Performance of Indexed and Nonindexed Estimated GFR
Silvia Titan,Shiyuan Miao,Hocine Tighiouart,Nan Chen,Hao Shi,Luxia Zhang,Zuo Li,Marc Froissart,Peter Rossing,Anders Grubb,Li Fan,Michael Mauer,Omran Bakoush,Christina Wyatt,Michael G Shlipak,Tariq Shafi,Lesley A Inker,Andrew S Levey
DOI: https://doi.org/10.1053/j.ajkd.2020.04.010
2020-01-01
Abstract:Glomerular filtration rate (GFR) is traditionally indexed to body surface area (BSA) to allow comparison among individuals of different body size, and normative values for GFR and the definitions of acute and chronic kidney disease are based on BSA-indexed GFR.1Kidney Disease: Improving Global Outcomes (KDIGO)KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease.Kidney Int Suppl. 2013; 3: 1-150Abstract Full Text Full Text PDF Scopus (1249) Google Scholar,2Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work GroupKDIGO clinical practice guideline for acute kidney injury.Kidney Int. Suppl. 2012; 2: 1-138Abstract Full Text Full Text PDF Scopus (1608) Google Scholar Modern GFR estimating equations were developed using measured GFR (mGFR) indexed for each participant’s BSA and report estimated GFR (eGFR) indexed to 1.73 m2.3Inker 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: 20-29Crossref PubMed Scopus (2086) Google Scholar,4Levey A.S. Stevens L.A. Schmid C.H. et al.A new equation to estimate glomerular filtration rate.Ann Intern Med. 2009; 150: 604-612Crossref PubMed Scopus (13355) Google Scholar However, some clinical settings may favor using nonindexed instead of indexed GFR; for example, for dosing of drugs eliminated by the kidney when BSA is very different from the index value of 1.73 m2 (eg, with obesity or anorexia nervosa), or for evaluation of change in GFR after a large change in BSA (eg, after bariatric surgery). Converting from indexed to nonindexed eGFR can be done by multiplying the indexed values by the patient’s BSA and dividing by 1.73 m2. However, the performance of nonindexed eGFR as an estimate of nonindexed mGFR has not been carefully evaluated. Our objective was to compare the performance of indexed and nonindexed eGFR from creatinine and cystatin C, overall and by BSA strata, in 3,506 participants from previously reported studies comprising a diverse population with a broad range of GFRs and BSAs. Data from 9 studies from the United States, China, and Europe with mGFR using exogenous filtration markers and serum concentrations of creatinine and cystatin C were analyzed, separately and pooled. Detailed methods for each study have been published elsewhere (Item S1; Table S1). The Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations3Inker 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: 20-29Crossref PubMed Scopus (2086) Google Scholar,4Levey A.S. Stevens L.A. Schmid C.H. et al.A new equation to estimate glomerular filtration rate.Ann Intern Med. 2009; 150: 604-612Crossref PubMed Scopus (13355) Google Scholar were used for calculating eGFRs using creatinine (eGFRcr), cystatin C (eGFRcys), and both markers (eGFRcr-cys) indexed for 1.73 m2 calculated using the Du Bois and Du Bois formula5Du Bois D. Du Bois E.F. A formula to estimate the approximate surface area if height and weight be known. 1916.Nutrition. 