Characteristics of Bleed-Free Patients on Every-5-Day Dosing in the Protect VIII (BAY 94-9027) Study

Background: BAY 94‐9027 is a B-domain-deleted recombinant factor VIII (FVIII) that is site-specifically PEGylated with a 60-kDa (2 x 30-kDa) polyethylene glycol (PEG) to extend its half-life. The efficacy and safety of BAY 94-9027 as prophylactic and on-demand therapy for patients with severe hemophilia A were demonstrated in the phase II/III PROTECT VIII trial, in which patients could receive prophylaxis every 7 days (E7D), every 5 days (E5D), or twice-weekly (2×W). Patients in the E5D arm had a low median annualized bleeding rate (ABR) of 1.9. This post hoc analysis was conducted to identify clinical predictors for 'best responders' (patients with ABR = 0) during E5D prophylaxis with BAY 94-9027 in PROTECT VIII.

Methods: PROTECT VIII was a partially randomized, open-label trial of 134 males aged 12-65 years with severe hemophilia A (FVIII < 1%) and ≥ 150 FVIII exposure days. Prophylaxis patients received BAY 94-9027 25 IU/kg 2×W for a 10-week run-in period. Patients with ≤ 1 spontaneous joint or muscle bleed during this period were randomized to 45‒60 IU/kg E5D or 60 IU/kg E7D for the main 26-week study period; patients enrolling after the randomization arms were full, or with ≥ 2 bleeds in the run-in period, received 30-40 IU/kg 2×W. All patients randomized to the E5D arm (n = 43) remained on this dose-frequency throughout the main study. This post hoc analysis presents a description of baseline characteristics of best responders to E5D dosing, defined as patients with ABR = 0 during the main study period; the study was not powered to support a formal statistical comparison.

Results: Of 43 patients randomized to E5D dosing, 24 (55.8%) had ABR ≥ 1 and 19 (44.2%) had ABR = 0, meeting the criterion for best responders. The median (95% CI) age was slightly higher for best responders, 38 years [30.0; 47.0] versus 31.5 [24.0; 41.0]), than patients with ABR ≥ 1. In the 12 months prior to the study, best responders had lower median (95% CI) bleeds (2.0 [0.0; 5.0] versus 10.0 [3.0; 21.0]) and joint bleeds (0.0 [0.0; 2.0] versus 8.0 [2.0; 12.0]) compared with patients with ABR ≥ 1. Most patients in both groups had previously received prophylaxis prior to study (84.2% and 70.8%, respectively). Fewer best responders had target joints (57.9%, versus 70.8% of patients with ABR ≥ 1); and those who did have target joints only 36.4% had ≥ 1 target joint, versus 47.1% of patients with ABR ≥ 1.

Conclusions: Overall, the 43 patients receiving BAY 94-9027 E5D prophylaxis had a low median ABR of 1.9. The 19 of these patients who had an ABR of 0 (best responders) had numerically fewer bleeds in the 12 months prior to treatment initiation, particularly joint bleeds, and numerically fewer target joints, vs patients with ABR ≥ 1. Though requiring validation in a powered study, taken together, these results suggest the applicability of E5D BAY 94-9027 dosing across patient profiles, and that patients with a similar bleed and joint profile to those analyzed here may be able to achieve ABR of 0 with E5D BAY 94-9027.