Factor VIII chromogenic assays can be used for potency labeling and postadministration monitoring of N8-GP

Transcription

1 Journal of Thrombosis and Haemostasis, 14: DOI: /jth ORIGINAL ARTICLE Factor VIII chromogenic assays can be used for potency labeling and postadministration monitoring of N8-GP W. PICKERING, M. HANSEN, M. KJALKE and M. EZBAN Novo Nordisk A/S, Novo Nordisk Park, 276 Maløv, Denmark To cite this article: Pickering W, Hansen M, Kjalke M, Ezban M. Factor VIII chromogenic assays can be used for potency labeling and postadministration monitoring of N8-GP. J Thromb Haemost 216; 14: Essentials Chromogenic assays may be less variable than one-stage clot assays for measuring modified factor VIII. Chromogenic assays were evaluated for N8-GP potency labeling and postadministration monitoring. There was no significant difference between chromogenic assay kits for measuring N8-GP potency. Postadministration monitoring of N8-GP was comparable to turoctocog alfa for all kits tested. Summary. Background: Factor VIII activity (FVIII:C) is commonly measured using one-stage activated partial thromboplastin time (aptt) based clot assays. Chromogenic assays are, however, an alternative, and potency assessment in Europe is performed using chromogenic assays. One-stage clot assays are in general associated with high variability, and modified FVIII products may add to this variability. FVIII chromogenic assays may be less affected. Objectives: To evaluate available chromogenic assay kits for potency labeling of polyethylene glycol glycoconjugated turoctocog alfa (turoctocog alfa pegol [N8-GP]) and to evaluate selected chromogenic kits for postadministration monitoring of N8-GP using turoctocog alfa (Novoeight Ò ) as comparator. Methods: Six FVIII chromogenic assay kits were adapted to the European Pharmacopeia guidelines for potency labeling, including assessment of time to 5% FX activation. Four kits were adapted for postadministration monitoring using an ACL Ò TOP 5 analyzer. Severe hemophilia A plasma was spiked with N8-GP or turoctocog alfa to simulate postadministration samples. The World Health Organization (WHO) 8th International Standard (IS) Correspondence: Mirella Ezban, Novo Nordisk Park, Novo Nordisk A/S, 276 Maløv, Denmark. Tel.: mie@novonordisk.com Received 18 November 215 Manuscript handled by: S. Kitchen Final decision: P. H. Reitsma, 21 April 216 FVIII concentrate was used as calibrator throughout. In addition, a plasma calibrator was used for postadministration samples. Results: When measuring N8-GP potency, no significant difference using a 1% significance level was observed between kits. In simulated postadministration samples, all test kits were highly accurate and precise, except at low concentrations, with no significant difference between FVIII:C (P >.5) measured using the different calibrators. However, values obtained using the WHO 8th IS were closer to labeled values. Conclusions: Chromogenic assay kits tested measured consistent FVIII: C for N8-GP potency and showed comparable results for N8-GP and turoctocog alfa in simulated postadministration samples. Keywords: blood coagulation tests; factor VIII; hemophilia A; polyethylene glycols; turoctocog alfa. Introduction Assays commonly used to measure factor VIII (FVIII) activity (FVIII:C) include one-stage activated partial thromboplastin time (aptt)-based clot assays and FVIII chromogenic assays that measure the amount of activated Factor X (FXa) in a sample proportional to FVIII: C in the reference material [1]. FVIII:C assays are used to determine FVIII concentrate potency and for the routine monitoring of FVIII levels in hemophilia A (HA) patients treated with FVIII products. Potency testing is performed to quantify FVIII:C for each FVIII concentrate product label. For products approved in Europe, the European Pharmacopoeia (Ph. Eur.) requires that potency for FVIII products is assigned using a chromogenic assay. Clinical monitoring of FVIII:C levels in hemophilia patients is important to ensure accurate diagnosis and assessment of clinical severity, appropriate detection of the presence of inhibitory alloantibodies, and efficient hemostatic monitoring of on-demand and routine prophylactic therapy. Recommendations and guidelines, including information on the optimal use of assays for measuring FVIII:C, have been issued by This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

