| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Clinical Biochemistry, Wythenshawe Hospital, Manchester M23 9LT, United Kingdom.
2
Clinical Applications Group, Micromass, UK Ltd, Manchester M23 9LZ, United Kingdom.
aAuthor for correspondence. Fax 44-161-291-2125; e-mail bkeevil{at}smuht.nwest.nhs.uk.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: CsA was measured in whole blood from 180 cardiac and lung transplant recipients by a liquid chromatography-tandem mass spectrometry (MS) assay, and the results were compared with the Dade Behring Emit assay. Proteins were precipitated with acetonitrile containing ascomycin as internal standard. We used isocratic elution on a Supelco CN column (33 x 3.0 mm; 3- µm bead size) with a mobile phase of 65% aqueous acetonitrile containing ammonium acetate (2 mmol/L) and formic acid (1 g/L), at a flow rate of 0.5 mL/min, with a sample injection volume of 6 µL. We used positive-ion electrospray MS to monitor the ammonium adducts of the compounds of interest decomposing under controlled conditions to the most dominant fragments of the individual molecules. Calibration curves used linear least-squares regression with 1/x weighting.
Results: Maximum sensitivity was obtained by monitoring fragmentation of the ammonium adducts m/z 1220
m/z 1203 for CsA and m/z 809m/z 765 for ascomycin. Sample throughput, including preparation time, was 30 samples in 1.5 h with an injection-to-injection cycle time of 1.5 min. The calibration curve was linear to 5000 µg/L, with a detection limit of 0.03 µg/L and a limit of quantification of 1 µg/L. Regression analysis [tandem MS method (y) and Emit assay (x)] yielded a slope of 1.09 (± 0.03), an intercept of 6.2 (± 4.5) µg/L, and Sy|x = 27 µg/L.
Conclusions: Tandem MS assay is a realistic alternative to immunoassay for the routine monitoring of CsA in transplant recipients. Its wide dynamic range has utility for pharmacokinetic studies of CsA.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In an attempt to standardize the practice of CsA measurement, the current recommendations for CsA monitoring involve analyzing trough concentrations of the drug in EDTA whole blood, using a method specific for the parent compound (2). Because of the need to turnaround results within one dosing interval, laboratories serving busy transplant units have opted for the faster immunoassays rather than the more specific, but time-consuming, HPLC methods. Immunoassay procedures have thus become increasingly popular and dominate the methods used for CsA measurement, whereas HPLC, which is used by a minority of laboratories, is still regarded as a reference procedure (4). Although the immunoassays fulfill the turnaround criteria, there is increasing concern that none of the currently available immunoassays actually fully satisfies the performance criteria recommended in the consensus documents for accuracy and specificity of CsA measurement (2). The main reason for this is the cross-reactivity of the immunoassay antibodies with inactive CsA metabolites, and this is particularly important in liver transplant recipients (2)(5). Indeed, HPLC is considered the method of choice for monitoring liver transplant recipients because of the accumulation of CsA metabolites in trough concentration blood samples (2).
Despite these concerns, HPLC in its current form is incapable of providing routine service to a busy transplant center, and the situation is likely to become more difficult because recent clinical studies have suggested that measurement of CsA concentrations at single or multiple time points in the early period (0–6 h) after CsA administration might improve clinical outcome compared with traditional trough (predose) measurements (6)(7)(8)(9). This approach will increase the workload and will require assay techniques with a concentration range far beyond that of the current immunoassays, i.e., up to at least 2000 µg/L (6). Thus there is a growing need for a rapid, sensitive, and specific assay for CsA. The introduction of mass spectrometry (MS) coupled to HPLC may solve some of these difficulties and provide an appropriate turnaround time. MS is often thought to be an expensive and difficult technique; nevertheless, it is finding increasing use in laboratories screening for inborn errors of metabolism (10)(11), where it is estimated that >1 000 000 neonates will be screened for inherited metabolic disorders worldwide in 2001.
Advances in mass spectrometer design have allowed the production of compact bench-top instruments that can easily be interfaced to HPLC systems and are more suited to clinical laboratories. In addition, improvements in software and instrument control have simplified the operation of these devices to the point where they can now be viewed as robust, highly specific detectors with a wide dynamic range. Liquid chromatography-tandem MS (LC-MS/MS) methods have already been developed for immunosuppressive drug measurement, but the assay times are still longer than immunoassay methods, and rigorous sample cleanup with either solid-phase or liquid-liquid extraction is advocated (12)(13)(14)(15).
