Simultaneous determination of mycophenolic acid and its glucuronide in human plasma using a simple high-performance liquid chromatography procedure.
Results from controlled clinical trials with renal transplant patients have shown that the addition of MMF to immunosuppression protocols with cyclosporine and steroids substantially reduces the early acute rejection rate and decreases the need for antirejection therapy while showing a favorable safety profile (6-8). However, the full clinical potential of the drug has not yet been realized. The use of MMF in heart, liver, and lung transplantation as well as in autoimmune diseases is not yet sufficiently established (3, 4, 9). Limited data are available regarding its pharmacokinetic properties in children or after multiple dosage in patients with renal or liver dysfunction. The role of therapeutic drug monitoring to improve the therapeutic efficacy and to minimize adverse side effects is still under investigation (10, 11).
To address these questions, accurate and sensitive methods are needed to determine the free and total MPA concentrations, as well as the concentration of its glucuronide metabolite (MPAG) in patient blood specimens. Because of the high binding to plasma albumin (>97%) (12), free MPA concentration (MPAf) measurement may be a better indicator of total drug exposure. Therefore, a suitably sensitive method with a very low detection limit is required.
Several gas chromatographic (13,14) and HPLC methods (15-17) have been developed for the determination of MPA. Previously described methods suffered from several disadvantages, such as a lack of sensitivity and specificity or the need for a large sample volume (13, 14). To date, only two methods allow the simultaneous measurement of MPA and MPAG (15,17). The analysis of MPA and MPAG as described by Tsina et al. (17) is relatively complex. MPA and MPAG are analyzed either sequentially in two separate analytical runs using one HPLC system or in parallel by splitting the extract into two portions and injecting them onto two separate HPLC systems. Alternatively, MPAG has been cleaved enzymatically and MPA determined with and without enzymatic pretreatment of the sample (15). Compared with the extraction procedure described here, both methods use time-consuming and laborious extraction protocols. The approach described by Tsina et al. (17) can be automated by use of a robotic system, thereby making it more convenient. However, this robotic system is not readily available in every laboratory.
Here, we describe a rapid and reliable reversed-phase HPLC method for the simultaneous determination of MPA and MPAG in plasma, as well as for the determination of the MPAf concentration in plasma water, based on simple sample preparation and gradient elution chromatography. The method has a lower detection limit than previously published procedures. This renders the method particularly suitable for measurement of the MPAf in plasma samples.
Materials and Methods
DRUGS AND REAGENTS
MPA, MPAG, and the carboxybutoxy ether of MPA (MPAC) were a gift from Hoffmann-La Roche (Grenzach-Wyhlen, Germany). Sodium tungstate dihydrate, potassium dihydrogen phosphate, sodium hydroxide, phosphoric acid, and perchloric acid were from Merck. Acetonitrile (HPLC-grade) was obtained from J. T. Baker B.V. Stock solutions of MPA and MPAC in acetonitrile, each at a concentration of 1 g/L, were prepared separately and stored at -20 [degrees]C. The MPAG stock solution (5 g/L) was prepared in a solution containing 800 mL/L acetonitrile and 200 mL/L water, and stored at -20 [degrees]C.
EDTA-blood was centrifuged at 30008; plasma was collected and either immediately analyzed or stored at -20 [degrees]C until measurement. For determination of the plasma concentrations of MPAG and total MPA, 200 [micro]L of the sample (calibration, control, or plasma) and 100 [micro]L of acetonitrile containing the internal standard MPAC (15 mg/L) were vortex-mixed in a 1.5-ml, polypropylene tube for 5 s. This was followed by the sequential addition of 20 [micro]L of 150 g/L perchloric acid and 20 [micro]L of 250 g/L sodium tungstate to the tube, with vortex-mixing for 15 s after each addition. The sample was then centrifuged for 5 min at 10 000g, and 50 [micro]L of the supernatant was removed for chromatography. The Centrifree[R] Micropartition System (Amicon) described by Nowak and Shaw (12) was used to obtain an ultrafiltrate for MPAf determination. Because impurities were found in some batches of the Centrifree filters, all filters were routinely washed immediately before use. The micropartition tubes were soaked in a mixture of 500 mL/L methanol and 500 mL/L water and sonicated for 10 min. They were subsequently flushed with a fresh mixture of 500 mL/L methanol and 500 mL/L water and centrifuged for 20 min at 20008, leading to the complete accumulation of the flush solution in the collection cups. The flush solution was then discarded, and the cups were carefully dried with a cotton tip. For the ultrafiltration procedure, 300 [micro]L of either calibration or control solution or plasma sample was added to the sample reservoir, and the tube was centrifuged for 40 min at 2000g at 20 [degrees]C, yielding ~150 [micro]L of ultrafiltrate. The internal standard MPAC (2.5 mg/L) was mixed with the ultrafiltrate at a ratio of 1:10 (by volume), and a 100 [micro]L aliquot was then injected directly onto the column.
