A Prospective Cross-Sectional Study on the Performance of the 2021 CKD-EPI Equations Without Race in a Multiracial Population of Adults With Solid Tumors in Brazil.

Accurate glomerular filtration rate (GFR) evaluation is important in patients with cancer to determine correct dosing of chemotherapy and eligibility for treatments and clinical trials. Estimated GFR (eGFR) based on serum creatinine (Scr) (eGFRcr) using the 2009 Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation is recommended as the “initial test” for GFR evaluation for routine clinical practice in adults, and eGFR based on serum cystatin C (Scys), with or without Scr (eGFRcr-cys or eGFRcys, respectively) using the 2012 CKD-EPI equations or clearance measurements are recommended as confirmatory tests, as appropriate for the clinical setting.1Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work GroupKDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease.Kidney Int Suppl. 2013; 3: 1-150https://doi.org/10.1038/kisup.2012.74Abstract Full Text Full Text PDF Scopus (0) Google Scholar Non-GFR determinants for Scr (sarcopenia) and Scys (smoking, inflammation, and use of glucocorticoids) might be more frequent in cancer patients. Using the CKD-EPI equations, we recently demonstrated that eGFRcr overestimated measured GFR (mGFR) in a multiracial cohort of patients with solid tumors in Brazil,2Costa e Silva V.T. Gil Jr., L.A. Inker L.A. et al.A prospective cross-sectional study estimated glomerular filtration rate from creatinine and cystatin C in adults with solid tumors.Kidney Int. 2022; 101: 607-614https://doi.org/10.1016/j.kint.2021.12.010Abstract Full Text Full Text PDF PubMed Scopus (8) Google Scholar not in active treatment, but was more accurate than using the Cockcroft-Gault (CG) equation2Costa e Silva V.T. Gil Jr., L.A. Inker L.A. et al.A prospective cross-sectional study estimated glomerular filtration rate from creatinine and cystatin C in adults with solid tumors.Kidney Int. 2022; 101: 607-614https://doi.org/10.1016/j.kint.2021.12.010Abstract Full Text Full Text PDF PubMed Scopus (8) Google Scholar; that eGFRcys underestimated mGFR; and that the eGFRcr-cys equation was the most accurate equation. The 2009 eGFRcr equation3Levey A.S. Stevens L.A. Schimd C.H. et al.A new equation to estimate glomerular filtration rate.Ann Intern Med. 2009; 150: 604-613https://doi.org/10.7326/0003-4819-150-9-200905050-00006Crossref PubMed Scopus (17079) Google Scholar and 2012 eGFRcr-cys equation,4Inker L.A. Schmid C.H. Tighiouart H. et al.Estimating glomerular filtration rate from serum creatinine and cystatin C.N Engl J Med. 2012; 367: 20-29https://doi.org/10.1056/NEJMoa1114248Crossref PubMed Scopus (2676) Google Scholar but not the 2012 eGFRcys equation,4Inker L.A. Schmid C.H. Tighiouart H. et al.Estimating glomerular filtration rate from serum creatinine and cystatin C.N Engl J Med. 2012; 367: 20-29https://doi.org/10.1056/NEJMoa1114248Crossref PubMed Scopus (2676) Google Scholar include a term for race (Black vs non-Black). Nephrology societies in the United States now recommend elimination of race from eGFR equations and use of 2021 CKD-EPI equations for eGFRcr and eGFRcr-cys without a race coefficient.5Delgado C. Baweja M. Crews D.C. et al.A unifying approach for GFR estimation: recommendations of the NKF-ASN Task Force on Reassessing the Inclusion of Race in Diagnosing Kidney Disease.Am J Kidney Dis. 2022; 79: 268-288https://doi.org/10.1053/j.ajkd.2021.08.003Abstract Full Text Full Text PDF PubMed Scopus (170) Google Scholar,6Inker L.A. Eneanya N.D. Coresh J. et al.New creatinine- and cystatin C-based equations to estimate GFR without race.N Engl J Med. 2021; 385: 1737-1749https://doi.org/10.1056/NEJMoa2102953Crossref PubMed Scopus (611) Google Scholar The race-free 2021 CKD-EPI equations lead to lower eGFR values in Black subjects and higher values in non-Black subjects than the 2009 and 2012 equations.6Inker L.A. Eneanya N.D. Coresh J. et al.New creatinine- and cystatin C-based equations to estimate GFR without race.N Engl J Med. 2021; 385: 1737-1749https://doi.org/10.1056/NEJMoa2102953Crossref PubMed Scopus (611) Google Scholar The 2021 equations have not been assessed in cancer patients. In addition, GFR estimating equations without race might be particularly relevant in countries with multiracial populations, such as Brazil. Here we evaluate the performance of the 2021 CKD-EPI eGFRcr and eGFRcr-cys equations in the same cohort. A complete description of methods is reported elsewhere2Costa e Silva V.T. Gil Jr., L.A. Inker L.A. et al.A prospective cross-sectional study estimated