Validation and evaluation of a newly developed time-resolved fluoroimmunoassay for cyclosporine quantitative analysis in human whole blood

Journal of Pharmaceutical and Biomedical Analysis 177 (2020) 112890

Contents lists available at ScienceDirect

Journal of Pharmaceutical and Biomedical Analysis journal homepage: www.elsevier.com/locate/jpba

Validation and evaluation of a newly developed time-resolved fluoroimmunoassay for cyclosporine quantitative analysis in human whole blood Xuebin Wang a,1 , Yuan gao b,1 , Yunyun Yang a,1 , Zhengyue Liu a,1 , Lihong Gao a,1 , Fengbo Wu c,1 , Xianmi Yang a , Xiaojian Xu a , Zhuo Wang a,∗,1 , Shusen Sun d,e,∗∗,1 a

Department of Pharmacy, Shanghai Changhai Hospital, Second Military Medical University, Shanghai, 200433, China Department of Clinical Pharmacy and Pharmaceutical Management, School of Pharmacy, Fudan University, Shanghai, China c Shanghai Genext Medical Technology Co., Ltd., Shanghai, 200433, China d College of Pharmacy and Health Sciences, Western New England University, Springfield, MA, 01119, USA e Department of Pharmacy, Xiangya Hospital Central South University, Changsha, Hunan, 410008, China b

a r t i c l e

i n f o

Article history: Received 26 June 2019 Received in revised form 20 August 2019 Accepted 20 September 2019 Available online 22 September 2019 Keywords: Time-resolved fluoroimmunoassay Cyclosporine Chemiluminescent microparticle immunoassay Therapeutic drug monitoring Drug analysis

a b s t r a c t Various immunoassay methods have been developed and used for the therapeutic drug monitoring (TDM) of cyclosporine (CsA). However, there is no report on the application of a time-resolved fluoroimmunoassay (TRFIA) in routine CsA TDM. The aim of this study was to evaluate the feasibility and validate the performance of a newly developed TRFIA method for CsA analysis in human whole blood. The TRFIA method was then compared with the method of chemiluminescent microparticle immunoassay (CMIA). The calibration range of the CsA-TRFIA method was 0–1000 ng/mL. The linear range and correlation coefficients were 30–1000 ng/mL and more than 0.990, respectively. The accuracy, precision, and inter-batch range were 90.0%–110.0%, less than 10%, and no more than 15%, respectively. The lowest limit of detection was less than 10 ng/mL. The linear regression equation was YCMIA = 0.961XTRFIA + 3.357, which showed that the measurements of CMIA and TRFIA were strongly correlated (r = 0.980). The results demonstrate that TRFIA is a precise and reproducible method for detecting the CsA concentration and can be used for routinely CsA TDM. © 2019 Elsevier B.V. All rights reserved.

1. Introduction Cyclosporine (CsA) is a potent immunosuppressant used in solid organ transplantation and autoimmune diseases [1,2]. CsA inhibits calcineurin and exerts its immunosuppressive effect by suppressing T-lymphocyte proliferation and blocking interleukin2 production [3]. Therapeutic drug monitoring (TDM) of CsA is recommended because of its narrow therapeutic index, high variations of pharmacokinetics, extensive drug-drug interactions, and potentially severe adverse drug reactions [4,5]. CsA TDM is accomplished by immunoassays or liquid chromatography coupled to tandem mass spectrometry (LC–MS/MS) [5,6]. A variety of immunoassays have been developed including chemiluminescent

∗ Corresponding author. ∗∗ Corresponding author at: College of Pharmacy and Health Sciences, Western New England University, Springfield, MA, 01119, USA. E-mail addresses: [email protected] (Z. Wang), [email protected] (S. Sun). 1 These authors contributed equally to this work. https://doi.org/10.1016/j.jpba.2019.112890 0731-7085/© 2019 Elsevier B.V. All rights reserved.

