Laboratory Assessment of Emicizumab Levels in Hemophilia A: Influence of Assay Selection on Reported Results

Abstract

Background

Emicizumab has emerged as a promising therapy for hemophilia A (HA), employing a bypassing mechanism to restore hemostasis. However, since the traditional factor assays cannot be used for measuring the effect of emicizumab, treatment monitoring is difficult.

Objectives

To assess the impact of emicizumab on global hemostatic potential (OHP), as well as on chromogenic and modified one-stage (mOSA) FVIII assays calibrated for emicizumab in patient samples.

Methods

Peripheral blood samples from patients with HA on emicizumab were utilized for a validation (n = 28) and a clinical cohort (n = 9). Emicizumab concentration was measured by chromogenic FVIII and mOSA (Actin FS, Actin FSL, and PTT-LA), while overall haemostatic potential (OHP) was used to assess global hemostasis.

Results

Significantly lower emicizumab concentrations were observed using chromogenic FVIII assay compared to mOSA (P < .005, P < .005, and P < .005 for mOSA Actin FSL, mOSA Actin FS, and mOSA PTT-LA respectively) in the validation cohort. In the clinical cohort, emicizumab concentrations were significantly lower with chromogenic FVIII assay compared to mOSA Actin FS (P < .005) and mOSA PTT-LA (P = .009), but not mOSA Actin FSL. In the clinical cohort, no correlation was seen between OHP and emicizumab.

Conclusion

Accurate measurement of emicizumab concentration may have the potential to understand assay variability, which is important for informed clinical decision-making in HA patients. Our findings underscore the need for introducing new methods for measuring emicizumab concentration and suggest chromogenic-based methodologies as a viable option for mitigating assay interference.

Introduction

Emicizumab has emerged as a highly effective therapeutic option for prophylaxis in people with hemophilia A (PwHA).^1,2 This bispecific monoclonal antibody bridges activated factor IX and factor X, providing a bypass mechanism for restoring hemostasis in patients with deficient or dysfunctional factor VIII (FVIII). Emicizumab was initially approved as prophylaxis for PwHA with inhibitors in February 2018 by the European Medical Agency (EMA), and in March 2019, even for patients with severe hemophilia A (HA) without inhibitors. ³ The lack of similarity between emicizumab and FVIII eliminates the risk of cross-reactivity or the development of inhibitors targeting FVIII, which is a significant complication associated with FVIII concentrate replacement therapy. ⁴

Due to its unique mechanism of action that bypasses the rate-limiting step of activation of FVIII by thrombin activation present in one-stage assays (OSA), emicizumab exhibits interference that overestimates the mimicked FVIII activity in these assays.^5–7 A modified FVIII OSA (mOSA) that uses emicizumab as a calibrator and a higher sample predilution (1:8) has been suggested as a convenient and more precise method for emicizumab monitoring.^8,9 By comparison, chromogenic assays (CSA) will be differently affected by emicizumab depending on the origin of the reagent, i.e., bovine or human-derived. While CSA containing human reagents are affected by the presence of emicizumab and are unsuitable for assessing residual/endogenous FVIII activity, they can be adjusted to reflect the concentration of emicizumab only. ¹⁰ When using CSA with bovine reagents, emicizumab will not be detected due to the antibody's inability to recognize non-human coagulation factors, whereas a chromogenic assay containing human FX and FIXa, such as the BIOPHEN FVIII:C assay (Hyphen Biomed, Neuville-sur-Oise, France), is responsive to emicizumab. Unlike mOSA, this CSA preactivates endogenous FVIII by thrombin, reducing the potency difference between FVIII and emicizumab. Recent studies have shown promising results in using CSA and mOSA in spiked samples, but less is known about real-world samples.^11–14

To address this evidence gap, we assess the impact of different coagulation assays on the reported concentration of emicizumab in patient material. We hypothesized that variations in assay methods would lead to differences in measured emicizumab concentrations differently correlating with clinical outcomes. In addition, given the limitations of standard coagulation assays in the presence of emicizumab, we studied a global hemostatic assay to provide an integrated assessment of thrombin generation and fibrin dynamics, and to explore whether overall hemostatic potential could offer complementary information beyond factor-based measurements which could further aid in patient evaluation and management. Understanding these assay-dependent variations is of importance for optimizing therapy and ensuring accurate monitoring of emicizumab levels in patients with hemophilia A.

