**9. Studies comparing reference methods**

42 Basic Nephrology and Acute Kidney Injury

clearance and constant infused rate Iohexol (HPLC) Plasma clearance, samples at 5, 10, 20, 30, 45, 60, 90, 120, 180, 240, 300, 450, 600 min

clearance and constant infused rate Iohexol (XRF) Plasma clearance: samples at 150, 195, 240 + 360 min if estimated GFR under 30 mL/min

clearance and constant infused rate Iohexol (HPLC) Urinary clearance and constant infused rate

Correlation Regression BA

Correlation Regression BA

0.97 =0.994x+2.339 1.02±7

=0.971x+7.65 r²=0.96 (Io-inulin) =8.67±7.21

Wilcoxon Not different

41 CKD 6 to 160 Inulin: urinary

31 intensive care ±10 to 130 Inulin: urinary

20 healthy 106 to 129 Inulin: urinary

Table 5. Studies comparing iohexol with inulin. NA: not available, CKD: chronic kidney disease subjects, BA: Bland and Altman analysis, BAr: Bland and Altman analysis re-

iohexol, SC: subcutaneous, XRF: X ray fluorescence.

**8.3 Strengths and limitations** 

calculated by us, BM: Brochner-Mortensen, HPLC: high pressure liquid chromatography, Io:

Iohexol is probably the easiest way to measure GFR. It can be used in all patients (except in patient with true allergy to contrast product). Its measurement by HPLC is probably one of the most precise compared to other cold method (inulin and iothalamate). Iohexol is the less expensive marker and the cost of HPLC is also low. More important, it must be underlined that an external quality control does exist for iohexol measurement (Equalis, Sweden). From unpublished data, it can be concluded that the inter-laboratory CV for iohexol measurement is very low (less than 5%). Such results don't exist for iothalamate and inulin, and, at least for inulin, we think that such good inter-laboratory results would not be reached (personal

(Gaspari et al., 1995)

(Erley et al., 2001)

(Sterner et al., 2008)

In Table 6, we resumed the results of studies comparing reference markers (other than inulin). We selected studies in adults. We focused on studies having used the best statistical methods to analyze the results, i.e. the Bland and Altman analysis. It is difficult to interpret results from studies having compared different markers but also different methods (for example, plasmatic clearance of iothalamate with urinary clearance of 51Cr-EDTA) because it is impossible to affirm that potential differences are due to difference in markers or to difference in methods. Another limitation of several studies is the relatively small sample of subjects included. If we take into account these two limitations, we can stress on some interesting results showing good concordance (bias±SD) between plasma clearances of 51Cr-EDTA and 99Tc-DTPA (1.91±6.1 mL/min), and between plasma clearances of 51Cr-EDTA and iohexol (-0.16±6.17 mL/min in (Brandstrom et al., 1998), 4±7.9 mL/min in (Bird et al., 2009), 2±9.2 (Lundqvist et al., 1997), and -0.6±3.6 mL/min in (Pucci et al., 2001)).


How Measuring Glomerular Filtration Rate? Comparison of Reference Methods 45

Io (XRF): samples at 120, 165 and 210 for Dt samples at 120, 150, 180, 210 and 240 min for Io + BM correction

Io (HPLC): samples at 5, 10, 15, 30, 60, 90, 120, 150, 180, 210, 240, 270, 300 + 360 and 420 if creatinine>2 mg/dL + 1440 min if>5mg/dL

Io (XRF): samples at 20, 40, 60, 120, 180 and 240 min

±15 to 140 Cp of Cr and

Regression Correlation BA Io-Dt

Regression Correlation

BA Cr-Io

BA Cr-Io 4±7.9

Io=0.9938Dt+4.91 6 0.97 4.3±7.7

Type 1 Io=0.978Cr+0.132 0.999 Type 2 0.987 Io=0.078Cr+2.352

BA :-0.42±3.69

21 Diabetic 50 to 145 Cp of Dt and

41 Diabetic 29 to 150 Cp of Cr and

Table 6. studies comparing different reference methods (other than inulin), NA = not available, BA: Bland and Altman, BAr: Bland and Altman recalculated by us, BM: Brochner Mortensen,Cr: 51Cr-EDTA, Dt: 99Tc-DTPA, Io: iohexol, It: iothalamate, Cp: plasma clearance,