microparticle immunoassay (CMIA), electrochemiluminescence immunoassay (ECLIA), enzyme-multiplied immunoassay technique (EMIT), fluorescence polarization immunoassay (FPIA), affinity column-mediated immunoassay (ACMIA), and the cloned enzyme donor immunoassay (CEDIA) [7–11]. Among these methods, CMIA, ECLIA, and EMIT assays are commonly used due to ease of operation and automation. However, the high cost of instruments and CsA kits prohibit their widespread use for routine CsA TDM, especially in developing countries. LC–MS/MS is a gold method for CsA analysis, and various LC–MS/MS assays have been developed [10,12,13]. However, the high cost of instruments and time also limits the application of LC–MS/MS in routine CsA monitoring. Therefore, there is a need to develop accurate and cost-effective analytical methods for routine CsA TDM. Time-resolved fluoroimmunoassay (TRFIA) has been proven to be very applicable for detecting viral protein antibodies and evaluating rabies vaccine potency due to its excellent precision, higher sensitivity, and broader detection range [14,15]. However, there has been no report on the application of TRFIA in the detection of

2

X. Wang, Y. gao, Y. Yang et al. / Journal of Pharmaceutical and Biomedical Analysis 177 (2020) 112890

drug concentration. In this paper, we validated the performance of the newly developed Quantitative Kits for the detection of CsA (TRFIA). Also, a total of 110 samples from recipients were analyzed by the CsA TRFIA and compared to that obtained by CMIA. The consistency between these two analytical methods was evaluated by a clinical trial [12,16–18]. This study aimed to evaluate the feasibility of TRFIA method for CsA analysis in human whole blood. 2. Experimental 2.1. The methods of TRFIA and CMIA for CsA analysis The CsA-TRFIA was performed strictly according to the kit instruction (Shanghai Genext Medical Technology Co., Ltd. Shanghai, China) using the semi-automatic TRFIA DR6608 analyzer (Guangzhou Darui Biotechnology Co.,Ltd. Guangzhou, China). The CsA-CMIA was conducted according to the ARCHITECT CsA kit instruction using the automatic CMIA ARCHITECT i1000SR analyzer (Abbott Park, IL 60064, USA) approved by the FDA. 2.2. Method validation 2.2.1. The limit of detection The limit of detection (LOD) was determined according to the guideline EP17 published by the Clinical and Laboratory Standards Institute (CLSI) [19]. The limit of blank (LOB) was first estimated by measuring replicates of a blank EDTA blood sample (n = 24) by CSA TRFIA on the DR6608 analyzer, and by calculating the mean result and the standard deviation (SD). The LOB was obtained according to the equation: LOB = mean blank + 1.645 (SD blank). LOD is then determined based on this equation: LOD = LOB + 1.645 (SD low CSA concentration), by utilizing both the measured LOB and the SD of test replicates of a blood sample containing 20 ng/mL of CSA prepared by adding pure CSA to a blank EDTA blood sample. 2.2.2. Linear range The main components of experimental samples were the highconcentration samples and the low-concentration samples, which were close to and beyond the detectable limit of the expected linear range. The concentration range of samples was 0–1,500 ng/mL. The following method was used to determine the sample concentration: a high-concentration stock solution of CsA calibrators was diluted by the zero calibrator, and the theoretical concentration of the diluted solution was calculated. There were two samples made by Shanghai Genext Medical Technology Co., Ltd. The whole blood CsA reserve solution of 10 ␮g/mL was diluted with the frozenthawed and EDTA-anticoagulated whole blood matrix of bovine. The high-concentration sample of 1500 ng/mL CsA was obtained. A low-concentration sample without CsA was obtained from the frozen-thawed and EDTA-anticoagulated whole blood matrix of bovine. Samples were preserved at 2–8 ◦ . Using the standard diluent as the matrix, a high-concentration CsA sample (1500 ng/mL) and a low-concentration CsA sample (0 ng/mL) were prepared. Both tthese samples were diluted to obtain eight samples with different concentrations by the method of double dilution. The sample preparation process was conducted according to the approximately equal concentration intervals. The volume of each sample was 1.50 mL. The concentrations of CsA samples were 30, 100, 200, 400, 800, 1000, 1200 and 1500 ng/mL. On the same day, linear analysis experiments were conducted by using the same batch of reagent kits and samples. All samples were measured for three times. The diluted concentration (Xi ) was the independent variable, and the mean value (Yi) of the test results was the dependent variable. The regression equation was derived from Xi and Yi. The linear correlation coefficient (r) should not be less than 0.990. The Xi was substituted into the linear regression

