The Irish Personalized Approach to the Treatment of Haemophilia (iPATH) - Determinants of Inter-Individual Variation in FVIII Pharmacokinetics

Introduction:

Standard factor VIII (FVIII) prophylaxis aims to minimise bleeding episodes in people with haemophilia A (PWHA) using a weight based dosing strategy. However, it is well recognised that FVIII half-life (FVIII t1/2) varies markedly between individual PWHA. Consequently, some PWHA continue to experience repeated bleeding episodes even after commencing standard dose prophylaxis therapy. In essence, these patients 'bleed their way' to eventually reaching their required optimal FVIII prophylactic regimen. Given that even a small number of joint bleeds have been shown to represent a risk factor for hemophilic arthropathy, there is a clear unmet need for the development of personalised treatment approaches for PWHA that include consideration of individual FVIII pharmacokinetics (PK) to guide prophylaxis dosing. With the introduction of limited plasma FVIII sampling and Bayesian analysis, evaluation of FVIII pharmacokinetics (PK) are now possible in routine clinical practice. However, real life clinical experience to date has been limited to small cohorts of PWHA. Critically, the influence of PK in modulating clinical phenotype has not been systematically studied. As part of the Irish Personalised Approach to the Treatment of Haemophilia (iPATH) study, we integrated extensive clinical data and individual PK profiles to investigate modulators of FVIII PK, the role of limited versus extended sampling FVIII PK and the impact of PK on FVIII prophylaxis and clinical phenotype.

Methods:

Written informed consent was obtained prior to recruitment to the iPATH study. All PK samples were obtained following administration of a single rFVIII product (antihemophilic factor [recombinant]; ADVATE®, Shire, Lexington, MA, USA) and using myPKFiT™ software (Shire, Lexington, MA, USA). Utilising limited sampling (2 timed FVIII:C levels, taken at 3‒6 and 24‒32 hours post rFVIII administration) and Bayesian analysis, PK curves were constructed using myPKFiT™. For each individual the FVIII t1/2, Von Willebrand Factor antigen (VWF:Ag), blood group, weight, age at PK and Haemophilia Joint Health Score (HJHS) was recorded. rFVIII prescribed over the preceding 5 years was analysed; all patients used antihemophilic factor (recombinant) throughout this time period. Mann-Whitney analyses were used for non-parametric comparisons and correlations were performed using the Pearson test on Prism 7.0c for Mac OSX.

Results:

Complete PK profiles were available for 53 patients with FVIII deficiency recruited to the iPATH study (3 moderate, 50 severe FVIII deficiency). FVIII t1/2 varied considerably, ranging from 7.7‒20.1 hours (median=11.4 hours). Patients with blood group O had a significantly shorter FVIII t1/2 (n=31, median=10.9 hours) than non-O patients (n=22, median=12.2 hours, p=0.014). Both increased age (r²=0.2932, p<0.0001) and higher plasma VWF:Ag levels (r²=0.3, p<0.0001) at time of PK were associated with a significantly longer FVIII t1/2. HJHS significantly increased with age (r²=0.494, p<0.0001) but no significant correlation between FVIII t1/2 and HJHS was observed, likely reflecting the multifactorial nature of arthropathy in PWHA.

In a subset of 10 PWHA we compared the FVIII t1/2 results generated from extended (10 point) to those from limited sampling (2 point) PK. Each limited sampling study included a 24 hour time point as well as either a 3, 4 or 6 hour sample. The 10 point results significantly correlated with limited sampling results using either the 3/24 hour (r²=0.9801, p<0.0001), 4/24 hour (r²=0.9762, p<0.0001) or 6/24 hour sampling combination (r²=0.9693, p<0.0001), underscoring the clinical utility of limited PK sampling.

Prior to the introduction of routine PK-guided dosing in 2018, prescribed rFVIII dose was altered according to bleed rate at clinical review. Analysing the prescribed rFVIII dose over the past 5 years, a clear relationship to rFVIII t1/2 was identified (r²=0.3517, p<0.0001), suggesting that bleeding events in PWHA guided them towards their PK dose.

Conclusions:

This study highlights the accuracy of simplified vs extended sampling for generation of PK profiles on antihemophilic factor (recombinant) using myPKFiT™. Bleed rate adjusted prophylaxis was found to correlate with FVIII t1/2 in an Irish population of PWHA, suggesting that early evaluation of PK and implementation of PK guided dosing of prophylaxis may be of benefit in minimising bleeds in PWHA.