Materials and Methods

Patients

Peripheral blood samples were collected as part of the clinical routine from PwHA followed up at the adult and pediatric Coagulation Department, Karolinska University Hospital, Stockholm, Sweden. Samples were identified at two intervals: an initial stage including 28 anonymized patient samples to evaluate assay methods (validation cohort), and a follow-up study with 16 samples drawn concomitantly with routine samples from 9 patients (clinical cohort) aged 6 years and older participating in a study (NCT02453542) on using global hemostatic methods to monitor and tailor therapy in patients with HA and inhibitors. In this study we did not obtain serial samples at specific timepoints but used consecutive samples for comparisons. Samples included in the validation cohort were pseudonymized and no written consent was collected in accordance with the ethical permit. All patients in the clinical cohort received written and oral information on the study and gave written consent prior to inclusion. For children aged 6–14 years written consent was provided by a legal guardian while children 15–17 years signed the written consent themselves. The Swedish Ethical Review Authority approved the study (validation cohort Dnr 215/03 and clinical cohort Dnr 01-003 with amendments 2015-02783 and 2019-02783) that was performed in accordance with the Helsinki Declaration.

Plasma Sampling

Samples were drawn into glass or plastic 3.2% sodium citrate tubes BD Vacutainer tubes (Becton Dickinson, Franklin Lakes, NJ, USA). Within one hour of collection, the samples were centrifuged once at 3000 × g for 10 min at 15 °C for samples drawn in plastic tubes, or at 2000 × g for 15 min at 15 °C for samples drawn in glass tubes, to obtain platelet-poor plasma. After centrifugation, the supernatant (platelet-poor plasma) was pipetted into a clean, labelled 0.4 ml Ellerman tubes (ThermoFisher Scientific, Waltham, Massachusetts, USA)).and frozen at −70°, and the samples were not freeze-thawed for multiple analyses.

Laboratory Assays for Emicizumab and Estimated FVIII (eFVIII) Concentration

Samples were thawed by total immersion in a water bath for 5 min at 37° and thoroughly mixed before analysis. The mOSA and the CSA were performed using the Siemens BCS XP (Siemens, Erlangen, Germany) for the validation cohort and the Sysmex CS-2500 (Sysmex, Kobe, Japan) for the clinical cohort. The CSA was performed using the BIOPHEN FVIII:C Hyphen Biomed, Neuville-sur-Oise, France) to quantify both emicizumab concentration and eFVIII and analyzed as a single point analysis. All reagents were diluted according to the manufacturer's instructions. The assay was calibrated separately for emicizumab concentration using emicizumab calibrator (R² Diagnostics Inc, South Bend, USA) and for eFVIII using the Biophen Plasma Calibrator (Hyphen Biomed, Neuville-sur-Oise, France). This approach allowed parallel quantification of both parameters using the same assay platform and sample dilution Calibration curves for eFVIII and emicizumab concentration are shown in Supplementary Figure 1A and B, respectively. The mOSA was performed using three different activated partial thromboplastin time (APTT) reagents, each with a distinct activator and phospholipid composition. Dade Actin FS (ellagic acid; soy and rabbit brain phospholipids, Dade Actin FSL (ellagic acid and soy phospholipids, both from Siemens Healthineers (Erlangen, Germany) and PTT-LA (silica and rabbit brain phospholipids (Diagnostica Stago (Asnières sur Seine, France). All mOSA were assessed using FVIII deficient plasma from Siemens Healthineers (Erlangen, Germany) and samples were diluted by 1:8. For both CSA and mOSA, calibration curves were generated up to a maximum concentration of 68.5 µg/mL to ensure precision and accuracy within the assay's dynamic range. In cases where the result was above the maximum concentration, samples were manually diluted by 1:2 and then reanalyzed. The clinical cohort was performed using the Sysmex CS-2500 system after transitioning from the BCS platform due to instrument replacement. Comparative control analyses confirmed consistency between platforms, and patient samples were analyzed with the same reagents on the new platform.