equation to calculate the estimated value of Yi, and the absolute deviation or relative deviation between Yi and the estimated value should meet the requirement of linear deviation [20–22]. 2.2.3. Accuracy Four samples with CsA at 50, 90, 300 and 600 ng/mL, respectively, were prepared by mixing the zero calibrator with a 5000 ng/mL CsA stock solution (prepared by dissolving pure CSA in zero calibrator) and then analyzed by the CSA TRFIA (n = 3). The mean value of the test results was taken respectively, and the relative deviation (B%) was calculated by the following formula: B%=(M-T)/T×100% (M, mean of the test results; T, the marked value of the reference material) [20–22]. 2.2.4. Precision With three batches of CsA-TRFIA kits and two CsA samples (150 and 600 ng/mL), two operators used the same set of instruments on the same day and repeatedly tested two times for the two samples, lasting 20 days, in order to evaluate the precision of intra- and interbatch, intra- and inter-day, and different operators. With three batches of CsA kits, one operator used the same set of instruments on the same day and repeated the test three to four times for the two samples. The average values and the relative range (R) of measured  results were  calculated by the formula: T =

1 + 2 + 3 /3 and R =

max − min /T ×

100% [i (i = 1,2,3)] [20–22]. 2.2.5. Specificity The CsA-TRFIA method was designed to have a mean recovery of 85.0%–115.0% in the pharmaceutical substances and potentially interfering endogenous substances. A study based on the guidance from the CLSI document EP7-A2 was performed for the CsA-TRFIA method at the CsA levels of 150 ng/mL and 600 ng/mL. Whole blood specimens which were spiked with CsA targeting concentrations of 150 ng/mL and 600 ng/mL were supplemented with the following potential pharmaceutical substances, endogenous substances and clinical conditions (including EDTA-K2 , heparin sodium, triglyceride, bilirubin, uric acid, prednisone, mycophenolic acid, vancomycin hydrochloride, amphotericin B, azathioprine, hydrocortisone acetate, allopurinol, cefotaxime sodium, glibenclamide, diltiazem hydrochloride, digoxin, gentamicin sulfate, gemfibrozil, tacrolimus and sirolimus). To evaluate whether the above 18 substances were interfering substances, every specimen was tested for three times by using three batches of CsA-TRFIA kits. If one of the above substances became an interfering substance, the next step was to evaluate the dose-effect of the interfering substance (Table 1). 2.2.6. Methods comparison A clinical trial was conducted to evaluate the consistency between TRFIA and CMIA for CsA analysis. This comparative study was approved by the Shanghai Changhai Hospital Ethics Committee (Ethical approval number: CHEC2017-004). The CsA-TRFIA method was designed to have a correlation coefficient of ≥ 0.90 for specimens between 30–1,000 ng/mL when compared to the ARCHITECTi1000 CsA-CMIA method. This study was performed to compare the TRFIA method to the CMIA method by using 110 human whole blood EDTA-K2 specimens from kidney transplant patients, hematopoietic stem cell transplant patients, and patients with aplastic anemia or nephrotic syndrome. Both the equation and the correlation coefficient of the two measurements were evaluated by the Passing–Bablok regression and the Pearson test. The agreement of the two methods was evaluated by the Bland-Altman plot. The values of the two measurements were evaluated by paired- samples t- test [16–18].

X. Wang, Y. gao, Y. Yang et al. / Journal of Pharmaceutical and Biomedical Analysis 177 (2020) 112890

3

Table 1 The interfering substances for validation of specificity. Test Compound

Concentration

Dissolved matrix

Test Compound

Concentration

Dissolved matrix

Sampl 1 Sampl 2 Sampl 3 EDTA-K2 heparin sodium triglyceride bilirubin uric acid tacrolimus sirolimus prednisone

0.9%NaCl 0.1 N NaOH Methanol or DMF 4 mmol/L (1.5 mg/mL) 3000U/L 1 g/dL 40 mg/dL 40 mg/dL 60 ng/mL 60 ng/mL 100 ␮g/mL

– – – 0.1 N NaOH 0.9% NaCl 0.9% NaCl 0.1 N NaOH 0.1 N NaOH methanol or DMF methanol or DMF methanol or DMF

mycophenolic acid Hydrocortisone acetate vancomycin hydrochloride azathioprine allopurinol cefotaxime sodium glibenclamide sulfate diltiazem hydrochloride digoxin gemfibrozil glibenclamide

500 ␮g/mL 1.2 ␮g/mL 6mg/dL 1mg/dL 5 mg/dL (50 ␮g/mL) 100 ␮g/mL 12 mg/dL 60 ␮g/mL 4.8 ng/mL 100 ␮g/mL 200 ng/mL

methanol or DMF methanol or DMF 0.9%NaCl 0.1 N NaOH 0.1 N NaOH 0.9%NaCl 0.9%NaCl methanol or DMF 0.9% NaCl methanol or DMF methanol or DMF

DMF: dimethylformamide.