The HPLC system consisted of a chromatographic pump (M480), an automatic injector (GINA 50), a diode array detector (UVD 340S), a computer interface system controller linked to a PC (Gyncotek) and a 250 mm x 4.6 mm [Symmetry-C.sub.18]. reversed-phase column (Waters Associates). The column was maintained at 38 [degrees]C in a column temperature compartment (Du Pont Instruments) to improve separation. The mobile phase consisted of solution A (250 mL of acetonitrile and 750 mL of 20 mmol/L phosphate buffer, pH 3.0) and solution B (700 mL of acetonitrile and 300 mL of 20 mmol/L phosphate buffer, pH 6.5) that formed the following gradient: 0-4.5 min, 3% B; 4.5-5 min, 30% B; 5-12 min, 30% B; 12-12.5 min, 100% B;12.5-14.5 min, 100% B;14.5-15 min, 3% B. The flow rate was 1.2 mL/min, and the compounds were quantified in parallel by absorbance at 254 and 215 nm. The Gyncosoft program, Ver. 5.42 (Gyncotek), was used for recording and calculating the data and also for recording the UV spectra when required. Calculations were made in the internal standard mode, using peak area ratios. Calibration and quality control were performed by use of in-house prepared solutions that were analyzed with each run. Calibration and control materials for total MPA and MPAG were prepared by adding the analytes to drug-free plasma. MPA was calibrated at a concentration of 3 mg/L; MPAG was calibrated at 200 mg/L. Solutions containing 0.2, 1.0, and 25.0 mg/L MPA and 10, 50, and 250 mg/L MPAG were used for control of those analytes. A protein-free NaCl (9 g/L) solution adjusted to pH 7.4 with phosphate buffer (67 mmol/L) and to which 0.05 mg/L MPA was added was used for calibration of the MPAf determination. For accuracy control of MPAf measurements, 0.025 and 0.1 mg/L MPA were added to the NaCl solution. The precision of MPAf measurement was established by use of drug-free plasma to which 1 and 5 mg/L MPA were added.
ASSESSMENT OF PERFORMANCE CHARACTERISTICS
The detection limit of the method was calculated using a signal-to-noise ratio of 3. For this purpose, the noise signal of the baseline was obtained from a segment of the chromatogram that preceded each MPA or MPAG peak, respectively. The linearity of the method was established using drug-free plasma with MPA and MPAG added to yield concentrations from 0.04 to 50 mg/L for MPA and 1-500 mg/L for MPAG. Within-run and between-run precision, as well as extraction efficiency, were studied with drug-free plasma to which MPA and MPAG were added to yield final MPA concentrations of 0.2, 1, 5, and 25 mg/L and final MPAG concentrations of 10, 50, and 250 mg/L.
Extraction efficiency was calculated by comparing peak areas obtained from the extracted plasma samples with added MPA or MPAG with peak areas obtained with acetonitrile/water (80:20, by volume) containing the same amounts of MPA and MPAG, which were directly injected onto the column without extraction. To obtain information on the accuracy of the method, including the influence of the plasma protein concentration, the recovery was determined by adding known amounts of MPA (0.2,1, 5, and 25 mg/L) and MPAG (10, 50, and 250 mg/L) to drug-free plasma pools containing 41, 67, and 91 g/L total protein. The percentage of recovery was calculated by comparing the measured concentrations with the expected concentrations.
