A field study evaluating the activity of N8‐GP in spiked plasma samples at clinical haemostasis laboratories

Abstract Aim N8‐GP (turoctocog alfa pegol) is a glycoPEGylated, extended half‐life human recombinant factor VIII (FVIII) shown to be an efficacious treatment for patients with haemophilia A. Accurate monitoring of replacement therapy helps ensure proper dosing, leading to better patient care. The objective of this field study was to evaluate the accuracy and intra‐ and inter‐laboratory variabilities of N8‐GP and rAHF (Advate®) FVIII activity (FVIII:C) measurements in clinical laboratories using their routine methods and reagents. Methods Laboratories measured plasma samples spiked with 0.03, 0.2, 0.6 and 0.9 IU/mL N8‐GP or rAHF. Samples were blinded, and laboratories were instructed to perform evaluations using their routine FVIII activity assays and calibrators. Results Of the 67 participating laboratories from 25 countries, 60 used a one‐stage assay, 36 used a chromogenic assay, and 29 used both one‐stage and chromogenic assays. Participating laboratories used nine different activated partial thromboplastin time (aPTT) reagents, the most common being SynthASil® and Actin® FS. Most aPTT reagents recovered N8‐GP close to target. Three silica‐based aPTT reagents (APTT‐SP, TriniCLOT™ and STA® PTT‐Automate) underestimated N8‐GP, recovering 40%‐83% of target concentration. For chromogenic assays, N8‐GP and rAHF recoveries were comparable at all concentrations, with overall mean recoveries for both products close to 130%. Assay variability was similar for both assay types and both products; inter‐laboratory variability was greater than intra‐laboratory variability and highest at 0.03 IU/mL. Conclusions Most clinical laboratories accurately measured N8‐GP and rAHF when using their in‐house one‐stage or chromogenic FVIII:C assays. However, three silica‐based aPTT reagents underestimated N8‐GP recovery.


| INTRODUC TI ON
cule by 1.6-fold in adults 1 and 1.9-fold in children. 2 Upon activation of N8-GP, thrombin cleaves the FVIII B-domain, with the attached PEG moiety, which is released from the remaining molecule, leaving the primary native structure of activated FVIII intact. 3 Results from clinical studies show that N8-GP is efficacious in the treatment of patients with HA and shows a favourable safety profile in children, 2 adolescents and adults. 1,4 Accurate monitoring of FVIII activity (FVIII:C) during replacement therapy helps determine the dosing regimen and maintain trough levels in the target range. To monitor FVIII:C levels in patients with HA, clinical laboratories currently use activated partial thromboplastin time (aPTT)-based one-stage clotting assays, chromogenic activity assays or both. The most common method used to monitor patients treated with replacement therapy in clinical laboratories is the one-stage clotting assay. 5 One-stage assays estimate FVIII:C during the clotting phase of the reaction using aPTT reagents that vary in the contact activators used to initiate clot formation. 6 In vitro evidence suggests that some aPTT reagents can influence FVIII:C measurement of EHL-FVIII products. 7 A two-centre study found that most aPTT reagents reliably recover N8-GP in spiked plasma samples. However, of the eight reagents evaluated in the study, one aPTT reagent (APTT-SP [Instrumentation Laboratory]) underestimated FVIII:C and was judged unsuitable for patient monitoring. 8 Another method used to measure FVIII:C in the clinical laboratory is the chromogenic assay. Although the European Pharmacopoeia (Ph. Eur.) recommends that manufacturers of FVIII products assign potency using a chromogenic assay, this assay is less commonly used in clinical laboratories. 9-11 N8-GP potency was assigned using the Coamatic ® (Chromogenix; Instrumentation Laboratory) FVIII chromogenic assay and an in-house reference material traceable to the World Health Organization (WHO) 8th international FVIII concentrate standard (IS) (National Institute of Biological Standards and Control [NIBSC]). 1,12 The potency assignment for N8-GP has since been verified using six different chromogenic kits from alternative vendors with no significant difference in recovery of N8-GP. 12 Accurate FVIII:C monitoring is necessary to ensure optimal patient care. Thus, it has been recommended that assays used to measure FVIII:C in patients treated with EHL-FVIII products should be validated at the individual laboratory prior to use. 13,14 In general, the objective of a haemophilia replacement product field study is to assess the methods and suitability of reagents currently in use for measuring FVIII:C in new products in order to give guidance to the clinical laboratories on the assay performance of specific reagents. 9,11,[15][16][17] This global comparative field study evaluates the accuracy and intra-and inter-laboratory variabilities of FVIII:C measurements in clinical laboratories when using their routine FVIII:C procedures for measurement of N8-GP and the unmodified rFVIII molecule, rAHF (Advate ® , Shire Plc).
Field study kits of congenital haemophilia A plasma samples spiked with a range of concentrations of N8-GP and rAHF were distributed to participating clinical laboratories. Participating laboratories responded to a survey about reagents and methods they routinely use to monitor FVIII:C and measured the field study kits using their routine one-stage clotting assay, chromogenic assays or both.

