PB0229 - Emicizumab: In-house method validation

Background: Emicizumab (Hemlibra®, ROCHE), is effective in the prevention of bleeding in severe Hemophilia A patients with (sHAinh) or without (sHA) inhibitors. To date, emicizumab plasma quantification (Epq) is not routinely recommended but, could be required to verify treatment adherence, suspected anti-Emicizumab antibodies. Epq method, calibrators and quality controls are not yet available.

Aims: To validate in-house Epq method using a modified one-stage assay (modOSA).

Methods: modOSA: 1:80 pre-diluted sample with Owren Koller Buffer(OKB). Calibration Curve: 1:1(1:80 pre-diluted home-made calibrator (hmCal) with OKB), 1:2,1:4,1:8,1:16,1:32. hmCal (150µg/mL of Emicizumab): FVIII deficient plasma (FVIIIdp) was spiked with Emicizumab. Home-made Controls: level 1=25µg/mL (hmL1), level 2=75µg/mL (hmL2). Reagents: Immunodef VIII, C.K. Prest, CaCl2 0.025M. STA Compact Max2, Diagnostica Stago.
Repeatability(CVr)= hmL1 and hmL2: 20 times in the same run. Within-laboratory Imprecision(CVwl)= hmL1 and hmL2 in triplicate for 5 days. Calibration curve: 6 levels in duplicate (range: 5–150µg/mL). Linearity: 11 levels in triplicate, using hmCal and FVIIIdp (0µg/mL). In addition, aPTT, One Stage FVIII Assay (OSA), FVIII Chromogenic Assay (CSA, Bovine reagent, Chromogenix) and OSA/Epc were measured in the 9 levels. Limit of quantification (LoQ): FVIIIdp, sHAinh and sHA samples were evaluated. Bias estimation: from an external quality assessment program (UK NEQAS). Interference: mixtures of Emicizumab with FVIII were evaluated (Unicalibrator, Diagnostica Stago as a source of FVIII). EP evaluator was used for statistical analysis.

Results: hmL1: CVr 4.8%(27µg/mL); CVwl 5.7%(26µg/mL), hmL2: CVr 3.1%(73µg/mL); CVwl 4.1%(74µg/mL). Calibration curve: r:-0.999, r2:0.998. Linearity: Table1, r2:0.997 (linear range 5-150µg/mL), Slope:1.011, Intercept:0.52. LoQ:5µg/mL. Bias:-4.0%. Interference: Table2.

Conclusion(s): modOSA showed a good performance in the Epq. aPTT was below the reference range and CSA did not interfere in the evaluated range of 5-100 ug/ml of Epq. FVIII surrogate activity is appoximately 7 times the level of Epq. In vitro FVIII acts as an interference in the method.