Potential chromatographic interference by commonly administered drugs was evaluated by analysis of patient specimens received for routine therapeutic drug monitoring, including transplant patients under immunosuppressive therapy without MMF, human therapeutic drug monitoring quality-control sera (Ciba Corning Diagnostics), and drug-containing methanol standards. In addition, the existence of endogenous chromatographic interferences was evaluated by separate analysis of 60 patient specimens, including transplant recipients without MMF therapy, sent to the laboratory for routine clinical chemical tests.
Twenty-five plasma specimens derived from pharmacokinetic studies in renal transplant recipients were used for method comparison with the HPLC procedure that has been used for the acquisition of extensive pharmacokinetic/pharmacodynamic data on MPA and MPAG (17,18). Frozen plasma specimens in which MPA and MPAG concentrations had been determined by the method of Tsina et al. (17) at the Philadelphia laboratory were shipped to Gottingen. After the samples were thawed, they were analyzed using the new HPLC procedure without prior knowledge of the results from Philadelphia.
The nonparametric regression procedure of Passing and Bablok (19) was used for comparison of the two HPLC methods (EVAPAK, Ver. 2.08, Boehringer Mannheim). The regression equations are given in Results, together with the 95% confidence intervals for the estimates of slope and intercept in parentheses. The dispersion of the residuals are documented as the 68% median distance. For comparison, the standard deviation of the residuals, [S.sub.y|x], calculated using the standard principal component procedure, is also given.
Representative chromatograms of a plasma sample, containing 2.11 mg/L MPA and 102.0 mg/L MPAG, obtained from a patient on MMF therapy immediately before a dose and a sample obtained from a renal transplant patient not on MMF therapy are shown in Fig. 1. Retention times of MPAG, MPAC, and MPA were 5.7 min ([+ or -] 0.4 min), 11.2 min ([+ or -] 0.6 min), and 12.7 min ([+ or -] 0.6 min), respectively. As depicted, the substances eluted fully baseline-separated as symmetric peaks (Fig. 1), which facilitated their subsequent quantification. Each chromatographic run was completed within 20 min. With the exception of one case, no interfering endogenous peaks were detected in patient samples. The exception was plasma obtained from a patient who later died, in which unidentified peaks that interfered with the quantification of MPAG were observed. The interferents were presumably produced by a combination of multiple drugs administered and a considerably altered drug metabolism in this individual. The specificity of the assay was further examined by analyzing various drugs that may be potentially administered with MMF (Table 1). None of the drugs listed in Table 1 caused any interference under the conditions of the assay described.
The detection limit (signal-to-noise ratio, 3) at 215 nm for plasma samples was 0.01 mg/L for total MPA and 0.03 mg/L for MPAG. The corresponding values for detection at 254 nm were 0.03 mg/L for total MPA and 0.1 mg/L for MPAG. In the case of the MPAf concentration, even lower detection limits of 0.005 mg/L at 215 nm and 0.015 mg/L at 254 nm could be achieved, because of the lower background noise in the chromatograms associated with the use of ultrafiltrates.
The linearity of the assay was verified up to 50 mg/L MPA and up to 500 mg/L MPAG (correlation coefficient >0.999). Performance characteristics of this method were tested at different concentrations of MPA and MPAG (Table 2). In the working ranges for MPA (0.2-25 mg/L) and MPAG (10-250 mg/L), the extraction efficiency was 56-59% for MPA and 73-77% for MPAG (n = 3). The extraction efficiency for MPAC was 65% (n = 3). The analytical recovery of MPA was in the range of 95-106%; for MPAG, analytical recovery was 96-106% (Table 3). There were no major differences in the recovery of MPA and MPAG, using plasma specimens with low, normal, and increased total protein concentrations (Table 3).
[FIGURE 1 OMITTED]
The within-day CVs for measurement of MPAf were 11.8% (at 17.9 [micro]g/L) and 6.8% (at 210.1 [micro]g/L) at 254 nm and 11.6% (at 18.4 [micro]g/L) and 6.5% (at 209.2 [micro]g/L) at 215 nm (n = 12). The between-day CVs were 15.8% (at 22.8 [micro]g/L) and 7.5% (at 197.2 [micro]g/L) at 254 nm and 14.6% (at 23.2 [micro]g/L) and 7.2% (200.6 [micro]g/L) at 215 nm (n = 12).
