underestimation of rat serum vancomycin concentrations measured by an enzyme-multiplied immunoassay...

Research articleDrug Testing

and Analysis

Received: 28 August 2012 Revised: 21 March 2013 Accepted: 22 June 2013 Published online in Wiley Online Library: 13 August 2013

(www.drugtestinganalysis.com) DOI 10.1002/dta.1511

350

Underestimation of rat serum vancomycinconcentrations measured by an enzyme-multiplied immunoassay technique andthe strategy for its avoidanceHiroki Konishi,a* Ikumi Igaa,b and Katsuhito Nagaia

An enzyme-multiplied immunoassay technique (EMIT) has been widely adopted for the measurement of serum concentrationsof vancomycin (VCM) in clinical practice. Because of the growing demand for its application to fundamental pharmacokineticstudies, we examined whether VCM concentrations in rat serum were accurately measured by EMIT. It was found thatmeasured values of known amounts of VCM spiked to rat serum were markedly underestimated with a large analyticalvariance. When ultrafiltrated rat serum was used as the sample matrix, interference was significantly improved, and thedegree of underestimation was attenuated also by diluting samples with physiological saline. These results suggest thatendogenous substances of a high molecular weight in rat serum interfere with the analysis of VCM concentrations by EMIT.However, measured values of rat serum VCM concentrations by EMIT were restored to theoretical levels by exposing samplesto 70°C for 3–7min. A likely explanation for the avoidance of interference is that an appropriate thermal force eliminated theimmunological function of endogenous substances falsely recognizing VCM without affecting the VCM molecule itself.Regarding serum samples collected from rats that were administered VCM, values measured by EMIT following theheat-treatment agreed well with those by the high performance liquid chromatography (HPLC) method. This is the first reportshowing interference by endogenous high-molecular substances in the measurement of drug concentrations in rat serumusing EMIT. Our findings will contribute to the appropriate use of VCM based on evidence provided by clinical-oriented ratexperiments requiring the measurement of serum VCM concentrations by EMIT. Copyright © 2013 John Wiley & Sons, Ltd.

Keywords: vancomycin concentration; enzyme-multiplied immunoassay technique; rat serum; underestimation; Viva-E analyzer

* Correspondence to: Hiroki Konishi, Laboratory of Clinical Pharmacy andTherapeutics, Faculty of Pharmacy, Osaka Ohtani University, 3-11-1 Nishikiori-kita,Tondabayashi 584-8540, Japan. E-mail: [email protected]

a Laboratory of Clinical Pharmacy and Therapeutics, Faculty of Pharmacy,Osaka Ohtani University, Tondabayashi, Japan

b Division of Pharmacy, National Hospital Organization, Shiga Hospital,Higashiomi, Japan

Introduction

Immunoassay is a common method used for the detection andquantification of low levels of specific substances in biologicalfluids containing numerous components. Among various immu-noassays, the enzyme-multiplied immunoassay technique (EMIT)involves a homogeneous processing system and thus allows au-tomatic analysis without requiring separation of the antibody-bound enzyme-analyte conjugate and free conjugate. Becauseof its analytical simplicity and rapidity, EMIT is widely used inthe measurement of blood drug concentrations for routine ther-apeutic monitoring and intoxication management of patients inclinical practice.[1–4] A variety of drugs including anti-epileptics,antibiotics, immunosuppressants, anti-arrhythmics, cardiac glyco-sides, antiasthmatics, and anticancer agents are analyzable usingEMIT; however, these agents have narrow therapeutic rangeseven though their clinical utility has been validated. Monitoringof blood drug concentrations thus plays an important role inachieving the maximum therapeutic outcome along withavoiding adverse events in pharmacotherapy, while fundamentalpharmacokinetic and toxicokinetic examinations using experimen-tal animals also contribute useful information for the assessment ofpharmacological and toxicological responses to administereddrugs. Although the measurement of drug concentrations in blood

Drug Test. Analysis 2014, 6, 350–356

obtained from experimental animals has been performed usingEMIT in several studies,[5–7] the reliability of measured values hasnot been fully examined despite the use of blood samples fromsources other than human origin.

Vancomycin (VCM) is a glycopeptide antibiotic that exerts itsbactericidal activity by inhibiting cell wall biosynthesis of Gram-positive bacteria and is widely used for the treatment of infectiousdiseases caused by methicillin-resistant Staphylococcus aureus.However, administration of VCM carries a considerable risk ofcausing adverse events including nephrotoxicity in spite of excel-lent antimicrobial efficacy, and this increased risk is dependent onthe characteristics of VCM's disposition associated with the degreeof its systemic exposure.[8] In addition, the pharmacokinetics of

Copyright © 2013 John Wiley & Sons, Ltd.

