development of a ce method for the determination of mycophenolic acid in human plasma: a comparison...

Filippo Carlucci1

Maurizio Anzini2

Michele Rovini2

Dario Cattaneo3

Simona Merlini3

Antonella Tabucchi1

1Dipartimento di MedicinaInterna,Scienze Endocrino-Metabolichee Biochimica,University of Siena,Siena, Italy

2Dipartimento Farmaco-Chimico-Tecnologico,Università degli Studi di Siena,Siena, Italy

3Unità diFarmacologia e Farmacogenetica,Istituto di RicercheFarmacologiche “Mario Negri”,Ranica, Italy

Received September 28, 2006Revised May 22, 2007Accepted June 3, 2007

Research Article

Development of a CE method for thedetermination of mycophenolic acid inhuman plasma: A comparison with HPLC

Mycophenolate mofetil (MMF) is a widely used drug for the maintenance of immunosup-pressive therapy in renal-transplant recipients. MMF is rapidly metabolized in vivo tomycophenolic acid (MPA), a reversible, noncompetitive inhibitor of inosine monopho-sphate dehydrogenase, which represents a limiting enzyme in lymphocyte proliferation.MPA shows large interindividual pharmacokinetic variability: its monitoring is therefore ofprimary importance to achieve adequate immunosuppression with minimized risk of graftrejection or toxicity. We developed a CE method for the determination of total MPA (tMPA)in plasma, based on easy sample preparation; CE evaluation of tMPA was performed in30 mmol/L sodium-borate with 10 mmol/L SDS (pH 10.00) at 257C using a 60 cm (54.5 towindow) uncoated capillary with UV detection at 254 nm wavelength. MPA was readilydetectable in plasma; the CE method was linear in the range of 0.7–120 mg/mL (r .0.992).Intra- and interassay imprecision was ,7% except for the lowest spiked MPA concentra-tion, which had an intra-assay RSD% of 14.7 compared to 18.3 interassay. Data by CE werecompared with results obtained by a validated HPLC method. CE assay of tMPA exhibited avery good correlation (r2 .0.988) with respect to HPLC; Bland–Altman difference versusaverage showed a mean of 20.18 mg/mL 6 1.14 SD. CE determination of tMPA is a robust,sensitive and reproducible method with the advantage over HPLC of being fast, simple andunexpensive, also enabling quick assessment of MPA for pharmacokinetic studies.

Keywords:

CE / HPLC / Immunosuppression / Mycophenolic acidDOI 10.1002/elps.200700190

3908 Electrophoresis 2007, 28, 3908–3914

1 Introduction

Organ transplantation represents the optimal treatment formany patients with end-stage organ failure. The currentsuccess of organ transplantation is largely correlated toadvances in immunosuppressive therapy and very few allo-grafts are now lost as a result of acute rejection. Mycophe-nolate mofetil (MMF) has become in recent years a morewidely prescribed drug, used for maintenance immunosup-pressive therapy both in adult [1] and pediatric renal-trans-plant recipients [2].

MMF and enteric-coated mycophenolate sodium (MPS)are the two commercially available prodrugs, which arerapidly converted in the liver to mycophenolic acid (MPA,(E)-6-(4-hydroxy-6-methoxy-7-methyl-3-oxo-1,3-dihydroisoben-zofuran-5-yl)-4-methylhex-4-enoic acid) which represents theactive compound (Fig. 1). The target of MPA is inosinemonophosphate dehydrogenase (IMPDH), the rate-limitingenzyme in the de novo synthesis of guanine nucleotideswhich are essential for DNA synthesis [3]. There are two iso-forms of the IMPDH enzyme, the type I found predomi-nantly on resting cells and the type II induced and expressedon activated lymphocytes [4, 5]. MPA preferentially inhibitsthe type II isoform of IMPDH: hence the blockade results inrelatively selective arrest of lymphocyte proliferation andimmunosuppression.

Despite its efficacy, MMF has been frequently associatedwith gastrointestinal side effects (including nausea, abdom-inal pain, vomiting, and diarrhoea), which occur in up to45% of patients [6]. The reduction, interruption or dis-continuation of MMF administration has proved to relievethese symptoms, but it may result in subtherapeutic dosingand an unfavorable clinical outcome [7].

