This post with its dialogue between Dr. Uljon and Dr. Sadeghi-Khomami describes how to minimize EHL measurement discrepancies. Read to the end!
Here’s a post from Sacha N. Uljon MD PhD, Massachusetts General Hospital (MGH) and Brigham and Women’s Hospital (BWH), Boston, US, describing their experience with Altuviiio measurement using a one-stage clot-based assay and a chromogenic substrate assay. Dr. Uljon references our 3-14-24 discussion “How to Measure Plasma Efanesoctocog Alfa (Altuviiio)” and writes:
Hello everyone. I’m posting our experience hoping to pool data. We obtained the manufacturer’s standards and made a calibration curve. We use the Coamatic Kit on the ACL TOP 750 for our chromogenic assay and HemosIL Recombiplastin 2G on the ACL TOP 750 for the clot-based assay. MGH/BWH labs have merged, so we are now running all specimens here at MGH—we are getting both pediatric and adult Altuviiio specimens. Here is the comparison data for the standards:
|
Altuviiio (%): X
|
FVIII clot based (%)
|
chromogenic (%)
|
|
100
|
59.7
|
244
|
|
80
|
48.4
|
194
|
|
20
|
16.2
|
52
|
|
5
|
4.8
|
13
|
|
||||||||||||||||||||||||||||||||||
This is going to keep happening as the gene therapy FVIII molecules are also modified and may require specialized testing.
Individual patients may vary in their response to factor VIII, demonstrating different half-lives and recoveries. It is recommended to use a validated one-stage clotting assay to determine plasma factor VIII activity of efanesoctocog alfa. Throughout the clinical development, an Actin FSL-based one-stage clotting assay was used. According to the findings of a comparative analysis of clinical study samples, results obtained using a chromogenic assay should be divided by 2.5 to approximate the patient’s factor VIII activity. In addition, a field study comparing different aPTT reagents indicated approximately 2.5-fold higher factor VIII activity levels when using Actin-FS instead of Actin-FSL in the one-stage clotting assay and approximately 30% lower results when using SynthASil. Therefore, there is a potential for dosing errors based on the assay type used for the measurement of FVIII levels. Relevant safety information concerning monitoring laboratory tests with the chromogenic assay or the one-stage clotting assay are included in SmPC sections 4.2 and 4.4.
From Dave McGlasson 12-6-24:
- If assessment of plasma FVIII activity is needed, it is recommended to use a validated one-stage clotting assay. The ALTUVIIIO FVIII activity level is overestimated by the chromogenic assay and a specific ellagic acid-based APTT reagent in a one-stage clotting assay by approximately 2.5-fold. If those assays are used, divide the result by 2.5 to approximate the patient’s ALTUVIIIO Factor VIII activity level.
- Use of a reference laboratory is recommended when a qualified one-stage clotting assay or chromogenic assay is not available locally.
- Monitor for the development of FVIII inhibitors. When bleeding is not controlled with ALTUVIIIO and the expected factor VIII activity plasma levels are not attained, perform an assay to determine if FVIII inhibitors are present (use Bethesda Units to titer inhibitors).
See Dr. Uljon’s comment below; she adds this comment on 12-7-24: Thank you! Unfortunately, it’s the same advice we already had–calibrate a single stage or use chromogenic and divide. Neither works. It may be that if clinical labs are not running both assays on the same sample, they have no idea that there is a problem using these correction factors.
Additional discussion beginning with Dr. Uljon on 12-11-24: Our hematologists just got back from ASH and said that other labs are using the chromogenic assay with the correction factor. It seems based on the info here that this is not a good idea for monitoring half-life and may lead to worries about inhibitor development when measured too long after Altuviiio administration. There is no way labs would know this if they were just following the instructions from Sanofi, run the chromogenic assay and divide–which honestly troubles me.
This figure appears in Ryu JH, Bauer KA, Schulman S. Periprocedural management of type 2N von Willebrand disease with efanesoctocog alfa. J Thromb Haemost. 2023;21:3508–10. doi: 10.1016/j.jtha.2023.09.009.
From George: The data are taken from a VWD Type 2 Normandy variant patient. The Altuviiio generates a slight increase in VWF concentration and activity, I suspect what is being measured is the VWF fragment portion of the Altuviiio molecule. The graph supports Ali’s observation that the OSA and CSA values draw close over 50 hours past administration. Do either of you speculate that FVIII activity kinetics may be different in VWD Type 2 Normandy from true hemophilia?
From Dr. Uljohn: Altuviiio shouldn’t generate an increase in VWF activity or antigen by itself since the antibody in the Ag assays isn’t to that part of VWF. I think what they are doing here is bypassing the need for VWF to bind to FVIII, which is the primary defect in 2N
Dr. Ali responds: The origin of the elevated vWF (both antigen and activity) observed in this patient remains unclear. Could it be stress induced by participation in the study? Without control subjects, it’s difficult to conclude. One possible explanation, consistent with the CSA/OSA decline, is that the D’D3 domain may have been released from ALTUVIIIO and detected in these assays as vWF. However, this remains speculative. This study was conducted on a single 2N-vWD patient with one replicate measurement, limiting its value as strong evidence. It underscores the need for a larger dataset to explore the OSA and CSA relationship over time in real-world hemophilia A patient scenarios. See also Pipe S, Sadeghi-Khomami A, Konkle BA, et al. A global comparative field study to evaluate the factor VIII activity of efanesoctocog alfa by one-stage clotting and chromogenic substrate assays at clinical haemostasis laboratories. Haemophilia. 2024;30:214–23. doi: 10.1111/hae.14831.