Very good agreement (Fig. 2A) was observed between the plasma MPA concentrations determined by our new HPLC procedure [y] and those obtained with the method of Tsina et al. [x] (17): y = 1.086 (1.042-1.119)x + 0.224 (0.066-0.323), 68% median distance = 0.136 mg/L, [S.sub.y|x] = 0.157 mg/L. This was also true for the comparison of plasma MPAG concentrations (Fig. 2B) determined with the two procedures: y = 1.033 (0.973-1.075)x + 1.016 (-5.935-8.896), 68% median distance = 6.278 mg/L, [S.sub.y|x] = 11.387 mg/L.
The results obtained using the new HPLC procedure for samples from an international MPA proficiency testing scheme are presented in Table 4. Excellent agreement was found between the measured MPA concentrations in the samples and the target concentrations.
The applicability of the method was proven by analyzing MPA and MPAG concentrations in >1500 plasma samples and MPAf in >600 samples collected from liver and kidney transplant recipients. The concentration-time profile of MPA and MPAG in a pediatric kidney recipient who was treated with 0.6 g/m2 twice per day MMF and who is participating in an ongoing pharmacokinetic study is shown in Fig. 3. The determination of MPAf in 305 plasma samples drawn for pharmacokinetic profiles in children after kidney transplantation revealed a median for protein-bound MPA of 98.1% (range, 92.9% to 99.3%). MPAf concentrations below the respective detection limits were observed in 27 of 305 (8.7%) samples at 215 nm and in 38 of 305 (12.6%) at 254 nm.
The purpose of the present study was to develop a method for the simultaneous determination of MPA and MPAG within one run and with an analytical sensitivity sufficient to allow the additional measurement of MPAf concentration in plasma water. In contrast to previously published HPLC protocols, these requirements were successfully met with the method presented here. Furthermore, the assay protocol is particularly convenient because of extremely simple sample preparation and no requirement for special equipment. According to our experience, a batch of 25 MPA and MPAG samples can be prepared within 30 min. In combination with an autosampling device, this allows the preparation and analysis of up to 50 samples during 24 h. This seems to be sufficiently rapid, even in laboratories with a high workload. The performance characteristics show that the assay is accurate and reproducible. Our participation in an international MPA proficiency testing scheme revealed a deviation from the target values of <7%. The method was further validated by comparison with an established HPLC method (17,18) that has been used to obtain much of the pharmacokinetic data on MPA. Very good agreement was observed between the two methods for both MPA and MPAG. Results with the new HPLC procedure were marginally higher, but the slopes of the regression lines were <10% from the line of identity.
[FIGURE 2 OMITTED]
The sensitivity of this procedure allows the quantification of MPAf concentration after separation of the protein-bound MPA, using the ultrafiltration procedure described by Nowak and Shaw (12). The observation of these authors that binding of MPA to the ultrafiltration system is insignificant was confirmed in the present investigation (data not shown). The sample preparation described with acetonitrile/perchloric acid/sodium tungstate (5:1:1, by volume) leads to an extraction efficiency of >56% and >72% with MPA and MPAG, respectively. The efficiency for MPA can be improved by increasing the volume of acetonitrile, with a concomitant decrease in the recovery of MPAG. The conditions described here represent a compromise to extract both MPA and MPAG simultaneously in one extraction step. With a detection limit of 0.03 mg/L at 254 nm and 0.01 mg/L at 215 nm for MPA, the extraction efficiency was sufficient for all plasma specimens investigated thus far. There was no need, therefore, to modify the protocol. Our equipment allowed us to measure the peaks at both 254 and 215 nm in parallel. Analysis of the spectra under our assay conditions revealed three absorption peaks at 215 nm, 251 nm, and 305 nm for MPA, whereas MPAG had three absorption peaks at 215 nm, 251 nm, and 294 nm. Both compounds displayed their absorption maximum at 215 nm. Therefore, it is also possible to follow MPA and MPAG at either 254 nm or 215 nm. The detection at 215 nm was superior to that at 254 nm, which has been used exclusively before. In our experience, measurement at 215 nm not only improved the detection limit and the sensitivity of the method but also the specificity. Because of the lower detection limit, detection at 215 nm is also recommended for determination of MPAf.
