Multi-center evaluation of analytical performance of the microparticle enzyme immunoassay for sirolimus

Long-term evaluation of analytical methods used in sirolimus therapeutic drug monitoring

Results of therapeutic monitoring of sirolimus blood concentrations are assay and laboratory dependent. This study compared performance over time of the IMx microparticle enzyme immunoassay (MEIA), Architect chemiluminescent microparticle immunoassay (CMIA), and liquid chromatography with mass spectrometric detection (LC/MS/MS) as part of a proficiency testing scheme. Pooled samples from sirolimus-treated patients and whole-blood samples spiked with known quantities of sirolimus were assayed monthly between 2004 and 2012. When results of pooled patient samples were compared with LC/MS/MS, the MEIA assay showed an overall mean percent bias of -2.3% ± 11.2% that, although initially positive, became increasingly negative from 2007 through 2009. The CMIA, which replaced the MEIA assay, had a mean percent bias of 21.9% ± 12.3%, remaining stable from 2007 through 2012. Similarly, for spiked samples, the MEIA showed an increasingly negative bias over time vs. LC/MS/MS, whereas CMIA maintained a stable positive bias. Based on comparison of immunoassay measurements on individual patient samples, CMIA values were more than 25% higher than MEIA values. These results highlight the importance of continued proficiency testing and regular monitoring of sirolimus assay performance. Clinicians must be aware of the methodology used and adjust target levels accordingly to avoid potential effects on efficacy and toxicity.

Development and validation of an HPLC method for the analysis of sirolimus in drug products

PURPOSE

The aim of this study was to develop a simple, rapid and sensitive reverse phase high performance liquid chromatography (RP-HPLC) method for quantification of sirolimus (SRL) in pharmaceutical dosage forms.

METHODS

The chromatographic system employs isocratic elution using a Knauer- C18, 5 mm, 4.6 × 150 mm. Mobile phase consisting of acetonitril and ammonium acetate buffer set at flow rate 1.5 ml/min. The analyte was detected and quantified at 278nm using ultraviolet detector. The method was validated as per ICH guidelines.

RESULTS

The standard curve was found to have a linear relationship (r(2) > 0.99) over the analytical range of 125-2000ng/ml. For all quality control (QC) standards in intraday and interday assay, accuracy and precision range were -0.96 to 6.30 and 0.86 to 13.74 respectively, demonstrating the precision and accuracy over the analytical range. Samples were stable during preparation and analysis procedure.

CONCLUSION

Therefore the rapid and sensitive developed method can be used for the routine analysis of sirolimus such as dissolution and stability assays of pre- and post-marketed dosage forms.

Identification and elimination of ion suppression in the quantitative analysis of sirolimus in human blood by LC/ESI-MS/MS

Ion suppression can negatively affect the performance characteristics of LC/ESI-MS/MS based methods, and we wished to identify sources of ion suppression in an assay for quantitating sirolimus in human whole blood. We first compared the peak areas of sirolimus and ascomycin added to human blood samples treated with and without extraction using octadecyl silyl (ODS)-silica gel after protein precipitation, and we found that water-soluble compounds cause the ion suppression for both drugs. Post-column infusion studies indicated that compounds retained in the sample after ODS extraction and protein precipitation caused ion suppression. MS analysis of these compounds suggested they were hydroxyl group-possessing phosphocholines, and this was confirmed using purified lysophosphatidylcholine variants. Therefore, we included a HybridSPE treatment step after the ODS extraction into the preanalytical workflow to remove phosphocholines, and this successfully eliminated the observed ion suppression for determining sirolimus concentration in human whole blood by LC/ESI-MS/MS. Sirolimus is a highly lipophilic molecule, and this study demonstrates the impact that preanalytical extraction and purification steps can have on a laboratory's ability to accurately detect and quantitate this and other lipophilic drugs.

A rapid liquid chromatography-tandem mass spectrometry analysis of whole blood sirolimus using turbulent flow technology for online extraction

BACKGROUND

Sirolimus is an immunosuppressant used in solid organ transplantation. Due to variable individual pharmacokinetics and narrow therapeutic ranges, therapeutic drug monitoring (TDM) is critical to the success of post-transplantation patient care. We developed a rapid method quantifying whole blood sirolimus using liquid chromatography-tandem mass spectrometry (LC-MS/MS) with an automated online extraction technology.

