Page 1

tap_948261..268

Simple Cystatin C Formula Compared to Sophisticated

CKD-EPI Formulas for Estimation of Glomerular Filtration

Rate in the Elderly

Sebastjan Bevc,1Radovan Hojs,1Robert Ekart,1Maksimiljan Gorenjak,2

and Ludvik Puklavec3

Departments of1Nephrology,2Clinical Chemistry, and3Nuclear Medicine, Clinic of Internal Medicine,

University Medical Center Maribor, Maribor, Slovenia

Abstract: Despite the fact that the serum creatinine level is

notoriously unreliable for the estimation of glomerular fil-

tration rate (GFR) in the elderly,the serum creatinine con-

centration and serum creatinine-based formulas, such as

the Modification of Diet in Renal Disease study equation

(MDRD) are the most commonly used markers to estimate

GFR.Recently,serum cystatin C-based formulas,the newer

creatinine formula (the Chronic Kidney Disease Epi-

demiology Collaboration formula (CKD-EPI creatinine

formula), and an equation that uses both serum creatinine

and cystatin C (CKD-EPI creatinine and cystatin formula)

were proposed as new GFR markers.The aim of our study

was to compare the MDRD formula, CKD-EPI creatinine

formula, CKD-EPI creatinine and cystatin formula, and

simple cystatin C formula (100/serum cystatin C) against

51Cr-EDTA clearance in the elderly. A total of 317 adult

Caucasian patients aged >65 years were enrolled. In each

patient,51Cr-EDTA clearance,serum creatinine,and serum

cystatin C were determined, and the GFR was calculated

using the MDRD formula, CKD-EPI formulas, and simple

cystatin C formula. Statistically significant correlations

between

found. In the receiver operating characteristic (ROC)

51Cr-EDTA clearance and all formulas were

curve analysis with a cut-off of GFR 45 mL/min/1.73 m2, a

higher diagnostic accuracy was achieved with the equation

that uses both serum creatinine and cystatin C (CKD-EPI

creatinine and cystatin formula) than the MDRD formula

(P < 0.013) or CKD-EPI creatinine formula (P < 0.01), but

it was not higher than that achieved for the simple cystatin

C formula (P = 0.335). Bland and Altman analysis for the

same cut-off value showed that the creatinine formulas

underestimated and the simple cystatin C formula overes-

timated measured GFR. All equations lacked precision.

The accuracy within 30% of estimated51Cr-EDTA clear-

ance values differ according to the stage of CKD.Analysis

of the ability to correctly predict GFR below and above

45 mL/min/1.73 m2showed a high prediction for all formu-

las. Our results indicate that the simple cystatin C formula,

which requires just one variable (serum cystatin C concen-

tration), is a reliable marker of GFR in the elderly and

comparable to the creatinine formulas, including the

CKD-EPI formulas.

Key Words:

Chronic kidney disease,Elderly,Glomerular filtration rate,

Serum creatinine-based equation, Serum cystatin C-based

equation.

51Cr-EDTA clearance,

Chronic kidney disease (CKD) is an important

public health problem. In response to the rising

prevalenceofCKD,nephrologists andother

physicians have focused research on the prevention

and early detection of renal failure. Estimation of

glomerular filtration rate (GFR) is essential for the

evaluation of patients with CKD, which is defined

as kidney damage or GFR <60 mL/min/1.73 m2for

three months or more, irrespective of cause, and

classified into stages according to the level of GFR.

Therefore, GFR estimation allows us to detect early

impairment of kidney function, prevent further

deterioration and complications, correct the dosage

of drugs cleared by the kidney so as to avoid poten-

tial drug toxicity, and to help us manage CKD

Received March 2011.

Address correspondence and reprint requests to Dr Sebastjan

Bevc, Univerzitetni klinic ˇni center Maribor, Klinika za interno

medicino, Oddelek za nefrologijo, Ljubljanska 5, 2000 Maribor,

Slovenia. Email: sebastjanb@yahoo.com

Presented in part at the Symposium Celebrating the 40th Anni-

versary of Chronic Dialysis and Kidney Transplantation in

Slovenia held 4–5 November 2010 in Bled, Slovenia.

Conflict of interest statement: None declared.