2 158 W. Pickering et al various advisory committees, including the Ph. Eur., the Scientific and Standardization Committee (SSC) of the International Society on Thrombosis and Haemostasis, and regulatory authorities [2 4]. Such recommendations have helped to reduce the interlaboratory variation that is often associated with FVIII:C assays. To ensure consistent labeling across products, it is important that methods to measure FVIII:C activity for potency determination follow the relevant guidelines and recommendations, which are linked to current World Health Organization (WHO) international standards (IS) for factor concentrates [5]. Furthermore, to ensure accurate potency labeling, assay conditions must be identified that correlate with biological activity. The recent SSC recommendations take into account that potency measurement for the new generation of modified recombinant FVIII products may be affected by the choice of assay or assay reagents and indicate that one-stage clot assays or chromogenic assays may be used but that the appropriate assay should be chosen individually for specific products and performed using the current WHO IS for FVIII concentrates [4]. Potency labeling for efmoroctocog alfa (Eloctate Ò ; Biogen Idec Inc., Cambridge, MA, USA), a rfviii-fc fusion protein that was approved by the U.S. Food and Drug Administration in mid-214, is performed using chromogenic assays [6]. The Ph. Eur. guidelines (version 8.2) for potency labeling of FVIII products to be released in Europe indicate that a chromogenic assay should be used and stipulate the recommended conditions for such assays [2]. These include predilution of samples/calibrators in von Willebrand factor (VWF)-containing FVIII-deficient plasma, using dilution buffer containing 1% albumin, diluting FVIII to preferably <.1 IU ml 1, and adjusting FX activation time to obtain 5% FX activation relative to maximal level. While the conditions for dilution can easily be adapted, the adjustment of activation time requires experimental evaluation. For clinical monitoring of FVIII levels in patient plasma, FVIII:C is most commonly monitored using onestage aptt-based clot assays. However, this method is associated with a high level of variability due to the large number of aptt reagents, instruments, calibrators, and factor-deficient plasma used [7 9]. In contrast, chromogenic assays show less variability in the measurement of FVIII activity levels, possibly due to a smaller number of available assay kits and reagents, and more accurately reflect the bleeding phenotype in patients with some genetic mutations associated with mild HA [1 13]. The Medical and Scientific Advisory Council within the National Hemophilia Foundation recently recommended the addition of chromogenic assays in diagnostic laboratories for clinical monitoring of FVIII:C to the use of one-stage aptt-based clot assays [14]. Given that variability in FVIII:C assay performance has been observed for the new generation of modified clotting factor molecules [15 17], many current recommendations may not necessarily be appropriate for these new products. Indeed, the recent SSC guideline acknowledges this point and specifically indicates that measurement of FVIII:C using one-stage clot assays for modified FVIII molecules may require testing with different aptt reagents [4]. Characterization of potency assays for various modified recombinant FVIII (rfviii) products has revealed that chromogenic assays are robust across different kits [18]. A polyethylene glycol (PEG)-conjugated version of rfviii, that is, turoctocog alfa pegol (N8-GP; Novo Nordisk, Bagsværd, Denmark), is currently in late-stage clinical development for the treatment of patients with HA [16,19]. The aims of the current study were to evaluate different FVIII chromogenic assay kits that are commercially available in Europe and in some cases also in the US for their accuracy in determining the potency of N8- GP and to investigate the potential of these kits for clinical monitoring using N8-GP containing plasma samples. Turoctocog alfa (rfviii; Novoeight Ò, Novo Nordisk) was chosen as the comparator for both potency assignment and simulated postadministration monitoring, as results of a field study have shown that it can be reliably measured in plasma [8]. The use of turoctocog alfa facilitated an assessment of the effect of the PEG moiety on FVIII chromogenic assays. Both a plasma calibrator and the concentrate WHO 8th IS were used for the analysis of simulated postadministration samples in order to link the potency assignment (using the WHO 8th IS) to clinical monitoring and to evaluate the effect of using different calibrators on FVIII:C measurement. Materials and methods Reagents N8-GP and the comparator (turoctocog alfa, Novoeight Ò ) were manufactured by Novo Nordisk A/S. The potency of N8-GP and turoctocog alfa was assigned using the Coamatic Ò (Chromogenix; Instrumentation Laboratory, Bedford, MA, USA) chromogenic assay with WHO 8th IS for FVIII concentrate as calibrator. The WHO 8th IS Factor VIII Concentrate (NIBSC7/35) was used as a calibrator to investigate both FVIII chromogenic assays and simulated postadministration samples. Plasma samples were also investigated using standard human plasma (SHP; Siemens, Marburg, Germany) as calibrator. The same calibrator was used for all kits in order to standardize the assays. Severe HA plasma (George King Bio-Medical, Inc., Overland Park, KS, USA) was used for preparation of simulated postadministration samples, and a 1-fold dilution in FVIII-deficient plasma (Siemens) was applied for N8-GP analyzed in the chromogenic assays.