Overall Haemostatic Potential Assay

Overall haemostatic potential (OHP) was measured using fibrin aggregation curves. Citrated plasma was recalcified and treated with thrombin (Sigma-Aldrich, St Louis, Missouri, USA) at 0.04 U/mL, with or without tissue plasminogen activator (t-PA) (Boehringer Ingelheim, Ingelheim am Rheim, Germany) at 300 ng/mL. The absorbance changes at λ = 405 nm were monitored in real-time every 12 s for one hour using a Multiskan™ FC Microplate Photometer (ThermoFisher scientific, Waltham, Massachusetts, USA). Overall coagulation and hemostasis potential were derived from fibrin formation and fibrinolysis curves, respectively, with the difference representing overall fibrinolytic potential (OFP). Fibrin clot density was assessed using a modified clot turbidity assay. Parameters such as “lag-time” (clotting time), “max absorbance” (clot density), “max-Abs time” (time to plateau), and “slope” (fibrin polymerization rate) were determined from the turbidimetric curve for overall clotting potential (OCP), as described previously. ¹⁵ As an internal normal control, citrated plasma from a healthy volunteer was included in the OHP assay and analyzed under the same conditions as the patient samples, to provide a reference for normal hemostatic potential.

Statistical Methods

Statistical analysis was performed using R (version 4.3.1 (2023-06-16), R Foundation for Statistical Computing, Vienna, Austria). The normality of the continuous variables was evaluated using the Shapiro-Wilk test. Descriptive statistics for continuous variables were summarized as mean and standard deviation (SD) for normally distributed data. Differences in assay results within each cohort were assessed using repeated measures analysis of variance (ANOVA), followed by Bonferroni-corrected post-hoc paired t-tests to identify specific pairwise differences. The Pearson correlation coefficient (r) was calculated to assess the correlation between eFVIII activity, emicizumab concentration, and global hemostatic assays. An r value >0.9 was classified as a very strong correlation, 0.7–0.9 as a strong correlation, 0.5–0.7 as a moderate correlation, and ≤0.5 as a weak correlation. A p-value < .05 was considered statistically significant.

Results

Measurement of Emicizumab Concentration Using CSA and mOSA in the Validation Cohort

A total of 28 anonymized samples from patients receiving emicizumab were analyzed by the Siemens BCS XP instrument. The overall coefficient of variation (CV), including intra- and inter-assay variability for controls, was 3.6%, and the duplicate sample CV was 1.6% for emicizumab concentration measurements. Repeated measures ANOVA revealed a significant overall difference in concentrations of emicizumab between the four assays (F (3.81) = 31.8 P < .001), post-hoc pairwise comparisons with Bonferroni correction showed significantly lower emicizumab concentration with CSA compared to all three mOSA (P < .001 for mOSA Actin FSL, mOSA Actin FS and mOSA PTT-LA) Table 1 and Figure 1A. No significant differences were observed between the mOSA assays.

Table 1.

Results from Emicizumab Measurements in the Validation and Clinical Cohorts.

	CSA	mOSA
		Actin FSL	Actin FS	PTT-LA
Validation	46.5 (22.0-80.4; 13.8)	54.7 (22.5-88.1; 13.9)	55.4 (21.1-94.7; 16.9)	56.7 (28.4-90.8; 14.3)
Clinical	44.7 (11.3-86.2; 23.0)	47.3 (13.0-97.1; 22.3)	49.6 (12.6-93.3; 22.6)	48.4 (12.6-87.9; 21.2)

Emicizumab concentration in µg/mL was measured by chromogenic assay (CSA) and three different modified one-stage clotting assays (mOSA). The validation cohort was assessed on a Siemens BCS XP instrument, and the clinical cohort was measured on a Sysmex CS-2500 instrument. Results presented as mean (range; SD).