glomerular filtration rate from creatinine and cystatin C in adults with solid tumors.Kidney Int. 2022; 101: 607-614https://doi.org/10.1016/j.kint.2021.12.010Abstract Full Text Full Text PDF PubMed Scopus (8) Google Scholar and is summarized in Item S1. The study population was composed of 1,200 patients. Mean (SD) age was 58.8 ± 13.2 years and mGFR was 78.5 ± 21.7 mL/min/1.73 m2; 50.9% of patients were male. Race distribution was Black participants 12.8% and non-Black participants 81.2% (including White 69%, Asian 2.2%, and Mixed 16%) (Table S1). For eGFRcr, the overestimation of mGFR was larger for the 2021 versus 2009 equation (median of −10.0 vs −8.1 mL/min/1.73 m2), resulting in lesser accuracy (1-P30 of 22.3% vs 19.1%) (Table 1). Nonetheless, the 2021 equation was more accurate than the CG equation (1-P30 of 22.3% vs 24.9%, P = 0.05). For eGFRcr-cys, the overestimation of mGFR was larger for the 2021 versus 2012 equation (median of −4.1 vs −2.0 mL/min/1.73 m2). The 2021 eGFRcr-cys equation was more accurate than the 2021 eGFRcr equation and the 2012 eGFRcys equation (1-P30 of 9.8% vs 22.3% and 12.3%, P values <0.001 and 0.01, respectively). Findings were similar in subgroups defined by age, sex, and clinical stage; as expected, overestimation of mGFR by 2021 eGFRcr and eGFRcr-cys equations was less for Black than non-Black participants (Fig 1 and Tables S2-S5). These results support implementation of the 2021 CKD-EPI eGFRcr equation for patients with solid tumors in Brazil rather than the CG equation as the initial test in GFR evaluation, and for greater implementation of Scys testing.7Miller W.G. Kaufman H.W. Levey A.S. et al.National Kidney Foundation Laboratory Engagement Working Group recommendations for implementing the CKD-EPI 2021 race-free equations for estimated glomerular filtration rate: practical guidance for clinical laboratories.Clin Chem. 2022; 68: 511-520https://doi.org/10.1093/clinchem/hvab278Crossref PubMed Scopus (30) Google ScholarTable 1Performance of Glomerular Filtration Rate Estimating EquationsFiltration Marker (eGFR)EquationUse of RaceBias (Median)(mL/min/1.73 m2)Precision (IQR)(mL/min/1.73 m2)Accuracy (1-P30)(%)Accuracy (RMSE)Creatinine (eGFRcr)CG 1976No−8.1 (−9.4 to −6.7)24.2 (22.4-25.8)24.9 (22.4-27.3)0.239 (0.229-0.249)Creatinine (eGFRcr)CKD-EPI 2009Yes−8.1 (−8.9 to −7.1)18.4 (17.1-19.6)19.1 (16.8-21.2)0.206 (0.197-0.215)Creatinine (eGFRcr)CKD-EPI 2021No−10.0 (−10.7 to −8.8)18.2 (17.0-19.6)22.3 (19.8-24.4)aP=0.05 compared to CG equation.0.220 (0.211-0.230)Cystatin C (eGFRcys)CKD-EPI 2012No4.6 (3.7-5.5)17.5 (16.3-19.2)12.3 (10.3-14.3)0.215 (0.204-0.225)Creatinine–cystatin C (eGFRcr-cys)CKD-EPI 2012Yes−2.0 (−2.6 to −1.1)15.9 (14.7-16.8)7.8 (6.3-9.4)0.165 (0.157-0.172)Creatinine–cystatin C (eGFRcr-cys)CKD-EPI 2021No−4.1(−4.8 to −3.3)15.7 (14.6-17.1)9.8 (8.0-11.4)bP<0.001 compared to CKD-EPI 2021 eGFRcr equation and P<0.01 compared to the 2012 eGFRcys equation.0.171 (0.163-0.179)Data are presented with 95% confidence intervals. Bias was calculated as the median value of (mGFR − eGFR); positive and negative values represent an underestimate and overestimate of mGFR, respectively. Precision is the IQR of the difference between mGFR and eGFR. Accuracy, which reflects absence of bias and precision, is defined as the percentage of estimates that differed by more than 30% from the measured GFR (1-P30) and the root mean squared error (RMSE) for the regression of log mGFR on log eGFR. For bias, IQR, 1-P30, and RMSE, values closer to zero indicate better performance. Selected values for 1-P30 were compared using the McNemar test for paired samples. Abbreviations: CG, Cockcroft-Gault, indexed for body surface area; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; eGFR, estimated glomerular filtration rate; mGFR, measured glomerular filtration rate; IQR = interquartile range.a P = 0.05 compared to CG equation.b P < 0.001 compared to CKD-EPI 2021 eGFRcr equation and P < 0.01 compared to the 2012 eGFRcys equation. Open table in a new tab Data are presented with 95% confidence intervals. Bias was calculated as the median value of (mGFR − eGFR); positive and negative values represent an underestimate and overestimate of mGFR, respectively. Precision is the IQR of the difference between mGFR and eGFR. Accuracy, which reflects absence of bias and precision, is defined as the percentage of estimates that differed by more than 30% from the measured GFR (1-P30) and the root mean squared error (RMSE) for the regression of log mGFR on log eGFR. For bias, IQR, 1-P30, and RMSE, values closer to zero indicate better performance. Selected