Therapeutic Apheresis and Dialysis 15(3):261–268

doi: 10.1111/j.1744-9987.2011.00948.x

© 2011 The Authors

Therapeutic Apheresis and Dialysis © 2011 International Society for Apheresis

261

Page 2

patients. The National Kidney Disease Education

Programrecommended

>60 mL/min/1.73 m2not as an exact number but

simply as

>60 mL/min/1.73 m2, and

?60 mL/min/1.73 m2, the exact numerical estimate

should be reported (1). For clinicians, a GFR value

<60 mL/min/1.73 m2is very important. The values

indicate the presence of CKD and represent an

increased risk of impaired kidney function, progres-

sion to kidney failure, and premature death caused

by cardiovascular events of patients with CKD

(2,3). Furthermore, in the elderly, an estimated

GFR < 60 mL/min/1.73 m2is very common and car-

diovascular disease increases and rises sharply once

the estimated GFR is <45 mL/min/1.73 m2. More-

over, some studies have reported the importance of

a cut-off value of GFR = 45 mL/min/1.73 m2in older

people as it identifies a potentially vulnerable group

of the elderly who could be considered for further

investigation, referral, or intervention (4,5).

To date, several different markers for the estima-

tion of GFR have been proposed.The ideal marker of

GFR should be an endogenous molecule that, being

produced at a constant rate, is cleared solely by the

kidneys via free glomerular filtration without being

either secreted by tubular cells or reabsorbed into the

peritubular circulation (6). The clearance of exog-

enous substances such as inulin,iohexol,51Cr-EDTA,

99mTcDTPA, and125I-iothalamate has been presented

as the “gold standard” for the estimation of

GFR. These techniques are time-consuming, labor-

intensive and expensive, and require administration

of substances that make them incompatible with

routine monitoring.

Despite all known disadvantages, the serum crea-

tinine concentration has become the most commonly

used marker to estimate GFR in clinical practice and

in most studies (7). Moreover, also in the elderly,

kidney function has traditionally been assessed using

serum creatinine,despite the fact that the serum crea-

tinine level is especially unreliable for the estimation

of GFR in this population (4,7).The current Kidney

Disease Outcomes Quality Initiative (K/DOQI)

guidelines emphasize the need to assess kidney func-

tion using predictive equations, such as the abbrevi-

ated modification of diet in renal disease (MDRD)

formula, rather than just serum creatinine (8,9).

Unfortunately,all these formulas are also limited by a

lack of validation in the full range of GFR to which

they are applied (10). To minimize some of these

limitations, such as imprecision and a systematic

underestimation of measured GFR with the MDRD

formula, new the Chronic Kidney Disease Epidemi-

ology Collaboration equation (CKD-EPI creatinine

reportingGFRvalues

for values

formula) was developed (11).The authors of the new

equation validated the CKD-EPI equation using data

pooled from several previous studies and showed

that the new formula is more accurate than the

widely-used MDRD formula (11). Common features

of these equations are reliance on serum creatinine

and demographic and anthropometric data, and the

accuracy of these formulas is still debated (8,10).

Serum cystatin C, a protease inhibitor with a low

molecular weight, was proposed as an endogenous

filtration marker. Serum cystatin C is produced at a

constant rate by all nucleated cells.It is freely filtered

across the glomerular membrane and is reabsorbed

and metabolized in the proximal tubule (12,13). Pre-

vious reports have suggested that the serum cystatin

C concentration is a better indicator of GFR than the

serum creatinine concentration in patients with

spinal injuries, liver cirrhosis, diabetes and mild-to-

moderate kidney function impairment,and in elderly

patients (14–16). Over the last few years, several

serum cystatin C-based equations (cystatin C formu-

las) have been developed and proposed to estimate

the GFR from serum cystatin C concentration as an

alternative filtration marker to serum creatinine-

based equations (17–23).Additionally, some authors

have also recently advanced the hypothesis that an

equation combining filtration markers (serum creati-

nine and serum cystatin C) may be useful.The CKD

Epidemiology Collaboration formula is one of the

proposed new equations that use both serum creati-

nine and serum cystatin C (CKD-EPI creatinine and

cystatin formula) for the estimation of kidney func-

tion (24,25).

The aim of our study was to compare two

creatinine-based formulas (MDRD formula and

CKD-EPI creatinine formula), the CKD-EPI creati-

nine and cystatin formula, and the simple cystatin C

formula (100/serum cystatin C) against

clearance in the elderly.

51Cr-EDTA

PATIENTS AND METHODS

In this study, 317 adult Caucasian patients aged

>65 years (170 women, 147 men) were included. All

patients were referred for51Cr-EDTA clearance by

nephrologists,diabetologists,cardiologists,or general

internists because of suspected or established renal

dysfunction. At the same time as51Cr-EDTA clear-

ance was estimated,both serum creatinine and serum

cystatin C were measured. Serum creatinine was

measured by using the kinetic method according to

the Jaffé method without deproteinisation (Roche

Diagnostics,Mannheim,Germany).This is a compen-

sated method based on manufacturer instructions

S Bevc et al. 262

© 2011 The Authors

Therapeutic Apheresis and Dialysis © 2011 International Society for Apheresis

Ther Apher Dial,Vol. 15, No. 3, 2011

Page 3

and has been described previously (26).Serum cysta-

tin C was measured by the particle-enhanced immu-

nonephelometric method (Dade Behring, Marburg,

Germany). The GFR was estimated from a single

51Cr-EDTA injection and three blood samples (120,

180, and 240 min after parenteral application of the

marker) according to the Committee on Renal Clear-

ance recommendations (27).

was calculated in milliliters per minute per 1.73 m2.