Figure 1.

Comparison of Emicizumab Concentration Measured with Chromogenic Assay (CSA) or Modified one-Stage Assays (Actin FSL, Actin FS and PTT-LA). A. Validation Cohort; B. Clinical Cohort. * p < 0.05 *** p < 0.001 **** p < 0.0001

We also examined the impact of calibration against emicizumab or eFVIII activity (eFVIII). In the cohort, the mean eFVIII was 38 IU/dL (SD 0.08). For CSA, a high correlation between emicizumab concentration and eFVIII was observed (Pearson coefficient (r) = 0.98), Supplemental Figure 2A. In the mOSA, a very strong correlation was seen between eFVIII and mOSA Actin FSL (r = 0.92) and mOSA PTT-LA (r = 0.88), while mOSA Actin FS had a strong correlation (r = 0.79), Supplemental Figure 3A–C.

Measurement of Emicizumab Concentration Using CSA and mOSA in the Clinical Cohort

Patient Characteristics

Nine patients with multiple blood samples drawn on different occasions were included in the study. Of those patients, seven were children, and two were adults when the treatment with emicizumab was initiated (median age 15 years (8-68)) and the blood samples drawn, Table 2. All patients had inhibitors against FVIII for at least eight years before treatment was started (median 12 years (range 8-35 years)). Seven patients were treated with factor concentrate, and two with a bypassing agent (activated prothrombin complex concentrate (aPCC)) before emicizumab treatment started. The median inhibitor titer at the start was 18 BU (range 2-1380 BU). All patients had received immunotolerance induction (ITI) prior to initiation of treatment (median ITI duration 10 years (4-14), missing data for one adult patient who had gone through multiple ITIs). Prior to treatment, the median annualized joint bleeding rate (AJBR) was 4.5 (0-28, three patients had 0), the Hemophilia Joint Health Score (HJHS) was 13 (0-28, two patients had 0), and the annualized bleeding ratio (ABR) 4.5 (0-30, one patient had 0). Following the initiation of treatment, the median scores were 0 (all patients), 13.5 (2-36), and 0 (all patients) for AJBR, HJHS, and ABR, respectively. The blood samples were drawn at a median 42 months (3-67 months) for the first sampling (n = 9), 53.5 months (12-72 months) for the second sampling (n = 6) and 74 months (n = 1) for the third sampling.

Table 2.

Clinical Cohort Description.

Pat	Age	Inhibitor Diagnosis	Age at Start Emi	Treatment at Start Emi	ABR Before Emi	ABR After Emi	HJHS Score Before Emi	HJHS Score After Emi	HEAD-US Before Emi	HEAD-US After Emi	First Evaluation*	Dosage at First Evaluation	Second Evaluation*	Dosage at Second Evaluation	Third Evaluation*	Dosage at Third Evaluation	Latest Emi-eFVIII (IU/dL)
1	68	N/A**	68	Elocta 4000U every other day	N/A	N/A	34	N/A	N/A	N/A	3	120 mg once weekly (1.4  mg/kg)					0.23
2	66	18	60	Feiba 3/wk and NovoSeven on demand	8	0	39	36	N/A	N/A	67	127 mg once weekly (1.5  mg/kg)	72	150 mg once weekly (1.9  mg/kg)	74	150 mg once weekly (1.9 mg/kg)	0.16
3	20	10	15	Elocta 6000U daily	0	0	11	17	15	15	48	105 mg, once weekly (1.5  mg/kg)	53	105 mg once weekly (1.4  mg/kg)			0.59
4	19	6	15	Elocta 3000U daily	1	0	0	2	2	3	34	105 mg, once weekly (1.4  mg/kg)	50	120 mg once weekly (1.6  mg/kg)			0.44
5	18	200	13	Elocta 3000U and NovoSeven 5 mg daily	27	0	13	16	10	10	46	105 mg, once weekly (1.8  mg/kg)					0.44
6	16	N/A**	15	Novoeight 1500U daily	0	0	9	7	18	20	5	80 mg, once weekly (1.5  mg/kg)	12	82 mg once weekly (1.5  mg/kg)			0.1
7	15	2	12	Elocta 3000U daily	1	0	9	10	4	12	4	95 mg, once weekly (1.8  mg/kg)					0.11
8	15	1000	10	Feiba 2000U 6/wk	30	0	25	18	13	14	45	63 mg, once weekly (1.5  mg/kg)	54	75 mg once weekly (1.7  mg/kg)			0.32
9	13	1380	8	Immunate 3000U × 7 F1000 U 3/wk	20	0	28	11	12	13	42	39 mg, once weekly (1.3 mg/kg)	55	105 mg once weekly (2.3 mg/kg)			0.35