| MATERIAL S AND ME THODS
Invitation letters for this study were sent to laboratories that participated in a previous field study 9 and to laboratories affiliated with the External quality Control for diagnostic Assays and Tests (ECAT) Foundation. Participating laboratories completed an online questionnaire, indicating their routine methods, kits and reagents used to measure FVIII:C, and were sent a field study kit. Laboratories were blinded to the product and exact concentration.

| Field study samples
However, samples were marked to indicate the expected FVIII:C as 'very low', 'low', 'medium' or 'high'. Three colour-coded replicates of each vial were provided, and laboratories were instructed to measure each sample on a separate day according to colour code. Esoterix Inc and the Laboratorium für Klinische Forschung GmbH verified factor activity in spiked sample sets prior to initiation of the study.
Laboratories were instructed to perform FVIII:C analyses using their routine FVIII:C procedures, reagents, calibrator and instruments.

| Statistical analysis
Each laboratory analysed the samples based on local practice.
When more than one analysis using the same assay and methodology was performed by a clinical laboratory, the average was calculated. FVIII:C measurements are reported as IU/mL or per cent of target concentration based on actual potency. FVIII:C levels were log-transformed and analysed separately by assay using a mixed effect model with the combination of trial drug and concentration as fixed effect and laboratory/assay as a random effect. The mean estimates of each concentration level together with the 95% confidence intervals (95% CI) were back-transformed and presented alongside the inter-and intra-laboratory variation. The same model was also used to analyse the concentration as percentage of target, based on the calculated percentages without any transformation. All data presented in histogram and scatter plots were prepared by per cent target concentration. The acceptable range of recovery was considered ±30% of the expected target concentration. All statistical analyses were performed using sas ® 9.4, with sas/stat ® 13.2 software.
Full details of the geographic distribution of laboratories by assay type are summarized in Table S1. Of the laboratories that used both one-stage and chromogenic assays, 17 (58.6%) used the same calibrator for both assays, while 11 (37.9%) used different calibrators.
Survey results for calibration method and sample dilution are summarized in Table 1.

| One-stage assay activity measurements
Three of the nine one-stage assay reagents (ie, APTT-SP, TriniCLOT™ and STA ® PTT-Automate, all containing silica-based contact activators) measured N8-GP activity at approximately 40%-83% of target concentration (full results are presented in Table 2) and were thus omitted from subsequent statistical analyses for N8-GP.
Both intra-and inter-laboratory variability was similar for N8-GP and rAHF. The highest inter-laboratory variability was observed in the 'very low' samples (0.03 IU/mL) for both products.