METHODS

Whole blood (100 microL) was mixed with a precipitation solution containing internal standard (32-desmethoxyrapamycin) and centrifuged at 15,634 x g for 10 min. The supernatant (50 microL) was injected onto a turbulent flow preparatory column and then a C18 analytical column. The mass spectrometer was set for positive electrospray to monitor the ammonium adducts.

RESULTS

Analytical time was 4 min/injection. Inter- and intra-assay variation coefficients across three concentration levels ranged from 2.3% to 7.4%. The method was linear from 1.0 to 100.0 ng/mL with an accuracy of 93.3%-100.0%. No carryover was detected from samples at 313.6 ng/mL. There was no obvious ion suppression from patient samples or interference from other commonly used immunosuppressants. Good correlation with an in-house commercial LC-MS was observed.

CONCLUSIONS

The LC-MS/MS method coupled with turbulent flow technology is rapid and efficient in TDM of whole blood sirolimus.

Characterization of sirolimus metabolites in pediatric solid organ transplant recipients

Potential age-dependent changes of sirolimus metabolite patterns in pediatric renal transplant recipients remain elusive. Thirteen pediatric solid organ transplant recipients (10 kidney, one combined liver-kidney, two liver, mean age 8.0 +/- 5.0 yr) underwent a sirolimus pharmacokinetic profile in steady-state with 10 samples drawn over 12 h post-intake to calculate the AUC(0-12 h). Concentrations of sirolimus and metabolite were quantified using a validated LC-MS/MS assay and metabolite structures were identified directly in blood extracts using LC-MS/iontrap. Average sirolimus AUC(0-12 h) was 64.9 +/- 29.7 ng h/mL. Median (range) AUC(0-12 h) for each metabolite (ng h/mL) was: 12-hydroxy-sirolimus 7.6 (0.2-18.8), 46-hydroxy sirolimus 3.1 (0.0-12.4), 24-hydroxy sirolimus 4.3 (0.0-12.6), piperidine-hydroxy sirolimus 3.5 (0.0-8.3), 39-O-desmethyl sirolimus 3.6 (0.0-11.3), 16-O-desmethyl sirolimus 5.0 (0.1-9.9), and di-hydroxy sirolimus 4.3 (0.0-32.5). The metabolites reached a median total AUC(0-12 h) of 60% of that of sirolimus. The range was 2.6-136%, indicating significant variability. In all, 77.5% of the metabolites were hydroxylated, while 39-O-desmethyl sirolimus accounted for only 8.4% of the AUC(0-12 h). This is clinically relevant as 39-O-desmethyl sirolimus shows 86-127% cross-reactivity with the antibody of the widely used Abbott sirolimus immunoassay. The metabolism of sirolimus in the children included in our study differed from that reported in adults, which should be considered when monitoring sirolimus exposure immunologically.

Determination of blood sirolimus concentrations in liver and kidney transplant recipients using the Innofluor fluorescence polarization immunoassay: comparison with the microparticle enzyme immunoassay and high-performance liquid chromatography-ultraviolet method

BACKGROUND

Although high-performance liquid chromatography (HPLC) is the method of choice for blood sirolimus determination, the microparticle enzyme immunoassay (MEIA) run on the IMx analyser is widely used in therapeutic monitoring of this immunosuppressant agent. The aim of our study was to evaluate the possible determination of sirolimus using the fluorescence polarization immunoassay (FPIA) commercialized for everolimus quantification.

METHODS

Sirolimus concentrations were determined in whole-blood samples from liver and kidney transplant recipients using the Innofluor Certican FPIA (Seradyn Inc.) run on a TDx analyser (Abbott Laboratories), Sirolimus MEIA run on an IMx analyser (Abbott Laboratories), and HPLC (UV detection) methods.

