Hamostaseologie
DOI: 10.1055/a-2717-3413
Original Article

On Factor VIII Assay Discrepancies in Post-infusion Samples Obtained from Patients Treated with Efanesoctocog Alfa

Authors

  • Jens Müller

    1   Institute of Experimental Hematology and Transfusion Medicine, University Clinic Bonn, University of Bonn, Bonn, Germany
  • Martin Büchsel

    2   Institute for Clinical Chemistry and Laboratory Medicine, Medical Center and Faculty of Medicine, University of Freiburg, Freiburg, Germany
  • Behnaz Pezeshkpoor

    1   Institute of Experimental Hematology and Transfusion Medicine, University Clinic Bonn, University of Bonn, Bonn, Germany
  • Heiko Rühl

    1   Institute of Experimental Hematology and Transfusion Medicine, University Clinic Bonn, University of Bonn, Bonn, Germany
  • Bernd Pötzsch

    1   Institute of Experimental Hematology and Transfusion Medicine, University Clinic Bonn, University of Bonn, Bonn, Germany
  • Claudia Klein

    1   Institute of Experimental Hematology and Transfusion Medicine, University Clinic Bonn, University of Bonn, Bonn, Germany
  • Georg Goldmann

    1   Institute of Experimental Hematology and Transfusion Medicine, University Clinic Bonn, University of Bonn, Bonn, Germany
  • Natascha Marquardt

    1   Institute of Experimental Hematology and Transfusion Medicine, University Clinic Bonn, University of Bonn, Bonn, Germany
  • Thilo Albert

    1   Institute of Experimental Hematology and Transfusion Medicine, University Clinic Bonn, University of Bonn, Bonn, Germany
  • Kerstin Herbst

    1   Institute of Experimental Hematology and Transfusion Medicine, University Clinic Bonn, University of Bonn, Bonn, Germany
  • Johannes Oldenburg

    1   Institute of Experimental Hematology and Transfusion Medicine, University Clinic Bonn, University of Bonn, Bonn, Germany

Funding There was no funding for this retrospective study.
 

Abstract

Efanesoctocog alfa (efa) is a recombinant coagulation factor VIII (FVIII) concentrate, engineered for improved extended half-life in hemophilia A treatment. Its design results in discrepancies in FVIII diagnostic tests, as has so far been demonstrated using spiked sample material (efa added to FVIII-deficient plasma). The aim of the present study was to evaluate FVIII assay discrepancies in post-infusion samples obtained from patients treated with efa. This retrospective analysis included 43 male patients on efa prophylaxis (26 on once weekly and 17 on twice weekly treatment scheme) summoned for determination of incremental recoveries. FVIII activity was measured at trough (3.5 or 7 days post-infusion) and peak levels (30 minutes post-infusion) using three diagnostic tests: two one-stage clotting assays (OSCAs) based on Actin FSL (AFSL) or Actin FS (AFS), and a chromogenic substrate assay (FVIII CSA). Incremental recoveries as determined by the recommended AFSL-based OSCA were found to be comparable between patient groups. At peak levels (64 to 214 IU/dL), comparable overestimation (1.9-fold) of plasma efa activity relative to Actin FSL-based OSCA was observed for both, the AFS-based OSCA and the FVIII CSA. In contrast, at trough levels (5 to 64 IU/dL), rate of overestimation relative to the AFSL-based OSCA results was found to be lower for the FVIII CSA (1.3-fold) when compared with the AFS-based OSCA (1.9-fold). Further analysis demonstrated different behavior of spiked and post-infusion samples within the FVIII CSA. Future studies will reveal underlying mechanisms and assess if drug-specific calibration will sufficiently correct for this phenomenon.


Introduction

Efanesoctocog alfa (efa) is a recombinant coagulation factor VIII (rFVIII) product characterized by multiple molecular modifications for further improved extended half-life in hemophilia A treatment.[1] [2] The complex structure of efa leads to notable discrepancies in factor VIII (FVIII) activity measurements depending on the monitoring test (assay) employed. In a global comparative field study that was based on spiked plasma samples (drug [efa] added to FVIII-deficient plasma [FVIII-dp]), significant inconsistencies have been observed not only between but also within the classes of activated partial thromboplastin time (aPTT)-based one-stage clotting assays (OSCAs) and FVIII chromogenic substrate assays (FVIII CSAs).[3] According to these findings, sufficient reproducibility of results between laboratories can be obtained using select OSCAs. Among these, OSCAs employing Actin FSL (AFSL) as the aPTT reagent were identified as the most accurate assays, owing to the fact that potency labeling of the product is also done using this particular kind of assay.[4] [5] [6] In contrast to OSCAs, FVIII CSAs were found to consistently overestimate efa activity by a factor of 2 to 3.[3] A comparable rate of overestimation was demonstrated for Actin FS (AFS)-based OSCAs, leading to introduction of a proposed correction factor (CF) of /2.5 (×0.4) when interpreting results obtained by AFS-based OSCAs or FVIII CSAs.[4] [5]

