Search for a command to run...
Haemophilia B is an X-linked recessive disorder in which affected individuals have deficient and/or dysfunctional coagulation factor IX (FIX),1 resulting in recurrent bleeding episodes, which have the potential to be life-threatening. Molecular genotyping in a large cohort, published by Johnsen in 2017, shows that a significant proportion of individuals with haemophilia B have missense variants, which leads to circulating, dysfunctional FIX,2 making these individuals cross-reactive material (CRM) positive. Previous works have shown that the majority of FIX is bound to type IV collagen (Col4) in the subendothelial extracellular matrix,3, 4 which maintains a dynamic equilibrium with plasma FIX (FIXp).5 Preclinical experiments suggest that extravascular FIX (FIXEV) contributes to maintaining haemostasis in the absence of FIXp, thus helping protect against micro bleeding events.6 Murine haemophilia B models have shown that haemostatic efficacy of FIX products is reduced at a faster rate in CRM+ mice as compared with CRM− mice.5 It may be that defective endogenous FIX is competing with therapeutic FIX for binding Col4 in the extravascular space, resulting in a higher plasma recovery and reduced availability of FIXEV to support haemostasis.3 Extravascular distribution has therefore been suggested as a factor to consider when tailoring patient therapy. Currently, there are several factor replacement therapies available for patients, including recombinant factor IX (rFIX) and the extended half-life (EHL) factors rFIX Fc fusion protein (rFIXFc), rFIX albumin fusion protein (rIX-FP), and nonacog beta pegol (N9-GP). rFIX and rFIXFc bind to Col4 with high affinity in murine models and display higher volumes of distribution at steady state (Vss) and lower incremental recovery (IR) compared with rIX-FP and N9-GP, suggesting that the latter may have limited access to extravascular Col4.6 rFIXFc was the first EHL FIX therapy approved and is composed of a single molecule of rFIX fused to the Fc domain of immunoglobulin G1, which confers a longer half-life as compared with standard half-life (SHL) FIX products.7 The safety, efficacy, and pharmacokinetics of rFIXFc were investigated in the B-LONG study (NCT01027364). This was a Phase 3, open-label, multicentre study in male subjects with haemophilia B (endogenous FIX level of ≤2 IU/dl [≤2%]) aged ≥12 years with ≥100 exposure days and no history of inhibitors.8 B-LONG had four treatment groups: (1) weekly prophylaxis, starting at 50 IU/kg rFIXFc dose and adjustments made as needed (ensuring a target trough of 1%–3% above baseline or higher); (2) individualized prophylaxis, starting at 100 IU/kg rFIXFc every 10 days and adjusting as needed (ensuring target trough levels of 1%–3% above baseline or higher); (3) on-demand treatment at a dose of 20–100 IU/kg rFIXFc for acute bleeds; and (4) perioperative management with rFIXFc.8 The maximum dose was defined as 100 IU/kg for individuals not undergoing surgery. A subgroup of subjects receiving weekly prophylaxis also underwent a comparative sequential pharmacokinetic analysis between rFIX and rFIXFc.8 Described here are the results of a post hoc analysis of the B-LONG study population that aimed to determine if pharmacokinetic parameters of rFIXFc at baseline are influenced by endogenous FIX antigen levels. Firstly, descriptive statistics were executed on the baseline characteristics of the population, including those who were treated on demand or who were on prophylaxis during the study (excluding individuals in the perioperative management arm). Correlation analyses between pharmacokinetic parameters of Vss, IR, and terminal half-life, obtained from non-compartmental analyses and FIX antigen levels, stratified either by baseline FIX activity level (<1% [<1 IU/dl] or 1%–2% [1–2 IU/dl]) or variant status (missense [CRM+] or non-missense [CRM−]), were performed. FIX antigen levels were determined at baseline by enzyme-linked immunosorbent assay (goat anti-human FIX antibody [Affinity Biologicals]), and FIX activity assessments were made with the activated partial thromboplastin time one-stage assay (aPTT; TriniCLOT, Trinity Biotech) following a washout period (120 h). Data from a single visit (Visit 2 [baseline visit] of B-LONG) were utilized, and linear correlation was assessed by Pearson's correlation coefficient (r) with significance testing. No imputations were made, and where FIX antigen values were below the lower limit of quantification (n = 13), they were included as ‘0′. The final part of the post hoc analysis involved a comparison of annualized bleed rate (ABR) for those on prophylaxis according to variant status. ABR and p-value were obtained using a negative binomial regression model with number of bleeding events as the response and variant status of subjects on prophylaxis as the independent variable. The natural logarithm of follow-up time (in years) was used as the offset value. A total of 117 subjects were included in this post hoc analysis. The median (range) age and weight were 29 (12–71) years and 72 (45–187) kg, respectively. Subjects had a median (range) of .0 (.0–8.0) and 11.0 (.0–100.0) target joints at baseline and bleeds per subject in the 12 months prior to study enrolment, respectively. The number (%) of subjects with FIX activity levels of <1 and 1 to 2 IU/dl