors are key to improving long-term outcomes. In the ATTOM study, which prospectively collected data from 6844 end-stage kidney failure patients across all renal units in the United Kingdom, detailed sociodemographic and clinical data have been compared versus outcomes; these results could provide clear evidence on ethnicity-stratified outcomes after kidney transplant.31
Our study has several limitations for data interpretation. Typical of epidemiologic analyses, it should be considered that there are likely to be confounders that affect patient and/or graft survival after kidney transplant that we were not able to include (eg, lifestyle factors, travel costs). Missing data (and misclassification bias) may have also affected our analyses, which is also an inherent bias in epidemiologic analyses. Our number of black kidney transplant recipients was relatively small, which is reflective of our local and regional demographics. Therefore, our study was likely to be underpowered, and of short duration, to robustly assess differences in some outcomes. Given the few graft losses between the 2 strata, multivariable analyses of retrospective data may not have captured the full spectrum of risk factors, as highlighted by the large 95% confidence intervals. Further maturation of the database may provide more definitive answers in the future. However, future larger prospective cohort studies could also provide more definitive answers for this important issue.
Our study demonstrated that black kidney transplant recipients have increased risk for posttransplant complications, including prolonged length of stay postoperatively, delayed graft function, 1-year rejection, 1-year creatinine level, and risk for graft loss. However, after adjustment for baseline factors (including demographics, immunologic status, and socioeconomic deprivation), black ethnicity itself was not a risk factor for graft loss after kidney transplant. Our study suggests that kidney transplant recipients of black ethnicity warrant additional support and heightened surveillance strategies posttransplant due to their increased risk for adverse outcomes, although black ethnicity itself does not increase risk. In the context of conflicting reports for ethnicity and posttransplant survival, further research is warranted to help distinguish outcomes stratified by ethnicity versus confounding variables and to reinforce the lack of generalizability of transplant epidemiology between countries. Although an understanding of disparities for kidney transplant outcomes between ethnicities remains important, long-term survival of black patients with end-stage kidney failure is likely to remain superior with kidney transplant versus dialysis, and our data should reassure black patients of their excellent outcomes posttransplant despite increased presence of baseline risk factors.
Volume : 16
Issue : 6
Pages : 682 - 689
DOI : 10.6002/ect.2018.0241
From the 1Department of Nephrology and Transplantation, Queen Elizabeth
Hospital, Edgbaston, the 2University of Birmingham, and the 3Institute of
Immunology and Immunotherapy, University of Birmingham, Birmingham, United
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare. This work was presented as an abstract at the UK Kidney Week in Harrogate, UK (June 2018).
Corresponding author: Adnan Sharif, Department of Nephrology and Transplantation, Queen Elizabeth Hospital, Edgbaston, Birmingham, B15 2WB, UK
Phone: +44 121 371 5861
Table 1. Baseline Demographics of Kidney Allograft Recipients Stratified by Black Versus White Ethnicity
Table 2. Postoperative Complications Stratified by Black Versus White Ethnicity
Figure 1. Unadjusted Kaplan-Meir Plots of Patient (a) and Graft (b) Survival After Kidney Transplant Stratified By Black Versus White Ethnicity