3 Measuring FVIII:C of N8-GP in chromogenic assays 1581 Chromogenic assays for potency assignment according to Ph. Eur. guidelines Potency determination was investigated using the following chromogenic assay kits: Coamatic Ò (Chromogenix; Instrumentation Laboratory), Coatest Ò SP (Chromogenix; Instrumentation Laboratory), Biophen FVIII:C (Hyphen BioMed, Neuville sur Oise, France), Technochrom Ò FVIII:C (Technoclone, Vienna, Austria), DG-Chrom FVIII (Grifols, Barcelona, Spain), and FVIII Chromogenic (Siemens). The key characteristics of the different chromogenic assay kits are compared in Table S1. The general procedure outlined below was used to perform FVIII chromogenic assays for potency determination. Briefly, reagents and substrates were reconstituted according to the manufacturer s instructions unless otherwise specified. The samples containing N8-GP or turoctocog alfa were diluted in HEPES-buffered saline (HBS, 2 mmol L 1 HEPES, 145 mmol L 1 NaCl, ph 7.4) containing 1% bovine serum albumin (BSA [HBS/BSA]) to 1 IU ml 1 followed by a 1-fold dilution in VWF-containing FVIII-deficient plasma (Siemens) to 1 IU ml 1. In cases where the diluent in the kit did not include 1% albumin (kits from Technoclone, Siemens, and Grifols), the diluents for further dilutions were exchanged with HBS/BSA. For each analysis set, a vial of the WHO 8th IS calibrator was equilibrated to room temperature (RT) for 15 min before reconstitution in 1 ml Milli-Q H 2 O. After reconstitution, the calibrator was left for an additional 15 min at RT and then transferred to low protein-absorbing tubes (Nunc- Immuno Tubes, Minisorp Ò, Thermo Fischer Scientific, Waltham, MA, USA) and stored on ice for a maximum of 3 h. Samples and calibrator were further diluted to 2 miu ml 1 in diluent from the kits or HBS/BSA. Subsequently, dilutions of.25,.5, 1, 2, 3, 4, and 5 miu ml 1 were prepared. All dilutions were prepared in two independent series, resulting in independent duplicate measurements. Calibrator/samples were mixed with reagents containing FX, FIXa, (pro)thrombin, phospholipids, and calcium, according to the manufacturer s instructions, and incubated for a specified time period before reagent-containing FXa substrate was added. All reactions took place at 37 C. Absorbance at 45 nm was measured continuously on a SpectraMax Ò Microplate Reader (Molecular Devices, Sunnyvale, CA, USA) for 5 min. The increase in absorbance at 45 nm over time (DA 45 min 1 ) was detected using SoftMax Pro software (Molecular Devices); DA 45 min 1 of a negative control (diluent) was subtracted from all measurements. A linear plot of DA 45 min 1 vs. FVIII:C of the calibrator was generated to calculate FVIII:C of the samples. The experiments were performed three times. Specific activity (given in IU mg 1 ) was determined based on protein concentrations determined by sizeexclusion high-performance liquid chromatography by quantification of the peak area for the samples relative to peak area for reference samples. The protein concentration of the reference samples were determined by quantitative amino acid analysis. Protein concentrations of.527 and.869 mg ml 1 were obtained for N8-GP and turoctocog alfa, respectively. Time course of FX activation for potency assignment Time courses of FX activation were performed in order to select an incubation time at which 5% of the maximal (plateau) level of FX activation was achieved for each chromogenic assay kit. N8-GP or turoctocog alfa were first diluted to 1 IU ml 1, as described, and then further diluted in either the diluent provided (Coamatic Ò, Coatest Ò SP, and Biophen FVIII:C) or HBS/BSA (Technoclone, Siemens, and Grifols) to concentrations corresponding to a % sample as recommended in the kits. Next, samples of 4% (and 2% and 8% for N8-GP) were prepared in diluent or HBS/BSA. The final FVIII concentrations are noted in Table 1. Incubation times ranging from to 3 min were investigated. Maximal FX activation is obtained at the plateau where FXa activity (DA 45 min 1 ) no longer increases (Fig. 1). In this study, this was defined as the highest measurement of DA 45 min 1. Linear regression using GraphPad Prism version 6.4 (GraphPad Software, Inc., La Jolla, CA, USA) of the portion of the FX activation curve with linear