serum iohexol analysis by micellar electrokinetic capillary chromatography

Zak K. ShihabiMark E. Hinsdale

Department of Pathology,Wake Forest University Schoolof Medicine,Winston-Salem, NC, USA

Received September 8, 2005Revised February 27, 2006Accepted February 28, 2006

Research Article

Serum iohexol analysis by micellarelectrokinetic capillary chromatography

A simple and rapid (,4 min) method for the measurement of iohexol in serum forassessing the glomerular filtration rate is described. It is based on direct seruminjection on the capillary by MEKC. The method is linear between 8 and260 mg/L, with an RSD of peak height of 2.9%. Several simple steps havecontributed to an improved daily precision, such as choosing a high pH buffer,increasing the SDS concentration, frequent standardization, and eliminating anysample pretreatment.

Keywords: Caffeine / Glomerular filtration rate / MEKC / Omnipaque / Radiographiccontrast DOI 10.1002/elps.200500667

1 Introduction

The main feature of CE in research is the high theoreticalplate number with ease of altering the analysis selectivity.On the other hand, the main but not well-appreciatedfeature of CE in routine work is the ease of analysis. TheMEKC mode of CE allows direct injection of a complexmatrix, like serum proteins, on the capillary withoutadverse effects. This is not easily achievable in HPLC.Several workers have demonstrated the ease and flex-ibility of the MEKC for analysis of neutral and weaklycharged compounds [1–4]. Here, we demonstrate howdirect sample injection without any treatment can beapplied to the routine analysis of iohexol in serum forassessing the glomerular filtration rate (GFR) and withvery good daily reproducibility.

Several tests are employed clinically for accessing therenal function as GFR. All these have some inherentproblems. For example, creatinine clearance is commonlyused as an endogenous test for GFR determination. It iswell recognized that it is not very accurate because crea-tinine is secreted by the tubule and also depends onmuscle mass and requires careful urine collection. Esti-mated GFR based on serum creatinine corrected forpatient age and race gives better but not accurate infor-

mation especially near the normal range. Inulin clearanceis considered the gold standard for GFR measurement [5].However, because it requires pump infusion and collec-tion of multiple urine and blood samples it is not suitablefor routine patient care. Several organic compoundslabeled with radioisotopes have been used to replace theinulin test such as the 125I-iothalamate, 99m Tc-diethylene-triaminopentaacetic acid, and 51Cr-labeled ethylene-dia-minetetraacetic acid. These latter tests which are basedon blood rather than urine are less demanding than theinulin test; however, they are expensive and involve theadministration of radioisotopes to the patient. Iohexol(N,N-bis(2,3-dihydroxypropyl)-5-[N-(2,3-dihydroxypropyl)-acetamido]-2,4,6-triiodoisophthalamide), eliminates thelatter problem and avoids the need for multiple samplecollection while giving results for GFR similar to thatobtained by the previous compounds [6–8] including inu-lin [5]. Iohexol, known also as Omnipaque, is a tri-iodi-nated, nonionic radiographic contrast medium. It is usedfrequently in cardiology for angiography. It is a non-metabolizable compound and does not bind to serumproteins. It is distributed mainly in the plasma and iscleared by the kidney. For this reason iohexol has beenadvocated as a good substance for measuring GFR [6–8].In some patients, high doses of this compound elicitsome side effects such as nausea, angina, bradycardia,and nephrotoxicity [9–11]. In order to be used routinely forGFR measurement a simple, rapid, and sensitive assaymethod for this compound is needed. Serum iohexol hasbeen analyzed previously by HPLC after precipitation withperchloric acid [12, 13] and also by measuring thereleased iodine by the ceric reaction [14]. A commercialinstrument dedicated solely for measurement of this

Correspondence: Professor Zak K. Shihabi, Department of Pathol-ogy, Wake Forest University School of Medicine, Winston-Salem, NC27157, USAE-mail: [email protected]: 11-336-716-9944

Abbreviation: GFR, glomerular filtration rate

2458 Electrophoresis 2006, 27, 2458–2463

© 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

Electrophoresis 2006, 27, 2458–2463 CE and CEC 2459

compound, based on X-ray fluorescence has beendescribed [5, 15] indicating further the importance of thistest. We measured previously iohexol by CZE after proteinprecipitation with ACN [16, 17]. Bergert et al. [18] de-scribed the analysis of iothalamic acid by CE for GFRusing both serum and urine. Here, we describe a simplerand rapid method for iohexol assay based on MEKC byCE and we describe its application for GFR measure-ment. The main advantage is that the sample is injecteddirectly without any treatment or dilution leading to a verygood reproducibility.

2 Materials and methods

2.1 Instrument

Model 2050 CE instrument (Beckman Instruments, PaloAlto, CA, USA) equipped with an untreated capillary27 cm650 mm (id) was set at 8 kV (normal polarity), 247C,and 254 nm.

2.2 Separation buffer

Boric acid 136 mmol/L, pH 9.3 containing 30 mg/mL ofSDS.