were 97 (83%) and 20 (17%), respectively. FIX antigen levels were recorded for 106 subjects, with the median (range) value being 96.0 (0–6287) ng/ml. For subjects with a missense variant (n = 60/106), the median (range) value was 378.0 (0–6287) ng/ml and, as expected, FIX antigen was lower in subjects with non-missense variants (n = 46/106) at 45.5 (0–1572) ng/ml. These data indicate we have adequate sample size for both missense and non-missense variants. Results from the non-compartmental analyses showed no correlation between FIX antigen level and any of three assessed pharmacokinetic parameters (Vss, IR, and terminal half-life) when stratified by FIX activity level (Figure 1A–C) or variant status (Figure 1D–F). The adequacy of the pharmacokinetic data range was demonstrated by a significant inverse correlation between Vss and IR in this population (data not shown). The efficacy of rFIXFc was also investigated to determine whether there is a difference in ABR between missense and non-missense subjects who were on rFIXFc prophylaxis. Overall, age, number of target joints per subject, and number of bleeds within the 12 months prior to enrolment were similar between the missense and non-missense groups at baseline (data not shown). Of the 60 subjects with a missense variant, 51 were on prophylaxis and had data available for comparison, and 36 of the 46 subjects with a non-missense variant were on prophylaxis and had data available (Table 1). ABR (95% confidence interval [CI]) values were 3.15 (2.33, 4.25) and 2.54 (1.76, 3.66) for the missense and non-missense groups, respectively. An ABR ratio (95% CI) of 1.24 (.77, 1.99) was determined for the missense group, when considering the non-missense group as reference (p-value = .3726). Annualized joint bleed rate (AjBR) values were 2.20 (1.53, 3.19) and 2.12 (1.37, 3.29) for the missense and non-missense groups, respectively. An AjBR ratio (95% CI) of 1.04 (.59, 1.84) was determined for the missense group, when considering the non-missense group as reference (p-value = .8992). Overall, there was no significant difference in clinical efficacy of rFIXFc, as evidenced by ABR or AjBR, in subjects with missense or non-missense F9 variants. It should be noted that the correlation analyses for FIX antigen levels versus pharmacokinetic parameters presented here were only carried out at one point in time, using data from a single visit since it was the only visit where all subjects were required to have the relevant pharmacokinetic samples taken. Moreover, the findings discussed here may not be applicable to other populations: for example, subjects outside of the age or weight range included in this assessment. Finally, the results presented here are only applicable to rFIXFc, which shares similar volume of distribution to SHL FIX, and may not be applicable to other FIX replacement therapies. Results presented here for rFIXFc do not support the findings from murine haemophilia B models, whereby the haemostatic efficacy of FIX products that display increased extravascular distribution is reduced at a faster rate in CRM+ mice as compared with CRM− mice.5 Importantly, this also appears to be the case when considering clinical efficacy (i.e., there was no significant difference in ABR in those with missense variants compared with non-missense variants). Overall, these findings suggest that CRM status alone does not necessarily influence the pharmacokinetics of rFIXFc and that measuring FIXp antigen levels is not warranted when considering rFIXFc therapy for people with haemophilia B arising from missense variants of the F9 gene. This study was funded by Sanofi (Cambridge, MA) and Swedish Orphan Biovitrum AB (Sobi; Stockholm, Sweden). Medical writing support for the development of this manuscript was provided by Naziyat Khan, PhD, and Ashleigh Pulkoski-Gross, PhD, CMPP, of Fishawack Communications Ltd., part of Fishawack Health, and was funded by Sanofi and Sobi. Robert Sidonio Jr has consulted for Sanofi, Sobi, Bayer, Hema Biologics, Octapharma, Genentech, Roche, Guardian Therapeutics, Star Therapeutics, Novo Nordisk, Pfizer, Grifols, Uniqure. IIS: Takeda, Octpaharma and Genentech. Sandra Casiano, Arman Altincatal, and Suresh Katragadda are employees of Sanofi and may hold shares and/or stock options in the company. Aletta Falk and Elena Santagostino are employees of Sobi and may hold shares and/or stock options in the company. Jerzy Windyga has received research support or lectures honoraria from Amgen, Bayer, CSL Behring, Novartis, Novo Nordisk, Octapharma, Roche, Sanofi, Siemens, Sobi, Swixx Biopharma, and Takeda. Qualified researchers may request access to data and related documents. Any patient-level data will be anonymized, and study documents will be redacted, including to protect the privacy of trial participants. Further details on Sanofi's data sharing criteria, eligible studies, and process for requesting access can be found at https://vivli.org/. The B-LONG study was registered at clinicaltrials.gov as NCT02234310. The original study required investigators to obtain ethics committee approval of the study protocol, all amemdments, informed consent and other study documents. The original study was perfomed in accordance with the Declaration of Helsinki and all local regulations. No additional study approval required for this post hoc analysis.