In the present investigation, we performed a retrospective analysis of coagulation factor VIII (FVIII) activity measurements in post-infusion samples from patients with severe hemophilia A receiving efa, utilizing three distinct diagnostic tests/methodologies: (i) an AFSL-based OSCA, (ii) an AFS-based OSCA, and (iii) a FVIII CSA employing bovine coagulation factors. All included subjects had transitioned to efa therapy and underwent routine assessment of incremental recovery 30 minutes post-infusion (IR30) at our institution.[7] [8] Importantly, FVIII assay discrepancies observed at efa trough levels (pre-infusion) deviated substantially from those anticipated based on data from the above-mentioned global field study.[3] Our findings highlight a pronounced divergence between assay discrepancies in spiked versus post-infusion samples that should be considered for accurate clinical interpretation of test results.


Materials and Methods

Patients and Determination of Incremental Recoveries (IR30)

This dedicated retrospective analysis included data from 43 male outpatients diagnosed with severe hemophilia A (HA; baseline FVIII activity <1 IU/dL) on routine visits for IR30 determination at the Institute of Experimental Hematology and Transfusion Medicine, University Clinic Bonn, Bonn, Germany, between August 2024 and December 2024. Among these patients, 26 were on a standard prophylactic regimen consisting of a once weekly (qw) dose of 50 IU/kg body weight (BW), while 17 patients followed a twice weekly (biw) dosing scheme to maintain higher trough levels due to manifest hypertrophic synovia diagnosed by ultrasound.[2] [4] [5] [9] The median (interquartile range, IQR) dose in the biw group was 35.0 (33.8–40.0) IU/kg BW.

Included patients were either previously untreated patients (PUPs, n = 2), already on study drug (BIV001, n = 5), or switched from other FVIII concentrates at least 9 days before visit. Patients switched from the following products to efa: efmoroctocog alfa (n = 30), rurioctocog alfa pegol (n = 4), damoctocog alfa pegol (n = 1), or plasma-derived factor VIII (Octanate®, n = 1). Determination of IR30 values was based on corresponding individual prophylactic dosings at trough levels 7 days (qw) or 3.5 days (biw) after the last injection. Blood samples were taken at trough levels before (pre-infusion) and 30 minutes post-infusion. Incremental recoveries (IR30) were determined as ratio between baseline (trough level)-corrected post-infusion peaks (IU/dL) as measured by the AFSL-based OSCA and the administered FVIII doses (IU/kg BW).


Sample Collection and FVIII Activity Assays

Venous blood samples were collected using a 21-gauge winged infusion set and non-vacuum tubes (closed system, Sarstedt, Nümbrecht, Germany). After initial collection into an EDTA tube, blood was drawn into citrate tubes (10.5 mmol/L final concentration, Sarstedt) and centrifuged at 2500 × g for 15 minutes to obtain plasma. FVIII activity assays were conducted within 4 hours of blood collection.

FVIII activity was measured on a CN-6000 coagulation analyzer (Sysmex/Siemens Healthineers [Siemens], Eschborn, Germany) using three different diagnostic tests (assays): two OSCAs, using either AFSL (lots 562735 and 562787A) or AFS (lot 562457A) as aPTT reagents, and one FVIII CSA (FVIII chromogenic) based on bovine factors (lots 02257 and 02804, all Siemens). All assays were calibrated using Standard Human Plasma (SHP, Siemens). Calibration was performed automatically by the analyzer, covering 0.7 to 129 IU/dL for AFSL- and AFS-based OSCA and 0.6 to 120 IU/dL for FVIII CSA. Samples exceeding the upper calibrator limit were automatically reanalyzed after a 1:4 predilution with Owens buffer. The validated analytical range for all assays was 1 to 480 IU/dL. Measurements were conducted in singlicate for AFSL- and AFS-based OSCA and in duplicate for FVIII CSA, according to the manufacturer's specifications. IR30 values were calculated based on AFSL-based OSCA measurements. The other two assays were included due to previously observed assay discrepancies and were employed as part of an extended laboratory monitoring.