The GFR was calculated according to the MDRD

(Eqn 1),CKD-EPI creatinine (Eqn 2),and CKD-EPI

creatinine and cystatin (Eqn 3) formulas, where the

MDRD formula and CKD-EPI formulas are stan-

dardized for a body surface area of 1.73 m2:

51Cr-EDTA clearance

• GFR calculated according to the MDRD formula:

GFR serum creatinine mg dL

age years

()

0.203

−

=×

()

×

−

186

1 154.

(1)

– A correction factor of 0.742 was used for

women.

• GRF calculated according to the CKD-EPI creati-

nine formula:

GFR serum creatinine mg dL

age

)

(2)

=×

()

() ×(

ab

c

0 993.

– The variable a takes on the following values on

the basis of race and sex:

• black women = 166

• black men = 164

• white/other women = 144

• white/other men = 141

– The variable b takes on the following values on

the basis of sex:

• women = 0.7

• men = 0.9

– The variable c takes on the following values on

the basis of sex and creatinine measurement:

• women:serum creatinine

-0.329; serum creatinine >0.7 mg/dL = -1.209

• men: serum creatinine ?0.7 mg/dL = -0.411;

serum creatinine >0.7 mg/dL = -1.209

• GRF calculated according to CKD-EPI creatinine

and cystatin formula:

?0.7 mg/dL =

GFR177.6

serum cystatin C mg

(

serum creatinine mg dL

(

0.65

−

=×

(

)

)

()

×

L Lage

0.57

−

0.2

−

)

×

(3)

– A correction factor of 0.82 was used for women.

The GFR was also calculated according to the pre-

viously published simple cystatin C formula (Eqn 4)

(28):

GFR 100 serum cystatin C mg L

=

()

(4)

Inthestatisticalanalysis,SPSSforWindowsversion

18.0.1 (SPSS, Chicago, IL, USA) and MedCalc for

Windows version 5.00.020 (MedCalc Software, Mari-

akerke, Belgium) were used.The mean values, range

and SD were calculated.Pearson’s correlation coeffi-

cient was used for defining the correlation between

51Cr-EDTA clearance and serum creatinine, serum

cystatin C, the GFR calculated from the serum

creatinine-based formulas, and the GFR calculated

from the cystatin C formulas. In order to determine

the diagnostic accuracy of the serum cystatin C-based

formula in comparison with the other markers of

GFR, receiver operating characteristic (ROC) plots

were constructed and analyzed. The area under the

curve describes the test’s overall performance and is

used to compare different tests.Sensitivity and speci-

ficitywerecalculated.TheGFRdeterminedwith51Cr-

EDTA clearance was used as the gold standard and

the cut-off value was set at 60 mL/min/1.73 m2for

CKD, as defined by the National Kidney Foundation

(6).According to some previous studies that suggest

the importance of the lower cut-off value to define

CKD in the elderly, statistical analysis was also per-

formed for the cut-off value of GFR = 45 mL/min/

1.73 m2

(4). To compare

estimations of the GFR (MDRD formula, CKD-EPI

creatinine formula,and CKD-EPI creatinine and cys-

tatin formula) with

serumcystatinC-basedestimation,BlandandAltman

plots were used (29). The mean difference between

estimated and measured GFR values estimates the

global bias. The width of the SD of the mean differ-

enceisanestimationofprecision.Theaccuracywithin

30% for different equations was measured as the

percentage of results that did not deviate more than

30% from the measured GFR with51Cr-EDTA clear-

ance.Accuracy within 30% of the different stages of

CKD was analyzed.The ability to correctly estimate

the patient’s GFR below and above 60 and 45 mL/

min/1.73 m2with different equations compared to the

“gold standard” was also analyzed.

