Editorial
GFR as the “Gold Standard”: Estimated, Measured, and True
Related Article, p. 49
G
lomerular filtration rate (GFR) has long been
considered the best overall index of kidney
function in health and disease. The rationale is that
GFR is a property of the kidney, has a large range,
and is affected by physiologic, pharmacologic, and
pathologic conditions. Furthermore, GFR decline is
associated with many physiologic and clinical consequences and is correlated with decline in other
excretory functions, such as tubular reabsorption and
secretion, as well as endocrine and metabolic functions. Decreased GFR is one criterion in the definition
and staging of acute and chronic kidney diseases, and
GFR estimating equations are recommended for
routine use for kidney function assessment in clinical
practice.1 For a number of reasons, some have questioned the appropriateness of GFR as the “gold
standard” for defining kidney disease and assessing
kidney function.2-4 While it can be important to
question accepted concepts, to date, no alternative
measure has been proposed to replace GFR as an
index of kidney function or decreased GFR as a criterion for defining kidney disease.
Glomerular filtration cannot be measured directly
in humans; thus “true” GFR cannot be known with
certainty. However, GFR can be assessed from
clearance measurements (measured GFR [mGFR]) or
serum levels of endogenous filtration markers (estimated GFR [eGFR]). Urinary inulin clearance, the
classic method for measuring GFR described by
Smith,5 is too difficult for clinical practice and
research purposes; consequently, alternative filtration
markers and clearance methods are used to measure
GFR (Table 1). In particular, 2 methods that have
been used in recent studies are of importance in
developing and validating GFR estimating equations.
Urinary iothalamate clearance is the basis for the
CKD-EPI (Chronic Kidney Disease Epidemiology
Collaboration) equations using creatinine and cystatin
C that are recommended by current guidelines.6,7
Meanwhile, plasma clearance of iohexol is now being used more widely in research studies, including
those to assess the CKD-EPI equations and develop
other equations.8-10 In this issue of AJKD, Seegmiller
and colleagues11 describe observed differences between the urinary clearances of iothalamate and
iohexol. The goals of this editorial are to provide
explanations for these observed differences and
explore the implications of these differences on the
utility of GFR as a gold standard.
Am J Kidney Dis. 2016;67(1):9-12
Seegmiller and colleagues11,12 developed a liquid
chromatography–tandem mass spectrometry method
to enable simultaneous determination of nonradioactive iothalamate and iohexol in serum and urine. They
then compared simultaneous urinary clearances of
both markers after bolus subcutaneous infusion in
150 patients with a wide range of GFRs. The mean
proportional ratio of iohexol to iothalamate clearance
was 0.85 (95% confidence interval, 0.83-0.88) across
the range of GFRs, indicating that GFR determined
using iohexol clearance is lower than that determined
using iothalamate clearance. To further explore this
finding, Seegmiller et al11 performed in vitro dialysis
experiments using plasma samples from 10 patients,
which demonstrated a lower ratio of dialysate to
plasma concentration of iohexol than iothalamate,
suggesting iohexol is less “filterable” than iothalamate, perhaps because of binding to plasma proteins.
The authors concluded that the nonequivalence of
iohexol and iothalamate clearances should be taken
into account in studies that use these filtration
markers to measure GFR.
In principle, explanations for observed differences
between urinary clearances of iohexol and iothalamate
could include any of the following: slower equilibration of iohexol than iothalamate between subcutaneous tissue and plasma, tubular secretion of
iothalamate, and protein binding, tubular reabsorption,
or extrarenal elimination of iohexol. A recent systematic review by Soveri and colleagues13 showed that
some but not all prior studies demonstrate that urinary
iothalamate clearances overestimate urinary inulin
clearance, whereas those of iohexol underestimate it.
Of interest, differences between plasma clearance of
iohexol and urinary clearance of inulin were smaller,
likely due to extrarenal elimination of iohexol.
Strengths of the study by Seegmiller and colleagues11 are a large study population, rigorous assay
methods, and simultaneous measurement of urinary
clearances. Expression of the results as clearance ratios eliminates possible errors in the measurement of
urine flow rate (eg, due to incomplete bladder
emptying) as a cause of the difference between
filtration markers. Limitations of the study are the
short equilibration period before the clearance
Address correspondence to Andrew S. Levey, MD, William B.