2.3 Method

Serum samples were introduced on the capillary directlywithout any treatment by low pressure injection (3447 Pa)for 6 s. After each sample, the capillary was washed with1 mol/L NaOH for 40 s and filled with fresh buffer by pres-sure injection for 40 s. Initially a new capillary was washedfor 5 min with sodium hydroxide 1 mol/L and with waterfor 5 min followed by the buffer for 5 min.

2.4 Standards (130 mg/L)

Iohexol, stock standard (Winthrop Pharmaceuticals, NewYork, NY) can be diluted in 1% NaCl or in pooled serumfree from this compound. This compound is very viscousrequiring special care in dilution.

3 Results and discussion

Small amounts of serum can be injected directly whenMEKC is used. The micelles dissolve serum proteinseliminating the need for their removal. However, the anal-ysis conditions should be chosen such that the drug ofchoice migrates far from the proteins as well as otherpossible interferences. This can be achieved by manip-

Figure 1. Separation of iohexol (I) by MEKC from: (top)serum of a patient free from iohexol, (middle) a patientinjected with 8 mL of iohexol; and (bottom) iohexolstandard (130 mg/L) prepared in 1% NaCl. Samplesinjected directly without any treatment using the buffer atpH 9.3.

ulating the pH, SDS concentration, and detector wave-length. Figure 1 is a representative electropherogram foranalysis of iohexol. The electropherogram is relativelyclean. Iohexol emerges rapidly at about ,4 min.


2460 Z. K. Shihabi and M. E. Hinsdale Electrophoresis 2006, 27, 2458–2463

An internal standard such as iothalamic acid (anothermarker for renal function [18]) can be added (Fig. 2). Internalstandards are most useful when the electropherogramcontains multiple peaks and peak areas are used for calcu-lation; however, this involves extra steps. Each additionalstep of sample manipulation can lead to slight imprecision.Thus, for routine analysis we elected to inject the serum di-rectly without the addition ofan internal standard. The effectof pH on peak height is presented in Fig. 3. There is a slightincrease in height with increase in the pH.

Precision of the analysis is of utmost importance in GFRmeasurement. It should be very good, i.e., ,3%, in orderto justify injecting exogenous compounds to the patient.At pH of 9.3, there was an improvement in the precision(RSD = 0.69%, n = 10) compared to pH 8.5 (RSD =1.63%, n = 10) although the patient values at the two pHrelative to the standards are similar, Fig. 4 indicating theabsence of interferences. At pH 8.5 the peak heightkeeps decreasing with subsequent sample injectionrequiring frequent standardization. The improvement inprecision at pH 9.3 probably is due to the decrease ofserum protein binding to the capillary walls at that high pH[19]. The high SDS concentration used here also con-tributes to the overall good precision because it alsoleads to better serum protein solubilization. For example,the precision at 5 mg/mL and 30 mg/mL SDS is 1.6 and1.1%, respectively. To further improve the precision indi-rectly we elected to inject after every ten samples astandard. Kitahashi and Furuta [20] measured recentlyiohexol in serum also by CE. In spite of long multiple washsteps (15 min vs. 80 s in this work) a higher RSD resulted;and in spite of the high voltages (25 vs. 8 kV in this work)used, the migration time was twice as long as in thepresent method. This stresses again that several stepsare important for the improved precision in CE; some ofthese have not been explored.

The recovery in serum is close to 100% compared to thatin 1% NaCl but it is 27% higher compared to that in water(Fig. 5). This indicates that salts (including those inserum), under the present analysis conditions, causeslight stacking but more importantly that the standardshould be prepared in 1% sodium chloride or serum andnot in water. The test was linear by peak height, between8 and 260 mg/L (r = 0.999). The linearity should be betterbased on peak area. The minimum detection level (at 36baseline noise) is 2 mg/L. An increase in the SDS con-centration from 10, 20 to 30 mg/mL increased the peakheight of iohexol slightly. It also increased slightly theretention time as expected [21].

Application of this method to the analysis of 36 serumsamples showed that 2 h after a dose of 8 mL of iohexoladministered intravenously to the patient for GFR, the

Figure 2. Use of Iothalamic acid as internal standard iniohexol determination; same conditions as in Fig. 1.



Figure 3. Effect of pH on peak height of iohexol (boricacid 136 mmol/L, at different pH containing 30 mg/mL ofSDS).

Figure 4. Comparison of iohexol analysis for 36 patientsperformed with boric acid 136 mmol/L containing 30 mg/mL SDS at both pH 8.5 and 9.3. Same patient was ana-lyzed at both the pH values.

mean level was 68 mg/L. Thus, the injected dose, which isexpensive, can be cut into half without much sacrifice insensitivity or precision. The precision, based on duplicatesamples (n = 36) over 7 days, was 2.9%. Using 30 mg/mLSDS concentration, we did not encounter any inter-ference from endogenous compounds or other common

Figure 5. Effect of the matrix of the standard on peakheight of iohexol. Standard was diluted in water, serum, or1% NaCl.

drugs such as phenytoin, phenobarbital, theophylline,etc. On the other hand, at lower SDS concentration someof these compounds migrated very close to the iohexolpeak.

