Population PK analysis of Pegaspargase in Japanese and non-Japanese patients with acute lymphoblastic leukemia
Chayan Acharya (1), Céline Sarr (1), Adrien Tessier (2)
(1) Pharmetheus, Uppsala, Sweden (2) Clinical Pharmacometrics - Quantitative Pharmacology, Servier, Saclay, France
Introduction:
Asparaginase is an important agent in the treatment of acute lymphoblastic leukemia. PEG-L-asparaginase (Pegaspargase) is Escherichia coli L-asparaginase (L-ASP) coated with covalently linked large polyethylene glycol (PEG) polymers. The pharmacokinetic (PK) of Pegaspargase is characterized by measurement of asparaginase activity (AA).
Objectives:
- Characterize the PK of AA
- Assess the impact of Japanese population in the PK of AA.
- Illustrate PK behavior of AA via simulations.
Methods:
Data from four studies (AALL07P4, CCG-1962, DFCI-11-001 and SHP674-201) were used to develop a population PK model for AA. In studies AALL07P4 and DFCI-11-001, Pegaspargase 2500 U/m2 was administered via intravenous (IV) infusion, while in study CCG-1962, Pegaspargase 2500 U/m2 was administered via a single intramuscular (IM) injection. In study SHP674-201, Pegaspargase 2500 U/m2 or 82.5 U/kg was administered via IV infusion.
A legacy PK model [1] using the data from studies AALL07P4, CCG-1962 and DFCI-11-001 was used as the starting point of the PK analysis of AA. The legacy model was a 1-compartment model with first-order absorption with two bioavailability (F) estimates for the first and later doses for the IM route of administration, and combined linear first-order and saturable elimination. The model included the following covariate-parameter relationships: BSA on CL, Vmax and Vc and Japanese population on CL.
Following the evaluation of the predictive performance of the legacy model using the analysis data, a base model was obtained after removing the Japanese patient effect on CL from the legacy model to re-evaluate the effect of Japanese population on CL, Km and Vc. Following a covariate search on the base model, a final AA model was obtained. The final AA model was used to simulate the proposed dosing recommendation based on BSA and assess % of patients with AA ≥ 0.1 U/mL after 14 days and 25 days post first dose.
Results:
The final AA model was similar to the legacy model, except one covariate. In addition to the covariates in the legacy model, the final AA model also included Japanese effect on Km. Model diagnostics for the final model indicated an acceptable predictive performance for AA, supporting its usability to derive individual predictions of exposure in a similar population.
The typical values of CL and Km were estimated to be 0.00238 L/h and 80.5 U/L, respectively, while the same in the Japanese patients were estimated to be 0.00141 L/h and 14.1 U/L, respectively. A simulation-based analysis demonstrated that 41% lower CL and 83% lower Km in the Japanese patient would result in 18.7% and 20.1% higher area under the activity-time curve up to day 25 (AUC0-25d) and day 46 (AUC0-46d) after single dose (SiD), respectively. Forest plots illustrated that Cmax after a SiD was similar in both Japanese and non-Japanese patients. Although, the Japanese patients may achieve higher AUC0-25d after a SiD compared to the non-Japanese patients, it was not considered clinically meaningful as the median change was included in 80-125% intervals.
Forest plot also showed that, given a fixed BSA adjusted dose level of 2500 U/m2, both Cmax and AUC0-25d after a SiD may differ in patients with higher or lower BSA compared to the reference patient with BSA of 0.835 m2.
Simulations based on the final AA model and relevant distributions of BSA for Japanese and US populations, where dose levels were based on age and BSA values, showed that more than 97% of the patients in Japanese and US population had AA of ≥ 0.1 U/mL.
Conclusion:
- The legacy AA model was successfully updated in application to the current analysis data set.
- The final AA model predicted that the Japanese patients would have 41% lower CL and 83% lower Km than that of the non-Japanese patients.
- The effects of Japanese population and lyophilized formulation of Pegaspargase were confounded in the current analysis data. Hence, the impact of Japanese covariate should be interpreted carefully.
- There is no clinically meaningful difference in Cmax and AUC0-25d after a SiD between the non-Japanese and Japanese patients.
- Simulations indicated that after the first and the seventh dose of Pegaspargase, all population subgroups in both Japanese and US populations had more than 97% of the patients with AA ≥ 0.1 U/mL.
References:
[1] Population PK report. REP-1-SER-ONC-PMX-1. Population PK analysis of Oncaspar in Japanese and non-Japanese patients with acute lymphoblastic leukemia. Pharmetheus; 2022.