Sixteen samples from the nine included patients were assessed on the Sysmex CS-2500. The overall coefficient of variation (CV), including intra- and inter-assay variability for controls, was 5.3% and 1.9% for low and high control, respectively. The duplicate sample CV was 1.58% for emicizumab concentration measurements. Similarly to the results from the validation cohort, repeated measured ANOVA showed significantly overall differences in emicizumab concentrations between the four assays (F(3.24 0 13.75), P < .0001). Bonferroni-corrected post-hoc test showed significantly lower emicizumab concentrations when analyzed with CSA compared to mOSA Actin FS (P < .001) and mOSA PTT-LA (P = .029) while no significant difference was observed between CSA and mOSA Actin FSL (P = .24), Figure 1B. No difference was observed between mOSA PTT-LA and mOSA Actin FSL or mOSA Actin FS. None of the patients had received any other treatment (ie factor VIII concentrate, bypassing agents) other than emicizumab at the time of the sampling.

The mean eFVIII was 39 IU/dL (SD 0.14) and showed a very strong correlation between CSA emicizumab concentration and eFVIII (r = 0.99), Supplemental Figure 2B. In contrast with the validation cohort, all three mOSA exhibited a very strong correlation between emicizumab concentration and eFVIII determined by CSA, with r = 0.98, 0.99, and 0.99 for mOSA Actin FSL, mOSA Actin FS, and mOSA PTT-LA, respectively, Supplemental Figure 3D–F.

Overall Haemostatic Potential

OHP was measured in samples from the clinical cohort (Table 3). OHP and OCP curves are shown for a sample from a healthy volunteer (non-PwHA) and for individuals with low (24.7 µg/mL), therapeutic (41.7 µg/mL), and high (83.7 µg/mL) emicizumab levels (Figure 2A–D). The comparison between these individuals in Figure 2E shows an increase and delay in the OHP, indicating an effect on fibrin generation. The low emicizumab level results in reduced OHP, while a therapeutic level demonstrates a balanced fibrin generation and degradation. For OCP (Figure 2F), a similar AUC is seen between the non-PwHA and therapeutic emicizumab level, while OCP is elevated at high emicizumab concentration and reduced at low levels. Four patients showed hypocoagulable profiles characterized by a reduced OHP (32-57 Abs sum, reference range 59-200 Abs sum) and increased OFP (73%-83%, reference range 31%-68%), although the OCP was normal in all four patients (190-221 Abs sum, reference range 159-352 Abs sum). Although variation in OHP and OCP curves were observed across different emicizumab concentrations (Figure 2A–F), no correlation was found within the cohort between eFVIII and OHP (r = 0.20), eFVIII and OFP (r = −0.08) or eFVIII and OCP (r = 0.27), Figure 3A–C, nor with emicizumab concentration (Supplemental Table 1).

Table 3.

Overall Heamostatic Potential (OHP) in the Clinical Cohort.

	Mean (Range, SD)	Reference Range
OHP (n = 16)
OHP	88 (32-141, 28)	59–200 Abs sum
OFP	62 (46-83, 10)	31% to 68%
OCP	232 (171-335, 36)	159–352 Abs sum

OCP - overall coagulation potential, and OFP - fibrinolytic potential. Results presented as mean (range, standard deviation)

Figure 2.