Underestimation of rat serum vancomycin concentration by EMIT

Drug Testing

and Analysis

VCM is susceptible to change depending on various physiologicalconditions especially renal impairment.[9] From these aspects, anumber of pharmacokinetic studies of VCM using rats have beenextensively conducted to provide supportive evidence for promot-ing the appropriate and improved use of this drug. In theseexperiments, VCM concentrations in rat serum or plasma weremeasured by means of various analytical techniques such asfluorescence polarization immunoassay (FPIA),[10–18] high perfor-mance liquid chromatography (HPLC),[19–25] liquid chromatogra-phy- mass spectrometry (LC-MS),[26,27] and bioassay.[28] However,it remains unclear whether or not EMIT is capable of making accu-rate measurements of serum VCM concentrations in rats.

This paper describes a marked underestimation of VCMconcentrations in rat serum when measured by EMIT using aViva-E analyzer. In addition, a strategy for avoiding this interfer-ence is also demonstrated.

351

Experimental

Drugs and reagents

Standard powder of VCM hydrochloride was purchased fromSigma-Aldrich Co. (St Louis, MO, USA). A commercial pharmaceuti-cal product of VCM for injection (Vancomycin hydrochloride 0.5gTM, containing 500mg VCM per vial, Shionogi & Co., Ltd, Osaka,Japan) was used for intravenous administration to rats. Six kindsof standard artificial human sera (standard sera) supplementedindividually with designated amounts of drugs (VCM, phenobarbi-tal, phenytoin, carbamazepine, valproic acid, theophylline) wereobtained from Abbott Japan, Co., Ltd (Tokyo, Japan) and used assample sources for evaluating the assay reliability of EMIT and FPIA.

Sample preparation

Standard VCM was weighed and dissolved in distilled water toprepare aqueous stock solution (1mg/ml). An aliquot of theVCM stock solution was pipetted into conical plastic tubes, anda small volume of water was spontaneously evaporated to dry-ness at room temperature with the lid left open. The dried VCMmatrix was dissolved in rat serum, human serum, filtrated ratserum, or a mixture of rat serum and physiological saline to makedefined concentrations. Rat and human serum were obtained asdescribed below. When the VCM concentration was anticipatedto exceed the upper limit of the quantifiable range, serum sam-ples were diluted with physiological saline prior to measurement.

To examine the analytical performance of EMIT in measuringvarious drugs in rat serum, standard sera supplemented withVCM, phenobarbital, phenytoin, carbamazepine, valproic acid,or theophylline at labelled concentrations were diluted withthree-fold volume of drug-free rat serum and applied to EMIT.

Immunoassay analyzers and operation

EMIT for determining concentrations of VCM and the five otherdrugs (phenobarbital, phenytoin, carbamazepine, valproic acid,and theophylline) in fluid samples mixed with rat serum and stan-dard sera was performed on a Viva-E analyzer (Siemens HealthcareDiagnostics, Liederbach, Germany). As required, the validity ofmeasured values was examined by applying the same samples toFPIA using a TDxFLx immunochemistry analyzer (Abbott Japan,Co., Ltd, Tokyo, Japan). Analyzers were operated according to eachmanufacturer's instruction. All reagents necessary for EMIT and

Drug Test. Analysis 2014, 6, 350–356 Copyright © 2013 John W

FPIA were packaged as dedicated assay kits for each analyzer,and surfactant-containing phosphate buffer, was separately pro-vided as a reaction medium for FPIA.

In EMIT, serum samples were mixed with the drug labelled withglucose-6-phosphate dehydrogenase (G6PDH), and then the re-agent composed of mouse monoclonal antibodies to the drug,a substrate glucose-6-phosphate (G6P), and a coenzyme NADwas subsequently added for competitive antibody binding be-tween the drug in serum and the labelled drug. As binding ofthe G6PDH-drug conjugate to the antibody renders G6PDH inac-tive, the serum drug concentration was determined on the basisof G6PDH activity after allowing the formation of immune com-plex. G6PDH activity was calculated by an absorbance changelinked to the reductive conversion of NAD to NADH. In FPIA, se-rum samples were mixed with the fluorescein-labelled drug andanti-drug antibodies. The serum drug concentration was calcu-lated from the fluorescence response, since decreased rotationalmotion of the fluorescein-drug conjugate bound to the antibodyleads to an increase in the polarization of fluorescent light. Theseoperational procedures were carried out automatically. Calibra-tion curves for determining the serum concentration of eachdrug were constructed by a nonlinear least-squares curve fit,using calibrators that included standard sera consisting of sixlevels of defined amounts of respective drugs, and were storedin the instrument's memory. Calibrators for EMIT and FPIA weresupplied by the corresponding manufactures (Abbott Japan andSiemens Healthcare Diagnostics).