FX activation was used to determine the time to 5% of maximal FX activation. The portion of the curves to be included in a linear regression was identified visually. Chromogenic assays for clinical monitoring of simulated postadministration samples Simulated postadministration samples were prepared by spiking severe HA plasma with N8-GP or turoctocog alfa at concentrations of.2,.6, and.9 IU ml 1. These samples were investigated using the following chromogenic assay kits: Coamatic Ò, Coatest Ò SP, Biophen FVIII:C, and FVIII Chromogenic. The key characteristics of the different chromogenic assay kits are compared in Table S1. All chromogenic assays were adapted to be performed on an ACL Ò TOP 5 analyzer (Instrumentation Laboratory, Bedford, MA, USA) at 45 nm, with input from the manufacturers. Two calibrators were used: a plasma calibrator (Siemens SHP) and an FVIII concentrate calibrator (WHO 8th IS). Standard curves were generated covering the range between.2 and 1.25 IU ml 1. All dilutions were made in two independent series, resulting in independent duplicate measurements. Incubation times were selected based on the manufacturer s recommendations. The experiments were performed three times. Note

4 1582 W. Pickering et al Table 1 Determination of optimal incubation times based on time to 5% FX activation for N8-GP and the comparator turoctocog alfa (N8) using different chromogenic assay kits Chromogenic assay kit Recommended incubation time (s) FVIII (IU ml 1 )* Time to 5% FX activation (s) N8-GP N8 Selected incubation time (s) Coamatic Ò (Chromogenix) ND ND Coatest Ò SP (Chromogenix) ND ND Biophen FVIII:C (Hyphen BioMed) ND ND Technochrom Ò FVIII:C (Technoclone) ND ND DG-Chrom FVIII (Grifols) ND ND FVIII Chromogenic (Siemens) ND ND ND, not determined. *FVIII concentrations correspond to 8%, 4%, and 2% of recommended FVIII levels in the kits. Each data value represents the mean SD of three experimental runs, except for the Siemens kit, where six experimental runs were performed. that postadministration samples were not investigated using the Grifols and Technoclone assays. The Grifols assay was not available at the time that this part of the study was performed, and the Technoclone assay method could not be transferred to the ACL Ò TOP 5 analyzer. Data analysis For potency labeling, the data are expressed as mean values including standard deviation (SD). The coefficient of variation (CV) was calculated as SD. The mean FVIII:C combined intra-assay and inter-assay CV was calculated using the following equation: ðinter-assay CVÞ 2 þðmean intra-assay CVÞ 2, where qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi mean intra-assay CV corresponds to the mean of the CVs for the individual assays, and inter-assay CV is calculated from the mean FVIII:C and SD between the assays. The combined CV was next converted to a combined SD as follows: mean FVIII:Ccombined CV. Statistical analysis was performed using Tukey s honest significance difference test. For the simulated postadministration samples, the data are expressed as mean values including SD, percent CV, and relative error (% RE). Between-run precision (% CV) was calculated from values obtained from three runs on different assay days. Statistical analysis of simulated postadministration samples was performed using the paired t-test. A two-tailed P value <.5 was considered to be statistically significant. Results Establishment of FVIII:C assay conditions for potency assignment: the time required to obtain 5% FX activation was similar for N8-GP and the comparator turoctocog alfa The time to achieve 5% of the maximal (plateau) level of FX activation was determined for N8-GP and turoctocog alfa for each chromogenic assay kit, in accordance with the Ph. Eur. guidelines (Table 1 and Fig. 1) [2]. For this, concentrations of N8-GP or turoctocog alfa of.3.1 IU ml 1 were analyzed depending on the recommended dilutions in each kit (Table 1). The time to 5% FX activation was determined by linear regression of the initial linear part of the curve of FXa generation (Fig. 1). For one kit (FVIII Chromogenic; Siemens), a plateau was not reached within the 3 min evaluated and consequently 5% of FXa plateau level could not be calculated. For the Biophen FVIII:C (Hyphen BioMed) kit, a distinct plateau was not reached, however, the data tended toward a plateau. Instead, the calculation of time to 5% activation for the Siemens and the Hyphen BioMed kit was based on the latest measured time point, under the assumption that this corresponds to the maximal level. The activation times needed to obtain 5% FX activation using the six commercial kits differed among kits, but the incubation times needed to obtain 5% FX activation for each individual kit were similar for samples containing the same level of N8-GP or turoctocog alfa (Table 1). The