Schwartz Division of Nephrology, Tufts Medical Center Box 391,
800 Washington St, Boston, MA 02111. E-mail: alevey@
tuftsmedicalcenter.org
Ó 2016 by the National Kidney Foundation, Inc.
0272-6386
http://dx.doi.org/10.1053/j.ajkd.2015.09.014
9
10
Table 1. Comparison of GFR Measurement Methods Used in Epidemiologic Studies and to Develop GFR Estimating Equations and Sources of Error Compared to the Classic Method
of Urinary Inulin Clearance
Urinary Clearance of Inulin
(Classic Method)
Urinary Clearance of 125I-Iothalamate (Method Used
for CKD-EPI Equations)
Description
Clearance method
Administration of
filtration marker
Equilibration
period
Urinary
Continuous IV infusion allows
stable plasma concentrations
during the clearance period
Urinary
Subcutaneous bolus
administration leads to
exponential decline in
plasma concentrations
during the clearance
period
1 ha
Plasma Clearance of Iohexol (Method Often Used for
Epidemiologic Studies)
Error in mGFR vs Classic
Method
Description
Error in mGFR vs Classic
Method
NA
Underestimates mGFR,
worse if linear decline is
assumed
Plasma
IV administration leads to
biexponential decline in plasma
concentrations during the
clearance period
NA
Underestimates mGFR, worse
for short sampling periods
Underestimates mGFR if
longer equilibration
period is required
NA
None
Underestimates mGFR in
extracellular fluid expansion
4-6 h (longer if very low GFR)a
Underestimates mGFR for
short sampling periods
More accurate if more frequent
Urine collection
Bladder catheterization
Spontaneous bladder
emptying (water loading to
increase urine flow rate)
Imprecision in mGFR due
to incomplete bladder
emptying
2 samples to estimate early (fast)
decline, 2-4 samples to
estimate late (slow) declinea
No urine collection; avoids errors
due to incomplete bladder
emptying
5,200 Da
No
No
614 Da
No
No
NA
None
None
821 Da
Probable
Probable
NA
Underestimates mGFR
Underestimates mGFR
No
Minimal
Probable
Minimal
Overestimates mGFR
None
No
Probably more than minimal
None
Overestimates mGFR
Difficult
Easy (for 125I counting);
difficult (for HPLC or
LC-MS/MS)
Cannot be used in iodine
allergy; kidney toxicity at
high doses
Yes
None
Difficult (uses HPLC)
None
NA
Cannot be used in iodine allergy;
kidney toxicity at high doses
NA
NA
Yes
NA
Filtration marker
Molecular mass
Protein binding
Tubular
reabsorption
Tubular secretion
Extrarenal
elimination
Assay
Side effects
None
Available in US
Yes, but not FDA approved for
human use
1.5-2 h (four 20- to 30-min
clearance periods)a
2 samples for each urine
collection perioda
NA
Overestimates urinary
clearance if extrarenal
elimination of the marker
Abbreviations: CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; FDA, US Food and Drug Administration; HPLC, high-performance liquid chromatography; IV, intravenous;
LC-MS/MS, liquid chromatography–tandem mass spectrometry; mGFR, measured glomerular filtration rate; NA, not applicable; US, United States.
a
In clinical practice, procedures are frequently abbreviated, including shortening the equilibrium period and obtaining fewer urine and blood samples over a shorter interval.
Levey and Inker
Am J Kidney Dis. 2016;67(1):9-12
Clearance period
duration
Blood sampling
Required to achieve stable
plasma concentration (longer in
extracellular fluid expansion)
1.5-2 h (four 20- to 30-min
clearance periods)
Frequent
Editorial
measurement, absence of simultaneous measurements
of urinary clearance of inulin and plasma clearance of
iohexol, and absence of in vivo or in vitro investigations for the cause of lesser filtration of
iohexol. Further clarification on these points could be
derived from a comprehensive study of urinary and
plasma clearance of commonly used exogenous
filtration markers compared to urinary inulin clearance. In addition, this would also allow “calibration”
among methods, which could be useful for clinical
practice and research.