RESULTS

The Innofluor FPIA has a similar cross-reactivity with everolimus and sirolimus, and the within- and between-run coefficients of variation obtained for sirolimus determination were 2.7%-13.3%. In analysing different blood samples from liver and kidney transplant patients the linear regressions obtained were: FPIA = 1.12 HPLC + 0.43 (n=104, r=0.874), MEIA = 1.14 HPLC (n=146, r=0.892), and FPIA = 1.00 MEIA + 0.29 (n=106, r=0.941). Better correlation coefficients were obtained between the methods in the liver transplant samples (r>or=0.900) than in the kidney transplant samples (r>or=0.849). No significant effect was found for sirolimus clearance or the blood hematocrit on the relationship between the results produced by both immunoassays and HPLC.

CONCLUSION

The Innofluor FPIA is a valid alternative with an analogous performance to the MEIA for the therapeutic monitoring of sirolimus.

Pharmacokinetics of mycophenolate mofetil and sirolimus in children

This review summarizes the pharmacokinetics in children and youths of 2 commonly used immunosuppressive drugs, mycophenolate mofetil (MMF) and sirolimus (Sir), as presented at the IATDMCT 2007 conference. The review focuses on the developmental changes of drug disposition during childhood and adolescence. Regarding mycophenolate mofetil, the authors were unable to demonstrate age dependency of MMF in combination with cyclosporine. By contrast, there was an inverse relationship between age and the dose-normalized mycophenolate (MPA) area-under-the-time-concentration curve (AUC) in children who received concomitant tacrolimus (Tac). Dose-normalized MPA AUCs were higher than commonly observed in adult patients. It can be hypothesized that the age dependency is related to developmental changes in the expression of the UDP-glucuronosyltransferases. Sirolimus half-life and mean residence time (MRT) are shorter than in adults. Similar to that in adults, there is a profound drug-drug interaction between cyclosporine and Sir. In our own experience, Sir was started at 0.13 +/- 0.05 mg/kg/day. The average Sir AUC was 64.9 +/- 29.7 ng*h/mL. The median (range) AUC for each metabolite was as follows: 12-hydroxy-Sir, 7.6 (0.2-18.8); 46-hydroxy-Sir, 3.1 (0.0-12.4); 24-hydroxy-Sir, 4.3 (0.0-12.6); piperidine-hydroxy-Sir, 3.5 (0.0-8.3); 39-desmethyl-Sir, 3.6 (0.0-11.3); 16-desmethyl-Sir, 5.0 (0.1-9.9); and di-hydroxy-Sir, 4.3 (0.0-32.5) ng*h/mL. Of the total metabolite AUC, 77.5% was due to hydroxylated metabolites, while 39-O-desmethyl Sir (the main metabolite in adults) comprised only 8.4% of the metabolites. This is clinically relevant, as 39-O-desmethyl Sir shows 86% to 127% cross-reactivity with the Sir immunoassay. Metabolites reached a median AUC of 60% of that of Sir, but the range was 2.6% to 136%. The age dependency of Sir metabolite formation was confirmed in a human liver microsome model. On the basis of the age dependency of piperidine-hydroxy Sir, the authors postulate that the ontogeny of the drug disposition can be largely explained by developmental changes of the CYP2C8 expression. In conclusion, both Sir and MMF drug disposition vary in children and adolescents from adult patients, most likely because of developmental changes of biliary transporters and metabolic enzymes.

Quantification of sirolimus and everolimus by immunoassay techniques: test specificity and cross-reactivity evaluation

The possible cross-reactivity of immunoassays with structurally-related drugs was investigated. Innofluor Certican (FPIA) calibrators were measured by using IMx Sirolimus assay (MEIA) and MEIA Sirolimus calibrators were analysed by using FPIA Certican assay. Drug concentrations were measured in 95 and 100 samples from renal transplanted patients (RTP) on sirolimus or everolimus treatment by using immunoassays and LC/ESI-MSMS. A high cross-reactivity was found both for MEIA and FPIA. High correlation degrees, confirmed by the Bland-Altman and the Eksborg tests, were found between drug concentrations measured in real samples by both immunoassays (r = 0.909 and r = 0.970, respectively). LC/ESI-MSMS analysis of samples containing sirolimus showed no positivity for everolimus. Similarly, samples from patients on treatment with everolimus resulted negative as far as regards sirolimus. MEIA and FPIA could be considered mutually reliable and accurate alternatives for the specific-drug immunoassay. It should be noticed that in patients switching from one drug to the other unreal overestimation of the blood levels of the current administered immunosuppressant can occur.