The Overall Hemostatic Potential (OHP) Curve Illustrates the Combined Effects of Fibrin Formation from Fibrinogen by Thrombin Generation, and Fibrin Digestion by Plasmin. The Overall Coagulation Potential (OCP) is Measured Without Added Tissue Plasminogen Activator (tPA), Representing Coagulation Capacity more Specifically. Each Absorbance (Abs) Value Reflects Fibrin Levels Over Time, with the Area Under the Curve (AUC) Indicating the Balance Between Fibrin Generation and Breakdown. From OCP and OHP, the Overall Fibrinolytic Potential (OFP) is Derived. A: OHP and OCP Curves from a Healthy Volunteer (non-PwHA), Included as a Normal Control, B: OHP and OCP in a Patient with Low Emicizumab (24.7 µg/mL), C: Mid-range emicizumab (41.7 µg/mL), D: High emicizumab (83.7 µg/mL). E: Overlapping OHP Results for A–D, F: Overlapping OCP Results for A–D.

Figure 3.

Correlation Between Estimated FVIII Levels (eFVIII) Measured with CSA and Overall Heamostatic Potential (A) Overall Fibrinolytic Potential (B) and Overall Coagulation Potential (C) in the Clinical Cohort. Dotted Lines Indicate the High and low cut-off for the Reference Range for Each Assay.

Discussion

Despite the efficacy of emicizumab in reducing bleeding in PwHA, its monitoring remains problematic. This study showed that while mOSA and CSA emicizumab concentrations correlated to eFVIII activity, using mOSA yielded higher emicizumab concentrations.

A mOSA is often recommended when assessing emicizumab, although these assays are impacted by emicizumab bypassing the rate-limiting steps in these assays.^8,9,16 On the other hand, CSA assays are less affected by pre-analytical variables and do not require FVIII deficient plasma, ¹⁷ but they are less available in clinical laboratories, often perceived as more expensive, and any residual FVIII will impact the results. In our validation cohort, CSA consistently yielded significantly lower emicizumab concentrations than all three mOSA methodologies. The mean relative difference between CSA and mOSA (Actin FSL, Actin FS and PTT-LA) were higher than expected total CV for emicizumab concentration measurement, with mOSA Actin FS and mOSA Actin FSL measuring emicizumab concentrations approximately 17% higher than CSA, and mOSA PTT-LA showing 21% higher levels than CSA. In line with the validation cohort, lower emicizumab concentrations were observed with the enzymatic method in the clinical cohort. However, the difference was only statistically significant compared to mOSA Actin FS and mOSA PTT-LA, but not mOSA Actin FSL. The smaller sample size and inclusion of individual-level variation may limit the interpretability of these results. The observed differences in measured emicizumab levels between assay platforms consistently exceeded the total coefficient of variation (CV) expected from analytical variability alone. Specifically, the total CV was 3.6% on the BCS XP and up to 5.3% on the Sysmex CS-2500. These findings underscore that the assay-dependent variation is of a magnitude that may influence clinical interpretation, particularly in settings where emicizumab concentrations are used to support therapeutic decisions. In the clinical context, those differences have to be viewed as a complement to the actual treatment effect, as demonstrated by efficacy in bleeding control and prevention. They can be highly relevant and helpful when evaluating discrepancies between the expected and the actual effect, as well as when composing a treatment plan prior to an invasive procedure. However, it must be kept in mind that statistical difference is not always interchangeable with clinical significance, and a rigorous patient evaluation must take place.