Correspondence: Professor Filippo Carlucci, Dipartimento diMedicina Interna, Scienze Endocrino-Metaboliche e Biochimica,University of Siena, Polo Scientifico-Universitario di San Miniato– Via Aldo Moro, I-53100 Siena, ItalyE-mail: [email protected]: 139-0577-234285

Abbreviations: IMPDH, inosine monophosphate dehydrogenase;MMF, mycophenolate mofetil; MPA, mycophenolic acid; MPAG,

mycophenolic acid glucuronide; PCA, perchloric acid; tMPA, totalMPA

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

Electrophoresis 2007, 28, 3908–3914 CE and CEC 3909

Figure 1. Structure of MPA. IUPAC name: (E)-6-(4-hydroxy-6-meth-oxy-7-methyl-3-oxo-1,3-dihydroisobenzofuran-5-yl)-4-methylhex-4-enoic acid.

MMF has an apparent half-life of 18 h; it is mainlymetabolized in liver to mycophenolic acid glucuronide(MPAG), primarily eliminated in the urine. Extra-hepatictissues, kidney in particular, may play a role in overallMPA glucuronidation with the formation of two furtherminor metabolites [8]. MPAG is secreted via bile into thegut where it exhibits a significant entero-hepatic recircula-tion, with a concomitant second peak in plasma of MPAconcentration 6–12 h after administration. This entero-hepatic circulation might contribute to the gastrointestinaltoxicity [9]. In plasma MPA is highly bound to albuminwith a minimal quote (3%) of free MPA, so the protein-bound fraction has to be released by sample deproteiniza-tion to determine total MPA (tMPA) concentration. ClinicalMPA pharmacokinetic is characterized by a large inter-individual and intraindividual variability [10, 11] deter-mined by different factors as renal allograft function,hepatic function or concomitant therapy with cyclosporine[12, 13]. All these conditions underline the need for MPAtherapeutic drug monitoring (TDM) in order to optimizeoutcomes, especially in patients with higher risk of rejec-tion [14].

Up to date the procedures utilized for MPA drug mon-itoring mainly concern numerous HPLC methods, with [15–19] or without [11, 20222] simultaneous MPAG measure-ment, EMIT immunoassays [23] and only one CE method[24].

The aim of the present study is to evaluate the potentialof CE in TDM of tMPA in comparison with HPLC.

2 Materials and methods

2.1 Drugs and reagents

CE-grade sodium borate and sodium hydroxide wereobtained from Merck (Darmstadt, Germany); MPA referencestandard and all other chemicals of analytical grade werefrom Sigma Chemical Corporation (St. Louis, MO, USA).Biomax Centrifugal filters were from Millipore Corporation(Bedford, MA, USA).

2.2 Patients

We studied 32 patients in the care of the Department ofImmunology and Clinical Organ Transplant at the OspedaliRiuniti di Bergamo, who had received primary cadavericrenal allograft 6–9 months before. All patients receivedimmunosuppressive therapy by MMF with coadministrationof cyclosporine/steroids or sirolimus/steroids. All the proce-dures were in accordance with the current revision of theHelsinki Declaration. The informed consent was obtainedfrom all patients. All samples were collected for routinehematological analysis and for determination of troughblood levels of cyclosporine/sirolimus. Blood samples werecollected in heparinized tubes and plasma stored at 2207Cuntil HPLC analysis. Plasma aliquots were shipped to ourlaboratory in dry-ice package within 24 h and stored at2207C until CE analysis, which was performed in two weeksfrom blood collection (the lag from HPLC and CE analysis,relative to method comparison, was within two weeks).Samples containing MPA are stable for at least 11 months,when stored at 2207C [25].

2.3 HPLC determination of tMPA

Samples were analyzed by HPLC at the Mario Negri Institutefor Pharmacological Research – Center for Research onOrgan Transplantation (Ranica, BG, Italy). Sample prepara-tion and HPLC determination were performed according toCattaneo et al. [26], which utilized the validated HPLC meth-od from Tsina et al. [25]. tMPA levels measured by HPLCassay were in the range of 0.48–23.32 mg/mL, median 1.84.

2.4 Sample preparation for CE

Protein removal was comparatively performed by perchloricacid (PCA) and ACN precipitation. For PCA protein precipi-tation, 30 mL of 2.7 mol/L PCA were added to 100 mL ofsamples, extracts were centrifuged (12 0006g for 5 min) in acooled microfuge and the supernatants neutralized with2.7 mol/L potassium hydroxide (KOH). For ACN deproteini-zation, 100 mL of sample and 100 mL of ACN were mixed byvortex-mixing for 5 s. Samples were then treated as reportedfor PCA protein removal. Both deproteinization methodswere followed by centrifugal ultrafiltration with 10K car-tridges to remove protein excess.