1989; 5 (discussion 312-303): 303-311PubMed Google Scholar (indexed eGFR). Nonindexed eGFRs were calculated as described in the previous paragraph; additional methods are in Item S1. All studies received ethics approval. Tables 1 and S2 show characteristics of the study population, overall and according to BSA quintiles. Mean age was 57 ± 19 years, 55% were men, 27% had diabetes, 8% self-reported as black, and mean mGFR was 68 ± 32 mL/min/1.73 m2. Characteristics differed by BSA quintile.Table 1Baseline Characteristics Among Participants of the 9 Studies, Overall and by BSA QuintileAllBSA QuintileP1 (≤1.65 m2)2 (1.66-1.77 m2)3 (1.78-1.89 m2)4 (1.90 - 2.02 m2)5 (>2.02 m2)No. of participants3,506702701701701701BSA, m21.84 (0.22)1.54 (0.09)1.71 (0.03)1.83 (0.03)1.95 (0.04)2.14 (0.11)Age, y57 (19)53 (20)59 (19)58 (19)57 (20)59 (18)<0.001Male sex1,914 (54.6%)107 (15.2%)256 (36.5%)405 (57.8%)533 (76.0%)613 (87.5%)<0.001Diabetes935 (26.7%)130 (18.5%)169 (24.1%)181 (25.9%)227 (32.4%)228 (32.5%)<0.001Black race273 (7.8%)30 (4.3%)40 (5.7%)44 (6.3%)64 (9.1%)95 (13.6%)<0.001Body mass index, kg/m225.9 (4.8)21.7 (3.2)24.4 (3.0)25.8 (3.7)27 (3.9)30 (5.0)<0.001Height, m168.4 (9.3)158.8 (5.8)164.0 (5.9)168.3 (6.4)173 (6.9)178 (7.2)<0.001Weight, kg73.6 (15.6)54.3 (5.9)65.2 (4.1)72.4 (5.0)81 (5.5)96 (11.3)<0.001Creatinine, mg/dL1.4 (1.1)1.5 (1.3)1.3 (0.9)1.4 (1.1)1.4 (1.0)1.4 (0.9)<0.001Cystatin C, mg/L1.4 (0.9)1.6 (1.1)1.4 (0.8)1.5 (0.9)1.4 (0.7)1.3 (0.6)0.007mGFRIndexed, mL/min/1.73 m267.6 (32.3)63.5 (34.6)65.7 (30.7)66.0 (31.3)69.6 (31.6)73.1 (32.4)<0.001Nonindexed, mL/min72.1 (36.3)56.7 (31.2)65.1 (30.6)69.8 (33.2)78.5 (35.7)90.6 (40.7)<0.001Indexed eGFReGFRcr, mL/min/1.73 m268.3 (30.8)66.6 (34.9)68.0 (30.2)67.3 (30.5)68.8 (29.7)70.9 (28.4)0.09eGFRcys, mL/min/1.73 m264.1 (31.0)60.0 (33.7)63.0 (30.7)63.5 (30.5)66.0 (30.4)67.9 (29.2)<0.001eGFRcr-cys, mL/min/1.73 m266.3 (31.0)63.2 (34.7)65.6 (30.5)65.4 (30.6)67.5 (29.8)69.7 (28.7)<0.001Nonindexed eGFReGFRcr, mL/min72.6 (34.0)59.4 (31.3)67.4 (30.0)71.1 (32.3)77.6 (33.5)87.8 (35.6)<0.001eGFRcys, mL/min68.3 (34.4)53.7 (30.4)62.4 (30.5)67.1 (32.3)74.4 (34.3)84.1 (36.4)<0.001eGFRcr-cys, mL/min70.6 (34.3)56.4 (31.3)65.0 (30.4)69.1 (32.4)76.1 (33.6)86.3 (35.9)<0.001Note: Values expressed as mean (standard deviation) for continuous and count (percent) for categorical variables. The minimum for BSA quintile 1 was 1.16 m2; the maximum for quintile 5 was 3.04 m2. There were 4 missing values for diabetes. P from Kruskal-Wallis for continuous and χ2 for categorical variables, between BSA quintiles. Open table in a new tab Note: Values expressed as mean (standard deviation) for continuous and count (percent) for categorical variables. The minimum for BSA quintile 1 was 1.16 m2; the maximum for quintile 5 was 3.04 m2. There were 4 missing values for diabetes. P from Kruskal-Wallis for continuous and χ2 for categorical variables, between BSA quintiles. The distribution of the differences between mGFR and eGFR populations appeared similar for indexed and nonindexed GFR (Fig 1). As expected, absolute values for extreme differences are nominally larger for nonindexed versus indexed eGFR for subgroups with BSA > 1.73 m2 and smaller for nonindexed versus indexed eGFR for the subgroup with BSA < 1.73 m2, but the cumulative frequency distribution plots of standardized differences showed nearly complete overlap (Fig S1). As expected, proportional metrics (RMSE, % bias, % IQR, and P30) are the same for indexed and nonindexed equations (Table S3; Fig S2). For the pooled population, differences in bias for indexed and nonindexed eGFRs were statistically significant but not clinically meaningful. For the lowest BSA quintile, both indexed and nonindexed eGFRcr overestimated