Begin typing your search above and press return to search.
Volume: 15 Issue: 1 February 2017 - Supplement - 1


Kinetic Glomerular Filtration Rate Estimation Compared With Other Formulas for Evaluating Acute Kidney Injury Stage Early After Kidney Donation

Objectives: Kinetic glomerular filtration rate estimation may have more power and versatility than the Modification of Diet in Renal Disease or Cockcroft-Gault formula for evaluating kidney function when plasma creatinine fluctuates rapidly. After kidney donation, glomerular filtration rate rapidly fluctuates in otherwise healthy patients. We compared 3 formulas for estimating glomerular filtration rate: kinetic, Modification of Diet in Renal Disease, and Cockcroft-Gault, for determining stages of acute kidney injury early after kidney donation.

Materials and Methods: In 42 living kidney donors, we measured serum creatinine, cystatin C, neutrophil gelatinase–associated lipocalin, and glomerular fil­tration rates before uninephrectomy and 3 days afterward. To estimate glomerular filtration rate, we used Cockcroft-Gault, Modification of Diet in Renal Disease, and kinetic equations. We sought the most accurate formula for staging acute kidney injury according to the risk, injury, failure, loss, and end-stage criteria.

Results: The kinetic glomerular filtration rate model found more cases of stage 3 acute kidney injury than did the Modification of Diet in Renal Disease or Cockcroft-Gault formula. Receiver operating charac­teristic curves showed that the kinetic glomerular filtration rate model had more sensitivity and specificity than the Cockroft-Gault formula for discriminating among risk, injury, failure, loss, and end-stage criteria stages of acute kidney injury, based on serum creatinine changes. On day 2 after donation, a more sensitive marker with a shorter half-life (serum neutrophil gelatinase–associated lipocalin) was more significantly correlated with kinetic glomerular filtration rate estimation.

Conclusions: The kinetic glomerular filtration rate model was able to discriminate stages of acute kidney injury early after kidney donation according to risk, injury, failure, loss, and end-stage criteria better than the Modification of Diet in Renal Disease or Cockcroft-Gault formulas. The kinetic model detected failure-stage acute kidney injury ≥ 1 to 2 days earlier than the MDRD formula, CG formula detected no failure.

Key words : Cockcroft-Gault formula, Modification of diet in renal disease equation, Neutrophil gelatinase–associated lipocalin, RIFLE criteria, Transplant donors


In most situations, glomerular filtration rate (GFR) is the best measure of kidney function. Serum creatinine, in contrast, does not reflect GFR changes in real time if it is changing rapidly.1,3 In addition, the measurement of GFR is essential in the selection of potential kidney donors.3 In the early stages of acute kidney injury (AKI), serum creatinine levels over­estimate actual GFR when the latter is markedly reduced, because there may not have been enough time for creatinine to accumulate; conversely, estimated GFR will underestimate true GFR during recovery from AKI, when serum creatinine is rapidly declining.3,4 Also, owing to the nonlinear relation between creatinine/cystatin C and GFR, relatively small changes in these markers can represent significant alterations in GFR.1,4

Nephrologists are always searching for com­prehensive answers to questions such as: What are GFR changes over time? Are the changes gradual or precipitous? What is the severity of AKI, and how should drug regimens be adjusted in this setting? Is kidney function improving, and if so, when and how rapidly? Proponents of so-called “kinetic estimation of GFR” believe that all these answers are hidden in data on serum creatinine changes over time and can be revealed by kinetic GFR estimation.1 By translating all creatinine measurements into kinetic GFR, they believe, these questions are answerable.

Kidney donation presents a rare opportunity for evaluating rapid changes in GFR in otherwise healthy subjects. Because living donors in Iran are selected from among healthy young individuals with normal kidney function and no medical problems, evaluating their kidney function provides a unique opportunity for assessing the impact of organ donation on kidney function biomarkers. Further, the impact of acute loss of ≥ 50% of GFR on kidney function biomarkers can be assessed in each model of GFR estimation. In this study, we evaluated the hypothesis that kinetic estimation of GFR out­performs Modification of Diet in Renal Disease (MDRD) and Cockcroft-Gault (CG) formulas for assessing rapid changes in GFR early after kidney donation.