Preparation and Analysis of Plasma Samples Spiked with Efanesoctocog Alfa

Congenital FVIII-deficient plasma (Technoclone, Vienna, Austria) was spiked with efanesoctocog alfa (Lot 4114401B, 1000 IU/vial nominal activity, 333 IU/mL after reconstitution), efmoroctocog alfa (Lot 3671301A, 1000 IU/vial nominal activity, 333 IU/mL), both Swedish Orphan Biovitrum (Sobi, Stockholm, Sweden), and octocog alfa (Kogenate, Lot ITA2IKI, 2000 IU/vial nominal activity, 400 IU/mL, Bayer, Leverkusen, Germany) at levels of 320, 160, 80, 40, 20, 10, or 5 IU/dL and analyzed in parallel by all three routine FVIII diagnostic tests.


Data Analysis

Data analysis and graphical presentation were done using Microsoft Excel® and Powerpoint® (Microsoft Office 365). If not otherwise stated, data are given as median values (interquartile ranges [IQRs]). A non-parametric test (Mann–Whitney U test) was applied for assessment of statistical significance (p < 0.05) of results between datasets using a corresponding online tool (https://www.statskingdom.com).



Results

Patient Characteristics and Incremental Recoveries (IR30)

Patient characteristics and details on applied treatment schemes are provided in [Table 1]. No statistically significant differences in age or body weight were observed between patients on qw and biw prophylactic regimens. As anticipated, efa trough levels, determined by AFSL-based OSCA, were significantly higher in the biw group (median [IQR]: 44.7 [40.1–49.8] IU/dL) compared with the qw group (13.1 [10.7–17.4] IU/dL) ([Fig. 1A], [Table 1]). Incremental recoveries (IR30) were calculated as ratio between increments in FVIII activity and administered efa doses and found to be comparable between patient groups (qw: 2.77 [2.32–3.17] IU/dL/IU/kg BW versus biw: 2.65 [2.26–2.98] IU/dL/IU/kg BW; p = 0.785; [Table 1]). These findings align well with previously published data.[1] [10] [11]

Zoom
Fig. 1 Results of FVIII:C measurements of efanesoctocog alfa trough and peak levels in the qw and biw patient groups by Actin FSL or Actin FS-based one-stage clotting assays (OSCAs) or FVIII chromogenic substrate assay (FVIII CSA). Results are shown as absolute values (A, Actin FSL-based OSCA, Actin FS-based OSCA; FVIII CSA) or as ratios to values found applying the Actin FSL-based OSCA (B, Actin FS-based OSCA; C, FVIII CSA). Figures show single data points (circles) within box plot whiskers (range, excluding outliers [circles outside whiskers]). Boxes represent median (within box horizontal line), as well as interquartile ranges (box horizontal top line: Q3; bottom line: Q1).
Table 1

Summary of patient characteristics and key findings

Patient groups (treatment qwa versus biwb)

Once weekly (qw); n = 26

Twice weekly (biw); n = 17

qw versus biw

Patient data/parameter

Unit

n c

Median

Q1

Q3

n

Median

Q1

Q3

p -value

Age

Years

26

42.0

23.5

53.5

17

49.0

44.0

53.0

0.064

Body weight (BW)

kg

26

86.0

71.0

90.0

17

84.5

70.5

96.8

0.951

Time on efanesoctocog alfa (EFA)d

Days

22

29.0

21.0

41.5

16

36.0

21.0

73.5

0.533

EFA dose

IU/kg BW

26

50.0

48.2

51.7

17

35.0

33.8

40.0

<0.001

FVIII:C trough (Actin FSL [AFSL])

IU/dL

26

13.1

10.7

17.4

17

44.7

40.1

49.8

<0.001

FVIII:C peak (30 minutes) AFSL

IU/dL

26

148.0

128.3

176.9

17

144.7

124.3

174.8

0.691

FVIII:C incremente

dIU/dL

26

136.6

116.9

159.6

17

101.7

83.0

115.6

0.001

Incremental recovery (30 minutes)

(dIU/dL)/(IU/kg BW)

26

2.77

2.32

3.17

17

2.65

2.26

2.98

0.785

FVIII:C trough (Actin FS [AFS])