2.5 CE analysis

The analyses were performed with a Biorad Biofocus 3000instrument (BioRad Laboratories, USA) using a Supelco bareCEIect-Fs column (Supelco, Bellafonte, PA, USA;60 cm675 mm id), with the window at a distance of 54.5 cm.Borate buffer (30 mmol/L) containing SDS (10 mmol/L) wasused. Conditions were pH 10.00, 25 kV and 3 psi?s hydro-dynamic load, at 257C. The electric field was 420 V/cm with acurrent of approximately 80 mA. EOF velocity was


3910 F. Carlucci et al. Electrophoresis 2007, 28, 3908–3914

4.8361024 cm2/V?s (calculated using methanol as neutralmarker), the CV measured for calibrators (n = 20) and sam-ples (n = 20) was 2.28%. The results were read over the rangeof 190–300 nm and analysed at 254 nm. Between runs, thecapillary was washed with 0.1 mol/L NaOH for 60 s andrunning buffer for 120 s.

2.6 Method validation

2.6.1 Calibration plots and linearity

The linearity of the method was tested by preparing metha-nol–water MPA calibrators for eight-point standard curve.The tested concentrations were 0.7, 1.4, 2.6, 6.4, 12.8, 20, 60,120 mg/mL. Each concentration level was injected threetimes. Regression curve (linear regression analysis asy = a 1 bx) was obtained by plotting relative peak area (ana-lyte corrected peak area) versus concentration.

The LOD in calibrators was evaluated as an S/N of 3. Thenoise was defined as the signal height divided by half thepeak-to-peak distance of the noise. The LOQ as defined fromEURACHEM guide [27] was identified as the lowest con-centration of the analyte that can be determined with anacceptable level of repeatability precision; we considered anRSD ,20% according to other authors [28].

2.6.2 Accuracy

The accuracy was evaluated by spiking MPA into controlhuman plasma (n = 6) (at three concentration levels: 1.0/5.0/50 mg/mL). MPA was dissolved in methanol–water (1:1 v/v,to 9:1 v/v) and added to the appropriate amount of controlhuman plasma, which was then swirled in a blood mixer for10 min before ACN and PCA protein removal in a mannersimilar to that used for clinical samples. The fortified sam-ples were then analysed with the proposed CE method.

Extraction efficiency was determined by comparing cor-rected peak area obtained from MPA-fortified plasma sam-ples with corrected peak area obtained with nonextractedmethanol–water standards. The percentage of recovery wascalculated by comparing the measured concentrations withthe expected concentrations.

2.6.3 Precision

MPA calibration standards in control drug-free plasma(n = 6) at three different concentrations (1.0/5.0/50 mg/mL)were analyzed for six replicates within the same day to obtainthe repeatability (intraday precision) and for six replicates onthree consecutive days to obtain the intermediate precision(interday precision) of analyte corrected peak area, expressedas RSD% values.

Intra- and interassay precision for the methanol–watercalibration standards (n = 6) was also tested.

2.6.4 Reproducibility

Method reproducibility was tested by univariate approachevaluating SDS concentration in sample, buffer pH value,capillary temperature, injection time and wavelength detec-tion.

2.6.5 Preparation of quality control samples

Quality control samples (QCs), were prepared by spikingMPA into control human plasma at the following three con-centrations: QC 1 (1.28 mg/mL), QC 2 (5.12 mg/mL), QC 3(20.48 mg/mL).

The bulk QCs were prepared as previously reported inSection 2.6.2. QCs at each of the three different concentra-tions were analysed with sample batch to monitor methodperformance.

2.6.6 Method comparison

Thirty-two plasma specimens derived from renal transplantrecipients were used for method comparison with the HPLCprocedure, that has been used for the acquisition of extensivepharmacokinetic/pharmacodynamic data on MPA [26] andextensively validated [25]. The HPLC method is routinelyused for tMPA and MPAG determination in patients at theMario Negri Institute for Pharmacological Research – Centerfor Research on Organ Transplantation, which participates inthe MPA International Proficiency Testing Scheme, with anestimated inaccuracy of 5%.