indexed and nonindexed mGFR, indexed and nonindexed eGFRcys underestimated indexed and nonindexed mGFR, and eGFRcr-cys was unbiased for both indexed and nonindexed values (Fig 1). At higher BSAs, there was a significant trend to larger underestimation of mGFR by indexed and nonindexed eGFRcr, eGFRcys, and eGFRcr-cys. Performance of eGFR differed across studies, but performance of indexed and nonindexed eGFR was similar in each study (Table S4; Fig S3). Our main finding is that the performance of indexed and nonindexed eGFRs (based on creatinine, cystatin C, or both) compared with indexed and nonindexed mGFR, respectively, was not substantially different overall or by study or across BSA categories. Thus, when clinical decision making requires nonindexed GFR, converting indexed to nonindexed eGFR may be reasonable. Another important finding is that higher BSA was associated with underestimation of indexed and nonindexed eGFRcr, eGFRcys, and eGFRcr-cys, which may reflect the consequence of simultaneously increased generation of creatinine and cystatin C related to higher muscle and fat mass, respectively. At the lowest BSA quintile, eGFRcr-cys appears more accurate than either eGFRcr or eGFRcys. Strengths of our study include the large diverse population with a wide range of mGFRs and BSAs, mGFR using accepted clearance methods of exogenous filtration tracers, standardized measurement methods for serum creatinine and cystatin C, and guideline-recommended eGFR equations. The main limitation is that the underlying studies did not contain a large number of participants with extreme values of BSA. Other limitations include the use of the Du Bois and Du Bois equation, which may overestimate BSA in obesity,6Verbraecken J. Van de Heyning P. De Backer W. Van Gaal L. Body surface area in normal-weight, overweight, and obese adults. A comparison study.Metab Clin Exp. 2006; 55: 515-524Abstract Full Text Full Text PDF PubMed Scopus (286) Google Scholar and use of several mGFR methods, which may vary in their differences from the classic method of inulin clearance.7Soveri I. Berg U.B. Bjork J. et al.Measuring GFR: a systematic review.Am J Kidney Dis. 2014; 64: 411-424Abstract Full Text Full Text PDF PubMed Scopus (250) Google Scholar In conclusion, we have shown no substantial differences in performances of indexed and nonindexed eGFRs from CKD-EPI equations compared with indexed and nonindexed mGFRs, respectively, giving support to the use of nonindexed eGFR in clinical settings in which nonindexed GFR is more appropriate. However, both indexed and nonindexed eGFRs tend to underestimate indexed and nonindexed mGFRs at higher BSAs. Silvia Titan, MD, PhD, Shiyuan Miao, MS, Hocine Tighiouart, MS, Nan Chen, MD, Hao Shi, MBBS, Luxia Zhang, MD, MPH, Zuo Li, MD, Marc Froissart, MD, PhD, Peter Rossing, MD, DMSc, Anders Grubb, MD, PhD, Li Fan, MD, PhD, Michael Mauer, MD, Omran Bakoush, MD, PhD, Christina Wyatt, MD, Michael G. Shlipak, MD, MPH, Tariq Shafi, MD, MBBS, Lesley A. Inker, MD, MS, and Andrew S. Levey, MD. Research idea and study design: ASL, LAI; data acquisition: NC, HS, LZ, ZL, MF, PR, AG, LF, MM, OB, CW, MGS TS, LAI, ASL; data analysis/interpretation: ST, SM, HT, LAI, ASL; statistical analysis: ST, SM, HT, LAI, ASL; supervision or mentorship: LAI, ASL. 