Comparison of the CEDIA® and MEIA® assays for the measurement of sirolimus in organ transplant recipients

OBJECTIVES

This study evaluated two immunoassays, the CEDIA assay and the MEIA assay, used for the measurement of whole blood levels of sirolimus in organ transplant recipients.

DESIGN AND METHODS

We report on the performance characteristics (total precision, limit of quantitation (functional sensitivity), limit of detection (analytical sensitivity), linearity, accuracy) for each assay. Patient correlation studies were performed, and the results were analyzed using Bland-Altman plots and Passing-Bablok analysis.

RESULTS

Total precision for the MEIA assay, corresponding to three mean concentrations of 5.0, 10.6 and 20.2 ng/mL, was 10.5, 8.5, and 6.7%, respectively. The limit of detection was determined to be 1.1 ng/mL and the limit of quantitation was 1.5 ng/mL. The mean recovery for CEDIA was 105.4%, and analysis of proficiency material demonstrated a large negative bias with respect to the mass spectrometry peer mean-later determined to be due to matrix interference. Results for the CEDIA assay showed a total precision, corresponding to a mean concentration of 5.4, 10.5 and 20.7 ng/mL, of 13.5, 5.6, and 4.1%, respectively. The limit of detection was found to be 4.8 ng/mL, with a limit of quantitation of 5.2 ng/mL. The mean recovery for MEIA was 110.1%, and analysis of proficiency material demonstrated good agreement with the mass spectrometry peer mean with a slight positive bias. Both assays were acceptably linear over the reportable range of the assay. Patient correlation studies demonstrated a positive average bias for both assays versus results from LC-MS measurement (0.9 ng/mL for MEIA, 2.1 ng/mL for CEDIA).

CONCLUSION

Based on this evaluation, the MEIA demonstrated acceptable performance for use in clinical monitoring of sirolimus. However, based on a higher limit of quantitation that falls within the therapeutic interval, the CEDIA is not recommended for clinical monitoring of sirolimus.

Simultaneous LC–MS–MS determination of cyclosporine A, tacrolimus, and sirolimus in whole blood as well as mycophenolic acid in plasma using common pretreatment procedure

The purpose of the study was to develop rapid and simple procedure for simultaneous determination of cyclosporine A (CsA), tacrolimus (TCR), and sirolimus (SIR) in whole blood and mycophenolic acid (MPA) in plasma. Ascomycin (ASCO), cyclosporine D (CsD), and desmethoxysirolimus (DMSIR) were used as internal standards (IS) for TCR, CsA and MPA, and SIR, respectively. In the method development, six-level blood calibrators were used for CsA (range 47-1725 ng/ml), TCR (range 2.1-38.8 ng/ml), and SIR (range 2.4-39.6 ng/ml). Four-level calibrators were used for MPA (range 0.15-5.48 microg/ml). Four levels of quality control (QC) standards were used for blood samples, together with two levels of QC standards in plasma. All QC standards and calibrators were obtained from commercial sources. Sample preparation based on precipitation of 50 microl of sample in zinc sulfate-methanol-acetonitrile mixture containing IS, followed by centrifugation. HPLC was performed on ChromSpher pi column, 30 mm x 3 mm, in ballistic gradient of ammonium formate buffer-methanol at 0.8 ml flow rate. Following gradient elution profile was applied: 0-1.2 min at 30% methanol (divert valve to waste), 1.21-3.1 min 97% methanol (divert valve to detector), 3.11-3.7 min 30% methanol (divert valve to waste). ESI-MS-MS (MRM) was done on TSQ Quantum instrument with ESI source in positive ion mode. Ammoniated adducts of protonated molecules were used as precursor ions for all analytes but MPA. For this compound sodium adduct was used. Following transitions were monitored: for CsA m/z 1220-1203; for CsD 1234-1217; for SIR 931.6-864.5 and 882.6; for DMSIR 902-834.5; for TCR 821.5-768.5 and 785.5; for ASCO 809.5-756; for MPA 343-211.6; for MPA-glucuronide 514-306 and 211.6. The limits of quantitation were: 1 ng/ml for TCR and SIR, 20 ng/ml for CsA, and 0.1 microg/ml for MPA. Post-column infusion experiments showed that no positive or negative peaks appeared after injection of matrix in the elution range of target compounds. General signal suppression caused by matrix ranged from 20-40%, and was caused mainly by zinc sulfate present in deproteinizing solution. Extracted samples were stable for 2 days at 4 degrees C and for at least 20 days at -20 degrees C. MPA was fully separated from its glucuronide, which was eluted at around 0.7-0.8 min and directed to the waste. Some mutual cross-contribution of CsD and CsA was observed (below 1%), other IS did not contribute to target compounds and vice versa. Observations of chromatograms from patients taken single therapy demonstrated that possible metabolites of CsA, TCR, or SIR did not interfere with target compounds or IS.