Previously reported methods for MS analysis of CsA have used structural analogs of CsA [cyclosporin C (12)(13) and cyclosporin D (CsD) (16)(17)(18)] as internal reference compounds for quantifying CsA. However, because of the restricted availability of cyclosporin analogs, we have explored the use of ascomycin as an internal standard (14). In addition, we have modified an existing LC-MS/MS method (13) in an attempt to reduce both sample preparation and analysis time to improve assay throughput, thus making the technique more suitable for routine CsA monitoring.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
sample analysis by emit
For comparative purposes, CsA concentrations were measured by the Emit assay (Dade Behring Ltd.). Whole-blood samples and calibrators were precipitated with the Emit Cyclosporin Sample Pre-treatment Solution, centrifuged at 10 000g for 5 min, and then analyzed using a Cobas Mira (ABX) with instrument settings recommended by Dade Behring. All samples for the comparison were analyzed on the same day, using freshly calibrated reagents.
internal standards and calibrators
CsA and ascomycin were purchased from Sigma-Aldrich Company Ltd. CsD was a gift from Novartis Pharma AG (Basel, Switzerland). A CsA stock solution (1 g/L) was prepared in methanol, and a series of calibrators (0, 1, 5, 10, 20, 100, 200, 500, 1000, 2000, 3000, and 5000 µg/L) was prepared by drying down dilutions of the stock solution and then reconstituting in CsA-free whole blood. Aliquots of the whole-blood calibrators were kept at 4 °C before use. A working internal standard/precipitating solution was prepared by adding ascomycin to acetonitrile to give a concentration of 45 µg/L. A similar internal standard/precipitating solution was also prepared with CsD (45 µg/L).
sample preparation for lc-ms/ms
Whole-blood samples or calibrators (100 µL) were added to precipitant (200 µL) and allowed to stand for 
5 min. The samples were then vortex-mixed vigorously for 30 s to disperse the precipitated material. After centrifugation at 10 000g for 3 min, the clear, colorless supernatant was transferred to an autosampler vial for further analysis. All samples investigated in this study were prepared as part of the routine work of the Wythenshawe Hospital clinical chemistry laboratory.
hplc
Chromatography was performed on a Waters 2790 Alliance HT LC system (Waters Ltd.). Processed whole-blood samples were analyzed by isocratic HPLC using a Supelco LC CN column (33 x 3.0 mm; 3-µm bead size; Sigma-Aldrich) at a flow rate of 0.5 mL/min with 65% aqueous acetonitrile containing ammonium acetate (2 mmol/L) and formic acid (1 g/L). The column was maintained at 30 °C, and optimum peak shape was achieved with an injection volume of 6 µL. Mobile phase was used as the needle wash and purge solution. The eluate was connected directly to the electrospray probe of the mass spectrometer with no splitting or solvent diversion.
ms
A Quattro LC tandem mass spectrometer fitted with a Z Spray ion source was used for all analyses (Micromass). The instrument was operated in electrospray positive-ionization mode and was directly coupled to the HPLC system (see above). System control and data acquisition were performed with MassLynx NT 3.4 software with automated data processing by the MassLynx Quantify program provided with the mass spectrometer. Calibration curves were constructed using linear least-squares regression with 1/x weighting.
To tune the mass spectrometer, a solution of CsA, ascomycin, or CsD (10 µmol/L) in mobile phase was infused into the ion source, and the cone voltage was optimized to maximize the intensity of the [M+NH4]+ precursor ions (m/z 1220, 809, and 1234, respectively). The collision energy was then adjusted to optimize the signal for the most abundant product ions (m/z 1203, 756, and 1217, respectively). Typical tuning conditions were as follows: electrospray capillary voltage, 1.0 kV; sample cone voltage, 20V; collision energy, 18 eV at a collision gas pressure 180 mPa of argon. Sample analysis was performed in the multiple-reaction monitoring mode of the mass spectrometer with a dwell time of 0.25 s/channel, using the following transitions: m/z12201203 for CsA; m/z809756 for ascomycin; and m/z12341217 for CsD.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Maximum sensitivity was achieved by monitoring the fragmentation of the CsA and CsD ammonium adducts [M+NH4]+, which undergo deamination at low collision energy to reform [M+H]+ in high abundance (Fig. 1
). Ascomycin also forms ammoniated molecules that undergo CID, but in this case the major product ion is [M+NH4 - 53]+.