Cu: urinary clearance, IC: constant infusion rate, IB: bolus injection, IV: intravenous, SC: subcutaneous, AUC: area under the curve, NS: not significant, HPLC: high pressure

In this chapter, we reviewed all the reference methods available in 2011 to measure GFR. Among these methods, inulin clearance can certainly be considered as the gold-standard because it is historically the first method used and because this marker is certainly the best characterized from a physiological point of view. However, inulin is expensive and commercial sources are limited (Gaspari et al., 1997). Due to its high molecular weight, there are doubts to use inulin in simplified plasma clearance (urinary clearances with constant infusion rate remain necessary but are very cumbersome). Measurement of plasma inulin is neither easy nor standardized. For all these reasons, the use of inulin is and will always be relatively marginal. In 2011, it is maybe time to move from the perfect physiological marker (inulin) to markers, maybe less perfect in the renal physiologic handling, but less costly, easier to use everywhere in the world and with a standardized measurement. From our point of view, iohexol is probably the best marker with the best balance between

Houlihan (Houlihan et al., 1999)

(Pucci et al., 2001)

Bird (Bird et al., 2009)

56

CKD

healthy

liquid chromatography, XRF: X ray fluorescence.

**10. Conclusions and perspectives** 

19


It (HPLC): bolus IV and samples at 5, 10, 20, 30, 40, 50, 60, 90, 120, 180, 240, 300, 450 and 600 min

Io (XRF): samples à 180, 210, 240 and 270 minutes+ BM correction day 1 and 2

Dt: samples at 60 and 180 min

It (HPLC)

Io (XRF): samples at 180, and 240 or 270 min+ BM correction

Io (HPLC and XRF): samples at 150, 195 and 240 min + BM correction

Io (HPLC): samples at 5, 10, 15, 30, 60, 90, 120, 150, 180, 210, 240, 270, 300 + 360 and 420 if creatinine>2 mg/dL + 1440 min if>5mg/dL

±70 to 130 Cp of Cr and

Correlation Regression BAr It-Cr

Regression BA Dt-Cr

Regression BAr It-Cr

Correlation Regression BA (Io-Cr)

Regression Correlation BA Cr-Io

Regression Correlation BA Cr-Io

0.995 It=1.007Cr-0.303 -0.1±4.7

Day 2 : +0.9±5.9

Dt=0.982Cr+3.2 1.91±6.1

It=0.979Cr-3.04 1.3±5

0.918 Io=0.892Cr+6.28 2±9.2

XRF Io=1.03Cr-1.79 0.97 0.58±4.95 HPLC Io=1.05Cr-4.43 0.96 -0.16±6.17

0.995 Io=0.978Cr+2.45 -0.6±3.6

BA Cr-Io Day 1 : +2.1±10.2

19 CKD 7 to 148 Cp of Cr and

50 NA ±15 to 160 Cp of Cr and

17 Diabetic 7 to 105 Cu of Cr and

77 Urography ±25 to 125 Cp of Cr and

49 GFR>40 ±40 to 125 Cp of Cr and

32 Diabetic 13 to 151 Cp of Cr and

(Gaspari et al., 1992)

(Lundqvist et al., 1994)

(Galli et al., 1994)

(Sambataro et al., 1996)

(Lundqvist et al., 1997)

(Brandstrom et al., 1998)

(Pucci et al., 1998)

31 Para or

tetraplegic


Table 6. studies comparing different reference methods (other than inulin), NA = not available, BA: Bland and Altman, BAr: Bland and Altman recalculated by us, BM: Brochner Mortensen,Cr: 51Cr-EDTA, Dt: 99Tc-DTPA, Io: iohexol, It: iothalamate, Cp: plasma clearance, Cu: urinary clearance, IC: constant infusion rate, IB: bolus injection, IV: intravenous, SC: subcutaneous, AUC: area under the curve, NS: not significant, HPLC: high pressure liquid chromatography, XRF: X ray fluorescence.