| Chromogenic kit activity measurements
N8-GP recovery in the chromogenic assays was similar to rAHF at all concentrations, with both products at concentrations of 0.2 IU/mL and above recovering around 130% of target concentration ( Figure 4; Table 4). N8-GP recovered consistently across all concentrations, and overall, mean recovery was within the upper bound of the acceptable range (129%, 95% CI: 123%; 136%). rAHF mean recovery was similar to that of N8-GP (127%, 95% CI: 121%; 134%).
Mean recoveries at 0.03 IU/mL were the closest to target concentration for both products, but, as seen with one-stage assays, the highest intra-and inter-laboratory variabilities in chromogenic kits were also observed in 0.03 IU/mL samples. Variability decreased with increasing concentration (Figure 4; Table 4). Inter-laboratory variability was similar between N8-GP and rAHF, ranging from 11.5% to 27.0% for N8-GP, and 11.5% to 29.1% for rAHF. Inter-laboratory variability for SSC lot #4 plasma was similar to high concentration samples of N8-GP and rAHF (Table 4).

| Laboratories that could accurately measure N8-GP
The three aPTT reagents that underestimated N8-GP were used in 16 different laboratories, nine of which used both one-stage and chromogenic assays, of which one of these also used multiple aPTT reagents. Thus, 60 of 67 laboratories (89.6%) routinely used at least one FVIII:C assay that could accurately measure N8-GP.  Note: Results from one-stage clotting assays that used one of the three aPTT reagents (APTT-SP, TriniCLOT™ and STA ® PTT-Automate) that substantially underestimated N8-GP recovery were omitted from N8-GP statistical analysis. Overall, most clinical laboratories participating in this study could accurately measure N8-GP with methods already available in their laboratory, and the recovery for SSC lot #4 obtained using both the chromogenic and one-stage clotting assays was similar (Tables   3 and 4). Laboratories that deviated most from the assigned value of the SSC lot #4 control sample also provided results that deviated most from the target values for the remaining field study samples (data not shown). Due to regional variability in the availability of many aPTT reagents on the market, local verification of reagents not covered by this survey may be prudent. Although there is currently no consensus about the magnitude of difference from target concentration that is clinically relevant in postinfusion monitoring, many studies define a ±30% difference as acceptable. 10,20 F I G U R E 2 One-stage aPTT-based clotting assay mean FVIII:C in plasma samples spiked with 0.03, 0.2, 0.6 and 0.9 IU/mL N8-GP or rAHF. A, Data points represent mean per cent of target concentration from individual laboratories. Each column of data points represents a different aPTT reagent. Dashed lines represent interquartile range. B, Bars represent the number of laboratories at the mean per cent of target concentration. Results from one-stage clotting assays that used one of the three aPTT reagents (APTT-SP, TriniCLOT™ and STA ® PTT-Automate) that underestimated N8-GP recovery were omitted from N8-GP statistical analysis, as were values of zero At 92.5% of target concentration, mean N8-GP recovery for six aPTT reagents in one-stage assays was well within the 30% limit. used for N8-GP potency assignment, and this potency assignment was confirmed using various chromogenic kits. 12 In the clinical laboratories, measurement of FVIII:C is often performed using a chromogenic assay that is calibrated to an in-house normal, pooled plasma calibrator that is traceable to the WHO 6th IS. In a previous study, we showed that the use of a reference standard traceable to WHO 8th IS FVIII could increase measurement accuracy of chromogenic kits for both N8-GP and unPEGylated turoctocog alfa (NovoEight ® ).

| D ISCUSS I ON AND CON CLUS I ON
A recent study has shown that chromogenic assays can be validated for use with N8-GP to give recoveries closer to the expected range when using a normal, pooled plasma calibrator. 24 Thus, the observed over-recovery of FVIII:C with chromogenic assays in both EHL and standard rFVIII products deems further and consideration from both clinical laboratory scientists and assay manufacturers.
Overall, most participating clinical laboratories could accurately measure N8-GP and rAHF using their in-house available one-stage clotting or chromogenic FVIII:C assays, without the need of a product-specific standard. Three silica-based aPTT reagents substantially underestimated N8-GP recovery and should not be used to monitor N8-GP activity.

ACK N OWLED G EM ENTS
We thank Esoterix Inc (Englewood, CO, USA) for preparing the sam-