2.7 Statistical analysis

Passing and Bablok [29] nonparametric regression proce-dure was used for comparison of the two methods; it repre-sents a procedure with no special assumption regarding thedistribution of the samples and the measurement errors.The result does not depend on the assignment of themethods (or instruments) to X and Y. The slope (B) andintercept (A) are calculated with their 95% confidenceinterval. These confidence intervals are used to determinewhether there is only a chance difference between B and 1and between A and 0.

Agreement between the HPLC and CE methods was alsoassessed by using nonparametric Bland–Altman analysis,plotting the absolute and relative method differences againstthe method means.

The Bland–Altman plot [30] is a statistical method tocompare two measurement techniques. In this graphicalmethod, the differences (or alternatively the ratios) be-tween the two techniques are plotted against the averagesof the two techniques. The Bland–Altman plot is useful toreveal a relationship between the differences and theaverages, to look for any systematic bias and to identifypossible outliers. If there is a consistent bias, it can beadjusted for by subtracting the mean difference from the



new method. If the differences within mean 6 1.96 SD arenot clinically important, the two methods may be usedinterchangeably.

3 Results and discussion

Our procedure shows that tMPA can be easily detected inplasma from patients undergoing immunosuppressivetherapy with MMF by CE without SPE (Fig. 2B). Noe et al.[24] reported a method for the determination of tMPAusing micellar electrokinetic capillary chromatography(MEKC) for rapid drug screening in patient associated toSPE procedures. We have elaborated a new CE method,optimizing buffer composition, LOQ and sample prepara-tion; we have compared the CE performances to a validateHPLC method.

Figure 2. Electropherograms of human plasma (detection at254 nm wavelength). (A) Control plasma, (B) patient plasma afterdeproteinization with ACN, (C) the same patient sample afterdeproteinization with PCA, MPA concentration was 10.48 mg/mL.

3.1 Optimization of electrophoretic parameters

MPA determination was investigated at different SDS con-centrations (0, 10, 20 and 30 mM) while borate bufferconcentration was maintained at 30 mM, pH 10.00, oper-ating at 25 kV. SDS in buffer does not affect MPA resolu-tion from other peaks, but it improves band broadeningand peak symmetry. No significant improvement in termsof electrophoretic mobility was obtained with SDS con-centration of 20 and 30 mM versus 10 mM, being MPAtotally ionized at pH value above 9.7 (according to Hen-derson–Hasselbach equation). We obtained complete reso-lution of MPA peak in the 9.20–10.00 electrolyte pH range.Higher resolution of MPA from contaminating peaks wasobtained with a buffer system at pH 10.00, operating at25 kV. MPA determination was possible between 24 and307C. An increase in temperature reduced the compoundmigration times and increased their peak symmetry; how-

ever, above 307C the short migration times do not permitgood resolution of MPA with respect to the other uni-dentified peaks in samples.

MPA was identified by spiking experiments with stand-ard and observing comigration and on the basis of UV ab-sorbance spectra; all compounds evaluable in the range of190–300 nm wavelength migrated within 15 min. No clearassociation was evident between unidentified peaks andimmunosuppressive drug administration since the electro-pherogram profile was reproducible in patient samples andcontrol human plasma spiked with MPA; moreover,cyclosporine-A does not interfere with the method since it isnot detectable at the wavelength of 254 nm used for MPAmonitoring [24].

The injection times showed a good linearity with respectto peak areas over a 2–6 psi?s range. A 3 psi?s injection timeshowed the best compromise solution in terms of reproduc-ibility and sensitivity. Separation efficiency was greater than90 000 theoretical plates in the tested concentration range.Under these conditions the electric field was 420 V/cm. Theincrement of the voltage to 30 kV does not affect selectivityand resolution in a significant way; on the contrary itenhances the current with a consequent increase in baselinenoise and loss of sensitivity.

We obtained electropherograms in which the MPA peakis clearly evaluable in the range of therapeutic concentrations[31, 32]; a better LOQ (0.7 mg/mL) with respect to previousreported CE method (4.0 mg/mL) [26] was obtained, while theLOQ of the HPLC method was 0.1 mg/mL.

