Introduction: Leukemia represents around 30% of cancer diagnoses among children aged less than 15 years [1][2]. Among childhood leukemias, 75% of cases are attributed to acute lymphoblastic leukemia (ALL) and the remainder to acute myeloid leukemia (AML) [3]. Isatuximab (Sarclisa®) is an IgG1 monoclonal antibody that targets a specific epitope on CD38 and is approved for treatment of relapsed and refractory (R/R) multiple myeloma in adults. Isatuximab has demonstrated significant anti-leukemic activity in preclinical models; therefore, it is being tested as combination therapy for pediatric patients with R/R ALL and AML in the ISAKIDS study. The study is ongoing for children and adolescents treated with isatuximab (20 mg/kg). Before study initiation, the dose was extrapolated from adult clearance and allometric scaling based on weight, and selected by matching the exposure (area under the curve) in adults with ALL receiving 20 mg/kg weekly for 4 administrations and every other week after that. The de-risking strategy implied that the first cohort of patients enrolled were ≥2 years old.
Objectives: The aim of this work was to confirm the dose in pediatric patients ≥2 years old and to help select the appropriate dose to be administered in infants (1–24 months old).
Methods: ISAKIDS (NCT03860844) is a Phase 2, single-arm, multicenter, open-label study evaluating the antitumor activity, safety, and pharmacokinetics (PK) of isatuximab in combination with standard salvage chemotherapies in children with R/R leukemia in first or second relapse. Based on an interim analysis of the first 18 pediatric patients (2–17 years old) enrolled, two approaches were used, including the following: 1) an empirical data-driven approach using population PK modeling, pooling data from adult (n=14) and pediatric (n=18) patients with ALL, and body weight allometric scaling; and 2) a physiology-based PK (PBPK) approach leveraging adult ALL data (for model calibration) and model prediction in children. The population PK analysis was performed using the stochastic approximation expectation-maximization algorithm for nonlinear mixed-effects models implemented in MONOLIX. The PBPK modeling was done using PK-Sim® (Open Systems Pharmacology). PK simulations stratified by body weight (or age) groups and according to different dosing scenarios were performed using the two approaches.
Results: For the population PK approach, a two-compartment PK disposition and linear elimination model with allometric coefficients fixed to 0.85 for clearance and 1.0 for volume of distribution parameters enabled the description of observed adult and pediatric data. Age-based maturation function was not necessary to include in the population PK model. The PBPK model is characterized by a two-pore distribution, including competition with endogenous IgG and Fc receptor (FcRn) recycling [4]. FcRn recycling was calibrated on adult data by estimating the FcRn dissociation constant. When translated to pediatric physiology, the PBPK model was found to describe pediatric data well. Both population PK and PBPK modeling yielded consistent results. Comparable exposure was achieved in patients ≥2 years old. A slight underexposure (less than 20% median decrease) was predicted if a 20 mg/kg dose was administered in pediatric patients 1–24 months old (4–12 kg) compared with adults at the same dose (51–100 kg).
Conclusion: The modeling and simulation strategy has driven dose selection in pediatric patients. In the ISAKIDS study, enrollment opened for patients 1–24 months old using the current dosing regimen for older patients. A dose of 20 mg/kg was selected for all age groups based on confirmation of comparable exposure between adults and children ≥2 years old and the prediction of slightly lower drug exposure in infants, which does not require a dose adjustment. Modeling and simulation enabled de-risking of the dose selection and the ability to pursue additional pediatric investigation in the youngest patient age group.
Claire Brillac, Donato Teutonico, Corina Oprea, Dorothée Sémiond and Laurent Nguyen
https://www.page-meeting.org/default.asp?abstract=9966
Introduction: Leukemia represents around 30% of cancer diagnoses among children aged less than 15 years [1][2]. Among childhood leukemias, 75% of cases are attributed to acute lymphoblastic leukemia (ALL) and the remainder to acute myeloid leukemia (AML) [3]. Isatuximab (Sarclisa®) is an IgG1 monoclonal antibody that targets a specific epitope on CD38 and is approved for treatment of relapsed and refractory (R/R) multiple myeloma in adults. Isatuximab has demonstrated significant anti-leukemic activity in preclinical models; therefore, it is being tested as combination therapy for pediatric patients with R/R ALL and AML in the ISAKIDS study. The study is ongoing for children and adolescents treated with isatuximab (20 mg/kg). Before study initiation, the dose was extrapolated from adult clearance and allometric scaling based on weight, and selected by matching the exposure (area under the curve) in adults with ALL receiving 20 mg/kg weekly for 4 administrations and every other week after that. The de-risking strategy implied that the first cohort of patients enrolled were ≥2 years old.
Objectives: The aim of this work was to confirm the dose in pediatric patients ≥2 years old and to help select the appropriate dose to be administered in infants (1–24 months old).
Methods: ISAKIDS (NCT03860844) is a Phase 2, single-arm, multicenter, open-label study evaluating the antitumor activity, safety, and pharmacokinetics (PK) of isatuximab in combination with standard salvage chemotherapies in children with R/R leukemia in first or second relapse. Based on an interim analysis of the first 18 pediatric patients (2–17 years old) enrolled, two approaches were used, including the following: 1) an empirical data-driven approach using population PK modeling, pooling data from adult (n=14) and pediatric (n=18) patients with ALL, and body weight allometric scaling; and 2) a physiology-based PK (PBPK) approach leveraging adult ALL data (for model calibration) and model prediction in children. The population PK analysis was performed using the stochastic approximation expectation-maximization algorithm for nonlinear mixed-effects models implemented in MONOLIX. The PBPK modeling was done using PK-Sim® (Open Systems Pharmacology). PK simulations stratified by body weight (or age) groups and according to different dosing scenarios were performed using the two approaches.
Results: For the population PK approach, a two-compartment PK disposition and linear elimination model with allometric coefficients fixed to 0.85 for clearance and 1.0 for volume of distribution parameters enabled the description of observed adult and pediatric data. Age-based maturation function was not necessary to include in the population PK model. The PBPK model is characterized by a two-pore distribution, including competition with endogenous IgG and Fc receptor (FcRn) recycling [4]. FcRn recycling was calibrated on adult data by estimating the FcRn dissociation constant. When translated to pediatric physiology, the PBPK model was found to describe pediatric data well. Both population PK and PBPK modeling yielded consistent results. Comparable exposure was achieved in patients ≥2 years old. A slight underexposure (less than 20% median decrease) was predicted if a 20 mg/kg dose was administered in pediatric patients 1–24 months old (4–12 kg) compared with adults at the same dose (51–100 kg).
Conclusion: The modeling and simulation strategy has driven dose selection in pediatric patients. In the ISAKIDS study, enrollment opened for patients 1–24 months old using the current dosing regimen for older patients. A dose of 20 mg/kg was selected for all age groups based on confirmation of comparable exposure between adults and children ≥2 years old and the prediction of slightly lower drug exposure in infants, which does not require a dose adjustment. Modeling and simulation enabled de-risking of the dose selection and the ability to pursue additional pediatric investigation in the youngest patient age group.