HPLC analysis of VCM concentrations in rat serum

Changes in the concentration of VCM in rat serum by heating wereexamined by HPLC. HPLC determination of VCM was conductedaccording to a previous report [29] with minor modifications. To100μl of rat serum samples, 150μl of a mixture of 4% zinc sulfatesolution and methanol (1:1) was added and vortexed for 30 sec.After centrifugation at 3000g at room temperature for 10min,20μl of supernatant was injected onto the HPLC. The HPLC system(Shimadzu Co., Ltd, Kyoto, Japan) consisted of a pump (Model LC-20AD), UV-VIS detector (Model SPD-20A; Kyoto, Japan), and a sys-tem controller (Model CBM-20A). Separation of VCM was achievedon an ODS column (Shim-pack CLC-ODS column; 150×4.6mm;Shimadzu GLC, Tokyo, Japan). The mobile phase consisted of50mM acetate buffer (pH4.5) and acetonitrile (92.5:7.5, v/v). Theflow rate and detection wavelength were 1ml/min and 220nm,respectively, and the column temperature was maintained at 40°C.

Rat treatment and blood collection

Male Wistar rats were obtained from Japan SLC Inc. (Hamamatsu,Japan). Rats were given standard rodent chow with free access totap water and acclimatized for at least 2 days prior to experimen-tal use. In application to the pharmacokinetic study, VCM wasadministered at a dose of 50mg/kg via a polyethylene cannulaplaced into the left femoral vein. At appropriate times afterVCM administration, blood was withdrawn through a polyethyl-ene cannula inserted into the left femoral artery. VCM-free bloodsamples were collected from rats using a similar procedurewithout administration of VCM, and those from humans wereobtained after receiving consent for blood collection. Collectedblood samples were centrifuged to separate the serum fraction.

Rats were handled in accordance with the Guidelines forAnimal Experimentation of Osaka Ohtani University, and the

iley & Sons, Ltd. wileyonlinelibrary.com/journal/dta

Figure 2. Effect of serum dilution with physiological saline on the mea-surement of VCM concentrations by EMIT. VCM was spiked to solutionscomprising rat serum and physiological saline. Mixing rates of physiolog-ical saline in the total volume were 0%, 50%, and 90%. The VCM concen-tration in each mixture was 25μg/ml, and was subjected to EMIT usingViva-E. Data were expressed as the mean± SD of four experiments.*p< 0.05, **p< 0.01

H. Konishi, I. Iga and K. Nagai

Drug Testing

and Analysis

352

experimental protocol was approved by the Animal Care and UseCommittee of this institution.

Serum ultrafiltration

An aliquot portion of VCM-free rat serum was added to an ultrafil-tration device (Nanosep 30K, Pall Corporation, Port Washington,NY, USA) with a molecular weight cut-off of 30 kDalton. Using thisdevice, samples were centrifuged with a fixed angle rotor at5000g for 10min, and the filtrated serum was used for theexperiment.

Heat-treatment of serum

To examine whether the heat-treatment of rat serum can influ-ence measured values on EMIT and HPLC, serum samples werespiked with known amounts of VCM and then kept at 50, 60,70, and 80°C for 0–10min prior to measurement.

Statistical analysis

Results were expressed as the mean± SD. As required, intergroupcomparisons were made by a Student's t-test or one way analysisof variance followed by the Tukey's test. A p value less than 0.05was considered significant.

Results

Assay reliability of rat serum VCM concentrations on EMIT

VCM was added to human and rat serum at a concentration of25μg/ml, and the assay reliability of EMIT was examined on Viva-E.The validity of EMIT was confirmed by the agreement betweenmeasured and theoretical values in the measurement of VCM con-centrations in human serum. However, the concentration of VCMin rat serumwas significantly underestimated (p< 0.01), and a largevariance was observed among individual measurements (Figure 1).To test the effect of dilution of serum samples with physiolog-

ical saline on assay reliability, measurements were made usingmixtures of rat serum and physiological saline as sample sources.Each mixture sample was prepared so as to contain 25μg/mlVCM. As shown in Figure 2, measured values were increased with

Figure 1. Underestimation of VCM concentrations in rat serum inmeasure-ments with EMIT. VCM was spiked to human serum and rat serum at a con-centration of 25μg/ml, and samples were measured by EMIT using Viva-E.Data were expressed as the mean±SD of four experiments. **p< 0.01

wileyonlinelibrary.com/journal/dta Copyright © 2013 Jo

an increasing percentage of physiological saline in the mixture,although they did not attain the theoretical level.

