Population Pharmacokinetics Profiling to Tailor Prophylaxis in Hemophilia a Patients: A Pilot Feasibility Study

Background. Assessing the individual pharmacokinetics (PK) profile by Bayesian estimation on a limited set of post infusion samples has been proposed as an efficient way to optimize prophylactic dosing of factor concentrates in persons living with hemophilia. Limited evidence exists on the actual impact of adopting this strategy in clinical practice. Aim of this study was to assess the feasibility of performing a clinical study testing the impact of a population PK based approach to prophylaxis.

Methods. Consecutive severe hemophilia A patients (factor VIII plasma activity level below 0.01 IU/mL) referring to the Munich University Hospital were screened for eligibility. Consenting patients underwent a limited sampling PK assessment (using a population PK modelling approach), and individual treatment recommendations were prepared combining PK-based regimen optimization with lifestyle and patients' preferences. A specific textual and graphical treatment recommendation sheet was provided to each patient, defining the risk of bleeding for each day of the week on the base of the predicted plasma factor concentrate activity levels; specifically, days with plasma factor activity levels above 0.15 IU/mL were defined as low risk of bleeding (green color coding); days between 0.15 and 0.03 IU/mL as intermediate risk of bleeding (yellow color coding), and below 0.03 IU/mL as high risk of bleeding (red color coding). During a follow-up period of up to 12 months, information about bleeding events and patient Health Related Quality of Life (HRQoL) was collected. A sparse sampling protocol was employed to draw confirmatory post estimation samples. The accuracy of the individual PK profile and dose optimization process was assessed by comparing predicted and prospectively measured levels; two assessments were used: the proportion of confirmatory samples falling in a matching risk-defined time window; and the average deviation of observed from expected factor VIII plasma activity levels. The study protocol was approved by the Munich University Hospital Ethical Review Board.

Results. One hundred severe hemophilia A patients were screened, of which 37 were enrolled in the study. Median age was 41 years (range 16 - 62); median body weight 78 kg (range 58 - 121). Ten patients were treated with Advate, 11 with Beriate, 2 with Fanhdi, 3 with Haemoctin, 8 with Kogenate, and 1 each with Haemate P, Immunate, and Refacto AF. No patient had a history of inhibitor development. The HJHS at study enrollment was 23 (range, 2-49). The bleeding rate was available for 29 patients; the median number of joint bleed per patient was 4 (range, 0-30) during the 12 months before study entry and 1 (range, 0-11) during the study; the median monthly bleeding rate was 0.3 (range, 0-2.5) before and 0.1 (range, 0-1.2) during the study, with a median absolute reduction of -0.3 (range, -1.9-0.1) bleeds per patient per month. Hemo-QoL-A scores median was 78.9 (range 32.7-100) before and 84.5 (range, 38.6-100) during the study, with a median increase of 0.8 (range, -9.2 to 34.5). Ninety confirmatory plasma samples were drawn during the study, and 91% of the observations was in the appropriate bleeding-risk time window. No measurement below 0.03 IU/mL was observed outside of the predicted high risk of bleeding time window. The mean (SD) difference of observed versus expected plasma factor activity levels was -1.9 (4.8) IU/mL; the median (IQR) difference was -1 (-4.4 to 0.5).

Conclusions. This pilot study showed the feasibility of implementing tailored prophylaxis using a Bayesian forecasting approach in a routine clinical practice setting. Reduction in bleeds and measurable impact on quality of life were observed. The accuracy of the Bayesian prediction approach was very high for all studied concentrates and associated population PK models.