Materials and Methods

We studied 42 living, unrelated kidney donors with a mean age of 30.3 ± 4.18 years in the Qaem and Imam-Reza Hospitals, Mashhad, Iran, from June 2014 to April 2015. The research protocol was approved by the Ethical Committee of Mashhad University of Medical Sciences (No. 89066). Written informed consent was obtained from all patients. There were no exclusion criteria for the study subjects. We collected 168 blood samples: 4 samples from each donor, with 1 sample was taken before nephrectomy, and 3 other samples taken on days 1, 2, and 3 after the operation. Serum was centrifuged and then stored at −20°C. For measuring serum cystatin C and neutrophil gelatinase–associated lipocalin (NGAL), we used commercially available kits from BioVendor Medical Laboratory (Brno, Czech Republic). The creatinine kit was picric acid–based and qualitative (Pars Azmoon Company, Tehran, Iran). The CG and MDRD formulas were used for estimating GFR in our study subjects.

As proposed by Chen,1 the following methods were used: “The kinetic GFR formula is estimated from the initial creatinine content, creatinine production rate, volume of distribution, and the quantitative difference between consecutive plasma creatinine measurements over a defined time. For that period, the calculated creatinine excretion then equals the creatinine clearance rate. The additional variables needed for [the] formula are any steady-state plasma creatinine, the corresponding [GFR estimated] by an empirical formula like MDRD or CG, and the maximum increment in creatinine per day if [AKI is] anuric.” For calculating the kinetic GFR, the following formula was used1:



where SSPCr × CrCl means the product of any steady-state plasma creatinine and the corresponding creatinine as obtained by the CG or MDRD formula. ΔTime(h) is the interval in hours between 2 consecutive creatinine measurements. MaxΔPCr/Day signifies the maximal increment in plasma creatinine that can occur per day if renal function is completely lost. In anuric AKI, it is presumed that plasma creatinine can rise by ≤ 1.0-1.5 mg/dL/day, but it is preferable that the numerical choice for MaxΔPCr/Day be chosen using actual patient data.1

For classifying different stages of AKI in donors, we used the risk, injury, failure, loss, and end-stage (RIFLE) classification system5:

  • R = risk of renal dysfunction and is defined as an increase in serum creatinine ≥ 1.5× baseline, or a decrease in GFR ≥ 25%, or urine output < 0.5 mL/kg/hour for 6 hours.
  • I = injury to the kidney and is defined as an increase in serum creatinine ≥ 2.0× baseline, or a decrease in GFR ≥ 50%, or urine output < 0.5 mL/kg/hour for 12 hours.
  • F = failure of kidney function and is defined as an increase in serum creatinine ≥ 3.0 × baseline, or serum creatinine ≥ 4.0 mg/dL in the setting of an acute rise (0.5 mg/dL), or a decrease in GFR ≥ 75%, or urine output < 0.3 mL/kg/hour for 24 hours, or anuria for 12 hours.
  • L = loss of kidney function and is defined as persistent kidney failure lasting > 4 weeks.
  • E = end-stage renal disease and is defined as persistent kidney failure lasting > 3 months.

For evaluating kidney function after living kidney donation, we measured the serum creatinine, cystatin C, and NGAL of 42 living donors before unineph­rectomy and in the 3 days immediately after. Glomerular filtration rate was calculated before uninephrectomy and in each of the 3 days using CG, MDRD, and kinetic GFR equations. We also evaluated the prevalence of different stages of acute kidney injury, classified according to the RIFLE criteria, and the accuracy of each of the 3 GFR estimation formulas for predicting these stages. The utility of each GFR estimation formula for predicting AKI stage was compared with that of other markers: serum creatinine, plasma cystatin C, and plasma NGAL level. The MDRD equation was indexed to a body surface area of 1.73 m2.