With respect to their implications for the concept
of GFR as an index of kidney function, the findings
reported in Seegmiller et al11 do not diminish the
pathophysiologic importance of GFR or its role in
clinical decision making. As the first step in urine
formation, glomerular filtration has a dominant influence in the excretory functions of the kidney, but alterations in other aforementioned kidney functions are
also associated with important clinical manifestations.
In addition, GFR does not reflect the cause, duration,
or likelihood of remission or progression of kidney
disease. Furthermore, markers of kidney damage (eg,
albuminuria, urine sediment abnormalities, or pathologic or imaging abnormalities) and complications
associated with decreased GFR (eg, anemia, hyperparathyroidism, or metabolic acidosis) also provide
important diagnostic and prognostic information.
Thus, evaluating GFR is essential but not sufficient
for the clinical assessment of kidney disease.
The findings of Seegmiller et al have implications
for the interpretation of mGFR. Both mGFR and
eGFR have error compared to true GFR. Even the
classic method of inulin clearance is imprecise, and
together, the systematic review by Soveri et al13 and
the new study from Seegmiller and colleagues11
suggest that clearances of both iothalamate and
iohexol are biased compared to true GFR. At the
present time, GFR measurement methods vary and are
not standardized across referral centers or research
studies. The work by Seegmiller et al11 reinforces the
importance of clinicians and investigators appreciating that mGFR results may vary across centers and
studies according to methods.
The findings of Seegmiller et al11 may also have
implications for the development and evaluation of
estimation equations. Bias and precision of eGFR
compared to mGFR may be affected by the GFR
measurement method used for developing the estimating equation. The MDRD (Modification of Diet in
Renal Disease) Study and CKD-EPI equations were
developed using urinary clearance of iothalamate, so
direct comparisons between them are unbiased.
However, comparisons between these 2 equations and
others developed using alternative mGFR methods
are subject to bias due to potential differences in the
Am J Kidney Dis. 2016;67(1):9-12
mGFR method, which may account in part for the
heterogeneity of findings in past studies evaluating
GFR estimating equations.14
The findings by Seegmiller et al may raise questions
about the use of mGFR as a confirmatory test for
decreased eGFR. The 2012 KDIGO (Kidney Disease:
Improving Global Outcomes) guideline for evaluating
and managing CKD recommends that clinicians use
creatinine-based eGFR (eGFRcr) as an initial test
and either cystatin C–based eGFR (eGFRcr-cys or
eGFRcys) or mGFR as a confirmatory test if a more
accurate assessment is required for clinical decision
making.1 The most important causes for inaccuracy in
eGFR are a nonsteady state (eg, acute kidney injury),
presence of non-GFR determinants of filtration
markers (eg, extremes of muscle mass for creatinine or
hyperthyroidism for cystatin C), and interference with
filtration marker assays (eg, ketoacids for alkalinepicrate assays for creatinine). In our view, mGFR remains helpful in these cases, and differences among
mGFR methods highlighted by Seegmiller et al11 are
probably small compared with errors in eGFR.
Finally, the study by Seegmiller et al11 may have
implications for interpretation of studies comparing
associations of eGFR and mGFR. Studies that
compare associations of eGFR and mGFR with
markers of kidney damage or with complications or
outcomes of kidney disease (eg, kidney failure or
mortality) do not evaluate the accuracy of eGFR
compared to mGFR. Instead, they are useful for
evaluating GFR versus non-GFR pathways to explain
associations with eGFR. In these studies, imprecision
in mGFR may obscure true differences between
GFR versus non-GFR associations, especially if
imprecision in mGFR is large in comparison to the
imprecision in other measures of disease.15 This is a
particular problem in studies of kidney disease
progression, in which the imprecision in mGFR is
large in comparison to its rate of decline.16,17
Overall, the study by Seegmiller and colleagues11
provides further information on GFR measurement
methods. In our view, the differences between iothalamate and iohexol clearances probably reflect small
and opposite biases compared to urinary inulin
clearance, so that true GFR is likely in between them
(Table 1). These differences are unlikely to be of
clinical importance at GFRs less than w90 mL/min/
1.73 m2, but may be important at higher GFRs. For
example, these differences might be relevant in
studies involving healthy populations or clinical
conditions associated with hyperfiltration or in evaluating people without known kidney disease (such as
kidney donors or people receiving cancer chemotherapy). GFR measurement remains an important
confirmatory diagnostic test for decreased eGFR.18
However, all mGFR methods are imprecise and may
11
Levey and Inker
be biased compared to true GFR. Standardization
and calibration across methods could be helpful in
clinical and research studies, including the development of more accurate eGFR equations.
