Steven W. Pipe, MD from the Department of Pediatrics and Pathology, University of Michigan speaks about 972 Relationship between Endogenous, Transgene FVIII Expression and Bleeding Events Following Valoctocogene Roxaparvovec Gene Transfer for Severe Hemophilia A: A Post-Hoc Analysis of the GENEr8-1 Phase 3 Trial.
Link to Abstract:
https://ash.confex.com/ash/2021/webprogram/Paper144508.html
a brief introduction –
The adeno-associated virus gene therapy valoctocogene roxaparvovec (AAV5-hFVIII-SQ) delivers a B-domain deleted factor VIII (FVIII) cDNA to hepatocytes. In 134 men with severe hemophilia A (HA), the open-label, single-arm, multicenter phase 3 GENEr8-1 study (NCT03370913) found that valoctocogene roxaparvovec reduced annualized treated bleeding rate (ABR) by 83.8 percent and was superior to FVIII prophylaxis (P 0.001). In congenital HA (van Uijl, et al. Haemophilia 2011;17[1]:41–4), a link between baseline FVIII activity and ABR was estimated; it is unknown if a similar relationship persists following gene transfer. Post-hoc assessments of transgene-derived FVIII activity and bleeding in GENEr8-1 are shown below.
Techniques –
Men 18 years of age with severe HA who had previously been treated with FVIII prophylaxis and were negative for FVIII inhibitors and anti-AAV5 antibodies received a 6×1013 vg/kg infusion of the valoctocogene roxaparvovec. The median FVIII activity in each 4- or 6-week window was examined using chromogenic substrate (CSA; lower limit of quantification [LLOQ], 3.0 IU/dL) and one stage assays (OSA; LLOQ, 1.0 IU/dL). After week 4, when standard prophylaxis was to discontinue, self-reported treated bleeds were counted. Negative binomial regression was used to model the connection between the number of treated joint bleeds and matched median FVIII activity levels in each 4- or 6-week timeframe.
Observations –
The average follow-up was 71.6 weeks as of the data cutoff date; one participant was lost to follow-up at week 66. 9 percent (12/134) had CSA FVIII activity 3 IU/dL, 3% (4/134) had median FVIII activity 3–5 IU/dL, 17 percent (23/134) had median FVIII activity 5–15 IU/dL, and 71 percent (95/134) had median FVIII activity 15 IU/dL at week 49–52 (latest time with data for all participants; intent-to-treat population
32 percent of subjects (43/134) had an ABR of 0 when on FVIII prophylaxis prior to gene transfer. 75 percent of subjects (101/134) were bleed-free through their last follow-up before the data cut following gene transfer. The remaining 33 patients reported a total of 149 treated bleeds, with 62% (93/149) being traumatic and 38% (56/149) being spontaneous. Joints accounted for 53 percent (79/149), muscle for 19 percent (28/149), soft tissue for 14 percent (21/149), and other or undefined areas for 14 percent (21/149).
When CSA FVIII was 3 IU/dL (LLOQ), 54 percent of treated bleeds (80/149) occurred, 12 percent (18/149) occurred when FVIII was 3–5 IU/dL, 23 percent (35/149) occurred when FVIII was 5–15 IU/dL, and 11 percent (16/149) occurred when FVIII was 15 IU/dL. Thirteen of the 16 treated bleeding occurred when FVIII was less than 15 IU/dL.
Treatment-related joint bleeding followed a similar pattern: 51% (40/79) occurred when FVIII was 3 IU/dL, 15% (12/79) when FVIII was 3–5 IU/dL, 27% (21/79) when FVIII was 5–15 IU/dL, and 8% (6/79) when FVIII was 15 IU/dL. For people treated with valoctocogene roxaparvovec with FVIII activity 15 IU/dL, a negative binomial regression model based on these data and matched FVIII activity levels predicts 1 treated joint bleed in 2 years (CSA; Figure).
The CSA’s clinical use at low FVIII levels is restricted by its 3 IU/dL LLOQ. At weeks 49–52, nine of the 12 participants with median FVIII levels below the CSA LLOQ had improved or had the same ABR after gene therapy compared to prophylaxis. With an LLOQ of 1 IU/dL, the OSA offered crucial information. 1 had FVIII 1 IU/dL, 5 had FVIII 1–5 IU/dL, and 3 had FVIII 5 IU/dL, according to OSA. The remaining three subjects with higher ABR following gene transfer had FVIII levels of 0 IU/dL, 2.1 IU/dL, and 4.8 IU/dL, respectively, according to OSA. The median (IQR) ABR was 1.2 (0–7.9) for participants with OS FVIII 5 IU/dL at week 52 (n = 11), which was similar to the median (IQR) ABR of 1.6 (0.6–3.5) observed for those with moderate HA. J Throm Haemost 2020;18[12]:3203–10) (Abdi, et al., J Throm Haemost 2020;18[12]:3203–10)
Final thoughts –
In this phase 3 experiment, gene transfer with the valoctocogene roxaparvovec resulted in prolonged endogenous FVIII production and lower ABR. The majority of bleeds after gene transfer were traumatic. Reports of treated bleeding, especially joint bleeds, were uncommon when FVIII levels were less than 15 IU/dL. CSA FVIII activity, like epidemiologic congenital HA findings, was predictive of bleeding risk after gene transfer. The OSA gave clinically useful information at FVIII levels below the CSA LLOQ; bleeding with OSA FVIII 5 IU/dL was similar to observations in persons with moderate HA, albeit more research with more data is needed.