|
Sample preparation with acetonitrile produced a clear, colorless supernatant that gave clean chromatograms with no interfering compounds present (Fig. 2
). Both ascomycin and CsA were clearly separated from the void volume (0.6 min) and eluted in <1.5 min, allowing an injection-to-injection cycle time of <2 min (Fig. 2
).
|
In the experiments performed, ascomycin eluted after 1.1 min, CsA eluted at 1.2 min, and CsD (when used) eluted at 1.2 min. Although the compounds of interest do not precisely coelute, the specificity of the detection method used means that the compounds of interest do not interfere with each other.
Carryover from the 1000 µg/L calibrator to the zero calibrator was <0.1%. Results were quantified by integrating the area under the extracted ion chromatograms for CsA and internal standard for a series of whole-blood calibrators. A calibration curve was constructed by plotting the CsA/internal standard peak-area ratio against CsA concentration (Fig. 3
). The curve was linear over the typical working range to 5000 µg/L (Fig. 3
) and showed good correlation with the stated values (y = 0.012x + 0.95; r = 0.997). The limit of detection was 0.03 µg/L (2.5 SD of zero calibrator) and the limit of quantification, derived from the precision profile curve, was <1.0 µg/L (Fig. 4
).
|
|
The stability of the extracted materials in the supernatant was tested by repeated injections of the extract from a pooled sample. Six-microliter injections were made every 10 min over a 14-h period. No systematic loss in sensitivity was observed in the absolute peak areas measured, and the CV for peak-area ratio (analyte/internal standard) was 3.4% (Fig. 5
).
|
The within- and between-day imprecision of the assay, assessed on three separate pools, are shown in Table 1
. Within-day imprecision (CV <5%) and between-day imprecision (CV <10%) were both acceptable and fulfilled the recommended performance criteria for imprecision (2).
|
The mean recovery of CsA at concentrations of 200-1000 µg/L was 110% (range, 102–119%). The time taken to process a batch of 30 samples plus controls and calibrators was 1.5 h, which was similar to the time taken to process the same batch of samples by the Emit assay.
A Bland-Altman plot (Fig. 6
) for the LC-MS/MS vs Emit comparison showed good agreement between the two methods and confirmed the expected lower results of the LC-MS/MS assay because of its higher specificity [mean (SD) difference = -8 (27) µg/L]. The equation for the Passing-Bablok regression line (Fig. 7
) was LC-MS/MS = (1.094 ± 0.03)Emit - (6.2 ± 4.5); r = 0.954; Sy|x =27 µg/L.
|
|
Previous studies on the analysis of CsA by MS have been performed using a structural analog of the analyte (cyclosporin C or CsD) as an internal standard (12)(13)(16)(17)(18). The use of structural analogs is logical because their properties will be more similar to the properties of CsA than are those of ascomycin. However, in the present comparison, there was no significant difference between the CsA results obtained for 70 patient samples when either ascomycin or CsD was used as the internal standard [CsD = (1.005)ascomycin + (1.63); r2 = 0.9957; Sy|x = 11 µg/L; Fig. 8
]. There were no significant outliers in the measurements, and with a slope approaching unity, the use of either CsD or ascomycin as an internal standard material is indicated. It is the intention of this study to propose the use of ascomycin rather than CsD as an internal standard material because cyclosporin structural analogs are not commercially available.
|
After the initial evaluation of the fragmentation of the pure compounds under investigation (10 µmol/L solutions), all subsequent analyses were performed using samples extracted from a blood matrix. Any suppression of the ionization process from extracted physiologic materials would have a similar effect on coeluting compounds, thus potentially reducing the observed signal for both the analyte and the internal standard but having no effect on the observed ratio between the two signals. To examine this effect in detail, we added CsA, at concentrations of 100, 500, and 1000 µg/L, to a series of six whole-blood samples and prepared the samples in the standard way. In addition, we prepared a series of aqueous calibrators in triplicate at identical concentrations. The area counts for the internal standard and CsA for the whole-blood and aqueous samples are shown in Table 2
, together with the CsA:internal standard ratios.