On the other hand, when filtrated rat serum was used as asubstitute for original rat serum, measured concentrations ofspiked VCM were pronouncedly improved and approachedtheoretical values (Figure 3).

Assay reliability of drug concentrations in rat serum on EMITand FPIA

To examine general applicability of EMIT and FPIA to assays of avariety of drugs, concentrations of phenobarbital, phenytoin, car-bamazepine, valproic acid, theophylline, and VCM in the standardfluid mixture containing rat serum were measured by theseimmunoassay techniques. Concentrations of drugs other thanVCM could be determined accurately on EMIT. In contrast, mea-surements using FPIA with a TDxFLx analyzer gave accurate

Figure 3. Effect of ultrafiltration of rat serum on the measurement ofVCM concentrations by EMIT. VCM was spiked to rat serum and filtratedrat serum at a concentration of 25μg/ml, and samples were measuredby EMIT using Viva-E. Data were expressed as the mean± SD of four ex-periments. **p< 0.01

hn Wiley & Sons, Ltd. Drug Test. Analysis 2014, 6, 350–356

Table 1. Comparison of assay reliability of drug concentrations in rat serum on EMIT and FPIA

Drugs Applied(μg/ml)

EMIT FPIA

Founda (μg/ml) % Founda (μg/ml) %

VCM 18.8 9.8 52.1 18.4 97.8

Phenobarbital 12.5 13.0 103.7 12.1 96.5

Phenytoin 7.5 6.9 92.4 7.1 95.1

Carbamazepine 4.0 4.1 102.5 4.3 106.7

Valproic acid 37.5 35.7 95.3 37.9 100.9

Theophylline 6.5 6.3 97.4 6.1 93.3

Standard sera supplemented with the drugs listed in the table were diluted with three-fold volume of drug-free rat serum and applied to EMITandFPIA.

aEach value represents the mean of multiple measurements.


Drug Testing

and Analysis

results without underestimation regardless of the type of drugs,even though rat serum was used as a sample source (Table 1).

Measurements after sample treatment

The stability of VCM in rat serum during the heat treatment wasexamined by HPLC. As a consequence, VCM concentrations weremaintained at the original level at least up to 10min even at 70°C.From this finding, we examined the effect of pretreatment of se-rum samples on measurements by EMIT on Viva-E. As shown inFigure 4, there were marked time-dependent changes in mea-sured values of serum VCM concentrations. Values measuredwithout heating were much lower than the theoretical concen-tration, but values were restored to theoretical levels by exposureto 70°C for 3–7min . However, measured values were reverselydecreased with prolongation of the heating time. In contrast to70°C, there was little improvement of assay accuracy in the casewhere serum samples were exposed to temperatures of 50°C,60°C, and 80°C prior to measurement (Figure 5).

When applying VCM-spiked rat serum samples to heat at 70°Cfor 3min, observed values plotted against VCM concentrationsexhibited linearity with a correlation coefficient of more than

Figure 4. Effect of heat treatment on rat serum VCM concentrations mea-sured byHPLC and EMIT. Rat serum samples spikedwith 25μg/ml VCMwereheated to 70°C for 0–10min, and then assayed by HPLC (○) and EMIT usingViva-E (●). Data were expressed as the mean±SD of four experiments.


0.999. Analytical recovery was satisfactory and independent ofthe amount added (Table 2).

Application to pharmacokinetic study

VCM was intravenously administered to rats and their serumsamples were sequentially collected. Measurements of serumVCM concentrations were made by EMIT and HPLC for the samesamples. As shown in Figure 6, there was a high correlationbetween the values measured by the two analytical methods,provided that serum was subjected to the heat treatment prior toEMIT. Figure 7 shows the time-course of serum VCM concentrationsmeasured by EMIT. It was confirmed that VCM disappeared in a bi-phasic fashion, as demonstrated in previous studies. [10,11,14,15,17,25]

Discussion

The major significance of examinations on the disposition behav-iour of VCM using a rat model is to offer fundamental evidencesupporting the proper use of VCM in clinical practice. Measure-ment of VCM concentrations in serum is an essential procedurefor pharmacokinetic analysis, and in many cases, serum VCM con-centrations were measured by FPIA using TDxFLx or its

Figure 5. Temperature-dependency of heat treatment of rat serum onVCM concentrations measured by EMIT. Rat serum samples spiked with25μg/ml of VCM were heated at 50°C, 60°C, 70°C, and 80°C for 3min,and then assayed by EMIT using Viva-E. Data were expressed as themean± SD of four experiments.