5 Measuring FVIII:C of N8-GP in chromogenic assays 1583 A FXa activity (ΔA 45 min 1 ) Coamatic, Chromogenix mu ml 1 5 mu ml mu ml 1 B FXa activity (ΔA 45 min 1 ) 15 5 Coatest SP, Chromogenix mu ml 1 5 mu ml mu ml 1 C Biophen FVIII:C, Hyphen BioMed D Technochrom FVIII:C, Technoclone FXa activity (ΔA 45 min 1 ) FXa activity (ΔA 45 min 1 ) E DG-Chrom FVIII, Grifols miu ml 1 1 miu ml 1 5 miu ml 1 FXa activity (ΔA 45 min 1 ) F miu ml 1 1 miu ml 1 5 miu ml 1 FXa activity (ΔA 45 min 1 ) 4 3 FVIII Chromogenic, Siemens miu ml 1 1 miu ml 1 5 miu ml 1 26 miu ml 1 13 miu ml miu ml 1 Fig. 1. Time course of FX activation using different FVIII chromogenic assay kits. The incubation time at which 5% of the maximal (plateau) level of FX activation was achieved was determined for samples containing N8-GP at the concentrations noted. Chromogenic assays including (A) Coamatic Ò (Chromogenix), (B) Coatest Ò SP (Chromogenix), (C) Biophen FVIII:C (Hyphen BioMed), (D) Technochrome Ò (Technoclone), (E) DG Chrom FVIII (Grifols), and (F) FVIII Chromogenic (Siemens) were performed. Incubation times, ranging from to 3 min, were investigated. Each data value represents the mean standard deviation of three experimental runs performed in duplicate, except for the Siemens kit, where six experimental runs were performed. incubation time used for potency assessment was selected based on the values for time to 5% activation for samples containing.1 IU ml 1 or lower concentration of N8- GP, provided the incubation time was within the linear part of the FXa generation curve (Fig. 1 and Table 1). In several cases, the selected incubation times were different from those recommended by the manufacturer. No difference between chromogenic assays adapted to the Ph. Eur. guidelines for N8-GP potency assignment The FVIII:C values of N8-GP and the comparator turoctocog alfa were analyzed using each chromogenic assay adapted to the Ph. Eur. guidelines. The incubation times were adjusted as described here earlier, and samples were diluted as recommended. For each of the six kits, all of the measured FVIII:C values fell within 1% ( IU ml 1 ) and within 7% ( IU ml 1 ) of the calculated means for N8-GP and turoctocog alfa, respectively (Table 2). When using a 1% significance level, no significant difference in N8-GP FVIII: C was observed among the six chromogenic assay kits assessed. When the specific activity was calculated by dividing FVIII:C by the protein concentration, values of 9 11 IU mg 1 were obtained for both N8-GP and turoctocog alfa (Table 2).

6 1584 W. Pickering et al Table 2 Potency assignment of N8-GP and turoctocog alfa using different FVIII chromogenic assay kits adapted to the guidelines from the European Pharmacopoeia N8-GP Turoctocog alfa Chromogenic assay kit Number of assay runs FVIII:C (IU ml 1 )* Specific activity (IU mg 1 )* FVIII:C (IU ml 1 )* Specific activity (IU mg 1 )* Coamatic Ò (Chromogenix) Coatest Ò SP (Chromogenix) Biophen FVIII:C (Hyphen BioMed) Technochrom Ò FVIII:C (Technoclone) DG Chrom FVIII (Grifols) FVIII Chromogenic (Siemens) *Data are presented as mean combined intra-assay and inter-assay SDs. In simulated clinical postadministration samples containing N8-GP, there was minimal variation between FVIII:C using different FVIII chromogenic assays and calibrators Postadministration samples were simulated by adding N8-GP or turoctocog alfa to severe HA plasma to FVIII:C levels of.9,.6, and.2 IU ml 1 based on the labeled potency. FVIII:C of the plasma samples was evaluated using two different calibrators (a plasma calibrator and the WHO 8th IS used for potency labeling) and four chromogenic assay kits adapted to perform on the ACL Ò TOP 5 instrument. The measured