Anouk Donners

55 LC-MS/MS-method development quantifying FVIII Streptavidin 96 well plate brand and capacity test Two high capacity streptavidin 96 well plates were compared. One plate was obtained from Sigma-Aldrich (SigmaScreen) and the other from Thermo Scientific (Streptavidin Coated High Capacity Plates). The test was performed according to the procedure described above using 25, 50 and 75 µL standards (500 ng/mL) in duplicate on both plates. Validation of FVIII LC-MS/MS method The validation was performed according to EMA guidelines which requires the evaluation of LLOQ, linearity, accuracy and precision, carry-over, auto sampler stability, freeze and thaw stability and matrix effect [28]. The acceptance criterion for LLOQ was that the signal of the QC LLOQ level (1 ng/mL) should be at least 5× that of the blank sample which consisted of FVIII deficient human citrated plasma. The calibration curve used to establish linearity consisted of 7 standards ranging from 1 to 500 ng/mL and was analyzed on 3 separate days. Within run and between run accuracy expressed as %bias and precision expressed as %CV were validated by measuring four QC levels (LLOQ (1 ng/mL), QC low (5 ng/mL), QC med (150 ng/mL) and QC high (300 ng/mL)) in five-fold during three days. The overall bias was calculated from the mean concentration and the within-run and between-run %CV was calculated from one-way ANOVA derived mean squares. Carry-over effect was tested by injecting a blank sample after the highest standard and comparing the signal intensity at the analyte retention time to the signal intensity of the LLOQ. Auto-sampler stability was evaluated by re-injecting the samples on the next day. Overnight stability was evaluated on QC low and QC high sample. Matrix effect was evaluated by spiking random samples with FVIII and calculating the spike recovery. RESULTS AND DISCUSSION Method Development Human FVIII consists of a light chain and a heavy chain held together by calcium ion and by the stabilizing protein von Willebrand factor that is bound to the light chain of FVIII. The antibody used for sample purification targets the light chain of FVIII and therefore dissociation of von Willebrand factor was necessary to obtain a high recovery. This was achieved by triggering the coagulation cascade to occur through the addition of calcium chloride and thrombin to the citrated plasma sample. The amount of antibody used per well to capture FVIII was 200 ng providing around 100× molar excess in relation to the highest standard used. This was based on previous work with similar interaction where a therapeutic antibody was purified by means of a ligand [29]. However, FVIII concentration range in plasma is around 100× lower compared to therapeutic monoclonal antibodies, and therefore, the selection of the signature peptide was primarily based on the peptide that had low background interference and delivered the highest signal to noise ratio to meet assay requirements. The peptide ‘GELNEHLGLLGPYIR’ was found to meet this criterion and was found to be unique to 3