353

Table 2. Precision of determination of VCM concentrations in ratserum after heat treatment

Added (μg/ml) Analytical recoverya (%) CVb (%)

10 97.1 9.72

25 98.9 4.32

40 96.9 4.13

Rat serum samples spiked with VCM were heated at 70 °C for 3 min,and then assayed by EMIT using Viva-E.

aEach value represents the mean of experiments (n=4).bCoefficient of variation

Figure 7. Disappearance curve of serum VCM concentrations in rats af-ter intravenous administration. Each serum sample was collected as de-scribed in the legends of Figure 6. Data represents the mean± SD ofvalues measured by EMIT (n = 3).


Drug Testing

and Analysis

354

forerunner system because of its analytical usefulness.[10–18] How-ever, commercial production of the TDxFLx series wasdiscontinued, and the global market share of Viva-E is forecastedto surpass that of TDxFLx in the field of analytical chemistry fortherapeutic agents and abused drugs. Therefore, it is of practicalimportance to examine the applicability of EMIT on the Viva-E sys-tem for the measurement of serum drug concentrations in exper-imental animals, although the reliability of EMIT has already beenestablished in application to therapeutic drug monitoring inpatients.In the present study, measurement of VCM concentrations in rat

serum on EMIT was found to be inaccurate due to underestimation.The presence of endogenous substances perturbing the quantita-tive determination of VCM in rat serum was strongly suspectedbecause ofmarked discrepancies with theoretical values of samplesspiked with known concentrations of VCM. The possibility of inter-ference was also supported by the observation that the degree ofdisparity was attenuated when physiological saline was used as adiluent for serum samples. Improvements in measured values wereascribable to the reduced volumetric ratio of rat serum.

Figure 6. Correlation between rat serum VCM concentrations measuredby EMIT and HPLC. Blood was collected from three rats 2.5, 5, 10, and30min, and 1, 2, 4, 6, and 8 h after intravenous administration of 50mg/kg VCM, and obtained serum samples were subjected to measurementsby EMIT and HPLC. Prior to application to EMIT, serum samples wereheated at 70°C for 3min. When the measured concentration exceededthe upper limit of the quantifiable range, re-measurements wereperformed after diluting serum samples with physiological saline. Eachpoint represents the actual values measured by EMIT and HPLC.


It is well known that endogenous constituents present inhuman serum can interfere with clinical immunoassays by bind-ing to reagent antibodies.[30 31] This phenomenon should alsobe true for rat serum, taking into account evidence showing thefailure of immunoassays in measuring rat plasma prolactinconcentrations due to such interference.[32] The analysis formatof EMIT is common to the drugs examined here, and all assayreagent sets include a mouse monoclonal antibody raised againstthe targeted drug, the G6PDH-labelled drug, G6P and NAD.[33]

Serum concentrations of five drugs measured by Viva-E almostcoincided with their theoretical values, but that of VCM did not.These observations strongly suggested that the underestimationdemonstrated in the VCM assay was not attributable to eitherdirect inhibition of G6PDH or functional modification of captureantibodies, allowing for acceptance of the same assay principle.Judging from the decreased interference with VCM measure-ments in filtrated serum, the major interfering components canbe interpreted to involve high-molecular substances with a highaffinity to VCM, because serum components with a molecularweight greater than 30 kDalton are almost entirely removed byultrafiltration. From this perspective, a γ-globulin such as hetero-phile antibodies can be regarded as a strong candidate for thesubstances falsely recognizing VCM as an antigen. However, anattempt to ultrafiltrate serum samples is inapplicable formeasuring the total VCM concentration, as a portion of VCMexists in the protein-bound form in systemic circulation andVCM bound to the serum protein fraction is concurrentlyexcluded before the measurement.[34]

In contrast to EMIT, FPIA showed no obvious interference withthe determination of VCM in rat serum. A crucial difference inVCM assay between FPIA and EMIT is the use of sheep polyclonalantibodies in FPIA performed on TDxFLx, along with the use offluorescein as a tracer for detection. In light of the absence of in-terpretive error in reading of measurements in FPIA, it is possiblethat dissimilarity in biological properties of capture antibodies isresponsible for the discordance of propensity for interference.Apart from antigenic specificity for VCM, the high binding affinityof anti-VCM antibody compared with that of endogenous anti-bodies may be a requisite factor contributing to assurance ofaccurate determinations. The likely mechanisms explaining inter-ference observed in this study would be: (1) the amount of



Drug Testing

and Analysis

355

capture antibody available for the formation of an immune com-plex with VCM-G6PDH conjugate is relatively increased due totrapping of VCM by endogenous antibodies, which results in aloss of active G6PDH; and/or (2) endogenous antibodies interactwith VCM-G6PDH conjugate, leading to a decrease in G6PDHactivity. In either of these two cases, the modes of interferencewould produce a false-negative signal relevant to underestimation.