MPA exhibits three absorption peaks close to 214, 254and 304 nm wavelength [33]. We tested MPA detection at 214and 254 nm; at 214 nm we obtained the best results in termsof sensitivity but the MPA resolution versus an unidentifiedpeak was not optimal (R = 2(t2 2 t1)/(w1 1 w2),1). Due to itsaromatic nucleus, MPA exhibits a more specific absorptionpeak close to 254 nm; at this wavelength no interfering peakswere detected. The sensitivity was nearly 42% with respect to214 nm (214/254 ratio for corrected area was 2.36 6 0.02).For these reasons we validated our method at 254 nm.

3.2 Assessment of performance characteristics

In the tested concentration range (0.7–120 mg/mL), the cor-relation coefficient (corrected peak area vs. concentration)exceeded 0.989 at 3 psi?s injection time. Regression curve(linear regression analysis as y = a 1 bx, where y is the cor-rected area and x is the concentration expressed as mg/mL)gave the following parameters: intercept a = 62.71, slope ofthe curve b = 485.21 and squared correlation coefficientr = 0.992. The LOD was 0.32 mg/mL.

In the evaluation of the extraction efficiency of MPAfrom albumine (n = 3), we obtained a minimum recovery of96.1 6 3.8%, and 87.8 6 4.3%, respectively, for ACN andPCA extraction procedure.

Ultracentrifugal filtration allowed a better MPA identifi-cation, the unfiltrated samples exhibited a more con-



taminated electropherogram profile for the presence of uni-dentified peaks (Tabucchi, A. et al. unpublished data). Novariations were evident in terms of migration time betweenthe two protein removal procedures (RSD ,5%), but as it canbe easily observed in Fig. 2, ACN protein removal allows abetter MPA detection from unidentified contaminatingpeaks.

Table 1 shows that the measured concentrations weregenerally in good agreement with the added amounts, espe-cially for ACN deproteinization with recoveries between 93.8and 96.7%. These results suggest that interferences by theother matrix components are not significant and the CEconditions are suitable to obtain adequate method accuracy.ACN protein removal was then utilized for all further studieson our samples.

Table 1. Accuracy for ACN and PCA deproteinization tested forfour replicates in the same batch

Spiked concen-tration (mg/mL)

Recovery (%) RSD (%)

ACN PCA ACN PCA

1.0 96.7 88.4 14.7 16.45.0 95.0 86.7 5.6 6.3

50.0 93.8 86.3 5.2 5.9

The results for precision are summarized in Table 2. Themethod had very good reproducibility for the 1.0/5.0/50 mg/mL calibration standards in acidified drug free plasma bothin terms of migration times and individual peak areas, forconsecutive runs (Table 2). Intra and interassay reproduci-bility was determined: RSD values were less than 7% exceptfor the lowest spiked MPA concentration, which had anintra-assay RSD% of 14.7 compared to 18.3 interassay. Pre-cision for methanol–water standard solutions, exhibited amaximum RSD% of 13.8, relative to the 1.0 mg/mL standardsolution for the interassay evaluation.

No significant variation in MPA peak migration time,symmetry and area, was observed between acidified drugfree plasma spiked with MPA and submitted to ACN depro-teinization, with respect to treated patients, for the sameMPA concentration (RSD ,5%). No peak was present atMPA migration time in the electropherogram of healthysubject at 254 nm (Fig. 2A) in plasma.

CE precision profile (Fig. 3) was evaluated for 32 patientsin the 0.53 6 0.09–22.78 6 1.05 mg/mL range of MPA con-centration, assayed in triplicate. The LOQ corresponded to aconcentration of 0.72 mg/mL, as previously defined. If weconsider a typical range of the 0–12 h AUC curve (7.3–102.3 mg/h/mL) [31], the obtained RSD value was below15%, comparable to those reported by other authors forHPLC methods [34].

Figure 3. Precision profile. Specimens, triplicated in the samebatch, were plotted for mean CV% vs. mean MPA concentration.Dotted lines identify the maximum CV% accepted for the LOQand the analytical sensitivity, as indicated.