values correlated well with expected FVIII:C across the tested concentrations, especially at.6 and.9 IU ml 1, for all kits and both calibrators (Fig. 2 and Table S2). The accuracy of the measurements (represented by % RE) was within 22.7% for samples containing either N8-GP or turoctocog alfa and using each kit with both calibrators (Table S2). The kits mostly performed well and demonstrated an appropriate level of precision (CV < 2%; Table S2), although the Siemens kit consistently demonstrated higher CVs compared with the other kits. At the lowest labeled value (.2 IU ml 1 ), three kits (Coamatic Ò, Coatest Ò, and Siemens) generated results for both N8-GP and turoctocog alfa that were within 9.1% (% RE) of the labeled value, while one (Hyphen) gave results that were > 59% of the expected value. When the WHO 8th IS FVIII concentrate was used as the calibrator, results were closer to labeled values for both N8-GP and turoctocog alfa for all kits assessed compared to results with the Siemens SHP plasma calibrator; however, there was no statistically significant difference in the FVIII:C measurements using the two different calibrators. Using the Siemens SHP calibrator, the measurements were generally slightly higher than the expected FVIII:C. Among the kits tested, the Coatest Ò kit produced results closest to the label value for N8-GP, with N8-GP FVIII:C slightly lower than turoctocog alfa. However, this difference was not significant (P >.5) and the correlation between N8-GP and turoctocog alfa results was.997. These data indicate that the PEG moiety does not significantly affect the results obtained using the different FVIII chromogenic kits. Further, the results suggest that the choice of calibrator has a greater influence on the measurement of N8-GP than the presence of the PEG moiety. Discussion This study demonstrates that chromogenic assays adapted to the Ph. Eur. guidelines can be used to establish the potency of N8-GP and that comparable potency is obtained using different commercially available chromogenic assay kits. Further, the data demonstrate that chromogenic assays from different manufacturers can be used to measure FVIII:C in plasma samples containing N8-GP or turoctocog alfa, suggesting that chromogenic assays may be used for clinical monitoring. Similar potency assignments for N8-GP were obtained using all six chromogenic assays investigated in this study, indicating that the different composition of the chromogenic kits (Table S1) had little impact on measured FVIII:C. The measured FVIII:C values for each individual kit were within 1% of the mean determined for all kits, indicating that these chromogenic kits are robust and reliable for the assessment of N8-GP. Specific activities between 9 and 11 IU mg 1 were obtained for both N8-GP and the FVIII comparator turoctocog alfa, demonstrating that N8-GP FVIII:C measured using chromogenic assays is not affected by PEG conjugation. The results obtained from the analysis of simulated postadministration samples using the chromogenic assay kits correlated well with expected FVIII:C, especially when the concentrate calibrator was used. The accuracy of FVIII:C measurement was generally good for all four chromogenic assay kits tested, although accuracy tended to be reduced at lower FVIII:C concentrations. The observed precision of FVIII:C was also appropriate (< 2%) for all kits tested. The use of two calibration curves (high and low) may improve the nonlinearity that

7 Measuring FVIII:C of N8-GP in chromogenic assays 1585 A B N8-GP Turoctocog alfa C 1. D E.4 F Fig. 2. FVIII:C in N8-GP simulated postadministration samples. N8-GP and the comparator turoctocog alfa were added to severe HA plasma and FVIII:C measured using (A, C, E) a plasma (Siemens SHP) or (B, D, F) a concentrate (WHO 8th IS) calibrator and different chromogenic assay kits, that is, (1) Coamatic Ò (Chromogenix), (2) Coatest Ò SP (Chromogenix), (3) Biophen FVIII:C (Hyphen BioMed) or (4) FVIII Chromogenic (Siemens) kits. The results are expressed as mean SD (n = 3) showing the comparative distribution at three different dose levels, (A, B).9 IU/mL, (C, D).6 IU/mL and (E, F).2 IU/mL. The thick solid line indicates the expected label value and the thin solid lines indicate 1%, 2% and 3%.

