Discordance in GFR Estimating Equations and Dosing Guidance by Body Mass Index and Age.

Different glomerular filtration rate (GFR) estimating equations are used in drug development and clinical practice. Traditionally, United States Food and Drug Administration (FDA) dosing recommendations for drugs cleared by the kidneys were based on the Cockcroft-Gault equation,1Cockcroft D.W. Gault H. Prediction of creatinine clearance from serum creatinine.Nephron. 1976; 16: 31-41https://doi.org/10.1159/000180580Crossref PubMed Scopus (13209) Google Scholar which is less accurate than the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation.2Myers G.L. Miller W.G. Coresh J. et al.Recommendations for improving serum creatinine measurement: a report from the Laboratory Working Group of the National Kidney Disease Education Program.Clin Chem. 2006; 52: 5-18https://doi.org/10.1373/clinchem.2005.0525144Crossref PubMed Scopus (1016) Google Scholar,3Delgado 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: 268-288.e1https://doi.org/10.1053/j.ajkd.2021.08.003Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar The CKD-EPI equation is indexed to body surface area to facilitate comparison across individuals of different body size. However, because kidney clearance of drugs within individuals correlates better with nonindexed estimated GFR (eGFR) than indexed eGFR, nonindexed eGFR is recommended to guide optimal drug dosing.3Delgado 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: 268-288.e1https://doi.org/10.1053/j.ajkd.2021.08.003Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar,4Food and Drug AdministrationPharmacokinetics in Patients with Impaired Renal Function — Study Design, Data Analysis, and Impact on Dosing and Labeling.https://www.fda.gov/regulatory-information/search-fda-guidance-documents/pharmacokinetics-patients-impaired-renal-function-study-design-data-analysis-and-impact-dosing-andGoogle Scholar We evaluated when these GFR estimates differ, and how they may compel differences in drug dosing, using a population with atrial fibrillation and rivaroxaban as an example drug. We included 30,261 patients from Geisinger Health System with a diagnosis of atrial fibrillation (Table S1) and available serum creatinine, weight, and height measurements (Fig S1, Item S1). GFR was estimated using estimated creatinine clearance (eClCr) by the Cockcroft-Gault equation (mL/min, actual body weight),1Cockcroft D.W. Gault H. Prediction of creatinine clearance from serum creatinine.Nephron. 1976; 16: 31-41https://doi.org/10.1159/000180580Crossref PubMed Scopus (13209) Google Scholar indexed eGFR by the CKD-EPI (2021) equation (mL/min/1.73 m2),5Inker 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 (475) Google Scholar and nonindexed eGFR (mL/min) by accounting for individual’s body surface area (Table S2) and compared by body mass index (BMI) and age groups. The recommended dose of rivaroxaban based on the 3 estimates was evaluated in patients with indexed eGFR 15-60 mL/min/1.73 m2. In secondary analysis, we also evaluated eClCr using adjusted body weight.6Park E.J. Pai M.P. Dong T. et al.The influence of body size descriptors on the estimation of kidney function in normal weight, overweight, obese, and morbidly obese adults.Ann Pharmacother. 2012; 46: 317-328https://doi.org/10.1345/aph.1Q374Crossref PubMed Scopus (31) Google Scholar Mean age of the cohort was 76.4 (SD, 12.4) years, 45.5% were female, and mean BMI was 29.0 (SD, 7.5) kg/m2 (Table S3). Patients with higher BMI category had higher eClCr and nonindexed eGFR, with more substantial increase for eClCr (P for trend <0.001 for both). The distribution of eClCr, indexed eGFR, and nonindexed eGFR differed by BMI (Fig 1) and age (Fig S2). Different GFR estimates had good agreement with BMI 25-39.9 kg/m2 and age 65-84 years. In contrast, among patients with low BMI, eClCr was lower than nonindexed eGFR, which was lower than indexed eGFR. Among patients with high BMI, mean eClCr was higher than indexed and nonindexed eGFR (110.0 mL/min vs 65.3 mL/min/1.73 m2 vs 86.9 mL/min, respectively). In patients aged ≥85 years, eClCr was lower than indexed and nonindexed eGFR (38.5 mL/min vs 51.6 mL/min/1.73 m2 vs 51.9 mL/min, respectively). There were large differences in recommended rivaroxaban dosing. Among 13,015 patients with eGFR 15-60 mL/min/1.73 m2, people with low BMI often required a standard dose of rivaroxaban by indexed and nonindexed eGFR and a reduced dose by eClCr (33%-46% by indexed eGFR and 5%-17% by nonindexed eGFR for BMI <18.5 kg/m2, depending on age; Fig 2). In contrast, people with high BMI often required a reduced dose of rivaroxaban by indexed and nonindexed eGFR and a standard dose by eClCr (46%-62% by indexed eGFR and 12%-21% by nonindexed eGFR for BMI ≥40 kg/m2). Patients with low BMI often required a standard dose of rivaroxaban by indexed eGFR and a reduced dose by nonindexed eGFR; conversely, high BMI required reduced dose by indexed and standard dose by nonindexed eGFR. The distribution of eClCr using adjusted body weight was more similar than eClCr using actual body weight when compared with indexed and nonindexed eGFR (Fig S3), but substantial differences in recommended rivaroxaban dosing remained (Fig S4). As demonstrated by our data and prior studies,7Andrade J.G. Hawkins N.M. Fordyce C.B. et al.Variability in non–vitamin K antagonist oral anticoagulants dose adjustment in atrial fibrillation patients with renal dysfunction: the influence of renal function estimation formulae.Can J Cardiol. 