Several studies have shown a high correlation between eFVIII activity and emicizumab concentrations using mOSA and CSA.^8,16,18 However, due to the differences in properties between emicizumab and FVIII, it is not an accurate equivalence.^19,20 Animal studies on mice with severe hemophilia A and primate models with acquired hemophilia A showed that therapeutic doses of emicizumab correspond to an FVIII equivalent concentration of 10–20 IU/dL. ²¹ In the HAVEN clinical trials, the overall steady-state emicizumab concentration was 40–80 µg/ml, which correlated to eFVIII between 20.3–25.2 IU/dL.^3,22,23 In patients with suboptimal bleeding control who received an up-titration of emicizumab to 3 mg/kg once weekly an increase in both the steady-state emicizumab concentration as well as the eFVIII to mean 31.9 IU/dL (SD 13.0) was observed. ²³ In clinical samples, near linear correlations between emicizumab concentration and eFVIII activity have been seen in emicizumab concentration between 25–80 µg/ml, ¹⁸ similar to the steady state levels reported in the HAVEN trials.^1,3 The clinical study by Adamkewicz showed higher eFVIII activity (20-50 IU/dL) than those measured in the Haven trials. ¹⁸ We observed a strong correlation, particularly between both CSA and mOSA emicizumab concentration and eFVIII in both the validation and clinical cohort. While CSA may offer advantages in accurately quantifying emicizumab concentrations, mOSA could provide alternative options for emicizumab monitoring in clinical settings. In the clinical context, and particularly for clinicians who are not very familiar with emicizumab and other novel treatments, values such as eFVIII can be easier to understand and relate to, as well as to use for patient follow-up. Additionally, those values can be easier to compare, even in the cases they are not an accurate equivalent, since they function as a more widespread surrogate marker. It is, however, important to recognize the limitations of the markers and their correlations and put the laboratory values in setting of the particular clinical conditions.

Global hemostatic assays are suggested as a means to better understand the individual effect of emicizumab compared to conventional FVIII activity or emicizumab concentration assays,^24–26 primarily in patients undergoing surgery.^27,28 Thrombin generation assays and modified clot waveform analysis have been more extensively evaluated but less is known about OHP. The OHP assesses the balance of coagulation and fibrinolysis and can detect hypercoagulable states. In patients with hemophilia A, OHP is generally undetectable or at negligible levels. ²⁹ Furthermore, OHP can differentiate between bleeding phenotypes, with severe cases demonstrating the most profound reductions. ³⁰ Adding emicizumab in FVIII-deficient plasma has been shown to improve OHP. ³¹ Our results support that PwHA with emicizumab treatment have normalized or near normalized OHP. It is important to note that OHP, OCP, and OFP are not validated assays in this clinical setting, and their use in this context remains exploratory. However, they have been shown to correlate with other global hemostatic methods and routine assays in various cohorts.

An important aspect is whether the improved in vitro results will correlate with in vivo haemostatic response and whether they have a predictive value for clinical outcomes, similarly to the correlation of bleeding phenotypes and OHP levels. ³⁰ The patients in the clinical cohort with improved OHP did not experience any bleeding events; this should be further investigated in a larger cohort. However, the lack of correlation between OHP levels and emicizumab concentration indicates variability in individual response to emicizumab, particularly concerning the hemostatic potential. However, clinically, we observed a mostly consistent improvement in the scores used to evaluate joint health (Table 2), and no bleedings. Except for two patients who received recombinant FVIII on one occasion each, there was no need for additional treatment because of a bleeding since the initiation of treatment with emicizumab. This means, essentially, that the clinical benefits are present even if the laboratory response is not optimal and confirms the current practice of taking both the clinical and laboratory improvement into account when evaluating a drug effect. In the cases of patients two, six, and seven (Table 2), the low concentrations could be explained by factors such as compliance (missing dose) and additionally weight increase in one of the pediatric patients.

Conclusion

Accurately measuring emicizumab concentration can help to optimize prophylaxis in hemophilia A patients. However, its unique mechanism of action poses challenges in traditional coagulation assay evaluation. Our study assessed the impact of different coagulation assays on emicizumab concentration in forty-four patient samples using CSA and mOSA methods. Our results indicated significantly lower emicizumab concentrations with CSA compared to three mOSA methods. Global hemostatic assessments in nine patients from the clinical cohort revealed normal overall coagulation potential but increased overall fibrinolytic potential in four patients, reducing OHP. Due to assay-dependent variations, these findings underscore the need for standardized protocols in emicizumab assessment.

Our findings indicate that chromogenic-based approaches can reduce the risk of interference from emicizumab observed in one-stage aPTT-based assays. These insights contribute to optimizing emicizumab monitoring and ensuring effective hemophilia A management.