The study protocol was in conformity with ethical

guidelines and informed consent was obtained from

each participant.

thecreatinine-based

51Cr-EDTA clearance and the

RESULTS

The patient characteristics for those included in

this study are presented in Table 1. The age of

patients ranged from 65 to 90 years, giving a mean

age of 72.7 ? 5.1 years.The mean body mass index of

patients was 27.0 ? 4.7 kg/m2(women 27.0 ? 5.0 kg/

m2; men 27.1 ? 4.4 kg/m2). The mean

clearance in our patients was 34.5 ? 22.6 mL/min/

51Cr-EDTA

GFR Estimation in the Elderly 263

© 2011 The Authors

Therapeutic Apheresis and Dialysis © 2011 International Society for Apheresis

Ther Apher Dial,Vol. 15, No. 3, 2011

Page 4

1.73 m2. The mean serum creatinine concentration

value was 271.9 ? 175.7 mmol/L, and ranged from 61

to 871 mmol/L. The serum cystatin C concentration

values were between 0.6 and 7 mg/L, and the mean

value was 2.9 ? 1.5 mg/L.

A statistically significant correlation was found

between51Cr-EDTA clearance and serum creatinine

(r = -0.716; P < 0.0001), serum cystatin C (r = -0.783;

P < 0.0001), and with a GFR calculated from the

MDRD formula (r = 0.887; P < 0.0001), CKD-EPI

creatinine formula (r = 0.889; P < 0.0001), CKD-EPI

creatinine andcystatin

P < 0.0001), and simple cystatin C formula (r = 0.893;

P < 0.0001). In a comparison of the correlation coef-

ficients, we found that the correlations between51Cr-

EDTA clearance and the CKD-EPI creatinine and

cystatin formula were superior to the correlations

between

formula,CKD-EPI creatinine formula,or simple cys-

tatin C formula (P = 0.0117 for the MDRD formula,

P = 0.0163 for the CKD-EPI creatinine formula, and

P = 0.0308 for the simple cystatin C formula). No

difference between correlation coefficients of the

MDRD or CKD-EPI creatinine formulas and the

simple cystatin C formula was found (P = 0.7176 and

P = 0.8076).

formula(r = 0.923;

51Cr-EDTA clearance and the MDRD

Diagnostic accuracy (area under the ROC curves,

sensitivity, and specificity) at the cut-off values for

GFR of 60 and 45 mL/min/1.73 m2of the different

creatinine-based equations and the simple cystatin

C-based equation are presented in Table 2.The ROC

curve analysis for a GFR cut-off of 60 mL/min/

1.73 m2showed no statistically significant difference

of diagnostic accuracy between formulas.In the ROC

curve analysis for a GFR cut-off of 45 mL/min/

1.73 m2showed a higher diagnostic accuracy for the

equation that uses both serum creatinine and cystatin

C (CKD-EPI creatinine and cystatin formula) than

the MDRD formula (P < 0.013) or CKD-EPI creati-

nine formula (P < 0.01), but it was not higher than

that for the simple cystatin C formula (P = 0.335)

(Table 2, Fig. 1).

Bland and Altman analysis for the GFR cut-off

value of 60 mL/min/1.73 m2showed that the creati-

nine formulas (MDRD bias: -20.2; CKD-EPI creati-

nine bias: -22.2 mL/min/1.73 m2) and the CKD-EPI

creatinine and cystatin formula (bias: -20.8 mL/min/

1.73 m2) underestimated the measured GFR,and the

simple cystatin C formula (bias: 7 mL/min/1.73 m2)

overestimated it. Bland and Altman analysis for the

GFR cut-off value of 45 mL/min/1.73 m2showed less

underestimation for the creatinine formulas (MDRD

bias: -14.9; CKD-EPI creatinine bias: -17.2 mL/min/

1.73 m2) and the CKD-EPI creatinine and cystatin

formula (bias: -15.8 mL/min/1.73 m2), and, again,

overestimation for the simple cystatin C formula

(bias: 10.4 mL/min/1.73 m2) compared to measure-

ment of GFR with the gold standard.Analysis of the

SD of the mean difference between the estimated

and measured GFR showed that all equations lacked

precision (Table 3).

The accuracy within 30% of the estimated

EDTA clearance values differ according to the stages

51Cr-

TABLE 1.

Patient baseline characteristics

CharacteristicValue

Number of cases

Males/females

Age (years)

Body mass index (kg/m2)

Serum creatinine (mmol/L)

Serum cystatin C (mg/L)

317

147/170

72.7 ? 5.1

27.0 ? 4.7

271.9 ? 175.7

2.9 ? 1.5

Data are presented as N or mean ? SD.

TABLE 2.

