Race-Based eGFR Assessment for Kidney Transplantation

The search for a reliable strategy to easily measure kidney function has been likened to the search for the holy grail in nephrology because the interpretation of the measured blood creatinine requires an understanding of the nonlinear and reciprocal relationship between renal function and creatinine. The Modification of Diet in Renal Disease Study and later the Chronic Kidney Disease Epidemiology Collaboration equations offered a simple strategy to translate the blood creatinine into an eGFR on the basis of readily available variables. These equations appear to offer more accuracy and precision on a population level compared with estimates of creatinine clearance. Because the eGFR can readily be included on the laboratory report, these eGFR equations were promptly endorsed by the nephrology community, first in guidelines and then in practice. Both equations feature a race coefficient that yields better kidney function at any given creatinine for patients who are Black. This practice was widely accepted, although there were no criteria for employing the coefficient and attempts to explain the need for this difference were deficient.

Despite the fact that the eGFR is simply an estimate and still subject to factors that affect the blood creatinine, such as volume, diet, and medications, the eGFR forms the framework for CKD stages and is commonly used to dictate care. Indeed, some transplant centers use this strategy to screen potential kidney donors, and most use them to register patients for kidney transplant when the eGFR is ≤20 ml/min per 1.73 m². The report by Doshi et al. (1) in this issue of CJASN, is, at first glance, a simple survey of practicing US transplant providers, yet this manuscript demonstrates the recent groundswell in the opinions of the nephrology community regarding the use of the Black race coefficient in the eGFR equations. Over the past year, with uprisings for Black Lives Matter, there has been a reckoning over the use of race in medical practice; this survey is testimony to that reckoning.

Doshi and a team of investigators queried kidney transplant providers, predominantly transplant nephrologists, regarding the Black race coefficient and their current practice. The respondents represent centers that provide care for 70% of the transplant volume in the United States and all of the United Network for Organ Sharing (UNOS) regions. The overwhelming majority of respondents in this survey felt the current equations that utilize race are inappropriate. Although such a survey has never before been undertaken and respondents were not asked if their opinions on this issue have changed, this report demonstrates how this issue has become an important part of the national conversation.

Despite the almost uniformity in opinion regarding the need for change and potential harm of the Black race coefficient, on further questioning, more than half of respondents thought the race variable yielded an improvement in the precision of GFR estimation. Many transplant nephrologists believe that our Black patients’ kidney function can be better understood by viewing it on the basis of race, despite the known inaccuracies of this approach. Interestingly, the survey revealed several assumptions that dropping the Black race variable could lead either to “no benefit” or “some harm.” In terms of no benefit, 16% thought dropping the race variable would have no clinical benefit to patients, despite multiple recent reports that calculate the percentage of Black candidates who could gain predialysis time on the list toward a deceased donor kidney transplant. In terms of “some harm,” almost one third thought removing the race variable would lead to premature initiation of dialysis, although initiation of dialysis should be on the basis of symptomatology rather than a particular creatinine or eGFR calculation.

Nearly half of survey respondents reported concern that removal of race-based eGFR equations in donor screening might lead to an underestimate of kidney function in Black people who are potential kidney donors, exclude these individuals from donation, and thus lead to harm to their potential recipients. This level of concern is likely on the basis of an over-reliance on the eGFR in the living donor evaluation, although all equations perform poorly at higher GFR. The inverse relationship between eGFR and creatinine means that significant variation in the eGFR can result from small changes in creatinine because of both laboratory and patient factors. Instead, another way to view this finding is that removal of the race-based eGFR can protect potential donors from harm by avoiding overestimation of the eGFR. Reports of poor correlation between eGFR and measured GFR in Black potential donors and the finding of a greater decline in eGFR after donation by Black patients should refute the notion that the race-based eGFR is helpful in this setting (2,3). UNOS policy 14.4.B requires that potential kidney donor function be assessed via “evaluation of glomerular filtration rates by isotopic methods or from a 24-hour urine collection” (4). Thus, the race variable and eGFR equations should not be part of the evaluation of potential kidney donors regardless of race.