Statistical analyses
In this study, the paired sample t test and Mann-Whitney U test were performed for comparing paired quantitative variables. Changes in serum levels of GFR before and in the 3 days after uninephrectomy were compared using repeated measurement. Receiver operating characteristic (ROC) curves for the 3 GFR estimation equations were used to determine the most accurate formula for evaluating the different stages of AKI, as depicted by serum creatinine changes. The data were analyzed using IBM SPSS Statistics software for Windows (version 19.0, IBM Corp., Armonk, NY, USA).


Repeated measurement analysis showed that after kidney donation, GFR was significantly reduced, as estimated by each of the 3 models of GFR calculation. However, the percentile reduction of GFR as estimated by the kinetic model (64% ± 0.20) was significantly greater on day 1 after donation compared with the CG estimate (34% ± 0.24) and the MDRD estimate (38% ± 0.20) (for CG and MDRD, P < .001; Figure 1). This means that most of the GFR reduction on day 1 after kidney donation was significantly greater when evaluated using the kinetic GFR model, rather than the CG and MDRD formulas.

On day 2 after kidney donation, the mean of percentile reduction in kinetically changing GFR was actually negative compared with day 1 after donation (–0.83% ± 3.02). The large standard deviation (3.02) may indicate that, in some patients, GFR fluctuations as estimated by the kinetic model may have been caused by ischemia, reperfusion injury, or anesthesia drug effects rather than simple loss of GFR due to uninephrectomy. On day 2, the mean percentile reduction in GFR compared with day 1, as estimated by the CG (0.32% ± 0.17) and MDRD (0.34% ± 0.20) formulas, showed continuing GFR reduction. Thus, from day 2 after kidney donation, patients were mostly experiencing kidney function recovery, but only according to kinetic GFR estimation.

Table 1 shows the prevalence of different stages of AKI based on the RIFLE criteria using different methods of GFR estimation. As can be seen, there was a greater incidence of stage 3 AKI (ie, kidney failure) as reported by kinetic GFR (12.8%) than was reported by the MDRD and CG formulas (0%), although this difference did not reach statistical significance (for chi-squared test, P > .1). On day 3, the MDRD and CG formulas only obtained 3.4% and 0% failure rates, so is this statement accurate? Superiority of kinetic model was limited for the first two days after donation.

Table 2 shows the sensitivity or specificity of GFR measurements using the kinetic, MDRD, and CG formulas for discriminating among RIFLE stages of AKI, based on serum creatinine changes. The ROC curves in Figures 2-3-4 show that in each of the 3 days after donation, kinetic GFR had better sensitivity or specificity for discriminating among RIFLE stages of AKI, based on serum creatinine changes. Interestingly, serum NGAL levels before donation were significantly and positively correlated with GFR using the kinetic formula on day 3 after nephrectomy (r = .615, P = .009), whereas serum NGAL levels on day 1 after donation were significantly correlated with kinetic GFR estimations on the same day (r = 0.492, P = .008). These results possibly indicate that the kinetic GFR formula detects kidney function failure and recovery earlier than a more sensitive biomarker with a shorter half-life, such as serum NGAL.