GFR cut-off values of 60 mL/min/1.73 m2and 45 mL/min/1.73 m2(results for the latter are in brackets) of calculated

clearance using the MDRD, CKD-EPI creatinine, CKD-EPI creatinine and cystatin, and simple cystatin C formulas

Diagnostic accuracy (area under the ROC curves, sensitivity, and specificity) and comparison of ROC curves at

Equation AUC Sensitivity (%) Specificity (%)

P value*

P value**

MDRD formula0.955†

(0.969)‡

0.957

(0.969)

0.963

(0.984)

0.948

(0.978)

86.5

(90.9)

92.1

(92.3)

92.1

(89.5)

85.3

(92.7)

94.1

(94.8)

88.2

(93.8)

90.2

(99.0)

96.1

(96.9)

0.393

(<0.013)

0.564

(<0.01)

0.722

(0.421)

0.641

(0.402)

0.246

(0.335)

CKD-EPI creatinine formula

CKD-EPI creatinine and cystatin formula

Simple cystatin C formula0.246

(0.335)

*P calculated according to the CKD-EPI creatinine and cystatin formula.**P calculated according to the simple cystatin C formula.†The

unbracketed results are for a cut-off value for GFR of 60 mL/min/1.73 m2.‡The bracketed results are for a cut-off value for GFR of

45 mL/min/1.73 m2.The GFR determined with51Cr-EDTA was used as the gold standard.AUC, area under the curve; CKD-EPI, Chronic

Kidney Disease Epidemiology Collaboration; GFR, glomerular filtration rate; MDRD, modification of diet in renal disease; ROC, receiver

operating characteristic.

S Bevc et al.264

© 2011 The Authors

Therapeutic Apheresis and Dialysis © 2011 International Society for Apheresis

Ther Apher Dial,Vol. 15, No. 3, 2011

Page 5

of CKD (Table 4). In patients with stage 2 CKD, a

statistically significant higher accuracy within 30%

was found for the simple cystatin C formula (86.7%)

compared to the accuracy of the MDRD formula

(48.9%) (P < 0.0005), CKD-EPI creatinine formula

(44.4%) (P < 0.0005) or CKD-EPI creatinine and

cystatin formula (48.9%) (P < 0.0005). In patients

with moderate kidney impairment or with stage 4 or

5 of CKD, a statistically significant higher accuracy

within 30% was found for the MDRD, CKD-EPI

creatinine, and CKD-EPI creatinine and cystatin for-

mulas compared to the accuracy for the simple cys-

tatin C formula (Table 4).

Analysis of the ability to correctly predict GFR

below and above 60 or 45 mL/min/1.73 m2showed

high prediction for all formulas (from 83.6% to

90.9%). No statistically significant differences in

ability to correctly predict GFR below and above 60

or 45 mL/min/1.73 m2between formulas were found

(Table 3).

DISCUSSION

The current guidelines emphasize the need to

assess kidneyfunction

creatinine-based equations rather than just serum

creatinine.Additionally,these guidelines recommend

the use of the MDRD formula (8). Moreover, the

MDRD formula has been widely applied and evalu-

ated in numerous previously published studies;

however,some studies reported limitation (3).There-

fore,new creatinine-based CKD-EPI equations were

developed (11).In the last years,many new cystatin C

formulas were also developed, compared with crea-

tinine formulas,and proposed as an alternative filtra-

tion marker to creatinine (18,19,22,23,30,31). In our

study,we compared two widely used creatinine-based

equations, one equation that use both serum creati-

nine and cystatin C concentrations (CKD-EPI crea-

tinine and cystatin formula), and one very simple

cystatin C-based equation in a population of older

people.

Epidemiological studies have highlighted the high

prevalence of CKD in the elderly (8). In practice,

estimation of renal function in the elderly is based on

the measurement of serum creatinine and predictive

using the predictive

FIG. 1.

diagnostic accuracy of calculated clearance from the modification

of diet in renal disease (MDRD) formula, the Chronic Kidney

Disease Epidemiology Collaboration (CKD-EPI) study formulas,

and the simple cystatin C formula. The glomerular filtration rate

(GFR) determined with51Cr-EDTA was used as the gold standard,

and the GFR cut-off value was set at 45 mL/min/1.73 m2.

Receiver operating characteristic curve analysis of the

TABLE 3.

are in brackets) and the ability to correctly predict a patient’s GFR according to51Cr-EDTA clearance for the same cut-off

values in 317 elderly patients

Bias and precision for GFR cut-off values of 60 mL/min/1.73 m2and 45 mL/min/1.73 m2(results for the latter

Equation

Bias

(mL/min/1.73 m2)

Precision

(mL/min/1.73 m2)

Ability to correctly predict a patient’s

GFR below or above 60 mL/min/1.73 m2

(or 45 mL/min/1.73 m2)

MDRD formula

-20.2†

(-14.9)‡

-22.2

(-17.2)

-20.8

(-15.8)

7.0

(10.4)

14.9

(13)

14.9

(12.8)

12.3

(17.4)

17.8

(16)

88%

(86.8%)

90.9%

(83.6%)

89%

(86.1%)

87.1%

(85.8%)

CKD-EPI creatinine formula

CKD-EPI creatinine and cystatin formula

Simple cystatin C formula

†The unbracketed results are for a cut-off value for GFR of 60 mL/min/1.73 m2.‡The bracketed results are for a cut-off value for GFR of

45 mL/min/1.73 m2.CKD-EPI,Chronic Kidney Disease Epidemiology Collaboration;GFR,glomerular filtration rate;MDRD,modification

of diet in renal disease.

