Background: Rivipansel is a pan-selectin inhibitor in development for the treatment of sickle cell disease (SCD) vaso-occlusive crisis (VOC). A single phase 3 study was planned to evaluate the efficacy and safety of rivipansel (GMI-1070) in the treatment of VOC in hospitalized patients with SCD. Selection of the dose to be used in this phase 3 study was based on the efficacy, safety, and pharmacokinetic (PK) results from the rivipansel phase 2 study. Modeling and Simulation (M&S) was used to determine the dosing for adults and children aged >6 years with SCD. To ensure that the predicted drug exposure is actually achieved in patients aged 6-11 years (cohort 2), we conducted a blinded review of the PK data obtained from several subjects enrolled in cohort 2. The methodology for this evaluation is presented here.
Methods: Pharmacokinetic data for 109 subjects receiving doses of 2 to 40 mg/kg in 4 previous studies (3 Phase 1 and 1 Phase 2) were integrated to build a 3-compartment model, which was used to perform simulations to aid dose selection. As rivipansel is almost entirely renally excreted, the simulations of clearance considered renal function as well as hyperfiltration, as this type of altered renal filtration is common in SCD patients. This model allowed for scaling of the PK exposure from the model developed in adults and older children (aged 12-17 years) (Tammara BK, Harnisch LO. CPT Pharmacometrics Syst Pharmacol 2017;6:845-54) into the model for younger children (aged 6-11 years). Simulations from the model in the adult and pediatric population are presented in Figure 1.
To confirm the predicted exposure in cohort 2, in particular their average concentration at steady-state (Cavg,ss), age- and weight-dependent cumulative distribution functions (CDFs) of parameters in the population PK model (eg, CL Vss) were derived for the first 6 actively treated children. During this blinded review, concentration time profiles for these children were generated and empirical Bayesian estimates (EBEs) for their corresponding Cavg,ss were derived. The percentiles in which the EBEs fell within the simulated CDFs from the model were used to guide decisions for a potential dose adjustment. This process is outlined in the flow chart in Figure 2 and in the percentile plot diagram shown in Figure 3.
Results: The PK data from the Phase 2 study supported fixed flat dosing for patient aged 12 years and older. Pharmacokinetic M&S predicted that a loading dose of 1680 mg followed by a maintenance dose of 840 mg every 12 hours would result in exposures similar to those observed with the lower dose used in the Phase 2 study, such that a minimum plasma concentration >10 μg/mL would be maintained throughout the dosing interval. For patients in cohort 2, weight-based dosing was considered, as it is the most conservative approach, given that children in this age range have not been studied previously. For each dosing regimen tested, the simulated demographic target distribution in pediatric SCD patients was used to derive concentration-time profiles. A summary of the Cavg,ss distribution across different ages is shown in Figure 1. The simulations showed that for children aged 6-11 years, a 40-mg/kg loading dose (maximum 1680 mg) followed by 20 mg/kg every 12 hours (maximum 840 mg) would likely result in concentrations similar to those observed with the 20/10 mg/kg dosing used in adults in the Phase 2 study. To confirm the validity of the exposure predictions, PK assessments were performed in the Phase 3 study and a blinded interim analysis of pediatric exposures was conducted.
Conclusions: A population PK model for rivipansel has been developed, applicable to the whole target SCD population of adults and children older than 12 years. Its applicability in children aged 6-11 years is under investigation as part of the ongoing Phase 3 study, and interim assessments of the PK exposure have been done in small cohorts of 6 patients. The corresponding decision framework has been implemented, exercised successfully, with the results of potential dose adaptations to be published once the Phase 3 study has been concluded.
Tammara:Pfizer Inc.: Employment. Shafer:Pfizer Inc.: Employment. Harnisch:Pfizer Inc.: Employment.
Asterisk with author names denotes non-ASH members.