[FIGURE 3 OMITTED]
By changing the gradient conditions to 0-4 min, 3% B; 4-4.5 min, 26% B; 4.5-7.5 min, 26% B; 7.5-9 min, 35% B; 9-14.5 min, 35% B; 14.5-15.5 min, 100% B; 15.5-18 min, 100% B; 18-18.5 min, 3% B, MMF can also be measured using this method (retention time ~10 min; extraction efficiency, 60%; data not shown). In addition, when no MPAG determination is required, MPA alone can be separated using isocratic conditions (mobile phase: 450 mL acetonitrile/550 mL 20 mmol/L phosphate buffer, pH 4.5), a flow rate of 1.2 mL/min, and a chromatography run time of 13 min.
The results show that the presented method is a reliable and convenient procedure for the quantification of plasma concentrations of total and free MPA, as well as of MPAG.
We thank Tanja Schneider for excellent technical assistance. M. Shipkova was supported by a grant from Hoffmann-La Roche.
Received August 18, 1997; revision accepted March 20, 1998.
(1.) Fulton B, Markham A. Mycophenolate mofetil. A review of its pharmacodynamic and pharmacokinetic properties and clinical efficacy in renal transplantation. Drugs 1996;51:278-98.
(2.) Sollinger HW. Update on preclinical and clinical experience with mycophenolate mofetil. Transplant Proc 1996;28(Suppl 1):24-9.
(3.) Lipsky JJ. Mycophenolate mofetil. Lancet 1996;348:1357-9.
(4.) Wu JC. Mycophenolate mofetil: molecular mechanisms of action. Perspect Drug Discov Des 1994;2:185-204.
(5.) Ransom JT. The mechanism of action of mycophenolate mofetil. Ther Drug Monit 1995;17:681-4.
(6.) Halloran P, Mathew T, Tomlanovich S, Groth C, Hooftman L, Barker C, for the International Mycophenolate Mofetil Renal Transplant Study Group. Mycophenolate mofetil in renal allograft recipients. Transplantation 1997;63:39-47.
(7.) Growth CG, for the European Mycophenolate Mofetil Cooperative Study Group. The European experience with mycophenolate mofetil. Transplant Proc 1996;28(Suppl 1):30-3.
(8.) The European Mycophenolate Mofetil Cooperative Study Group. Placebo-controlled study of mycophenolate mofetil combined with cyclosporin and corticosteroids for prevention of acute rejection. Lancet 1995;345:1321-5.
(9.) Nowack R, Birck R, van der Woude F.I. Mycophenolate mofetil for systemic vasculitis and IgA nephropathy. Lancet 1997;349:774.
(10.) Shaw LM, Nowak I. Mycophenolic acid: measurement and relationship to pharmacologic effects. Ther Drug Monit 1995;17: 685-9.
(11.) Shaw LM, Sollinger HW, Halloran P, Morris RE, Yatscoff RW, Ransom J, et al. Mycophenolate mofetil: a report of the consensus panel. Ther Drug Monit 1995;17:690-9.
(12.) Nowak I, Shaw LM. Mycophenolic acid binding to human serum albumin: characterization and relation to pharmacodynamics. Clin Chem 1995;41:1011-7.
(13.) Bopp RJ, Schirmer RE, Meyers DB. Determination of mycophenolic acid and its glucuronide metabolite in plasma. J Pharm Sci 1972;61:1750-3.
(14.) Gainer FE, Wesselman HJ. GLC of mycophenolic acid and related compounds. J Pharm Sci 1970;59:1157-9.
(15.) Sugioka N, Odani H, Ohta T, Kishimoto H, Yasumura T, Takada K. Determination of a new immunosuppressant, mycophenolate mofetil, and its active metabolite, mycophenolic acid, in rat and human body fluids by high-performance liquid chromatography. J Chromatogr B 1994;654:249-56.
(16.) Li S, Yatscoff RW. Improved high-performance chromatographic assay for the measurement of mycophenolic acid in human plasma. Transplant Proc 1996;28:938-40.
(17.) Tsina I, Chu F, Hama K, Kaloostian M, Tam Y, Tarnowski T, Wong B. Manual and automated (robotic) high-performance liquid chromatography methods for the determination of mycophenolic acid and its glucuronide conjugate in human plasma. J Chromatogr B 1996;675:119-29.