IU/dL

26

23.1

20.0

29.0

17

87.7

69.5

95.3

<0.001

FVIII:C trough: AFS/AFSL

-

26

1.72

1.65

1.83

17

1.90

1.77

1.96

0.030

FVIII:C peak (30 minutes) AFS

IU/dL

25

286.0

257.5

342.7

15

274.8

227.8

309.0

0.335

FVIII:C peak (30 minutes): AFS/AFSL

-

25

1.92

1.87

2.06

15

1.87

1.76

1.95

0.074

FVIII:C trough (FVIII CSA [CSA])

IU/dL

26

16.7

12.9

21.6

16

61.8

49.3

71.0

<0.001

FVIII:C trough: CSA/AFSL

-

26

1.24

1.17

1.31

16

1.33

1.20

1.42

0.183

FVIII:C peak (30 minutes) (CSA)

IU/dL

25

285.2

229.6

348.8

17

253.5

200.3

306.3

0.131

FVIII:C peak (30 minutes): CSA/AFSL

-

25

1.98

1.81

2.09

17

1.74

1.67

1.89

0.038

Note: aqw: once weekly; bbiw: twice weekly; cnumber of (available) datasets, da total of 5 patients (4 × qw and 1 × biw) already on EFA study drug (BIV001) excluded; ebaseline corrected shifts; Actin FSL [AFSL]: Actin FSL-based OSCA; Actin FS [AFS]: Actin FS-based OSCA.



Assay Discrepancies in Post-infusion Samples

Given reported FVIII assay discrepancies for efa, patient samples were analyzed not only by the AFSL-based OSCA but also by an AFS-based OSCA and an FVIII CSA. The trough and peak values obtained by all three methods across both patient groups are presented in [Table 1] and [Fig. 1A].

FVIII activities measured using either the AFS-based OSCA or the FVIII CSA was consistently higher than those obtained using the AFSL-based OSCA, at both trough and peak levels. Interestingly, the degree of overestimation (ratio to AFSL-based OSCA results) by the FVIII CSA was more pronounced at peak levels than at trough levels in both the qw and biw patient groups ([Fig. 1C]). However, such an effect was much less pronounced or absent when comparing AFSL- versus AFS-based OSCA results ([Fig. 1B]).

When combining data from both patient groups (qw and biw), linear interpolation (y-intercept set to 0) revealed similar slopes for trough and peak levels as determined by AFSL-based or AFS-based OSCA (1.81 [trough] versus 1.93 [peak], respectively; [Fig. 2A]). In contrast, determined slopes for FVIII CSA versus AFSL-based OSCA differed substantially between trough (1.26) and peak (1.86) values ([Fig. 2B]). These findings are further illustrated by corresponding relative difference plots ([Fig. 2C, D]).

Zoom
Fig. 2 Correlation of absolute FVIII:C values and relative differences (ratios) at overall (combined results of qw and biw patient groups) efanesoctocog alfa trough and peak levels as determined by Actin FSL- or Actin FS-based one-stage clotting assays (OSCAs) or FVIII chromogenic substrate assay (FVIII CSA). (A, B) Linear correlation (slope, y-intercept set to 0) of Actin FSL-based OSCA versus Actin FS-based OSCA (A) and versus FVIII CSA-based absolute FVIII:C results (B). (C, D) Relative difference plots (mean FVIII:C values versus ratios) of Actin FS-based OSCA (C) and FVIII CSA results (D) versus Actin FSL-based results. Light gray symbols show trough levels and dark gray symbols show peak levels.

Factor VIII Assay Discrepancies in Spiked versus Post-infusion Samples

In contrast to our data obtained from analysis of post-infusion samples by the FVIII CSA, previous studies using spiked plasma samples reported more consistent and higher discrepancies of FVIII CSA- versus AFSL-based OSCA results across the covered efa activity range.[3]