3.3 Methods comparison

We compared the present CE procedure with a validatedHPLC-UV method [26]; Table 3 reports the analytical perfor-mance characteristics of CE and HPLC using UV detection.We utilized plasma samples from 32 patients undergoingMMF immunosuppressive therapy. In four patients, tMPAlevels measured by HPLC were below the LOQ of CE(0.72 mg/mL) and were not plotted. The Bland–Altman dif-ference plot method for tMPA revealed good agreement be-tween the two methods (Figs. 5A and B) over the clinicallyrelevant ranges. The tMPA concentrations measured by

Table 2. Intra- and interassay precision (n = 6) for the MPA spiked into control human plasma

Spiked concen-tration (mg/mL)

Mean concentration (mg/mL) RSD (%) Biases %

Intra-assay Interassay Intra-assay Interassay Intra-assay Interassay

1.0 1.16 1.18 14.7 18.3 216.8 218.05.0 4.90 4.87 5.8 6.6 8.5 8.7

50.0 54.1 44.3 3.4 4.1 1.42 1.51



HPLC-UV and plotted were in the range of 0.67–23.32 mg/mL (median, 1.75 mg/mL); those measured by CE were 0.76–21.84 mg/L (median, 1.84 mg/mL). The mean absolute differ-ence between the two methods was 0.18 mg/mL, while themean relative difference was 5.60% over the range of 0.67–23.32 mg/mL (Figs. 4A and B, respectively) as measured withHPLC-UV. The equation for the Passing–Bablok (Fig. 5)regression line was: y = 0.88x 1 0.35 mg/mL (r2 = 0.988;95%, Sy)x = 0.79 mg/mL).

Table 3. Analytical performance characteristics of CE and HPLCusing 254 nm UV detection

Parameter CE HPLCa)

Accuracy (mean recovery %) 95.14 92.0Intra-assay precision in the range

of 2–5 mg/mL (RSD%),6 ,2

Linearity range tested (mg/mL) 0.7–120.0 0.1–40.0LOQ (mg/mL) 0.72 0.10Analysis time (min) 15 15

a) Data from Tsina and co-workers [26].

Figure 4. Bland–Altman plots of the relative method differences(CE-HPLC). The solid line represents the mean deviation of thetwo methods: the dotted lines represent the 95% confidenceintervals. (A) Mean absolute differences between the two meth-ods; (B) mean relative differences over the 0.67–23.32 mg/mLrange (from HPLC assay).

Figure 5. Passing–Bablok linear regression illustrating the rela-tionship between CE and HPLC in 28 patients receiving MMFtherapy. The continuous line is the line of identity; the dottedlines represent confidence intervals. The regression techniquegave the equation: CE = 0.886HPLC 1 0.35. r2 = 0.988.

4 Concluding remarks

The run time for the determination of MPA by our CE iscomparable to that obtained by HPLC, being the analyteevaluable in 15 min with both methods; however, CE offersthe advantage of a fast and simple column conditioning. CEis also a particularly versatile technology, with limitedreagent consumption (3 mL every 20–30 runs), theoreticallyindefinite column life, fast and simple instrument con-ditioning and maintaining. Because of the high-mass sensi-tivity of the CE procedure, only few nanoliters of samples, asopposed to 100 mL for HPLC-UV, are needed for tMPA anal-ysis; a rerun of the sample is therefore possible. Never-theless, ACN deproteinization could be carried out with only20 mL of plasma: this is of particular interest in the case ofpediatric samples.

In contrast, the well-known HPLC technique might havea better LOD but often involves SPE for sample preparationand high-priced solvents. With respect to the only one pub-lished CE method for MPA determination in human plasma[24], our CE method exhibits a much higher sensitivity,mainly related to the sample volume injected and to the cap-illary internal diameter.

In comparison with EMIT® 2000 immunoassay method,CE allows the specific determination of tMPA without inter-ference from metabolites [35]. In addition, among the dis-advantages of the commercially available immunoassay areits very high costs. Only with a great sample throughput,immunoassay is a worthwhile rapid method with acceptablecosts. The precision, accuracy and sensitivity of the CE assay



is suitable for both routine TDM applications and pharma-cokinetic studies: the described CE assay provides thereforeanother convenient and accurate method for the measure-ment of tMPA concentrations in plasma.

5 References

[1] van Gelder, T., Shaw, L. M., Transplantation 2005, 15, 244–253.

[2] Payen, S., Zhang, D., Maisin, A., Popon, M., et al.,Ther. DrugMonit. 2005, 27, 378–388.

[3] Behrend, M., Drug Saf. 2001, 2, 645–663.

[4] Millan, O., Oppenheimer, F., Brunet, M., Vilardell, J., et al.,Clin. Chem. 2000, 46, 1376–1383.

[5] Pelletier, R. P., Akin, B., Henry, M. L., Bumgardner, G. L., etal.,Clin. Transplant. 2003, 17, 200–205.

[6] Millan, O., Brunet, M., Campistol, J. M., Faura, A., et al., Clin.Chem. 2003, 49, 1891–1899.