Determination of everolimus in whole blood using the Abbott IMx® sirolimus microparticle enzyme immunoassay

OBJECTIVES

Everolimus (Certican) is a new immunosuppressant derived from sirolimus (Rapamune) with a 2-hydroxyethyl chain at position 40 of the macrolide ring. The aim of our study was to evaluate the possible determination of everolimus in whole blood using a commercialized microparticle enzyme immunoassay for sirolimus determination.

DESIGN AND METHODS

Everolimus concentrations were determined in blood samples from 11 kidney transplant patients (n=51) and different control materials (n=35) using the Seradyn Innofluor Certican fluorescence polarization immunoassay (FPIA) and the Abbott IMx sirolimus microparticle enzyme immunoassay (MEIA).

RESULTS

The MEIA gave a concentration-dependent cross-reactivity with the everolimus, with a linear regression between the assigned values (y) for the Innofluor Certican calibrators and those obtained (x) using this immunoassay: y=0.96x+0.67 (ma68=0.21 microg/L, r=0.974, p<0.001). The within- and between-run coefficients of variation using the MEIA were <or=7.3%. In analyzing the different blood samples from patients and control materials using the MEIA and FPIA, a linear regression was found: MEIA=0.99FPIA-0.46 (ma68=0.32 microg/L, r=0.967, p<0.001). Correcting the MEIA results by means of the linear regression equation found between the assigned values for the Innofluor Certican calibrators and those obtained using this immunoassay led to a reduction in the deviation with respect to the FPIA values. The possible effect of the hematocrit on the results is analogous for both immunoassays.

CONCLUSIONS

The Abbott IMx sirolimus MEIA permits the simple and precise determination of everolimus in whole blood, making it a valid alternative for the therapeutic monitoring of this immunosuppressant agent.

Comparison of the Innofluor certican assay with HPLC-UV for the determination of everolimus concentrations in heart transplantation

OBJECTIVES

Therapeutic monitoring of everolimus with chromatographic methods (HPLC) enabled effective immunosuppression while limiting the incidence of drug-related adverse events. A fluorescence polarization immunoassay (FPIA) has been recently developed for the assessment of everolimus levels. The present study was designed to evaluate FPIA performance and to compare it to HPLC.

DESIGN AND METHODS

The performance of HPLC and FPIA was initially tested using drug-free whole blood spiked with different amount of everolimus concentrations and, subsequently, by analyzing 113 trough blood samples from heart transplant recipients chronically given everolimus as part of their immunosuppressive regimen.

RESULTS

Inaccuracy and imprecision of both methods were below 15%. The correlation between everolimus concentrations and measured FPIA and HPLC was good, with a Pearson coefficient of 0.9118. The FPIA gave a mean overestimation of 24.3% as compared with HPLC. As additional analysis, cross-reactivity ranging from 85.4% to 138.0% was found with sirolimus, an immunosuppressant with a chemical structure close to everolimus.

CONCLUSION

The FPIA demonstrated acceptable performance for therapeutic drug monitoring of everolimus, and is a viable alternative to HPLC-based methods.