To further assess these findings by the reagent/analyzer combination used in the present study, FVIII-dp was spiked with efa (5 to 320 IU/dL) and analyzed in parallel using the three FVIII diagnostic tests employed in this study. When plotting the obtained quantitative values of the AFS-based OSCA or the FVIII CSA (320 IU/dL efa yielded a result above the upper limit of quantification of the assay [ULQ, 480 IU/dL]) against those of the AFSL-based OSCA and comparing them to the values of the trough post-efa infusion samples (5 to 64 IU/dL, see [Fig. 2A/B]), the pattern shown in Fig. (for better visualization, in order not to overload [Fig. 2A, B], the trough level data from [Fig. 2A, B] are shown again here) emerged. According to the slopes of linear interpolations, analysis by the AFS-based OSCA yielded similar characteristics of spiked versus post-infusion samples (slopes: 2.22 for spiked and 1.81 for post-infusion samples; [Fig. 2A], [Fig. 3A]). In contrast, this was not true for the results of the FVIII CSA, which demonstrated a pronounced difference between spiked (slope: 1.97) and post-infusion samples (1.26, [Fig. 2A], [Fig. 3B]). At post-efa peak levels, slopes of spiked and post-infusion samples were quite comparable for both AFS-based OSCA and FVIII CSA analysis ([Fig. 2A, B] and [Fig. 3A, B]).

Zoom
Fig. 3 Correlation of absolute FVIII:C values found in spiked and post-infusion samples as determined by Actin FSL- or Actin FS-based one-stage clotting assays (OSCAs) or FVIII chromogenic substrate assay (FVIII CSA). Linear correlation (slope, y-intercept set to 0) of Actin FSL-based OSCA versus Actin FS-based OSCA (A) and versus FVIII CSA-based absolute FVIII:C results (B). Light gray circles show post-infusion trough levels (also see [Fig. 2A, B]); dark gray squares show spiked samples.

Although efa spiking confirmed the linearity of the FVIII assays employed in this study, additional experiments were conducted with two other FVIII concentrates for comparison: efmoroctocog alfa (rFVIIIFc) and octocog alfa (Kogenate), a recombinant full-length FVIII molecule. The data shown in [Supplementary Fig. S1] (FVIII input activity [5–320 IU/dL] versus FVIII:C found [IU/dL]) confirmed linearity across the entire defined dynamic range for all three FVIII assays applied. When compared with efa, assay discrepancies were markedly less pronounced for the other two products.

For further investigation, results obtained with the AFS-based OSCA and FVIII CSA were again expressed as ratios relative to the AFSL-based OSCA. As shown in [Fig. 4], across the tested activity range, comparable ratios were observed for efmoroctocog alfa and octocog alfa for both assays. AFS-based OSCA: efmoroctocog alfa [mean (SD)]: 1.07 (0.06), octocog alfa: 1.09 (0.05); FVIII CSA: efmoroctocog alfa: 1.11 (0.05), octocog alfa, 1.16 (0.08). In contrast, ratios for EFA were, as expected, substantially higher (AFS-based OSCA: 2.12 (0.17); FVIII CSA: 1.75 (0.24)). Moreover, for the FVIII CSA, decreasing EFA input activity was associated with a downward trend in ratios, ranging from approximately 2 at 160 and 80 IU/dL to approximately 1.5 at 10 and 5 IU/dL ([Fig. 4]). This assay-inherent trend likely contributed to the assay discrepancies observed between trough and peak post-infusion samples. However, it does not fully explain these differences, as the ratio between AFSL-based OSCA and FVIII CSA was substantially lower for post-infusion samples at trough levels, averaging 1.26 (slope; median [IQR] 1.28 [1.17–1.40]) within a range of 5 to approximately 60 IU/dL ([Fig. 2A, B] and [Fig. 3A, B]).

Zoom
Fig. 4 Results of FVIII spiking experiments relative (ratio) to that found by the AFSL-based OSCA. (A) AFS- versus AFSL-based OSCA. (B) FVIII CSA versus AFSL-based OSCA. White bars show octocog alfa; gray bars show efmoroctocog alfa; black bars show efa(nesoctocog alfa).

Reagent-lot–dependent Assay Discrepancies in the Analysis of Post-efa Infusion Samples

Two different reagent lots of AFSL and FVIII CSA were applied during the study period ([Supplemental Fig. S2]). Lot-specific stratification of the results shown in [Fig. 2A, B] revealed significant differences for both trough and peak post-infusion samples when comparing FVIII CSA with AFSL-based OSCA results (ratio: 1.21 [1.15–1.26] versus 1.37 [1.29–1.42] [p < 0.001] at trough [overall: 1.28 (1.17–1.40)] and 1.74 [1.63–1.89] versus 2.02 [1.75–2.12] [p = 0.004] at peak levels [overall: 1.85 (1.70–2.02]; [Supplementary Fig. S2]). In contrast, comparisons between AFSL- and AFS-based OSCA showed more minor lot-related effects, which reached significance only at peak levels (ratio: 1.77 [1.66–1.87] versus 1.82 [1.69–1.96] [p = 0.305] at trough [overall: 1.77 (1.68–1.91)] and 1.88 [1.80–1.95] versus 1.98 [1.91–2.08] [p = 0.011] at peak levels [overall: 1.90 (1.82–2.01)]; [Supplementary Fig. S2]). These findings demonstrate that reagent-lot–specific differences, albeit limited and deemed acceptable with respect to the further aspects addressed below, contributed to the overall variability observed in this study. However, available data did not allow for conclusions regarding the extent to which these lot-dependent effects were specific to efa.