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. CKD- EPI was supported by grants U01DK067651 and U01DK35073 . AGES-Kidney was supported by a grant from NIH ( R01 DK082447 and supplement 01A1S1 to ASL), contract from the NIA ( N01-AG-1-2100 and HHSN27120120022C ), the Icelandic Heart Association (Hjartavernd) and the Icelandic Parliament (Althingi), and the American Recovery Act (3R01DK082447-01A1S1). MESA -Kidney was supported by R01DK087961 . HIV Study was funded by Gilead Sciences , Inc under an investigator-initiated protocol ( NCRR L1RR025752 ) and supported by the NCRR and the National Center for Advancing Translational Sciences , NIH , through grants UL1 RR025752 (Tufts Medical Center) and UL1 RR029887 (Mount Sinai School of Medicine) and also the Center for Clinical and Translational Sciences through grant UL1 RR025777 ( University of Alabama at Birmingham ). Support was also received from international collaborative research grant no. 13430720800 from the Science and Technology Commission of Shanghai Municipality to Shanghai Jiao Tong University School of Medicine, Multi-Center Clinical Research Project (No: DLY201510 ); grant ID 201502010 , National Health and Health Council (previously Ministry of Health), from 2015 through 2018; National Natural Science Foundation of China (grants 81771938 , 91846101 , 81301296 ), from Peking University (grants BMU2018MX020 , PKU2017LCX05 ), the National Key Technology R&D Program of the Ministry of Science and Technology of the People's Republic of China ( 2016YFC1305400 ); NIDDK grant R01DK097020 , and Siemens Healthcare grant to Tufts Medical Center. Dr Inker reports grants from the NIH and the National Kidney Foundation, other from Retrophin, other from Omeros, other from Dialysis Clinic Inc, and other from Reata, outside the submitted work. Dr Levey reports grants from NIH, outside the submitted work. Drs Inker and Levey have a patent “Precise estimation of glomerular filtration rate from multiple biomarkers” patent number PCT/US2015/044567 pending. Dr Zhang reports grants from AstraZeneca, outside the submitted work. Dr Shafi reports personal fees from Siemens, University of California Irvine, University of Mississippi Medical Center, Hershey Medical Center, and University of Tennessee, outside the submitted work. Dr Shlipak is a Scientific Advisor for TAI Diagnostics. Dr Rossing reports grants, personal fees, and other from Novo Nordisk; grants and personal fees from Astra Zeneca; and personal fees from Astellas, Boehringer Ingelheim, Bayer, Eli Lilly, and Gilead, outside the submitted work. The other authors declare that they have no relevant financial interests. The authors appreciate the collaboration of Vilmundur Gudnason (University of Iceland, Reykjavik, Iceland & Icelandic Heart Association, Kopavogur, Iceland). The authors acknowledge the valuable contributions of the other investigators, staff, and participants of the MESA Study (supported by contracts HHSN268201500003I, N01-HC-95159, N01-HC-95160, N01-HC-95161, N01-HC-95162, N01-HC-95163, N01-HC-95164, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168 and N01-HC-95169 from NHLBI, and by grants UL1-TR-000040, UL1-TR-001079, and UL1-TR-001420 from NCATS). A full list of participating investigators and institutions can be found at www.mesa-nhlbi.org. Received January 31, 2020. Evaluated by 3 external peer reviewers and a statistician, with direct editorial input from an Associate Editor, who served as Acting Editor-in-Chief. Accepted in revised form April 7, 2020. 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 (.72 MB) Help with pdf files Supplementary File (PDF)Figures S1-S3, Item S1, Tables S1-S4.