8 1586 W. Pickering et al was observed for some kits across the tested concentrations (.9.2 IU ml 1 ). Similar to the data on potency measurements, the results of the analysis of the simulated postadministration samples for the kits tested showed that the presence of the PEG moiety did not influence the assay, suggesting that chromogenic assays may serve as an adjunct to existing clinical monitoring assays. However, since thrombin activation of N8-GP results in removal of the PEG molecule and the truncated B-domain of the turoctocog alfa molecule [16], leaving FVIIIa similar to that of the FVIIIa calibrator in the assay, the lack of influence of PEG conjugation on the specific activity of N8-GP may not extend to other modified FVIII molecules. Previous field studies have assessed the consistency of FVIII:C measurements using routine FVIII:C assays in clinical laboratories with FVIII compounds added to severe HA plasma [8,2]. In general, FVIII:C measured using chromogenic assays for turoctocog alfa, efmoroctocog alfa (Eloctate Ò ), and octocog alfa (Advate Ò ; Baxalta, Bannockburn, IL, USA) showed a uniform deviation from the expected values across tested FVIII concentrations (between.3 and.9 IU ml 1 ), that is, a 7 24% overestimation of FVIII:C was observed at all FVIII levels. Further, in both of these studies, chromogenic assays appeared to be more accurate in the lower FVIII concentration range compared to one-stage clot assays, which overestimated FVIII:C at the lowest FVIII concentrations (.3.5 IU ml 1 ) [8,2]. In a study comparing the pharmacokinetic profiles of different FVIII products (octocog alfa [Advate Ò and Kogenate Ò ], turoctocog alfa, and plasma-derived FVIII) using either the chromogenic assay or the one-stage clot assay, it was reported that all recombinant products showed a discrepancy between the two assays, with the chromogenic assay measuring a higher value across different activity levels [21] (unpublished data). This was observed for both clinical and simulated postadministration samples. There was little to no difference between assays for plasma-derived FVIII products. These findings may be attributable to the similarity of the tested plasma-derived FVIII products and the plasma calibrator used. In the present study, both a plasma calibrator and the WHO 8th IS for FVIII concentrates were included for analysis of simulated postadministration samples. The choice of calibrator influenced the measured FVIII:C in the simulated postadministration samples more than the presence of the PEG moiety on N8-GP. The WHO 8th IS FVIII concentrate calibrator produced results that were closer to labeled values and approximately 1% lower than those obtained using the plasma calibrator. This is likely related to the potency labeling of N8-GP (and turoctocog alfa in Europe) linked directly to the WHO 8th IS FVIII concentrate calibrator, while there is no direct link to the plasma calibrator. The results of this study support the use of chromogenic assays for the measurement of N8-GP FVIII:C for both potency labeling and clinical monitoring as an addition to aptt-based clot assays. However, despite increased community understanding of the applicability of the FVIII chromogenic assays, several challenges are associated with the implementation of chromogenic assays in routine clinical laboratory practice that may inhibit their routine use in clinical monitoring. These include a perception of higher associated costs compared to the one-stage clot assay and greater technical complexity. However, a recent cost analysis shows that the costs associated with FVIII chromogenic assays and one-stage clot assays are actually very similar [22]. This analysis suggests that chromogenic assay costs could be improved with the use of aliquoted, frozen chromogenic kit reagents and batching of samples. Technical issues may also be overcome by automation following initial setup and validation. The benefits of using chromogenic assays in addition to one-stage clot assays may also extend to more accurate diagnosis of HA patients, as some FVIII mutation variants lead to discrepancies between measured FVIII:C and HA phenotype, and are not detected using only one assay [1,11]. In summary, the present study supports the use of chromogenic assays for both potency assignment and clinical monitoring of N8-GP. Importantly, PEG conjugation does not affect the determination of potency for N8-GP or the measured FVIII:C in simulated postadministration samples. Further, the data demonstrate that the choice of calibrator for clinical monitoring influences measured FVIII:C more than the presence of the PEG moiety on the truncated B-domain in N8-GP. Addendum W. Pickering performed and analyzed the experiments involving simulated postadministration samples and M. Hansen and M. Kjalke performed and analyzed the experiments involving potency measurements. All authors contributed to the overall design of the studies and interpretation of results. All authors also contributed to the writing and review of the manuscript and approved the final version. Acknowledgements Lone Odborg and Mette Schmidt, Novo Nordisk A/S, are thanked for their expert technical assistance. Medical writing support was provided by Physicians World Europe GmbH, Mannheim, Germany and was financially supported by Novo Nordisk A/S, Denmark. Disclosure of Conflict of Interests All authors are employees of Novo Nordisk A/S, Maløv, Denmark. All authors hold shares in Novo Nordisk A/S. The authors have no other competing interests.