In this study, using a well-validated formula for calculating kinetic GFR, we found that only kinetic GFR measurement, but neither the CG nor the MDRD formula, was able to diagnose stage 3 AKI in the 48 hours after kidney donation. Since the introduction of the term “kinetic GFR” by Winter6 and Chen,1 it has been vehemently disputed—even described as a misnomer7—that kinetic GFR can calculate absolute creatinine clearance, not just GFR. However, 2 recently published articles reported that kinetic GFR improved risk prediction in a clinical model of delayed graft function.8,9 Other researchers have shown that, in cardiac surgery patients, kinetic GFR modeling enabled the early identification of patients with AKI before significant incremental increases in serum creatinine appeared.10

Our results showed the superiority of kinetic GFR estimation to MDRD and CG GFR estimation for predicting stages of AKI according to the RIFLE criteria, when creatinine-defined AKI staging is used. In addition, there was a stronger correlation between plasma NGAL as an early marker of AKI with kinetic GFR, compared with MDRD and CG formulas of GFR. This introduces the possibility that kinetic GFR is a superior measure when using a more sensitive marker such as plasma NGAL, instead of serum creatinine, when evaluating its performance. Regarding the relative superiority of MDRD to the CG formula for detecting various stages of AKI early after kidney donation, other researchers have shown the advantage of MDRD over both the CG formula and the Chronic Kidney Disease Epidemiology Collaboration equation for estimating GFR and diagnosing impaired renal function.11

Owing to the nonlinear relation between creatinine/cystatin C and GFR,4 relatively small initial incremental increases in these markers correlate with significant decreases in GFR. This nonlinearity is more pronounced when using kinetic GFR as a measure, as reflected by the greater percentile reduction in kinetic GFR compared with CG and MDRD GFR formulas on day 1 after donation (Figure 1). The relatively higher occurrence of failure in stage 3 AKI reported by kinetic GFR underscores this relation as well (Table 1). The superior power of the kinetic GFR model for discerning various stages of AKI may result from an exaggeration of this nonlinearity, as well as the fact that a kinetic model measures dynamic, not static, changes in serum creatinine.

The stronger correlation our study found of plasma NGAL as an early marker of AKI with kinetic GFR, compared with the MDRD and CG formulas, is similar to findings of other studies. The latter have shown that NGAL can be used as an early biomarker, for example, contrast-induced nep­hropathy in patients with normal serum creatinine levels undergoing percutaneous coronary inter­ventions,12 kidney transplant recipients,13 and hypertensive patients with coronary artery diseases.14 These results show that NGAL has the potential for detecting kidney injury earlier than either serum creatinine or, perhaps, GFR estimation formulas that use static measurements of serum creatinine. In conclusion, it has been suggested that the discriminating ability of kinetic GFR is equal to that of biomarkers, such as urinary NGAL. Further, given that there is no extra charge for performing kinetic GFR tests and less delay in measuring creatinine than for measuring these biomarkers, studies should be conducted to determine whether kinetic GFR is superior to these biomarkers.15

Our study had several limitations. First, it restricted GFR evaluation to the first 3 days after kidney donation. Although we were able to detect superior performance of kinetic GFR for discriminating among AKI stages according to the RIFLE criteria, perhaps by continuing the study for 1 to 2 more days, we would have witnessed other formulas emerge as the best estimators of GFR. Moreover, we were not able to use a more precise method, such as radioisotope scanning, as the criterion standard for comparing various GFR estimator models.


Kinetic GFR is better able to discriminate among stages of AKI according to the RIFLE criteria than are other creatinine-based formulas (MDRD and CG) in the first 3 days after kidney donation. Also, when using new markers of kidney function, such as NGAL, kinetic GFR measurements show stronger correlation with NGAL changes than do those derived from the MDRD or CG formulas.