Establishment and Use of Conversion Factors

Assuming that the AFSL-based OSCA yields the most accurate results, conversion factors (CFs) were derived from determined linear interpolation slopes from data of post-infusion samples as shown in [Fig. 2A, B]. For the AFS-based OSCA, an approximated CF of /1.9 was defined for both trough and peak efa plasma levels ([Fig. 2A]). For the FVIII CSA, approximated CFs were /1.3 for trough and /1.9 for peak levels ([Fig. 2B]).

These CFs are in contrast to the currently recommended generic CF of /2.5 for both AFS-based OSCAs and FVIII CSAs.[4] [5] When applying the generic CF of /2.5 to AFS-based OSCA results obtained during the present study, corrected values mainly approached or even exceeded the lower boundary of a defined ± 30% acceptance range relative to AFSL-based OSCA results ([Fig. 5A]). An even higher degree of overcorrection by the generic CF of /2.5 was found for FVIII CSA results, especially at efa trough levels ([Fig. 5C]).

Zoom
Fig. 5 Relative difference plots of corrected FVIII:C results. Mean (corrected) FVIII:C values versus ratios to Actin FSL-based OSCA results are shown. (A, B) Actin FS-based OSCA; (C, D) FVIII CSA. Correction factors (CF) were applied as stated. Light gray symbols show trough levels; dark gray symbols show peak levels. Dashed horizontal lines represent a defined acceptance range of ± 30% of 100% ratio (recovery).

When applying assay- and level (efa trough versus peak levels)-specific CFs as defined during the present study, corrected values closely aligned with AFSL-based OSCA results, with the majority of corrected values lying well within the ± 30% acceptance range ([Fig. 5B, D]). Although these CFs were applied on the respective underlying “training datasets,” raising concerns of circular reasoning, the data shown in [Fig. 5B, D] further illustrate that such specific CFs may offer better clinical utility than the current generic recommendation.



Discussion

Currently, AFSL-based OSCAs are the preferred method for determining efa activity in post-infusion samples. This is due to the fact that an AFSL-based OSCA is used for potency labeling of the drug and, besides, an FVIII CSA was used as the main evaluative diagnostic test during clinical studies on efa.[4] [5] [6] In the absence of this assay, alternative aPTT-based OSCAs or chromogenic assays may be employed, provided that one is aware of their potential to either under- or overestimate plasma efa activity.[3]

According to the efa product information, the recommended generic CF of /2.5 for interpreting AFS-based OSCA or FVIII CSA results is derived from comparative analyses of clinical study samples (FVIII CSA versus AFSL-based OSCA) as well as corresponding data from the global field study. In both instances, the generic CF of /2.5 is referenced to the results obtained using the AFSL-based OSCA, consistent with the approach applied in the present study.[4] [5]

This raises the question of why, based on the parallel analysis of post-infusion samples, the present study yielded divergent CFs. Several explanations are conceivable. One plausible explanation is that the FVIII CSA employed in the clinical studies (Coatest SP4 CSA [based on bovine factors, Werfen, Munich, Germany][6]) yielded systematically higher activity measurements relative to the AFSL-based OSCA than the FVIII CSA used in the present study (FVIII chromogenic, Siemens). Indeed, when considering the results of the global field study, a corresponding trend can be observed.[3] Furthermore, the global field study revealed substantial discrepancies even when identical reagents were used across different analytical platforms.[3] This factor may also have contributed to the overall lower CFs determined in the present investigation (/1.9 and /1.3 versus /2.5).