Other compounds can also be measured in the serum bythe same method. For example, caffeine which is used fortreatment of apnea in the neonate can be measured di-rectly in the same buffer and by the same method simplyafter changing the wavelength to 280 nm (Fig. 6).

4 Concluding remarks

This method illustrates the simplicity of the analysis ofiohexol in serum for routine measurement of GFR by usingMEKC. Precision is of great priority in routine analysis.This method has a very good precision, rapid analysistime, and simplicity, which makes it suitable for routineanalysis. The combination of several simple steps, someof these not reported before, contributed to the improvedprecision of this method, such as eliminating any sampletreatment, keeping the pH of the analysis high, anincrease in the SDS concentration, and running standardsevery few samples.

Based on our experience, iohexol on the HPLC columnelutes as a wide peak on most of the columns, Fig. 7,probably due to the multiple functional groups, leading topoor linearity by peak height. On the other hand, in the


2462 Z. K. Shihabi and M. E. Hinsdale Electrophoresis 2006, 27, 2458–2463

Figure 6. Caffeine (50 mg/L, final concentration) analysisby MEKC using the same buffer and electrophoresisconditions as of iohexol except wavelength was set at280 nm: (top) serum blank (free from caffeine), and (bot-tom) patient on caffeine.

Figure 7. Analysis of iohexol(I) by HPLC; (A) Standard130 mg/L and (B) patient 102 mg/L. Analysis conditions:Microspher C18 column, 5 mm, 4.66250 mm (Varian,Palo Alto, CA, USA) at 254 nm. The column was elutedwith 0.1% phosphoric acid. Sample was deproteinizedwith two volumes of ACN and diluted five times withpump solvent.

MEKC it elutes as a very sharp peak with good linearity bypeak height. None of the common drugs or endogenoussubstances interfered with the analysis. The method canbe applied for other compounds in serum such as caf-feine. The speed and ability to inject serum directly on thecapillary in MEKC without dilution or deproteinization is agood example where analysis by CE outperforms that byHPLC.



5 References

[1] Pappas, T. J., Gayton-Ely, M., Holland, L. A., Electrophoresis2005, 26, 719–734.

[2] Kunkel, A., Wätzig, H., Electrophoresis 1999, 20, 2379–2389.

[3] Terabe, S., Anal Chem. 2004, 76, 241A–246A.

[4] Welsch, T, Michalke, D., J. Chromatogr. A 2003, 1000, 935–951.

[5] Brown, S. C. W., O’Reilly, P. H., J. Urol. 1991, 146, 675–679.

[6] Stake, G., Monclair, T., Scand. J. Clin. Lab. Invest. 1991, 51,335–342.

[7] Stake, G., Moon, E., Rootwelt, K., Monclair, T., Scand. J.Clin. Lab. Invest. 1991, 51, 343–348.

[8] Eriksson, C.-G., Kallner, A., Clin. Biochem. 1991, 24, 261–264.

[9] Campbell, D. R., Flemming, B. K., Mason, W. F., Jackson, S.A. et al., Can. Assoc. Radiol. J. 1990, 41, 133–137.

[10] Hofmeister, R., Bhagava, A. S., Gunzel, P., Toxicol. Lett.1990, 50, 9–15.

[11] Steinberg, E. P., Moore, R. D., Powe, N. R., Gopalan, R. etal., New Engl. J. Med. 1992, 36, 425–430.

[12] Krutzen, E., Back, S. E., Nilsson-Ehle, P., Scand. J. Clin. Lab.Invest. 1990, 50, 279–283.

[13] Krutzen, E., Buck, S. E., Nilsson-Ehlle, I., Nillson-Ehlle, P., J.Lab. Clin. Med. 1984, 104, 955–961.

[14] Back, S. E., Masson, P., Nilsson-Ehle, P., Scand. J. Clin. Lab.Invest. 1988, 48, 825–829.

[15] Gronberg, T., Sjoberg, S., Almen, T., Golman, K., Mattsson,S., Invest. Radiol. 1983, 18, 445–452.

[16] Shihabi, Z. K., Constantinescu, M. S., Clin. Chem. 1992, 38,2117–2120.

[17] Rocco, M. V., Buckalew, V. M., Moore, L. C., Shihabi, Z. K.,Am. J. Kidney Dis. 1996, 28, 173–177.

[18] Bergert, J. H., Liedtke, R. R., Oda, R. P., Landers, J. P., Wil-son, D. M., Electrophoresis 1997, 18, 1827–1835.

[19] Graf, M., García, R. G., Wätzig, H., Electrophoresis 2005, 26,2409–2417.

[20] Kitahashi, T., Furuta, I., J. Pharm. Biomed. Anal. 2004, 34,153–158.

[21] Shihabi, Z., Electrophoresis 2004, 25, 1648–1651.


serum iohexol analysis by micellar electrokinetic capillary chromatography

Documents