Several strategies have been recommended to eliminate inter-ference by endogenous antibodies. A major approach is theexclusion of the interfering antibody and addition of a blockingagent derived from the same animal species.[30,31,35] Heat treat-ment is the most convenient method among the proposedstrategies, although there is great concern about the degradationof analytes by thermal force. VCM is chemically decomposedmainly to crystalline degradation products (CDP), and thisprocess involves rearrangement of an asparagine residue toisoaspartate, followed by the formation of CDP. [36] It wasreported that VCM concentrations were reduced by approxi-mately 70% at 80°C for 16 h or longer in aqueous solution,[36,37]

but our examination on HPLC analysis demonstrated the com-plete maintenance of VCM concentrations in rat serum at 70°Cfor up to 10min. However, susceptibility to heat varies widelyamong protein species including antibodies,[38–40] and exposureto high temperature can alter the molecular conformation of an-tibodies. Their thermal denaturation is directly linked to the lossof immunological recognition ability toward antigens, and de-pends on the rates of unfolding and aggregation stages. As itmay be likely that there is a difference in thermal stability be-tween VCM and suspected endogenous antibodies, we measuredVCM concentrations after heat treatment of serum samples andfound that VCM concentrations could be accurately measuredby exposing serum samples to a temperature of 70°C for 3min.Good quantitative performance was validated by assessinganalytical recovery and the precision of repeated measurementsunder this condition. In addition, there was a close agreementbetween VCM concentrations estimated by HPLC and EMIT,where VCM was intravenously administered to rats and theirserum samples were used for measurement. The eliminationbehaviour of VCM observed in the present study coincided withthat of previous reports of rat examinations.[10,11,14,15,17,25] There-fore, these results indicate that the proposed method of treatingrat serum prior to EMIT is applicable to fundamental pharmacoki-netic studies. On the other hand, there is a possibility that exces-sive thermal exposure induces tertiary structural changes orstereoisomeric interconversion of VCM without chemical decom-position, considering the presence of a polypeptide backbonewith many chiral centres.[37] This assumption provides one possi-ble explanation for the disagreement in measured values after a10-min heat treatment between EMIT and HPLC analyses, as thestability of VCM is probably affected by ionic strength andcoexisting constituents as well as pH in the sample fluid.[41]

To our knowledge, this is the first paper showing interfer-ence by endogenous high-molecular weight substancesduring measurement of the concentration of therapeuticdrugs in rat serum using EMIT. Our findings indicate thatcareful interpretation of obtained data is necessitated whenEMIT is used to assay endogenous and exogenous substancesin serum from experimental animals. It has been reported thatcross-reactivity with CDP in serum from patients with renaldysfunction on VCM therapy is less with EMIT than withpolyclonal antibody-based FPIA,[42] suggesting that EMIT issuperior to FPIA using TDxFLx in analytical accuracy and


precision for clinical therapeutic monitoring of VCM. In thisregard, it is of particular interest that a contrasting resultwas produced with respect to assay reliability when rat serumwas used instead of human serum.

Conclusions

In the present study, it was demonstrated that VCM concentrationsin rat serum are markedly underestimated on EMIT assay due tointerference by endogenous substances of a high molecularweight, and that this interference can be avoided by heattreatment of serum samples. These findings will contribute to theappropriate use of VCM based on evidence provided by ratexperiments in which serum VCM concentrations were measuredby EMIT.

Acknowledgements

We appreciate Takuya Jimura, Yukiko Morita, and Miyo Mizumurain our laboratory for their assistance in this study.

References[1] H.S. Chung, S.T. Lee, S.Y. Lee. Evaluation of Viva-E drug testing sys-

tem. Korean J. Lab. Med. 2007, 27, 330.[2] L. Yargui, A. Berhoune. Monitoring of C0 and C2 blood concentra-

tions of cyclosporine on the Roche Hitachi 902 analyzer. Transplant.Proc. 2009, 41, 3713.

[3] D. Moscato, A. Nonnato, R. Adamo, M. Vancheri, A. Caropreso. Ther-apeutic monitoring of tacrolimus: aberrant results by an immunoas-say with automated pretreatment. Clin. Chim. Acta 2010, 411, 77.

[4] N. Rebollo, M.V. Calvo, A. Martín-Suárez, A. Domínguez-Gil. Modifica-tion of the EMIT immunoassay for the measurement of unboundmycophenolic acid in plasma. Clin. Biochem. 2011, 44, 260.

[5] P.A. Miglioli, F. Pea, M. Mazzo, T. Berti, P. Lanzafame. Possibleinfluence of assay methods in studies of the pharmacokinetics ofantibiotics. J. Chemother. 1993, 5, 27.

[6] K. Janus, J. Antoszek. The effect of sex and age on caffeine pharma-cokinetics in cattle. Res. Vet. Sci. 2000, 69, 33.