Ah-HA! that makes perfect sense.
I looked at a few patients and this holds true–the tests converge over time.
It would have been kind to have put that in the package insert in the first place, knowing that labs all over would be getting repeat specimens from clinicians worried about half lives and inhibitor formation. Thank you Fritsma Factor and Dr. Ali Sadeghi-Khomami
A reference OSA is used for ALTUVIIIO potency assignment, clinical trials, and dose recommendations. Consequently, converting CSA results to the reference OSA equivalent is necessary.
In clinical samples, the CSA conversion factor is expected to align most closely with the reference OSA when blood samples are collected shortly after ALTUVIIIO injection. This is because circulating proteases, such as thrombin, can gradually detach the protecting peptides (XTEN-Fc-D’D3 vWF) from FVIII molecules over time, resulting in a time-dependent decrease in the CSA-to-OSA conversion factor (see Figures JTH-2023-3508, JTH-2024-214 for a visual representation).
This phenomenon also explains why chromogenic assays, which contain excess external thrombin in their reagents, tend to overestimate ALTUVIIIO potency by 2–3 times compared to the potency assigned by the reference OSA procedure.
Accurately predicting these conversion rates over time in patients, if necessary, would require a robust dataset, including pharmacokinetic profiles, distribution compartments, half-life, and other key parameters. For these reasons, I suggested reaching out to Sanofi’s medical experts for specific guidelines and recommendations regarding ALTUVIIIO assay.
This is very helpful. Can you clarify: “CSA conversion factor in clinical samples decreases as the time between injection and blood sampling increases.” Can you explain why?
None of the specimens listed were on the same day as injection- in some cases, the patients self-inject, so we can’t know for sure.
I don’t think we exceed the onboard stability for the chromogenic assay but I will look at the SOP.
Thanks again for your help with this. If we can figure out how the correction factor changes with time, we could compensate.
Thank you for your clarification. I was not questioning the integrity of the plasma samples. My query was whether the blood samples from the three patients were collected immediately after ALTUVIIIO injection. It wouldn’t be surprising if the CSA conversion factor in clinical samples decreases as the time between injection and blood sampling increases. As for Coamatic (CSA) variation, while it is not a 510(k)-cleared product on the ACL-TOP system, reproducible results can only be achieved within a short time after reagent reconstitution. A formulation flaw could lead to time-dependent result drift caused by reagent instability.
Hello, thank you so much for taking the time to reply.
Truthfully, I know nothing about the proprietary assay–these were run offline as a favor because of my real concern for patient safety.
For a while, Sanofi was offering a free test through Esoterix (I can’t find the linked website now–the instructions were to open an account and send samples to them–this was where I obtained contact info originally.) My assumption was that this was their reference assay and should therefore be regarded as correct.
Regarding our clot-based assay: The FVIII assay is PTT-based–I had incorrectly cut and pasted from our website. It’s HemosIL SynthASil on ACL TOP 750. We have a link to our reagents here in case you are curious
Regarding sample integrity, Our satellite sites spin and decant and send us the plasma on dry ice and we have several aliquots. So, although there were days between the measurements, all 3 had a single freeze-thaw cycle, (we used a “fresh” aliquot for the send-outs) and we store at -70. So we can’t explain it based on specimen integrity. There is a formal possibility that there is a difference between aliquots because of an underfilled/overfilled tube at the draw site. However, this discrepancy has happened regardless of where the patients are drawn.
We have hematologists with connections at Sanofi that are certainly closer to the “top” than mine, and I did ask them to bring this up at ASH. But it really is my job, so I am asking as many places as possible. If no one else has the discrepancy between the chromogenic and clot-based, at least we know we need to fix something in house.
In truth, I don’t love our chromogenic assay with Altuviiio. We once had a tube measure 56% and a few minutes later the very same tube measured approximately 100%, both with no error message. Bubble? Who knows. At least with the clot-based one gets a few dilutions to compare as a reality check.
Dear Dr. Sacha Uljon,
I have a few questions regarding your study:
1. Could you clarify what the proprietary assay listed in the table above refers to? I would appreciate it if you could elaborate further.
2. Why did you choose a PT-based FVIII assay instead of an aPTT-based assay for your study?
3. What was the time difference between ALTUVIIIO injection and blood sampling for patients 1, 2, and 3?
You made an excellent point regarding the assay variations observed in the ALTUVIIIO field study (Haemophilia-2024-214), noting that these data were not derived from clinical samples. I agree that further complementary studies are needed to evaluate these in vitro findings in patient samples. Have you considered reaching out to the Sanofi Medical Advisory Team? They may have post-marketing surveillance data, a larger clinical comparison dataset between OSA and CSA, or additional recommendations to share with you. Your Sanofi representative or account manager should be able to assist in connecting you.
Thank you for your responses. We did exactly what was recommended. We “validated” or one-stage assay (see graph) and we also used a chromogenic assay. My point is that the number you get when you divide the chromogenic by 2.5 is nowhere near the number you get when using the validated one-stage clotting assay. When compare to the altuviiio-specific proprietary assay, the clotting looks better but there are patients for whom neither value is correct (see #3 above).
The correction factors suggested were based on spiking studies, not patient samples- so maybe it shouldn’t surprise me that the calibration doesn’t work–even the spiking studies were all over the map (lab dependent, even WITHIN the same reagent/system). There is no way a hospital can get a fast TAT from an assay that is offered centrally.
Does anyone else run both the chromogenic and clot-based on each specimen? How do yours compare? Because we would never even have known there was a problem if we weren’t running both.