2018; 34: 1010-1018https://doi.org/10.1016/j.cjca.2018.04.019Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar,8Delanaye P. Guerber F. Scheen A. et al.Discrepancies between the Cockcroft-Gault and Chronic Kidney Disease Epidemiology (CKD-EPI) equations: implications for refining drug dosage adjustment strategies.Clin Pharmacokinet. 2017; 56: 193-205https://doi.org/10.1007/s40262-016-0434-zCrossref PubMed Scopus (23) Google Scholar the decision to use eClCr or eGFR, and whether to use indexed or nonindexed eGFR, affects dosing recommendations, particularly among people with higher BMI and older age—a subset of the population that is growing. Determining the optimal dose is critical, as inappropriate dosing of rivaroxaban may lead to increased mortality.9Camm A.J. Cools F. Virdone S. et al.Mortality in patients with atrial fibrillation receiving nonrecommended doses of direct oral anticoagulants.J Am Coll Cardiol. 2020; 76: 1425-1436https://doi.org/10.1016/j.jacc.2020.07.045Crossref PubMed Scopus (63) Google Scholar Large-scale outcome studies could inform these decisions as to which method better differentiates benefit and risk of adverse drug events, but are not likely to occur for all medications and clinical settings. Consistent with past recommendations,3Delgado 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: 268-288.e1https://doi.org/10.1053/j.ajkd.2021.08.003Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar,4Food and Drug AdministrationPharmacokinetics in Patients with Impaired Renal Function — Study Design, Data Analysis, and Impact on Dosing and Labeling.https://www.fda.gov/regulatory-information/search-fda-guidance-documents/pharmacokinetics-patients-impaired-renal-function-study-design-data-analysis-and-impact-dosing-andGoogle Scholar use of the CKD-EPI equation, the more accurate and widely used GFR estimating equation, may facilitate harmonization of GFR estimates between drug labeling and clinical practice. Draft industry guidance from the FDA now recommends use of the CKD-EPI equation for pharmacokinetic studies.4Food and Drug AdministrationPharmacokinetics in Patients with Impaired Renal Function — Study Design, Data Analysis, and Impact on Dosing and Labeling.https://www.fda.gov/regulatory-information/search-fda-guidance-documents/pharmacokinetics-patients-impaired-renal-function-study-design-data-analysis-and-impact-dosing-andGoogle Scholar Indeed, new drugs, such as dapagliflozin, have incorporated indexed eGFR in pharmacokinetic studies and include eGFR-based dosing recommendations in the FDA-approved label.10Food and Drug Administration. FARXIGA (dapagliflozin) tablets. 2014. https://www.accessdata.fda.gov/drugsatfda_docs/label/2014/202293s003lbl.pdf. [Accessed 30 March 2022].Google Scholar The strengths of our study include a systematic assessment of GFR distributions and dosing discordance by 3 estimating equations in a large real-world patient population. However, we did not examine actual drug clearance or drug safety and efficacy, and our results may not be fully generalizable to other populations. In conclusion, among patients with atrial fibrillation, eClCr, indexed eGFR, and nonindexed eGFR differ by body size and age, especially among people with very high BMI and older age, resulting in substantial drug dosing discordance. Using the most accurate method to assess GFR in individual patients may be a preferred approach. Research idea and study design: BL, LAI, JC, MEG, J-IS; data acquisition: ARC, JC, MEG, J-IS; data analysis: BL, YX, J-IS; data interpretation: BL, LAI, TDN, JC, MEG, J-IS. 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. Dr Lyu was supported by grant R01DK115534; Dr Grams, by grants R01DK115534 and K24HL155861 (both as Principal Investigator); Dr Shin, by grant K01DK121825 (as Principal Investigator); Dr Inker, by grant R01DK115534; all grants from the National Institute of Diabetes and Digestive and Kidney Disease. The funders of this study did not have any role in study design; collection, analysis, and interpretation of data; writing the report; and the decision to submit the report for publication. The authors declare that they have no relevant financial interests. Received August 21, 2022. Evaluated by 2 external peer reviewers, with direct editorial input from a Statistics/Methods Editor, an Associate Editor, and a Deputy Editor who served as Acting Editor-in-Chief. Accepted in revised form January 26, 2023. The involvement of an Acting Editor-in-Chief was to comply with AJKD’s procedures for potential conflicts of interest for editors, described in the Information for Authors & Journal Policies. Download .pdf (1.44 MB) Help with pdf files Supplementary File (PDF)Figures S1-S4, Item S1, Tables S1-S3.