[7] Atcheson, B. A., Taylor, P. J., Mudge, D. W., Johnson, D. W.,et al., Br. J. Clin. Pharmacol. 2005, 59, 271–280.

[8] Shipkova, M., Strassburg, C. P., Braun, F., Streit, F., et al., Br.J. Pharmacol. 2001, 132, 1027–1034.

[9] Sollinger, H. W., Clin. Transplant. 2004, 18, 485–492.

[10] Pisupati, J., Jain, A., Burckart, G., Hamad, I., et al., J. Clin.Pharmacol. 2005, 45, 34–41.

[11] Streit, F., Shipkova, M., Armstrong, V. W., Oellerich, M., Clin.Chem. 2004, 50, 152–159.

[12] El Haggan, W., Ficheux, M., Debruyne, D., Rognant, N., et al.,Transplant. Proc. 2005, 37, 864–866.

[13] Cattaneo, D., Merlini, S., Zenoni, S., Baldelli, S., et al., Am. J.Transplant. 2005, 5, 2937–2944.

[14] van Gelder, T., Meur, Y. L., Shaw, L. M., Oellerich, M., et al.Ther. Drug Monit. 2006, 28, 145–154.

[15] Yau, W. P., Vathsala, A., Lou, H. X., Chan, E., J. Chromatogr. B2004, 805, 101–112.

[16] Aresta, A., Palmisano, F., Zambonin, C. G., Schena, P., et al.Chromatogr. B 2004, 810, 197–202.

[17] Westley, I. S., Sallustio, B. C., Morris, R. G., Clin. Biochem.2005, 38, 824–829.

[18] Indjova, D., Kassabova, L., Svinarov, D., J. Chromatogr. B2005, 817, 327–330.

[19] Shipkova, M., Niedmann, P. D., Armstrong, V. W., Schutz, E.,et al., Clin. Chem., 1998, 44, 1481–1488.

[20] Teshima, D., Kitagawa, N., Otsubo, K., Makino, K., et al. J.Chromatogr. B 2002, 780, 21–26.

[21] Zambonin, C. G., Aresta, A., Palmisano, F., J. Chromatogr. B2004, 806, 89–93.

[22] Shen, J., Jiao, Z., Yu, Y. Q., Zhang, M., et al., J. Chromatogr.B 2005, 817, 207–213.

[23] Premaud, A., Rousseau, A., Picard, N., Marquet, P., Ther.Drug Monit. 2006, 28, 274–277.

[24] Noe, S., Bohler, J., Frahm, A. W., J. Pharm. Biomed. Anal.2000, 22, 197–202.

[25] Tsina, I., Chu, F., Hama, K., Kaloostian, M., et al., J. Chroma-togr. B.,1996, 675, 119–129.

[26] Cattaneo, D., Gaspari, F., Ferrari, S., Stucchi, N., et al., Clin.Transplant., 2001, 15, 402–409.

[27] Ellison, S. L. R., Rösslein, M., Williams, A., Eurachem/CitacGuide, 2nd Edn., LGC Ltd., Teddington, Middlesex, UK 1998.

[28] Shah, V. P., Midha, K. K., Dighe, S., McGilveray, I. J., et al., J.Pharm. Sci. 1992, 81, 309–312.

[29] Passing, H., Bablok, W., J. Clin. Chem. Clin. Biochem. 1983,21, 709–720.

[30] Bland, J. M., Altman, D. G., Lancet 1986, 1, 307–310.

[31] Jain, A., Venkataramanan, R., Hamad, I. S., Zuckerman, S., etal., J. Clin. Pharmacol. 2001, 41, 268–278.

[32] Cattaneo, D., Perico, N., Gaspari, F., Gotti, E., et al., KidneyInt., 2002, 62, 1060–1067.

[33] Shipkova, M., Armstrong, V. W., Wieland, E., Niedmann, P.D., et al., Br. J. Pharmacol., 1999, 126, 1075–1082.

[34] Ceglarek, U., Casetta, B., Lembcke, J., Baumann, S., et al.,Clin. Chim. Acta., 2006, 373, 168–171.

[35] Weber, L. T., Shipkova, M., Armstrong, V. W., Wagner, N., etal., Clin. Chem., 2002, 48, 517–525.


development of a ce method for the determination of mycophenolic acid in human plasma: a comparison...

Documents