GFR Estimation in the Elderly265

© 2011 The Authors

Therapeutic Apheresis and Dialysis © 2011 International Society for Apheresis

Ther Apher Dial,Vol. 15, No. 3, 2011

Page 6

equations based on it.Age-related sarcopenia causes

a fall in creatinine production. Predictive equations,

including age and gender, partially take this factor

into account.The MDRD formula,however,can only

take into account the mean fall in muscle mass and

creatinine associated with age. Inflammation, malnu-

trition and loss of muscle bulk (often associated with

chronic diseases, such as heart failure and bronchop-

neumonia) can further accentuate the muscle meta-

bolic abnormalities and influence the value of serum

creatinine-based formulas (32,33). Overall, serum

cystatin C appears to be less sensitive to metabolic

and extrarenal factors than creatinine in the elderly

(33,34).Therefore,serum cystatin C-related formulas

seem to be promising markers for the estimation of

renal function in the elderly.

We have shown that the simple cystatin C formula

achieved at least as good a diagnostic performance as

the creatinine formulas, including the newer CKD-

EPI creatinine and CKD-EPI creatinine and cystatin

formulas.The ROC curve analysis for a GFR cut-off

of 60 mL/min/1.73 m2showed no statistically signifi-

cant difference of diagnostic accuracy between the

observed formulas. On the contrary, the ROC curve

analysis for a GFR cut-off of 45 mL/min/1.73 m2

clearly showed a higher diagnostic accuracy for the

equation that uses both serum creatinine and cystatin

C concentrations (CKD-EPI creatinine and cystatin

formula) than the MDRD and CKD-EPI creatinine

formulas, but not for the simple cystatin C formula.

Bland and Altman analysis for the same cut-off

values found out that all three creatinine-based

formulas (including the CKD-EPI creatinine and

cystatin formula) underestimated the measured

GFR (51Cr-EDTA clearance). The simple cystatin C

formula, however, always overestimated measured

GFR.The accuracy within 30% of the estimated gold

standard values demonstrated the superiority of the

simple cystatin C formula compared to the MDRD

and CKD-EPI formulas, but only in patients with

mildly impaired kidney function. On the contrary, in

patients with moderately-impaired kidney or with

CKD stage 4 or 5, a statistically significant higher

accuracy within 30% was found for the MDRD,

CKD-EPI creatinine, and CKD-EPI creatinine and

cystatin formulas, compared to the accuracy for the

simple cystatin C formula.Furthermore,in the analy-

sis of the ability to correctly predict a patient’s GFR

below or above 60 or 45 mL/min/1.73 m2, a very high

prediction for all formulas was found.

The results of the present study suggest that the

cystatin C-based prediction equation, which requires

just one variable (serum cystatin C concentration),

achieved a diagnostic performance that was at least

as good as the creatinine formulas using more vari-

ables and even formula combining creatinine and

cystatin C. The newest sophisticated CKD-EPI for-

mulas, as with the MDRD formulas, require addi-

tional calculator equipment, which is absolutely

unnecessary when using the simple cystatin C

formula.

Our study has some potential limitations. First, it

analyzes only an older, Caucasian population, while

the creatinine-based equations were developed from

studies that involved participants of all races. Thus,

direct comparison of equations can only be per-

formed between the CKD-EPI creatinine formula

(the equation includes a variable on the basis of race)

and the simple cystatin C formula. Second, previous

studies have been performed with different GFR ref-

erences and gold standards. In the original studies,

the MDRD and CKD-EPI formulas were compared

with125I-iothalamate as the “gold standard” for the

measurement of the GFR, while in our study,51Cr-

TABLE 4.

clearance values for different stages of chronic kidney disease (CKD)

Accuracy of formulas within 30% of the measured51Cr-EDTA

Stages of chronic kidney disease

CKD stage

No. of patients

GFR (mL/min/1.73 m2)

Stage 1

6

?90

Stage 2

45

60–89

Stage 3

95

30–59

Stage 4

113

15–29

Stage 5

58

<15

Equation

Accuracy within 30% of measured

51Cr-EDTA clearance (%)

MDRD formula

CKD-EPI creatinine

formula

CKD-EPI creatinine and

cystatin formula

Simple cystatin C formula

16.7

16.7

48.9

44.4

77.9

71.6

56.6

54.0

55.2

39.7

33.348.976.860.2 63.8

50.086.747.428.3 6.9

CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; GFR, glomerular fil-

tration rate; MDRD, modification of diet in renal disease.