Assuming that the functional behavior of the efa molecules remains consistent whether circulating in patient plasma or added exogenously to FVIII-dp, one would expect comparable assay responses of both types of samples, also at different efa plasma activity (input) levels. Although this was found virtually true for AFSL- versus AFS-based OSCA results, significant differences were observed for spiked versus post-infusion samples at patient trough levels (5–64 IU/dL) when applying the FVIII CSA. Two explanations for this finding appear plausible: (i) matrix effects, e.g., differences in sample compositions (FVIII-dp versus patient plasma samples) or between assay calibrators and samples, or (ii) in vivo molecular modifications of efa (over time), altering its interaction with assay components.[12] [13] Indeed, such effects, whether independent or overlapping, appear particularly pivotal in application of different assay principles, such as FVIII OSCAs versus FVIII CSAs.[14] [15]

Our study suggests that specific CFs are more reliable than the generic one. However, such CFs must be validated locally. As demonstrated in the efa field study, variability may arise from differences in analyzers and test settings.[3] Moreover, there is currently insufficient evidence to rule out the need for re-assessing established CFs with each change of reagent lots.

FVIII CSAs utilizing bovine coagulation factors are of particular importance for monitoring FVIII substitution in patients receiving concurrent emicizumab therapy.[16] [17] Consequently, a discussion has arisen regarding the potential benefit of product-specific calibration of bovine FVIII CSAs to ensure accurate quantification of efa activity in the presence of emicizumab.[18] Given the generally observed assay discrepancies in efa measurements, this approach appears justified, extending beyond the context of emicizumab interference, to generally improve inter-laboratory reproducibility of results.[6] [19] The data presented in this study also suggest that product-specific calibration appears to be a reasonable approach. However, the differences observed between spiked samples and post-infusion samples suggest that assay discrepancies may still persist when analyzing post-EFA infusion samples, even under product-specific calibration. Although the latter was not within the scope of the present retrospective analysis of patient data obtained in routine diagnostics, indeed, to date, product-specific calibration has only been evaluated using spiked sample material.[18] [19] Further research is required to elucidate the underlying biochemical mechanisms of the observed assay discrepancies and to develop appropriate methodological strategies to address them.

What is Known About this Topic?

  • The long-acting recombinant factor VIII (rFVIII) product efanesoctocog alfa is used in the treatment of individuals with hemophilia A.

  • Previous studies using spiked FVIII-deficient plasma samples have shown that assay results for FVIII activity vary depending on the measurement method.

  • Due to a 2- to 3-fold overestimation of FVIII activity by chromogenic substrate assays (CSA), a generic conversion factor of /2.5 (multiplication by 0.4) has been recommended to approximate actual FVIII levels.

What Does this Paper Add?

  • This study presents a retrospective analysis of FVIII activity in post-infusion samples from patients treated with efanesoctocog alfa.

  • The observed deviations in FVIII CSA results differed significantly from those anticipated based on spiked sample analyses.

  • The discrepant behavior between artificial spiking samples and real patient samples underscores the need for adapted conversion factors rather than relying on a fixed, generic approach.



Conflict of Interest

J.M. has received institutional grants for research from Novo Nordisk and Pfizer and honoraria for lectures and consultancy from Octapharma, Siemens Healtineers, Swedisch Orphan Biovitrum, and Pfizer. M.B. has received institutional grants for research and studies from Bayer, CSL Behring, Roche, Swedisch Orphan Biovitrum, and Takeda and honoraria for lectures from Bayer, Chugai, and Roche. B.Pe. has received institutional grants for research from Biotest, Octapharma, and NovoNordisk as well as honoraria for lectures or consultancy from NovoNordisk and Octapharma. G.G. has received advisory board honorarium from Bayer, BioMarin, Biotest, Chugai Pharmaceutical Co, Ltd, CSL Behring, Grifols, LFB, Novo Nordisk, Octapharma, Pfizer, F. Hoffmann-La Roche Ltd, Swedisch Orphan Biovitrum, and Takeda. N.M. received honoraria/consultation fees from Bayer, Chugai, CSL Behring, LFB, NovoNordisk, Octapharma, Pfizer, Roche, Takeda, and Swedisch Orphan Biovitrum. T.A. received grants for patient support association from Bayer, Biotest, Chugai, CSL-Behring, Grifols, Novo Nordisk, Octapharma, Roche, Swedish Orphan Biovitrum, and Takeda, as well as personal fees for advisory board meetings, and consulting from Bayer, Biomarin, Biotest, CSL-Behring, Grifols, Novo Nordisk, Octapharma, Pfizer, Swedish Orphan Biovitrum, and Takeda. J.O. has received research funding from Bayer, Biotest, CSL Behring, Octapharma, Pfizer, Swedish Orphan Biovitrum, and Takeda and has received consultancy, speakers bureau, honoraria, and scientific advisory board honorarium from Bayer, Biogen Idec, Biomarin, Biotest, CSL Berhing, Chugai, Freeline, Grifols, LFB, Novo Nordisk, Octapharma, Pfizer, Roche, Sanofi, Spark Therapeutics, Swedish Orphan Biovitrum, and Takeda. All other authors have no conflicts of interest to declare that are relevant to the content of this manuscript.