Table 2 The lowest detectable limit of CsA-TRFIA kits. Lot: 20131202

Lot: 20131203

Lot: 20140101

Calibrator A-F (2 times)

Calibrator A (20 times)

CalibratorA-F (2 times)

Calibrator A (20 times)

Calibrator A-F (2 times)

Calibrator A (20 times)

436358 434928 329808 313148 234428 239210 165146 165273 119600 113081 66155 65239 X±S (cps) X-2S (cps) LOD (ng/mL)

– – 432288 418042 430847 390729 424912 414411 420891 434845 416823 414589 419907 425448 422394 424554 409536 384121 428424 415774 412823 399874 – – 417062 ± 13121 390820 9.6

436882 439637 323552 316538 230512 237404 162289 179122 106752 114305 68144 66114 X±S (cps) X-2S (cps) LOD (ng/mL)

– – 415593 426548 429161 423495 414070 414325 423429 414546 430427 435466 413826 417688 430084 421546 425574 421657 410282 420933 409588 409639 – – 420394 ± 7625 405145 5.6

483491 492871 357331 358070 255105 268955 179603 194881 133701 130808 73168 76986 X±S(cps) X-2S (cps) LOD (ng/mL)

– – 477080 459930 480379 453495 471804 464544 469548 463049 462243 422034 475968 479484 484177 473484 479388 470238 475419 470933 481787 465771 – – 469038 ± 13707 441625 7.8

2.3. Statistical analysis Statistical analysis was performed by using SPSS (version 21, SPSS, Chicago, USA) and MedCalc 15.8 (MedCalc Software bvba, Ostend, Belgium). Two-sided nominal P values < 0.05 were considered as statistically significant [12].

dard for CsA measurement, in China, there is no commercial reagent available for CsA monitoring by LC–MS/MS, and immunoassays are predominantly used as clinical routine. For convenience, we chose Abbott CsA-CMIA as the reference method for the comparison study due to an excellent correlation between CsA measurements by CsACMIA and LC–MS/MS [25].

3. Results and discussion

3.1. Limit of detection

The mechanism of the new TRFIA CsA analysis could be described as follows [23,24]. Similar to other quantitative immunoassays of whole blood CsA, CsA-TRFIA needs to process the whole blood sample to release CsA from its binding proteins. After the sample pretreatment, the obtained clear supernatant with CsA extracted was added into the microwells coated with CsAprotein conjugates and competed with the surface CsA to bind the limited amount of biotin-labeled anti-CsA monoclonal antibodies. The immune complex formed on the inner surface of the microwell was traced by Eu3+ -labeled streptavidin based on a wellestablished fluorescence enhancement technique, of which, the Eu3+ on the immune complexes formed highly bright fluorescent complexes through Eu3+ dissociation and fluorescence enhancement. The standard curve of CsA concentration was established by plotting the fluorescence intensity against the CsA concentration, and the concentration of CSA in samples was determined via a standard curve. In this study, we validated the performance of the TRFIA for human whole blood CsA concentration analysis to evaluate whether its performance was following the requirements of the design of CsA-TRFIA kit [20–22]. We then performed a single-center clinical study to compare the agreement between TRFIA and CMIA for CsA analysis. Although LC–MS/MS is considered as the gold stan-