Andrew S. Levey, MD
Lesley A. Inker, MD, MS
Tufts Medical Center
Boston, Massachusetts
ACKNOWLEDGEMENTS
The authors thank Aghogho Okparavero, MD, for assistance
with manuscript preparation.
Support: None.
Financial Disclosure: Drs Levey and Inker report involvement
as principal investigator and clinical director, respectively, in the
CKD-EPI research group, which developed the CKD-EPI equations for GFR estimation. Both authors have applied for a patent
for precise estimation of GFR using a panel of filtration markers.
REFERENCES
1. Kidney 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):1-150.
2. Murphy DP, Hsu CY. Estimating glomerular filtration rate:
is it good enough? And is it time to move on? Curr Opin Nephrol
Hypertens. 2013;22(3):310-315.
3. Rule AD, Glassock RJ. GFR estimating equations: getting
closer to the truth? Clin J Am Soc Nephrol. 2013;8(8):1414-1420.
4. Moynihan R, Glassock R, Doust J. Chronic kidney disease
controversy: how expanding definitions are unnecessarily labelling
many people as diseased. BMJ. 2013;347:f4298.
5. Smith HW. Measurement of the filtration rate. In: The Kidney: Structure and Function in Health and Disease. New York,
NY: Oxford University Press; 1951:39-62.
6. Levey AS, Stevens LA, Schmid CH, et al. A new equation to
estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):
604-612.
12
7. Inker LA, Schmid CH, Tighiouart H, et al. Estimating
glomerular filtration rate from serum creatinine and cystatin C.
N Engl J Med. 2012;367(1):20-29.
8. Schaeffner ES, Ebert N, Delanaye P, et al. Two novel
equations to estimate kidney function in persons aged 70 years or
older. Ann Intern Med. 2012;157(7):471-481.
9. Grubb A, Horio M, Hansson LO, et al. Generation of a new
cystatin C-based estimating equation for glomerular filtration rate
by use of 7 assays standardized to the international calibrator. Clin
Chem. 2014;60(7):974-986.
10. Fan L, Levey AS, Gudnason V, et al. Comparing GFR
estimating equations using cystatin C and creatinine in elderly
individuals. J Am Soc Nephrol. 2015;26(8):1982-1989.
11. Seegmiller JC, Burns BE, Schinstock CA, Lieske JC,
Larson TS. Discordance between iothalamate and iohexol urinary
clearances. Am J Kidney Dis. 2016;67(1):49-55.
12. Seegmiller JC, Burns BE, Fauq AH, Mukhtar N, Lieske JC,
Larson TS. Iothalamate quantification by tandem mass spectrometry to measure glomerular filtration rate. Clin Chem. 2010;56(4):
568-574.
13. Soveri I, Berg UB, Bjork J, et al. Measuring GFR: a systematic review. Am J Kidney Dis. 2014;64(3):411-424.
14. Earley A, Miskulin D, Lamb EJ, Levey AS, Uhlig K.
Estimating equations for glomerular filtration rate in the era of
creatinine standardization: a systematic review. Ann Intern Med.
2012;156(11):785-795.
15. Hsu CY, Propert K, Xie D, et al. Measured GFR does not
outperform estimated GFR in predicting CKD-related complications. J Am Soc Nephrol. 2011;22(10):1931-1937.
16. Xie D, Joffe MM, Brunelli SM, et al. A comparison of
change in measured and estimated glomerular filtration rate in
patients with nondiabetic kidney disease. Clin J Am Soc Nephrol.
2008;3(5):1332-1338.
17. Ku E, Xie D, Shlipak M, et al. Change in measured GFR
does not outperform change in estimated GFR in predicting
adverse outcomes in CKD: results from the CRIC Study [ASN
abstract FR-OR033]. J Am Soc Nephrol. 2014;25:53A.
18. Stevens LA, Levey AS. Measured GFR as a confirmatory
test for estimated GFR. J Am Soc Nephrol. 2009;20(11):
2305-2313.
Am J Kidney Dis. 2016;67(1):9-12