  1. Chen S. Retooling the creatinine clearance equation to estimate kinetic GFR when the plasma creatinine is changing acutely. J Am Soc Nephrol. 2013;24(6):877-888.
    CrossRef - PubMed
  2. Kannapiran M, Nisha D, Madhusudhana Rao A. Underestimation of impaired kidney function with serum creatinine. Indian J Clin Biochem. 2010;25(4):380-384.
    CrossRef - PubMed
  3. Rabito C, Halpern EF, Scott J, Tolkoff-Rubin N. Accurate, fast, and convenient measurement of glomerular filtration rate in potential renal transplant donors. Transplantation. 2010;90(5):510-517.
    CrossRef - PubMed
  4. Pasala S, Carmody JB. How to use… serum creatinine, cystatin C and GFR. Arch Dis Child Educ Pract Ed. 2016; pii: edpract-2016-311062.
    CrossRef - PubMed
  5. Edelstein CL, Faubel S. Biomarkers in acute kidney injury. In: Edelstein CL, ed. Biomarkers of Kidney Disease. Amsterdam: Elsevier Press; 2011:177-232.
    CrossRef - PubMed
  6. Winter ME. Basic Clinical Pharmacokinetics. 5th ed. Philadelphia: Wolters Kluwer/Lippincott Williams & Wilkins Health; 2009.
  7. Buchkremer F. Estimating the dynamic GFR when creatinine concentrations are changing rapidly. @swissnephro. May 19, 2015, in Nephrology, Calculator. Accessed December 1, 2016.
  8. Pianta TJ, Endre ZH, Pickering JW, Buckley NA, Peake PW. Kinetic estimation of GFR improves prediction of dialysis and recovery after kidney transplantation. PLoS One. 2015;10(5):e0125669.
    CrossRef - PubMed
  9. Endre ZH, Pianta TJ, Pickering JW. Timely diagnosis of acute kidney injury using kinetic eGFR and the creatinine excretion to production ratio, E/eG – creatinine can be useful! Nephron. 2016;132(4):312-316.
    CrossRef - PubMed
  10. Seelhammer TG, Maile MD, Heung M, Haft JW, Jewell ES, Engoren M. Kinetic estimated glomerular filtration rate and acute kidney injury in cardiac surgery patients. J Crit Care. 2016;31(1):249-254.
    CrossRef - PubMed
  11. Agoons DD, Balti EV, Kaze FF, et al. Performance of three glomerular filtration rate estimation equations in a population of sub-Saharan Africans with type 2 diabetes. Diabet Med. 2016;33(9):1291-1298.
    CrossRef - PubMed
  12. Bachorzewska-Gajewska H, Malyszko J, Sitniewska E, et al. Could neutrophil-gelatinase-associated lipocalin and cystatin C predict the development of contrast-induced nephropathy after percutaneous coronary interventions in patients with stable angina and normal serum creatinine values? Kidney Blood Press Res 2007;30:408-415.
    CrossRef - PubMed
  13. Malyszko J, Malyszko JS, Bachorzewska-Gajewska H, Poniatowski B, Dobrzycki S, Mysliwiec M. Neutrophil gelatinase-associated lipocalin is a new and sensitive marker of kidney function in chronic kidney disease patients and renal allograft recipients. Transplant Proc. 2009;41(1):158-161.
    CrossRef - PubMed
  14. Malyszko J, Bachorzewska-Gajewska H, Malyszko JS, Pawlak K, Dobrzycki S. Serum neutrophil gelatinase-associated lipocalin as a marker of renal function in hypertensive and normotensive patients with coronary artery disease. Nephrology (Carlton). 2008;13(2):153-156.
    CrossRef - PubMed
  15. Seelhammer TG, Maile MD, Heung M, Haft JW, Jewell ES, Engoren M. Kinetic estimated glomerular filtration rate and acute kidney injury in cardiac surgery patients. J Crit Care. 2016;31(1):249-254.
    CrossRef - PubMed

Volume : 15
Issue : 1
Pages : 104 - 109
DOI : 10.6002/ect.mesot2016.O104

PDF VIEW [260] KB.

From the Department of Nephrology, Ghaem Hospital, Mashhad University of Medical Sciences, Mashhad, Iran
Acknowledgements: The authors declare that they have no sources of funding for this study, and they have no conflicts of interest to declare.
Corresponding author: Reza Hekmat, Department of Nephrology, Ghaem Hospital, Ahmad-Abad Street, Mashhad, Iran
Phone: +98 511 801 2829
Mobile: +98 9153 012 669