As shown in Table 2, the lowest detection limits of three batches of CsA-TRFIA kit on the DR6608 TRFIA instrument were 9.6, 5.6 and 7.8 ng/mL respectively, which met the design requirements of the LOD (≤20 ng/mL) of the CsA-TRFIA kit [20–22]. The lowest detection limits of the TRFIA can well meet the requirement for clinical CsA monitoring since the CsA blood level is generally more than 50 ng/mL in clinical samples. 3.2. Linear range As shown in Table 3, linearity and linear range performance were evaluated by using three batches of CsA kit on the DR6608 TRFIA instrument. Within the detection range of 30–1,000 ng/mL, the results of this reagent were linear (r > 0.990). The relative deviation was not more than ±15% within 100–1,000 ng/mL and less than ±30.0 ng/mL within 30–100 ng/mL, which met the design requirements of the linear and linear range of CsA-TRFIA kit [20–22]. Compared to the Abbott CsA-CMIA which offers a wider measuring range up to 1500 ng/L of CsA, the CsA-TRFIA can measure CsA only to 1000 ng/mL. Although the TRFIA is potent enough for the CsA trough level (C0 ) measurement, the relatively narrow measuring range of the TRFIA is inconvenient for the peak level (C2 ) monitoring since CsA over 1000 ng/mL in C2 samples is common. A

4

X. Wang, Y. gao, Y. Yang et al. / Journal of Pharmaceutical and Biomedical Analysis 177 (2020) 112890

Table 3 The linearity and range of CsA-TRFIA kits. Detection range of CsA concentrations (ng/ml)

Lot: 20131202 Regression equation r value Lot: 20131203 Regression equation r value Lot: 20140101 Regression equation r value

30–1500

30–1200

30–1000

Y = 1.2043X-80.206, R2 = 0.9734 0.9866

Y = 1.0775X-37.053, R2 = 0.9808 0.9904

Y = 0.9568X-6.4328, R2 = 0.9986 0.9993

Y = 1.1769X-79.849, R2 = 0.9726 0.9862

Y = 1.0518X-37.281, R2 = 0.9797 0.9898

Y = 0.9468X-10.625, R2 = 0.9886 0.9943

Y = 1.0685X-38.842, R2 = 0.9856 0.9856

Y = 0.969X-4.9896, R2 = 0.9984 0.9992

Y = 0.9397X + 2.4683, R2 = 0.9996 0.9998

Table 4 The Recovery of CsA-TRFIA kits. The test result of sample B after adding A (ng/ml)

Lot: 20131202 Lot: 20131203 Lot: 20140101

Rrecovery

First test

Second test

Third test

X

482.3 511.2 463.2

475.6 498.2 458.8

484.6 513.9 491.9

480.8 507.8 471.3



CV = s/ × 100% ; R recovery = 1 /2



96.2% 101.6% 94.3%

× 100%; CV, coefficient of variation; s, standard

deviation; R recovery , average recovery rate; , mean of the measured values.

Table 5 The relative deviation or absolute deviation of CsA-TRFIA kits. T (ng/ml)

Lot: 20131202 M (ng/ml) B% or M-T Lot: 20131203 M (ng/ml) B% or M-T Lot: 20140101 M (ng/ml) B% or M-T

600

300

150

90

596.1 −0.7%

275.3 −8.2%

143.9 −4.1%

85.6 −5.7

587 −2.2%

279.6 −6.8%

136.1 −9.3%

84.3 −5.7

648.8 8.1%

313.5 4.5%

144.5 −3.7%

88.2 −1.8

M, measured value; T, the target value; B%, relative deviation; M-T, absolute deviation.

further improvement of the TRFIA to widen its measurement range is necessary to obviate the need for sample dilution when applied for C2 monitoring. It is known that the measurement range and the lowest detection limits of a competitive immunoassay are two parameters somewhat conflicting from each other, the CsA-TRFIA has an advantage to extend its upper limit of CsA measurement to more than 1500 ng/mL because of the high sensitivity of the CsA-TRFIA. 3.3. Accuracy The Rrecovery was within the range of 90.0%–110.0% (Table 4). The relative deviation (B%) between the measured value(M) and the target value(T) was less than ±10.0% within 100–1,000 ng/mL, and the absolute deviation (M-T) was no more than ±20 ng/mL within 30–100 ng/mL (Table 5), which met the design requirements of the accuracy of CsA-TRFIA kit.

day, inter-day, and different operators were 5.3%, 6.0% and 0.26% for 600 ng/mL CsA samples. Additionally, the inter-batch relative ranges were 4.6% and 8.0% for 150 and 600 ng/mL CsA samples. Considering the CsA-TRFIA was manually performed with relatively short incubation time, the precision of the CsA-TRFIA was regarded as acceptable. Overall, the CV was ≤ 10.0% and the inter-batch range ≤ 15%, which met the design requirements of repeatability and inter-batch difference for CsA-TRFIA kit [20–22]. 3.5. Specificity The CsA-TRFIA was designed to have a mean recovery of 100 ± 15% in the presence of the potentially interfering pharmaceutical substances and endogenous substances. The above 18 substances (Table 1) did not affect the test results for CsA-TRFIA kit [20–22]. We did not evaluate the possible interference by CsA metabolites, including M17, M18, M21, M1, and M8, because these metabolites were not commercially available for the crossreactivity testing. 3.6. Methods comparison The measurements of both CMIA and TRFIA were abnormally distributed (P < 0.01 for the Kolmogorov-Smirnov test). The measurements were 107.60 (84.30–167.38) ng/mL for CMIA and 102.70 (82.23–168.43) ng/mL for TRFIA. When evaluated by paired-samples t-test, the values of the two measurements were not different (P > 0.05). The linear regression equation was YCMIA = 0.961XTRFIA + 3.573 (r = 0.98) and there was no significant deviation from linearity (P > 0.05) (Fig. 1). When evaluated by the Passing–Bablok regression and the Pearson test, the 95% CIs of the intercept and slope were −2.545 to 8.610 and 0.9153 to 1.0220, respectively (A). There was good agreement between the two methods evaluated by the Bland-Altman plot (B) [12,16,17]. Due to the LLOQ of the CMIA method was 30 ng/mL, the samples with CsA ≥ 30 ng/mL by CMIA were used for the method comparison. Between the two measurements, the Pearson correlation coefficient was high (r = 0.98), which indicated that data points scattered along a straight line [12]. However, the linear range of CsA-TRFIA kit was 30–1,000 ng/mL, which was different from that of CsA-CMIA kit (30–1,500 ng/mL). Therefore, when CsA concentration was within the range of 1000–1,500 ng/mL, manual dilution might be required by using the zero calibrator of CsA-TRFIA kit before detection, but not for the CsA-CMIA kit. 4. Conclusion

3.4. Precision The total % coefficient of variation (CV) for CsA measurement by different batches of CsA-TRFIA kits ranged between 5.0%–6.7%. The CVs of intra-day, inter-day, and different operators were 6.1%, 2.7% and 0.05% for 150 ng/mL CsA samples, while the CVs of intra-

The results obtained from this evaluation demonstrate that the CsA TRFIA is a precise, reproducible method for monitoring CsA concentration in whole blood specimens. Owning to good accuracy and low cost of this immunoassay, the CsA TRFIA may be a promising alternative for TDM of CsA where expensive tests, i.e., the

X. Wang, Y. gao, Y. Yang et al. / Journal of Pharmaceutical and Biomedical Analysis 177 (2020) 112890

5

Fig. 1. Passing-Bablok regression and Bland-Altman plot of CMIA and TRFIA for cyclosporine analysis in human whole blood. (A) Passing-Bablok regression; (B) Bland–Altman plot.

ABBOTT CMIA, Roche ECLIA, and Siemens EMIT, are not available, especially in small hospitals of developing countries. However, due to the limited sample numbers analyzed in this study, further investigations are needed to identify its applicability for routine clinical CsA monitoring. Funding This study was funded by the National Natural Science ¨ Foundation of China (81870520), the Shanghai Rising Stars of ¨ Development Program (Shanghai Municipal Medical TalentYouth Health Commission Planning Personnel Matters [2019]72), the Shanghai Municipal Commission of Health and Family PlanningConstruction of clinical pharmacy service system (2016ZB0303); and the Military construction of the national key clinical specialist—clinical pharmacy. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.jpba.2019. 112890. Declaration of Competing Interest Authors have no conflicts to declare. References [1] T. Coelho, M. Tredger, A. Dhawan, Current status of immunosuppressive agents for solid organ transplantation in children, Pediatr. Transplant. 16 (2012) 106–122, http://dx.doi.org/10.1111/j.1399-3046.2012.01644.x. [2] J.U. Lee, L.K. Kim, J.M. Choi, Revisiting the concept of targeting NFAT tocontrol T cell immunity and autoimmune diseases, Front. Immunol. 9 (2018) 2747, http://dx.doi.org/10.3389/fimmu.2018.02747. [3] A.L. Taylor, C.J. Watson, J.A. Bradley, Immunosuppressive agents in solid organtransplantation: mechanisms of action and therapeutic efficacy, Crit. Rev. Oncol. Hematol. 56 (2005) 23–46, http://dx.doi.org/10.1016/j.critrevonc. 2005.03.012. [4] X. Wang, Y. Yang, Z. Liu, C. Xiao, L. Gao, W. Zhang, W. Zhang, Z. Wang, Switching immunosuppression from cyclosporine to tacrolimus in kidney transplant recipients based on CYP3A5 genotyping, Ther. Drug Monit. 41 (2019) 97–101, http://dx.doi.org/10.1097/FTD.0000000000000579. [5] Y. Zhang, R. Zhang, Recent advances in analytical methods for the therapeutic drug monitoring of immunosuppressive drugs, Drug Test. Anal. 10 (2018) 81–94, http://dx.doi.org/10.1002/dta.2290. [6] A.J. McShane, D.R. Bunch, S. Wang, Therapeutic drug monitoring of immunosuppressants by liquid chromatography-mass spectrometry, Clin. Chim. Acta 454 (2016) 1–5, http://dx.doi.org/10.1016/j.cca.2015.12.027.

[7] E.M. Brate, D.M. Finley, J. Grote, S. Holets-McCormack, P.F. Ozaeta, D. Pacenti, J.E. Peart, R.E. Piktel, C.S. Ramsay, K.R. Rupprecht, S.C. Saldana, T.G. Spring, S.Y. Tetin, B.C. Trudeau, P. Wang, H. Xie, Development of an Abbott ARCHITECT cyclosporine immunoassay without metabolite cross-reactivity, Clin. Biochem. 43 (2010) 1152–1157, http://dx.doi.org/10.1016/j.clinbiochem. 2010.06.007. [8] X. Qin, J. Rui, Y. Xia, H. Mu, S.H. Song, R.E. Raja Aziddin, G. Miles, Y. Sun, S. Chun, Multi-center performance evaluations of tacrolimus and cyclosporine electrochemiluminescence immunoassays in the Asia-Pacific Region, Ann. Lab. Med. 38 (2018) 85–94, http://dx.doi.org/10.3343/alm.2018.38.2.85. [9] W. Li, R. Li, H. Liu, X. Guo, A.S. Shaikh, P. Li, B. Wang, R. Guo, R. Zhang, A comparison of liquid chromatography-tandem mass spectrometry (LC-MS/MS) and enzyme-multiplied immunoassay technique (EMIT) for the determination of the cyclosporin A concentration in whole blood from Chinese patients, Biosci. Trends 11 (2017) 475–482, http://dx.doi.org/10. 5582/bst.2017.01121. [10] Y. Kagawa, M. Yanagawa, Y. Muraki, T. Iwamoto, H. Mizutani, Y. Sugimura, M. Kojima, Comparison of cyclosporine concentrations in renal transplant recipients using ACMIA and mFPIA methods, Clin. Biochem. 37 (2004) 1016–1021, http://dx.doi.org/10.1016/j.clinbiochem.2004.07.004. [11] W.C. Korte, H. Engler, E. Aldebert, W.F. Riesen, Heparinized blood provides equivalent results to EDTA in the CEDIA and FPIA cyclosporine immunoassays, thus facilitating routine cyclosporine determination, Clin. Chim. Acta 390 (2008) 138–140, http://dx.doi.org/10.1016/j.cca.2007.12.021. [12] S. Mei, J. Wang, D. Chen, L. Zhu, M. Zhao, X. Hu, L. Yang, Z. Zhao, Ultra-high performance liquid chromatography tandem mass spectrometry for cyclosporine analysis in human whole blood and comparison with an antibody-conjugated magnetic immunoassay, Ther. Drug Monit. 40 (2018) 69–75, http://dx.doi.org/10.1097/FTD.0000000000000477. [13] G. Deslandes, M. Grégoire, C. Renaud, C. Monteil-Ganière, C. Azoulay, A. Pineau, P. Jolliet, E. Dailly, Comparison between an automated and manual extraction for the determination of immunosuppressive drugs whole blood concentrations by liquid chromatography tandem mass spectrometry, J. Clin. Lab. Anal. 30 (2016) 924–929, http://dx.doi.org/10.1002/jcla.21958. [14] J.W. Zhou, L.M. Lei, Q.N. Liang, T.C. Liu, G.F. Lin, Z.N. Dong, R.L. Liang, Z.H. Chen, Y.S. Wu, Dual-labeled time-resolved immunofluorometric assay for the determination of IgM antibodies to rubella virus and cytomegalovirus in human serum, Clin. Biochem. 48 (2015) 603–608, http://dx.doi.org/10.1016/j. clinbiochem.2015.01.009. [15] G. Lin, S. Chen, H. Zhao, J. Liang, Q. Deng, R. Liang, X. Guo, Z. Chen, B. Chen, T. Liu, Y. Wu, A time-resolved fluoroimmunoassay to assay the rabies virus glycoprotein: application for estimation of human rabies vaccine potency, Sci. Rep. 7 (2017) 7288, http://dx.doi.org/10.1038/s41598-017-07687-7. [16] V. Chhapola, S.K. Kanwal, R. Brar, Reporting standards for Bland-Altman agreement analysis in laboratory research: a cross-sectional survey of current practice, Ann. Clin. Biochem. 52 (2015) 382–386, http://dx.doi.org/10.1177/ 0004563214553438. [17] M.J. Lu, W.H. Zhong, Y.X. Liu, H.Z. Miao, Y.C. Li, M.H. Ji, Sample size for assessing agreement between two methods of measurement by Bland-Altman method, Int. J. Biostat. 12 (2016), http://dx.doi.org/10.1515/ijb2015-0039, pii: /j/ijb.2016.1. [18] E. Olofsen, A. Dahan, G. Borsboom, G. Drummond, Improvements in the application and reporting of advanced Bland-Altman methods of comparison, J. Clin. Monit. Comput. 29 (2015) 127–139, http://dx.doi.org/10.1007/s10877014-9577-3. [19] Clinical and Laboratory Standards Institute, in: Protocols for Determination of Limits of Detection and Limits of Quantitation, Approved Guideline. CLSI Document EP17, CLSI, Wayne, PA USA, 2004.

6

X. Wang, Y. gao, Y. Yang et al. / Journal of Pharmaceutical and Biomedical Analysis 177 (2020) 112890

[20] FDA, Guidance for Industry: Bioanalytical Method Validation [web site], 2013, Available at: http://www.fda.gov/ucm/groups/fdagov-public/@fdagovdrugsgen/documents/document/ucm368107.pdf. (Accessed 21 July 21 2014). [21] B. Booth, M.E. Arnold, B. DeSilva, L. Amaravadi, S. Dudal, E. Fluhler, B. Gorovits, S.H. Haidar, J. Kadavil, S. Lowes, R. Nicholson, M. Rock, M. Skelly, L. Stevenson, S. Subramaniam, R. Weiner, E. Woolf, Workshop report: Crystal City V – quantitative bioanalytical method validation and implementation: the 2013 revised FDA guidance, AAPS J. 17 (2015) 277–288, http://dx.doi.org/10.1208/ s12248-014-9696-2. [22] J.F. Bower, J.B. McClung, C. Watson, T. Osumi, K. Pastre, Recommendations and best practices for reference standards and reagents used in bioanalytical method validation, AAPS J. 16 (2014) 352–356, http://dx.doi.org/10.1208/ s12248-014-9566-y.

[23] A. Papanastasiou-Diamandis, V. Bhayana, E.P, A simple time-resolved fluoroimmunoassay of total thyroxine in serum, Ann. Clin. Biochem. 26 (1989) 238–243, http://dx.doi.org/10.1177/000456328902600305. [24] Q. Mao, X. Liu, C. Chen, J. Ye, H. Liang, B. Li, X. Sun, L. Sun, Y. Sun, X. Mu, L. Li, A dual-label time-resolved fluorescence immunoassay for the simultaneous determination of carcinoembryonic antigen and squamous cell carcinoma antigen, Biotechnol. Appl. Biochem. 65 (2018) 816–821, http://dx.doi.org/10. 1002/bab.1663. [25] P. Wallemacq, G.T. Maine, K. Berg, T. Rosiere, P. Marquet, G. Aimo, G. Mengozzi, J. Young, K. Wonigeit, R. Kretschmer, B. Wermuth, R.W. Schmid, Multisite analytical evaluation of the Abbott ARCHITECT cyclosporine assay, Ther. Drug Monit. 32 (2010) 145–151, http://dx.doi.org/10.1097/FTD. 0b013e3181d46386.