[7] D.L. Williams. Lack of effects on lymphocyte function from chronictopical ocular cyclosporine medication: a prospective study. Vet.Ophthalmol. 2010, 13, 315.

[8] T.P. Lodise, B. Lomaestro, J. Graves, G.L. Drusano. Larger vancomycindoses (at least four grams per day) are associated with an increased inci-dence of nephrotoxicity. Antimicrob. Agents Chemother. 2008, 52, 1330.

[9] A. Marsot, A. Boulamery, B. Bruguerolle, N. Simon. Vancomycin: a re-view of population pharmacokinetic analyses. Clin. Pharmacokinet.2012, 51, 1.

[10] M. Kusama, K. Yamamoto, H. Yamada, H. Kotaki, H. Sato, T. Iga. Effectof cilastatin on renal handling of vancomycin in rats. J. Pharm. Sci.1998, 87, 1173.

[11] H. Mori, T. Nakajima, A. Nakayama, M. Yamori, F. Izushi, Y. Gomita. In-teraction between levofloxacin and vancomycin in rats-study of se-rum and organ levels. Chemotherapy 1998, 44, 181.

[12] M. Kajita, M.Morishita, K. Takayama, Y. Chiba, S. Tokiwa, T. Nagai. Enhancedenteral bioavailability of vancomycin using water-in-oil-in-water multipleemulsion incorporating highly purified unsaturated fatty acid. J. Pharm.Sci. 2000, 89, 1243.

[13] R.B. Greenwald, H. Zhao, J. Xia, A. Martinez. Poly(ethylene glycol)transport forms of vancomycin: a long-lived continuous release de-livery system. J. Med. Chem. 2003, 46, 5021.

[14] N. Hodoshima, Y. Nakano, M. Izumi, N. Mitomi, Y. Nakamura, M. Aoki,A. Gyobu, S. Shibasaki, T. Kurosawa. Protective effect of inactive in-gredients against nephrotoxicity of vancomycin hydrochloride inrats. Drug Metab. Pharmacokinet. 2004, 19, 68.

[15] M. Yoshida, S. Matzno, H. Namba, M. Nishikata, K. Matsuyama. Statis-tical analysis of the adverse effects of glycopeptide antibiotics,based on pharmacokinetics and toxicokinetics (PK/TK). J. Infect.Chemother. 2006, 12, 114.



Drug Testing

and Analysis

356

[16] O. Murillo, A. Doménech, A. Garcia, F. Tubau, C. Cabellos, F. Gudiol,J. Ariza. Efficacy of high doses of levofloxacin in experimentalforeign-body infection by methicillin-susceptible Staphylococcusaureus. Antimicrob. Agents Chemother. 2006, 50, 4011.

[17] N. Hodoshima, S. Masuda, K. Inui. Decreased renal accumulation andtoxicity of a new VCM formulation in rats with chronic renal failure.Drug Metab. Pharmacokinet. 2007, 22, 419.

[18] J. Pieróg, B. Kubisa, M. Droździk, J. Wójcik, M. Bielewicz, J. Pankowski,K. Safranow, T. Grodzki. Vancomycin lung concentration in acute andhyperacute rejection models of lung transplantation in rats. Eur. J.Cardiothorac. Surg. 2010, 38, 456.

[19] T. Nakamura, Y. Hashimoto, T. Kokuryo, KI. Inui. Effects of fosfomycinand imipenem/cilastatin on nephrotoxicity and renal excretion ofvancomycin in rats. Pharm. Res. 1998, 15, 734.

[20] K.E. Anderson, L.A. Eliot, B.R. Stevenson, J.A. Rogers. Formulation andevaluation of a folic acid receptor-targeted oral vancomycin liposo-mal dosage form. Pharm. Res. 2001, 18, 316.

[21] M.S. Luer, K.K. Neill, B.J. Gurley, M.L. Shannon, A.D. Killian, K.A. Rodvold.Fluctuations in vancomycin CNS tissue concentrations followingintermittent and continuous infusions in the rat. Neurol. Res.2004, 26, 312.

[22] D.M. Chilukuri, J.C. Shah. Local delivery of vancomycin for the prophy-laxis of prosthetic device-related infections. Pharm. Res. 2005, 22, 563.

[23] M.J. de Jesús Valle, F.G. López, A.D. Hurlé, A.S. Navarro. Pulmonary versussystemic delivery of antibiotics: comparison of vancomycin dispositionsin the isolated rat lung. Antimicrob. Agents Chemother. 2007, 51, 3771.

[24] J. Beckmann, F. Kees, J. Schaumburger, T. Kalteis, N. Lehn, J. Grifka,K. Lerch. Tissue concentrations of vancomycin and moxifloxacin inperiprosthetic infection in rats. Acta Orthop. 2007, 78, 766.