(18.) Shaw LM, Korecka M, Van Breeman R, Nowak I, Brayman KL. Analysis, pharmacokinetics and therapeutic drug monitoring of mycophenolic acid. Clin Biochem (in press).
(19.) Passing H, Bablok W. A new biometrical procedure for testing the equality of measurements from two different analytical methods. Application of linear regression procedures for method comparison studies in clinical chemistry, part I. J Clin Chem Clin Biochem 1983;21:709-20.
MARIA SHIPKOVA,  * PAUL DIETER NIEDMANN,  VICTOR WILLIAM ARMSTRONG,  EKKEHARD SCHUTZ,  EBERHARD WIELAND,  LESLIE M. SHAW,  and MICHAEL OELLERICH 
 Abteilung Klinische Chernie, Georg-August-Universitat Gottingen, D-37075 Gottingen, Germany.
 Department of Pathology & Laboratory Medicine, University of Pennsylvania Medical Center, Philadelphia, PA 19104.
* Address correspondence to this author at: Abteilung Klinische Chemie, Zentrum Innere Medizin, Georg-August-Universitat, Robert Koch Strasse 40, D-37075 Gottingen, Germany. Fax 49-551-398551; e-mail firstname.lastname@example.org.
 Nonstandard abbreviations: MPA, mycophenolic acid; MMF, mycophenolate mofetil; MPAf, free mycophenolic acid; MPAG, mycophenolic acid glucuronide metabolite; and MPAG, mycophenolic acid carboxybutoxy ether.
Table 1. Drugs found not to cause an interference with the quantification of MPA or MPAG. Acetaminophen Cyclosporine A Phenobarbital Amikacin Desipramine Phenytoin Amitriptyline Diazepam Prednisolone Amoxicillin Digoxin Prednisone Amphotericin B Disopyramide Primidone Caffeine Dopamine Quinidine Carbamazepine Ethosuximide Rapamycin Cefazolin Gentamicin Salicylate Chloramphenicol Imipramine Tacrolimus Cimetidine Lamotrigine Theophylline Clemastine Lidocaine Tobramicin Clonazepam Netilmicin Vancomycin Table 2. Imprecision of the HPLC method. Analyte added, Within-run Between-run UV detection mg/L CV, % (n = 12) CV, % (n = 12) 254 nm MPA 25 1.7 4.0 5 2.2 3.8 1 3.2 5.3 0.2 8.4 9.2 MPAG 250 2.6 4.9 50 2.5 5.5 10 4.9 6.1 215 nm MPA 25 1.7 3.3 5 2.1 4.3 1 2.3 5.8 0.2 4.4 6.2 MPAG 250 2.7 4.1 50 3.0 5.2 10 3.4 5.9 Table 3. Analytical recovery of MPA and MPAG and influence of the plasma protein concentration. Analyte recovery, % (n = 3) UV detection Analyte added, 41 g/L 67 g/L 91 g/L mg/L protein protein protein 254 nm MPA 25 100 103 100 5 104 101 99 1 103 100 102 0.2 103 104 98 MPAG 250 98 104 99 50 102 100 98 10 104 104 96 215 nm MPA 25 100 99 98 5 103 98 97 1 102 101 104 0.2 102 95 106 MPAG 250 102 101 99 50 106 98 100 10 104 96 96 Table 4. Comparison of the MPA concentrations measured using the new HPLC procedure with the target concentrations in five samples (a) to which known concentrations of MPA were added. Sample Target HPLC results concentration, mg/L (Gottingen), mg/L 1A 8.00 8.53 1B 2.00 2.01 2A 0.00 <0.05 2B 5.00 5.05 3B 3.00 3.07 (a) Samples were from the Mycophenolic Acid Proficiency Testing Scheme of Analytical Services International.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Drug Monitoring and Toxicology|
|Author:||Shipkova, Maria; Niedmann, Paul Dieter; Armstrong, Victor William; Schutz, Ekkehard; Wieland, Eberha|
|Date:||Jul 1, 1998|
|Previous Article:||Cross-reactivity of fosphenytoin in two human plasma phenytoin immunoassays.|
|Next Article:||Sensitive enzymatic assay for erythrocyte creatine with production of methylene blue.|