Page 134 - 2021_07-Haematologica-web

A. Tiede et al.
each of the five predicted FVIII:C categories for each patient, using negative binomial regression and predictions of a paramet- ric model. Relationships between ABR and mean FVIII:C were evaluated for the overall dataset (pivotal and extension trials), and by trial phase (pivotal vs. extension) and patients’ age (adults/adolescents [≥12 years] vs. children [0-11 years]). The proportion of time in which patients were predicted to achieve a FVIII:C level was calculated by taking the patient-years of exposure (PYE) for a FVIII:C category range, dividing it by the overall PYE, and multiplying it by 100%.
The risk of bleeding episodes and FVIII:C levels was also ana- lyzed using survival analysis. For each patient and trial phase, the time on a given factor level to the first bleed was calculated. The time to first bleed is presented in Kaplan-Meier plots dis- playing the proportion of patients without a bleed at a given time. For each trial phase, the difference between the FVIII:C levels was tested using a log-rank test for statistical significance.
Results
Patients and exposure to turoctocog alfa
The demographics of the PK subgroups in the pivotal trials (n=50) and of the patients participating in these trials and from whom post-dose FVIII:C measurements were used in the analysis are shown in Table 1. The analysis population, which was limited to patients receiving prophylaxis, comprised 232 patients across the pivotal and extension trials. One patient was excluded from the analysis as over 50% of bleeds recorded by this patient lacked a time recording, leaving 231 patients (63 children, 168 adults). Total PYE for adults/adolescents and children in the pivotal trials were 77 and 23, respectively, and 498 and 217, respectively, in the extension trial.
Population pharmacokinetic modeling
The total pool of PK data in the population PK model comprised the PK profiles from the PK subgroups in the pivotal trials supple- mented with post-dose measurements from routine clinical visits for patients participating in these trials (n=231). The PK of turoc- tocog alfa was found to align well with a one-compartmental model (i.e., the elimination of turoctocog alfa after a single intra- venous bolus dose was approximately log-linear) as described by the equation:
where D is dosage, t is the time since dose, V is the volume of distribution, CL is clearance, k is the elimination rate constant, and e is the base of the natural logarithm (~2.72). One-compartmental behavior is indicative of the compound being distributed mainly within the bloodstream itself. The population PK parameter esti- mates are shown in Table 2.12 Therefore, the one-compartmental model was selected as the basis for the population PK analysis, using turoctocog alfa data derived from the one-stage clot assay for all patients.12
The volume of distribution for an individual of weight W, V(W), was calculated according to the equation below, where Wref repre- sents the body weight of the reference individual (70 kg):
Clearance for an individual with a weight W and age A (CL(W, A)), was calculated according to the formula below, where Aref rep- resents the age of the reference individual age (20 years):
Predicted factor VIII activity - bleeding relationship
Overall, there were 1,237 spontaneous bleeds (324 during the pivotal phase and 913 during the extension phase), of which 1,063 occurred in adults/adolescents and 174 in children (Table 3). The vast majority (n=1,055; 85.29%) of spontaneous bleeds were joint bleeds; more adults/adolescents than children suf- fered joint bleeds. The proportion of time at FVIII:C levels >1% and the PYE that were used to calculate this parameter across patient populations and trial phases are shown in Table 3. It was predicted that overall FVIII:C levels ≥1% were achieved for 85.64% of the time.
Mean spontaneous bleeds, including joint bleeds, decreased as FVIII:C increased (Tables 4 and 5), indicating an exposure–bleeding relationship. This relationship was evident in both the pivotal and extension trial phases for both types of bleed. However, for each FVIII:C activity category, lower ABR were observed in the exten- sion phase than in the pivotal phase. A FVIII:C–bleeding relation- ship was also apparent for the two age groups of adults/adoles- cents and children, although children had a lower ABR than adults/adolescents within each FVIII:C activity category (Table 5).
At low FVIII:C, ABR were lower during the extension phase than in the pivotal phase for both adults/adolescents and chil- dren (Figure 2). The difference in ABR between adults/adoles- cents and children was particularly evident for the pivotal phase; during the extension phase, the ABR for each age group was similar for spontaneous bleeds, with slightly lower ABR in chil- dren compared with adults/adolescents. Similar relationships between FVIII:C and ABR were evident for joint bleeds (Online Supplementary Figure S1).
Using Kaplan-Meier plots, analyzed over a fixed period of time (60 days), the time to first spontaneous bleeding was longer during times at higher FVIII:C levels (Figure 3). A similar pattern was apparent with joint bleeds (Online Supplementary Figure S2). Hence, patients were more protected during times at higher FVIII:C levels, particularly during the extension phase. In gener- al, patients remained bleed-free for longer at any level of FVIII:C in the extension phase, particularly at the lower FVIII:C levels. Log-rank calculations showed a significant difference between each FVIII:C category in the proportion of patients without a bleed during the pivotal phase (P<0.0001) and the extension phase (P<0.0001).
.
Table 1. Demographics of patients included in the exposure–response analysis.
Adults/adolescents
Children
PK All*
subgroup
28 63
5.96 6.08
Parameters
Number of patients
Age on entering
program, years;
mean (SD)
Body weight,
kg; mean (SD)
PK subgroup
22
24.00
(7.88)
71.84
(12.44)
All*
168
28.98
(12.15)
73.5
(18.13)
(2.76)
(2.91)
24.43 24.6
(10.50)
(10.03)
*Adults/adolescents and children from the pivotal trials or patients who completed the phase I pharmacokinetic trials could enter the extension trial. Patients were excluded from the exposure–response analysis if they switched from prophylaxis to on-demand treatment during the extension trial, received only on-demand treatment throughout the extension trial, had missing diary returns, or failed to document time of bleeding for >50% of recorded bleeds.PK:pharmacokinetic;SD:standard deviation.
1904
haematologica | 2021; 106(7)