values for 1-P30 were compared using the McNemar test for paired samples. Abbreviations: CG, Cockcroft-Gault, indexed for body surface area; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; eGFR, estimated glomerular filtration rate; mGFR, measured glomerular filtration rate; IQR = interquartile range. Removing race from eGFR equations represents an important advance in GFR evaluation. Although race correlates with ancestry, a biological variable, race varies with numerous social determinants of health, which may affect the relationship between Scr and GFR and lead to heterogeneity in the non-GFR determinants of Scr.8Levey A.S. Titan S.M. Powe N.R. et al.Kidney disease, race, and GFR estimation.Clin J Am Soc Nephrol. 2020; 15: 1203-1212https://doi.org/10.2215/CJN.12791019Crossref PubMed Scopus (124) Google Scholar Self-reported race correlates imperfectly with genomically determined ancestry9Banda Y. Kvale M.N. Hoffmann T.J. et al.Characterizing race/ethnicity and genetic ancestry for 100,000 subjects in the Genetic Epidemiology Research on Adult Health and Aging (GERA) Cohort.Genetics. 2015; 200: 1285-1295https://doi.org/10.1534/genetics.115.178616Crossref PubMed Scopus (206) Google Scholar; and in mixed-race populations such as in Brazil, race ascertainment has been difficult to establish and is probably influenced by social, environmental, and cultural background, which play a higher role when race barriers are not so clearly established. For these reasons, removing race from eGFR equations may represent an advance in cancer care in Brazil by eliminating a source of bias. Strengths of our study include a mixed and multiracial study population, in which race ascertainment has been difficult to establish and which does not differ substantially from the general population in Sao Paulo; a prospective design; using a reference method for mGFR; and standardized assays for Scr and Scys. Limitations include conduct in a single-center study, lack of repeat evaluations after beginning treatment, and lack of comparability to the general population of Brazil. We conclude that the 2021 CKD-EPI equations without race can be incorporated in cancer care in Brazil. Although 2021 eGFRcr was slightly less accurate compared to 2009 eGFRcr and 2012 eGFRcr-cys, the difference was not substantial and could be acceptable in the framework of current understanding of use of race and ethnicity in medical decision making. It is important for oncologists and nephrologists in cancer centers to engage with clinical laboratory pathologists to discuss implementation of these new equations in clinical practice and to provide greater availability of Scys. Research idea and study design: VTCS, LAI, ASL, EAB; data acquisition: VTCS, LAG, RAC, EC, GCF, MTS, MDPED, GC, LA, LM; data analysis/interpretation: VTCS, PM, LAI, ASL, EAB; statistical analysis: VTCS, HT, DMTZ, SM; supervision or mentorship: VTCS, LAI, ASL, EAB; ASL and EAB contributed equally to this work. Each author contributed important intellectual content during manuscript drafting or revision and agrees to be personally accountable for the individual’s own contributions and to ensure that questions pertaining to the accuracy or integrity of any portion of the work, even one in which the author was not directly involved, are appropriately investigated and resolved, including with documentation in the literature if appropriate. This research has been funded by FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo, Project number 2014/19286-4), a local public research agency, and was undertaken by the first author as part of the Master’s Program in Translational Science from Tufts University. Funders had no role in the study design, data collection, data analysis, data reporting, or the decision to submit this manuscript for publication. Dr Levey and Dr Inker report grant support from the National Institute of Diabetes, Digestive and Kidney Disease (R01DK116790-01A1, R01DK097020). The remaining authors declare that they have no relevant financial interests. Received October 7, 2022. Evaluated by 4 external peer reviewers and a statistician, with editorial input from an Acting Editor-in-Chief (Editorial Board Member Shuchi Anand, MD, MS). Accepted in revised form January 6, 2023. The involvement of an Acting Editor-in-Chief to handle the peer-review and decision-making processes was to comply with AJKD’s procedures for potential conflicts of interest for editors, described in the Information for Authors & Journal Policies. Download .pdf (.36 MB) Help with pdf files Supplementary File (PDF)Item S1, Tables S1-S5.