Acknowledgments

The authors thank Simone Gasper for expert technical assistance.

Ethics Statement

For this retrospective analysis, no approval by the local Ethics Committee was required.


Authors' Contributions

J.M. and J.O.: designed the concept of this retrospective analysis; J.M., C.K., G.G., N.M., T.A., K.H., and J.O.: performed data curation and data analysis; M.B., B.Pe., H.R., and B.Pö.: analyzed the data and provided critical expertise; J.M.: wrote the manuscript. All authors reviewed, revised, and finally agreed on the final version of the manuscript.



Address for correspondence

Jens Müller, PhD
University Hospital Bonn
Venusberg Campus 1, 53127 Bonn
Germany   

Johannes Oldenburg, MD, PhD
University Hospital Bonn
Venusberg Campus 1, 53127 Bonn
Germany   

Publication History

Received: 18 August 2025

Accepted: 06 October 2025

Article published online:
19 January 2026

© 2025. Thieme. All rights reserved.

Georg Thieme Verlag KG
Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany


Zoom
Fig. 1 Results of FVIII:C measurements of efanesoctocog alfa trough and peak levels in the qw and biw patient groups by Actin FSL or Actin FS-based one-stage clotting assays (OSCAs) or FVIII chromogenic substrate assay (FVIII CSA). Results are shown as absolute values (A, Actin FSL-based OSCA, Actin FS-based OSCA; FVIII CSA) or as ratios to values found applying the Actin FSL-based OSCA (B, Actin FS-based OSCA; C, FVIII CSA). Figures show single data points (circles) within box plot whiskers (range, excluding outliers [circles outside whiskers]). Boxes represent median (within box horizontal line), as well as interquartile ranges (box horizontal top line: Q3; bottom line: Q1).
Zoom
Fig. 2 Correlation of absolute FVIII:C values and relative differences (ratios) at overall (combined results of qw and biw patient groups) efanesoctocog alfa trough and peak levels as determined by Actin FSL- or Actin FS-based one-stage clotting assays (OSCAs) or FVIII chromogenic substrate assay (FVIII CSA). (A, B) Linear correlation (slope, y-intercept set to 0) of Actin FSL-based OSCA versus Actin FS-based OSCA (A) and versus FVIII CSA-based absolute FVIII:C results (B). (C, D) Relative difference plots (mean FVIII:C values versus ratios) of Actin FS-based OSCA (C) and FVIII CSA results (D) versus Actin FSL-based results. Light gray symbols show trough levels and dark gray symbols show peak levels.
Zoom
Fig. 3 Correlation of absolute FVIII:C values found in spiked and post-infusion samples as determined by Actin FSL- or Actin FS-based one-stage clotting assays (OSCAs) or FVIII chromogenic substrate assay (FVIII CSA). Linear correlation (slope, y-intercept set to 0) of Actin FSL-based OSCA versus Actin FS-based OSCA (A) and versus FVIII CSA-based absolute FVIII:C results (B). Light gray circles show post-infusion trough levels (also see [Fig. 2A, B]); dark gray squares show spiked samples.
Zoom
Fig. 4 Results of FVIII spiking experiments relative (ratio) to that found by the AFSL-based OSCA. (A) AFS- versus AFSL-based OSCA. (B) FVIII CSA versus AFSL-based OSCA. White bars show octocog alfa; gray bars show efmoroctocog alfa; black bars show efa(nesoctocog alfa).
Zoom
Fig. 5 Relative difference plots of corrected FVIII:C results. Mean (corrected) FVIII:C values versus ratios to Actin FSL-based OSCA results are shown. (A, B) Actin FS-based OSCA; (C, D) FVIII CSA. Correction factors (CF) were applied as stated. Light gray symbols show trough levels; dark gray symbols show peak levels. Dashed horizontal lines represent a defined acceptance range of ± 30% of 100% ratio (recovery).