S Bevc et al. 266

© 2011 The Authors

Therapeutic Apheresis and Dialysis © 2011 International Society for Apheresis

Ther Apher Dial,Vol. 15, No. 3, 2011

Page 7

EDTA clearance was used to estimate GFR; there-

fore, an exact direct comparison between these

studies is difficult. Finally, in our study the serum

cystatin C and the serum creatinine were measured

only once and so we cannot rule out known intrapa-

tient variability of serum creatinine or serum cystatin

C concentration,which can be present due to produc-

tion and/or extrarenal elimination.

CONCLUSION

Our study demonstrated that the simple cystatin C

formula could be a useful tool for the evaluation of

renal function in the elderly in daily clinical practice

in hospital and in outpatient clinics,and especially for

those with mild impaired kidney function. Despite

the advantages of the simple cystatin C formula, cys-

tatin C-based equations cannot completely replace

the “gold standard” for estimation of the GFR in a

population with CKD, but may contribute to a more

accurate selection of patients requiring such invasive

and costly procedures.

Acknowledgment: This study was supported by a grant

(L3-0328) from the Slovenian Research Agency (ARRS).

REFERENCES

1. Myers GL, Miller WG, 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;52:5–18.

2. Go AS, Chertow GM, Fan D, McCulloch CE, McCulloch CE,

Hsu CY. Chronic kidney disease and the risk of death, cardio-

vascular events, and hospitalisation. N Engl J Med 2004;351:

1296–305.

3. Stevens LA, Coresh J, Green T, Levey AS. Assessing kidney

function—measured and estimated glomerular filtration rate.

N Engl J Med 2006;354:2473–83.

4. Roderick PJ, Atkins RJ, Smeeth L et al. Detecting chronic

kidney disease in older people;what are the implications? Age

Ageing 2008;37:179–86.

5. Royal College of Physicians, Royal College of General Prac-

titioners Renal Association. Identification, Management and

Referral of Adults with Chronic Kidney Dusease. Guidelines

for General Physicians and Practitioners. London: Royal

College of Physicians, 2006.

6. Donadio C, Lucchesi A, Ardini M. Cystatin C, b2 microglobu-

lin, and retinol binding proteins as indicators of glomerular

filtration rate: comparison with plasma creatinine. J Pharm

Biom Anal 2001;24:835–42.

7. Levey AS, Berg RL, Gassmann JJ, Hall PM, Walker WG.

Creatinine filtration, secretion and excretion during progres-

sive renal disease. Kidney Int 1989;36(Suppl 27):S73–80.

8. National Kidney Foundation.K/DOQI clinical practice guide-

lines for chronic kidney disease: evaluation, classification, and

stratification. Am J Kidney Dis 2002;39(Suppl 1):S1–266.

9. Froissart M, Rossert J, Jacquot C, Paillard M, Houillier P.

Predictive performance of the Modification of Diet in Renal

disease and Cockcroft-Gault equations for estimating renal

function. J Am Soc Nephrol 2005;16:763–73.

10. Poggio ED,Wang X,Greene T,Van Lente F,Hall PM.Perfor-

mance of the Modification of Diet in Renal disease and

Cockcroft-Gault equations in the estimation of GFR in health

and in chronic kidney disease.JAm Soc Nephrol 2005;16:459–

66.

11. Levey AS, Stevens LA, Schmid CH et al. A new equation to

estimate glomerular filtration rate;CKD-EPI (Chronic Kidney

Disease Epidemiology Collaboration). Ann Intern Med 2009;

150:604–12.

12. Randers E, Erlandsen EJ. Serum cystatin C as an endogenous

marker of renal function: a review. Clin Chem Lab Med 1999;

37:389–95.

13. Coll E, Botey A, Alvarez L et al. Serum cystatin C as a new

marker for noninvasive estimation of glomerular filtration rate

and as a marker for early renal impairment. Am J Kidney Dis

2000;36:29–34.

14. Dharnidharka VR, Kwon C, Stevens G. Serum cystatin C is

superior to serum creatinine as a marker of kidney function: a

meta-analysis. Am J Kidney Dis 2002;40:221–6.

15. Hojs R, Bevc S, Antolinc B, Gorenjak M, Puklavec L. Serum

cystatin C as an endogenous marker of renal function in the

elderly. Int J Clin Pharm 2004;24:49–54.