Taken together, these findings reveal that a significant proportion of respondents hold the idea that the Black race variable adds accuracy to eGFR estimates when applied broadly to people identified as African American. This ongoing assumption is as flawed as the notion that there are distinct human races. Differences in individuals more commonly reflect differences in environment and other external factors rather than differences in core biology evident by skin color. Furthermore, the race-based eGFR has never accounted for individuals who self-identify as mixed race, a growing proportion of our population. Indeed, the 2020 US Census data reveal an increase in nearly 300% of individuals who self-identify with more than one race or ethnicity (5). Ignoring people who identify as multiracial in these equations is an example of the ongoing influence of poorly substantiated assumptions about the biologic nature of race on our scientific and medical reasoning (6).

Investigators also provide a snapshot of the practice patterns of transplant centers with respect to the use of the eGFR and found that although only a small minority of centers had already dropped the Black race coefficient, nearly half were considering this action. In 2017, our institution removed race from the eGFR report, in favor of providing a range of values and with the caveat that several variables can affect the eGFR. This has positively affected our use of creatinine in achieving a qualifying eGFR for transplant listing wait time in a small number of Black candidates who are evaluated predialysis (7).

An additional and often overlooked outcome of applying a Black race variable to eGFR calculation relates to UNOS policy 3.6.B.i, which delineates rules for waiting time reinstatement: this can occur if the kidney is removed or has failed in the first 90 days after transplant. Kidney graft failure definition includes GFR ≤20 ml/min. The disparity in access to a second transplant when a first has failed for Black transplant recipients, as compared with a recipient of any other race with the same poor function and the same serum creatinine, is hard to justify. Any 50-year-old patient with CKD stage 4 on the basis of serum creatinine of 3.4 mg/dl at 90 days post-transplant will get all their wait time back, unless they are Black. Equity in access to wait time reclamation would be supported by using a race neutral means of assessing GFR for purposes of reinstatement of wait time.

A strong defense of the use of the Black race variable in eGFR in any circumstance requires that we accept the underlying assumption that there are fundamentally two types of people: Black people and everyone else. Although this assumption fails as overly simplistic, it has been challenging to expunge; this assumption has been taught in higher education, reinforced during medical training, and actively used for patient-level decisions regarding diagnosis and management. It is time to more than question. Even if we accept the above premise as simplistic and potentially harmful when applied to individual patients, when pushed further, it is clear from the report by Doshi et al. that the assumption lingers in our collective thinking. The National Kidney Foundation and the American Society of Nephrology Task Force on Reassessing the Inclusion of Race in Diagnosing Kidney Diseases recently recommended immediate implementation of the CKD-EPI equation refit without the race variable (8), and the Organ Procurement and Transplantation Network Minority Affairs and Kidney Transplant committees are requesting public comment on the use of race in eGFR calculation (9). We recommend all transplant programs use the new equations without race for all transplant listing GFR determinations. These steps will be the first in the elimination of race in medical decision-making tools, but they will not be enough. Disparities in access to kidney transplantation precede the use of the race-based estimating equations for GFR and are likely to postdate it (10,11). Next, we must actively work to eliminate the disparities in health equity that have finally garnered the attention they deserve.

Disclosures

M.P. Hoenig reports receiving honoraria from the Pri-Med conference. M. Pavlakis reports consultancy agreements with Transplant Solutions; reports receiving research funding from Trugraf Genomics and CareDx; reports serving as a scientific advisor or member of the Merck Global Advisory Board and Moderna Study Safety Review Committee; and serving as a content writer for EBSCO Industries Inc, serving on the New England Board of Directors for the Leukemia and Lymphoma Society, and serving as the Organ Procurement and Transplantation Network Kidney Transplantation Committee Chair.

Funding

None.