[25] I. Shimada, C. Iwata, S. Taga, H. Teramachi, M. Nomura, K. Miyamoto,H. Tsuciya, T. Wada, K. Kimura, R. Matsushita. Enhanced renal clear-ance of vancomycin in rats with carcinogen-induced osteosarcoma.Anticancer Res. 2012, 32, 823.

[26] Y.V. Prasad, S.P. Puthli, S. Eaimtrakarn, M. Ishida, Y. Yoshikawa,N. Shibata, K. Takada. Enhanced intestinal absorption of vanco-mycin with Labrasol and D-alpha-tocopheryl PEG 1000 succinatein rats. Int. J. Pharm. 2003, 250, 181.

[27] N. Shibata, M. Ishida, Y.V. Prasad, W. Gao, Y. Yoshikawa, K. Takada.Highly sensitive quantification of vancomycin in plasma samples usingliquid chromatography-tandem mass spectrometry and oral bioavail-ability in rats. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2003,789, 211.

[28] P. Vergidis, M.S. Rouse, G. Euba, M.J. Karau, S.M. Schmidt, J.N.Mandrekar, J.M. Steckelberg, R. Patel. Treatment with linezolid or


vancomycin in combination with rifampin is effective in an animalmodel of methicillin-resistant Staphylococcus aureus foreign bodyosteomyelitis.Antimicrob. Agents Chemother. 2011, 55,1182.

[29] G. Ye, X. Cai, B. Wang, Z. Zhou, X. Yu, W. Wang, J. Zhang, Y. Wang, J.Dong, Y. Jiang. Simultaneous determination of vancomycin andceftazidime in cerebrospinal fluid in craniotomy patients by high-performance liquid chromatography. J. Pharm. Biomed. Anal. 2008,48, 860.

[30] L.J. Kricka. Human anti-animal antibody interferences in immunolog-ical assays. Clin. Chem. 1999, 45, 42.

[31] J. Tate, G. Ward. Interferences in immunoassay. Clin. Biochem. Rev.2004, 25, 105.

[32] D.R. Grattan, J.H. Miller, R.L. Averill. Failure of an indirect competitiveenzyme-linked immunosorbent assay to measure prolactin in ratplasma samples. Proc. Soc. Exp. Biol. Med. 1989, 191, 170.

[33] EMIT 2000 Assay Application Sheet (package insert), Siemens JapanK.K., 2008.

[34] B.H. Ackerman, E.H. Taylor, K.M. Olsen, W. Abdel-Malak, A.A. Pappas.Vancomycin serum protein binding determination by ultrafiltration.Drug Intell. Clin. Pharm. 1988, 22, 300.

[35] A.A. Ismail. A radical approach is needed to eliminate interfer-ence from endogenous antibodies in immunoassays. Clin. Chem.2005, 51, 25.

[36] C.M. Harris, H. Kopecka, T.M. Harris. The stabilization of vancomycinby peptidoglycan analogs. J. Antibiot. (Tokyo), 1985, 38, 51.

[37] P.J. Bonnici, M. Damen, J.C. Waterval, A.J. Heck. Formation and effi-cacy of vancomycin group glycopeptide antibiotic stereoisomersstudied by capillary electrophoresis and bioaffinity mass spectrome-try. Anal. Biochem. 2001, 290, 292.

[38] A. Demonte, I.Z. Carlos, E.J. Lourenço, J.E. Dutra de Oliveira. Effect ofpH and temperature on the immunogenicity of glycinin (Glycine maxL.). Plant Foods Hum. Nutr. 1997, 50, 63.

[39] R.D. Soltis, D. Hasz, M.J. Morris, I.D. Wilson. The effect of heat inacti-vation of serum on aggregation of immunoglobulins. Immunology1979, 36, 37.

[40] A.J. Law, J. Leaver. Effect of pH on the thermal denaturation of wheyproteins in milk. J. Agric. Food Chem. 2000, 48, 672.

[41] A.S. Antipas, D.G. Vander Velde, S.D. Jois, T. Siahaan, V.J. Stella. Effectof conformation on the rate of deamidation of vancomycin in aque-ous solutions. J. Pharm. Sci. 2000, 89, 742.

[42] L. Anne, M. Hu, K. Chan, L. Colin, K. Gottwald. Potential problem withfluorescence polarization immunoassay cross-reactivity to vancomy-cin degradation product CDP-1: its detection in sera of renally im-paired patients. Ther. Drug Monit. 1989, 11, 585.


underestimation of rat serum vancomycin concentrations measured by an enzyme-multiplied immunoassay...

Documents