16. Hojs R, Bevc S, Ekart R, Gorenjak M, Puklavec L. Serum

cystatin C as an endogenous marker of renal function in

patients with mild to moderate impaired kidney function.

Nephrol Dial Transplant 2006;21:1855–62.

17. Grubb A, Nyman U, Bjorak J et al. Simple cystatin C

based prediction equations for glomerular rate compared

with modification of diet in renal disease prediction equa-

tion for adults and the Schwartz and the Counhan-Barratt

prediction equations for children. Clin Chem 2005;51:1420–

31.

18. Hoek FJ,Kemperman FA,Krediet RT.A comparison between

cystatin C, plasma creatinine and the Cockcroft and Gault

formula for the estimation of glomerular filtration rate.

Nephrol Dial Transplant 2003;18:2024–31.

19. Larsson A, Malm J, Grubb A, Hansson LO. Calculation of

glomerular filtration rate expressed in mL/min from plasma

cystatin C values in mg/L.Scand J Clin Lab Invest 2004;64:25–

30.

20. MacIsaac RJ, Tsalamndris C, Thomas MC et al. Estimating

glomerular filtration rate in diabetes: a comparison of cystatin

C and creatinine-based methods. Diabetologia 2006;49:1686–

89.

21. Rule AD, Bregstraalh EJ, Slezak JM, Bergert J, Larson

TS. Glomerular filtration rate estimated by cystatin C

among different clinical presentations. Kidney Int 2006;69:

399–405.

22. Hojs R, Bevc S, Ekart R, Gorenjak M, Puklavec L. Serum

cystatin C-based equation compared to serum creatinine-

based equations for estimation of glomerular filtration rate in

patients with chronic kidney disease. Clin Nephrol 2008;70:

10–7.

23. Tidman M, Sjostrom P, Jones I. A comparison of GFR esti-

mating formulae based upon s-cystatin C and s-creatinine and

combination of the two. Nephrol Dial Transplant 2008;23:154–

60.

24. Ma YC, Zuo L, Chen JH et al. Improved GFR estimation by

combined creatinine and cystatin C measurements.Kidney Int

2007;72:1535–42.

25. Stevens LA,Coresh J,Schmid CH et al. Estimating GFR using

serum cystatin C alone and in combination with serum

creatinine: a pooled analysis of 3,418 individuals with CKD.

Am J Kidney Dis 2008;51:395–406.

26. Mazzachi BC, Peake MJ, Ehrhardt V. Reference range and

method comparison studies for enzymatic and Jaffé creatinine

assay in plasma and serum and early morning urine. Clin Lab

2000;46:53–5.

27. Blaufox MD, Aurell M, Bubeck B et al. Report of the Radio-

nuclides in Nephrourology Committee on renal clearance. J

Nucl Med 1996;37:1883–90.

GFR Estimation in the Elderly 267

© 2011 The Authors

Therapeutic Apheresis and Dialysis © 2011 International Society for Apheresis

Ther Apher Dial,Vol. 15, No. 3, 2011

Page 8

28. Perkins BA, Nelson RG, Ostrander BE et al. Detection of

renal function decline in patients with diabetes and normal or

elevated GFR by serial measurements of serum cystatin C

concentration: results of 4-year follow up study. J Am Soc

Nephrol 2005;16:1404–12.

29. Bland JM, Altman DG. Statistical methods for assessing

agreement between two methods of clinical measurement.

Lancet 1986;1:307–10.

30. Hojs R, Bevc S, Ekart R, Gorenjak M, Puklavec L. Serum

cystatin C formulas for prediction of glomerular filtration rate

in patients with chronic kidney disease. Nephron Clin Pract

2009;114:118–26.

31. Stevens LA, Coresh J, Feldman HI et al. Evaluation of the

modification of diet in renal disease study equation in a large

diverse population. J Am Soc Nephrol 2007;18:2749–57.

32. Coresh J, Astor B. Decreased kidney function in the elderly:

clinical and preclinical, neither benign. Ann Intern Med 2006;

145:299–301.

33. Seronie-Vivien S, Delanaye P, Pieroni L, Marita C, Froissart

M, Cristol JP. Cystatin C: current and future prospects. Clin

Chem Lab Med 2008;46:1665–86.

34. Wasen E,Isoaho R,Mattila K,Vahlberg T,Kivela SL,Irjala K.

Serum cystatin C in the aged: relationships with health status.

Am J Kidney Dis 2003;42:36–43.

S Bevc et al.268

© 2011 The Authors

Therapeutic Apheresis and Dialysis © 2011 International Society for Apheresis

Ther Apher Dial,Vol. 15, No. 3, 2011