9 Measuring FVIII:C of N8-GP in chromogenic assays 1587 Supporting Information Additional Supporting Information may be found in the online version of this article: Table S1. Summary of the key characteristics of the various commercially available FVIII chromogenic assay kits used in this study. Table S2. The effect of different chromogenic assay kits and different calibrators on the observed FVIII:C for N8-GP or turoctocog alfa (N8) at label value.* References 1 Moser KA, Adcock Funk DM. Chromogenic factor VIII activity assay. Am J Hematol 214; 89: European Pharmacopoeia 8.2: Assay of Human Coagulation Factor VIII. 214: Barrowcliffe TW, Hubbard AR, Kitchen S. Standards and monitoring treatment. Haemophilia 212; 18(Suppl. 4): Hubbard AR, Dodt J, Lee T, Mertens K, Seitz R, Srivastava A, Weinstein M. Recommendations on the potency labelling of factor VIII and factor IX concentrates. J Thromb Haemost 213; 11: Hubbard AR. Potency labeling of novel factor VIII and factor IX concentrates: past experience and current strategy. Semin Thromb Hemost 215; 41: ELOCTATE: Full Prescribing Information, fda.gov/downloads/biologicsbloodvaccines/bloodbloodproducts/ ApprovedProducts/LicensedProductsBLAs/FractionatedPlasma Products/UCM4192.pdf. Accessed 25 March Kitchen S, Signer-Romero K, Key NS. Current laboratory practices in the diagnosis and management of haemophilia: a global assessment. Haemophilia 215; 21: Viuff D, Barrowcliffe T, Saugstrup T, Ezban M, Lillicrap D. International comparative field study of N8 evaluating factor VIII assay performance. Haemophilia 211; 17: Makris M, Peyvandi F. Assaying FVIII activity: one method is not enough, and never was. Haemophilia 214; 2: Srivastava A, Brewer AK, Mauser-Bunschoten EP, Key NS, Kitchen S, Llinas A, Ludlam CA, Mahlangu JN, Mulder K, Poon MC, Street A. Guidelines for the management of hemophilia. Haemophilia 213; 19: e Oldenburg J, Pavlova A. Discrepancy between one-stage and chromogenic factor VIII activity assay results can lead to misdiagnosis of haemophilia A phenotype. Hamostaseologie 21; 3: Potgieter JJ, Damgaard M, Hillarp A. One-stage vs. chromogenic assays in haemophilia A. Eur J Haematol 215; 94 (Suppl. 77): Trossaert M, Lienhart A, Nougier C, Fretigny M, Sigaud M, Meunier S, Fouassier M, Ternisien C, Negrier C, Dargaud Y. Diagnosis and management challenges in patients with mild haemophilia A and discrepant FVIII measurements. Haemophilia 214; 2: MASAC statement regarding the use of various clotting factor assays to monitor factor replacement therapy. National Hemophilia Foundation, New York, NY, Gu JM, Ramsey P, Evans V, Tang L, Apeler H, Leong L, Murphy JE, Laux V, Myles T. Evaluation of the activated partial thromboplastin time assay for clinical monitoring of PEGylated recombinant factor VIII (BAY ) for haemophilia A. Haemophilia 214; 2: Stennicke HR, Kjalke M, Karpf DM, Balling KW, Johansen PB, Elm T, Ovlisen K, Moller F, Holmberg HL, Gudme CN, Persson E, Hilden I, Pelzer H, Rahbek-Nielsen H, Jespersgaard C, Bogsnes A, Pedersen AA, Kristensen AK, Peschke B, Kappers W, et al. A novel B-domain O-glycoPEGylated FVIII (N8- GP) demonstrates full efficacy and prolonged effect in hemophilic mice models. Blood 213; 121: Till S, Tippl S, Palige M, Boehm E, Schrenk G, Koehn J, Gritsch H, Knappe S, Hartmann R, Scheiflinger F, Dockal M. Functional characterization of different factor VIII molecules with focus on phospholipid and platelet binding. J Thromb Haemost 215; 13(Suppl. S2): abstract PO Dodt J, Hubbard AR, Wicks SJ, Gray E, Neugebauer B, Charton E, Silvester G. Potency determination of factor VIII and factor IX for new product labelling and postinfusion testing: challenges for caregivers and regulators. Haemophilia 215; 21: Tiede A, Brand B, Fischer R, Kavakli K, Lentz SR, Matsushita T, Rea C, Knobe K, Viuff D. Enhancing the pharmacokinetic properties of recombinant factor VIII: first-in-human trial of glycopegylated recombinant factor VIII in patients with hemophilia A. J Thromb Haemost 213; 11: Sommer JM, Moore N, McGuffie-Valentine B, Bardan S, Buyue Y, Kamphaus GD, Konkle BA, Pierce GF. Comparative field study evaluating the activity of recombinant factor VIII Fc fusion protein in plasma samples at clinical haemostasis laboratories. Haemophilia 214; 2: Tiede A, Ezban M, Krogh-Meibom T, Saugstrup T, Møss J. Factor VIII assessment using one-stage clot and chromogenic assay in trials investigating pharmacokinetics of different FVIII products. J Thromb Haemost 213; 11(Suppl. 2): abstract PB Kitchen S, Blakemore J, Friedman KD, Hart DP, Ko RH, Perry D, Platton S, Tan-Castillo D, Young G, Luddington RJ. A computer-based model to assess costs associated with the use of FVIII and FIX one-stage and chromogenic activity assays. J Thromb Haemost 216; 14: