The guideline C24 (now in its 4th edition) issued by the Clinical and Laboratory Standards Institute (CLSI) recommends that for purchased quality control (QC), the laboratory should never use the manufacturer's declared target (T) value but that identified by testing the product at least ten times with each new batch (1). Of course, verifying T is quite different from estimating the variability of the analytical process, albeit both are used for the statistical process control (SPC) since T should be the production target of the QC material on which the manufacturer has complete control.

Background: A new homogeneous affinity chrome-mediated immunoassay (ACMIA) "EVRO" from Siemens Healthcare was evaluated for therapeutic drug monitoring of everolimus (EVL) with automated sample pretreatment and compared with quantitative microsphere system (QMS) "EVER" from Thermo Fisher Scientific.Methods: Imprecision, inaccuracy, and limit of quantitation (LoQ) of ACMIA/EVRO were verified using both hemolysate quality control (QC) samples and pooled whole blood specimens. The interchangeability of methods and the agreement of results were analyzed using 72 specimens (from 38, 30, and 4 kidney, liver, and lung transplant recipients, respectively).Results: Within-run imprecision ranged within %CV = 2.81-2.53 with pooled whole blood specimens and within %CV = 2.88-2.53 with QCs; total imprecision with QCs was within %CV = 2.14-1.51. Inaccuracy with value assigned QC was %O = 5.36 at the 5.6 ng/ mL level and %O = 5.56 at the 11.7 ng/mL level. LoQ was 0.93 ng/mL (%CV = 10). Passing-Bablok regression showed a constant bias of 0.679 ng/mL (95% CI: 0.216-1.026) and a proportional bias of 1.326 (95% CI: 1.240-1.425). Bland-Altman analysis showed 5/ 72 (6.9%) paired differences exceeding the limits of agreement and 1/72 (1.4%) paired differences exceeding 1.96 SD to a combined bias of 39.9% after detrending.Conclusions: ACMIA/EVRO shows satisfactory analytical performances that comply with recommendations, but it does not fulfill requirements for interchangeability with QMS/EVER. Particularly, this new assay using sirolimus-specific antibody shows a sizable proportional bias versus the more specific comparator, which may be because of EVL metabolites. This is supported by the lack of agreement for individual differences in most samples collected at the peak concentration (C2). Therefore, further evidence is needed to support the transition of EVL level monitoring from QMS/EVER to ACMIA/EVRO without making extensive changes to both reference interval and patient's baseline.

Objectives This study aims to evaluate the interchangeability between the Siemens Healthineers’ “EVRO” new affinity chrome-mediated immunoassay (ACMIA/EVRO) and Thermo Fisher Scientific’s “EVER” Quantitative Microsphere System (QMS/EVER) with Chromsystems’ CE-IVD-certified “MassTox” liquid-chromatography/tandem-mass spectrometry (LC-MS/MS) assay for the therapeutic drug monitoring of everolimus. Methods A single lot of reagent, calibrators and controls were used for each assay. A total of 67 whole blood samples (n=67) from patients receiving solid organ transplant were analyzed (n=31 with kidney transplant and n=36 with liver transplant); Passing-Bablok regression and Bland-Altman difference plot were used to evaluate bias and individual agreement; LC-MS/MS analysis was used to measure the actual concentrations of calibrators and controls compared to the assigned value. Results ACMIA/EVRO did not show any systematic bias compared to LC-MS/MS (intercept=0.244 ng/mL, 95% CI: −0.254 to 0.651 ng/mL). Nevertheless, significant proportional bias (slope=1.511, 95% CI: 1.420 to 1.619) associated to a combined bias of 44.8% (95% CI: 41.2–48.3%) was observed. Conversely, QMS/EVER did not show any bias at both systematic (intercept=−0.151 ng/mL, 95% CI: −0.671 to 0.256 ng/mL) and proportional level (slope=0.971, 95% CI: 0.895 to 1.074) with a non-statistically significant combined bias of −3.6% (95% CI: −8.4–1.1%). Based on a concentration of calibrators and controls above the assigned value for both the analytical methods, in the ACMIA/EVRO a correction which was approximately one-third of the correction for the QMS/EVER was observed. Conclusions ACMIA/EVRO but not QMS/EVER shows a lack of interchangeability with the CE-IVD-certified LC-MS/MS assay. We hypothesize that, as the ACMIA/EVRO uses an anti-sirolimus antibody, the under-corrected assigned value in the assay calibrators was not sufficient to reproduce the everolimus metabolites cross-reactivity occurring in real samples.