|
To examine the effects of other drugs on the analytical performance of the assay, samples from 20 patients on a tacrolimus (FK506) treatment regime were analyzed before and after the addition of CsA. In addition to tacrolimus, the patient cohort was also being treated with one or more of the drugs shown in Table 3
according to "standard" treatment regimes for the drugs in question. The samples should, therefore, have contained both parent drug and metabolites. However, one potential limitation of this part of the study was that the concentrations of the additional drugs were not determined. The empirical molecular formulas and monoisotopic molecular masses have also been included in Table 3
for comparison with the compounds investigated in this study. The whole-blood samples were then prepared in the described method. There were no interfering peaks present in the CsA-free samples because of the specificity of the detection method (see above). The CV for the peak-area ratio (CsA/ascomycin, i.e., the concentration of CsA) in the 20 samples was 6.7%. The CV for ascomycin as an absolute response was 5.3%. Suppression of ionization of either CsA or ascomycin by any of the drugs present was therefore not observed.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The use of ammoniated cyclosporin species increases the assay sensitivity by effectively concentrating all the usual fragment ions observed from the fragmentation of the protonated molecule into a single product-ion species. Although the fragmentation of both CsA and CsD appears very simplistic (loss of ammonia from an ammoniated species), we found that there is no interference from endogenous compounds and thus the transformation is highly specific. We have shown that the specificity of the MS/MS detector makes lengthy sample preparation using solid-phase or liquid-liquid extraction unnecessary for the assay of CsA in routine samples. We have therefore simplified the sample preparation stage and decreased the chromatography time to increase the speed of the assay. This has produced a faster assay than that reported previously (13) with no detrimental effect on sample carryover. There are therefore several advantages of this method over existing LC-MS/MS methods, including a much faster and easier sample preparation stage with fewer steps, coupled with a shorter analysis time on the LC-MS/MS to allow a faster throughput of samples. This has made the assay much more suitable for the analysis of large numbers of clinical samples, where a quick turnaround is required. The assay time is similar to that of the Emit assay, but the LC-MS/MS assay offers a much wider assay range in addition to a linear calibration (response) curve. The accuracy of the test for CsA concentrations >500 µg/L is thus improved, and the need for sample dilution is reduced when concentrations higher than this are encountered in routine testing. This is of particular importance because it has been demonstrated that clinical laboratories have shown large variability when diluting external quality assurance samples (19). The increased sensitivity of the assay also improves the accuracy of the assay for CsA concentrations <50 µg/L. This wider dynamic range should be useful for pharmacokinetic studies, especially if blood samples containing peak concentrations are used for profiling to replace measurements of trough concentrations (6).
The study of samples from patients on different drug treatments was used to assess the effect of other compounds present in the blood sample on efficiency of ionization. The ionization efficiency data suggest that there is no appreciable interference in the signal response of the targeted compounds from interfering substances. In addition, because the CsA and the ascomycin coelute from the LC column, any general ion suppression mechanism would affect the observed signals for both the analyte and the internal standard and thus not affect the accuracy of the CsA measurement. The effect of the matrix (whole blood) on the ionization of CsA and the internal standard was assessed by comparing the analysis of CsA added to whole blood and water. Because whole blood contains 220 g/L protein, the sample preparation method used here would produce a final supernatant 7% lower in volume compared with water, and the analytes (CsA and internal standard) would therefore appear to be 7% more concentrated. The results of the study shown in Table 2
indicate that, on average, the area of the internal standard was 6% higher in the whole-blood samples than in the aqueous samples. Furthermore, there is no apparent difference in the CsA:ascomycin peak-area ratio between the whole blood and the aqueous samples. These data suggest that the sample preparation and analytical conditions used do not cause any suppression of ionization for either CsA or ascomycin.
The comparison between the LC-MS/MS and Emit assays showed constant lower values for the LC-MS/MS assay and is in agreement with other published studies that showed a similar relationship, caused by interference with CsA metabolites in the Emit assay (4). The LC-MS/MS assay does need recalibrating with every batch because of changes in detector response, but because of the excellent curve linearity, only two or three calibrators are required for routine monitoring of trough CsA concentrations.
Ascomycin has been used previously as an internal quantifying standard for the determination of tacrolimus (14) with a solvent system similar to that used in this assay. It appears to make little difference that ascomycin is not a structural analog of CsA because the cyclosporin determinations were similar with both ascomycin or CsD as internal standard. Ascomycin also has the advantage of being commercially available, unlike CsD. The good correlations observed for both the weighed-in amounts of CsA and the existing Emit assay suggest that ascomycin is viable as an internal standard in this situation.
The method was developed to enable rapid preparation of patient samples coupled with a short analysis time. During this study, the LC columns used showed no appreciable decrease in performance after >1000 analytical injections of the supernatant. The design of the mass spectrometer ion source is intended for the analysis of physiologic extracts. In a typical analytical scenario, 
150 samples would be analyzed on a daily basis. Preventative source maintenance (<5 min, with no need to vent the mass spectrometer and no need to reoptimize the ion-source tuning conditions) was undertaken daily, although minimal visible debris was observed on the critical ion-source components.
Although the capital cost of the LC-MS/MS equipment is relatively high, typically 5–10 times the cost of conventional HPLC, the annual running costs are extremely low compared with immunoassays. There is therefore scope to purchase or lease equipment based on reagent savings, particularly for laboratories that perform large numbers of cyclosporin tests. Introduction of LC-MS/MS into a busy transplant center would also confer other benefits, because immunosuppressive drugs such as tacrolimus, rapamycin, and SDZ-RAD can also be measured by this technique (14)(15)(20)(21).
Although several studies have shown a definable bias between HPLC and immunoassay, it has recently been demonstrated that assay bias cannot be predicted in individual samples. The reason for this is that interindividual differences greatly exceed the influence of either type of organ transplant or degree of hepatic dysfunction; this therefore suggests that there is still a need for an easy, fast, and truly specific assay with high precision and sensitivity (19). We have shown that LC-MS/MS provides us with an assay that meets all of these requirements and will provide us with a valuable routine laboratory tool.
In conclusion, a method has been developed for the quantitative analysis of CsA that uses minimal sample preparation and a rapid LC-MS/MS method with internal standardization. The data collected showed good agreement with existing methodology and did not suffer from interferences from substances coadministered with CsA. The method is robust and can be readily incorporated into the day-to-day operations of a routine testing laboratory. The method has also been demonstrated to provide a linear response up to a whole-blood concentration of 5000 µg/L for CsA with an internal standard (ascomycin) at a single concentration. Although the cost-effectiveness of MS/MS in a routine clinical testing environment is a moot point, the sensitivity, specificity, and versatility of such methodologies must be considered when making such a judgment.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  J. Kuhn, C. Prante, S. Schon, C. Gotting, and K. Kleesiek Measurement of Fibrosis Marker Xylosyltransferase I Activity by HPLC Electrospray Ionization Tandem Mass Spectrometry Clin. Chem., December 1, 2006; 52(12): 2243 - 2249. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. J. Taylor, S. R. Brown, D. P. Cooper, P. Salm, M. R. Morris, P. I. Pillans, and S. V. Lynch Evaluation of 3 Internal Standards for the Measurement of Cyclosporin by HPLC-Mass Spectrometry Clin. Chem., October 1, 2005; 51(10): 1890 - 1893. [Full Text] [PDF]
|
|
|
|
 |
|
 R. G Morris Immunosuppressant Drug Monitoring: Is the Laboratory Meeting Clinical Expectations? Ann. Pharmacother., January 1, 2005; 39(1): 119 - 127. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. M. Annesley and L. Clayton Simple Extraction Protocol for Analysis of Immunosuppressant Drugs in Whole Blood Clin. Chem., October 1, 2004; 50(10): 1845 - 1848. [Full Text] [PDF]
|
|
|
|
|
|
M. Groschl, M. Uhr, and T. Kraus Evaluation of the Comparability of Commercial Ghrelin Assays Clin. Chem., February 1, 2004; 50(2): 457 - 458. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
