Andrew Brandon1, Hinke Huisman-Siebinga2, Shelby Barnett1, Kayode Ogungbenro3, Alwin Huitema2,4,5, Gareth Veal1
1Newcastle University Centre for Cancer, Newcastle University, 2Department of Pharmacy and Pharmacology, The Netherlands Cancer Institute, 3Centre for Applied Pharmacokinetic Research, University of Manchester, 4Department of Pharmacology, Princess Máxima Center for Pediatric Oncology, 5Department of Clinical Pharmacy, University Medical Center Utrecht
Objectives: Mitoxantrone is a well-established chemotherapy drug used to treat paediatric acute myeloid leukaemia (AML) [1]. Both adults and older children are typically dosed at 10-12 mg/m², with reduced mg/kg dosing applied to infants and small children. Such chemotherapy dose reductions in the youngest patients commonly lack standardisation and are supported by limited scientific rationale [2]. Information on the pharmacokinetics (PK) of mitoxantrone in children is lacking, and its safety and effectiveness in a paediatric setting have not been formally described [3]. Previous studies reporting mitoxantrone PK in children have commonly been multi-drug studies involving few patients [4,5]. In a recent review of cytotoxic drugs in neonates and infants, it was concluded that evidence-based dosing guidance could not be provided for mitoxantrone due to a lack of published data [6]. There is a clear need for a more data-informed assessment of mitoxantrone dosing in childhood cancer. Here we present the development of a mitoxantrone population PK model in paediatrics and explore the relationships between dosing approach, PK and exposure, with the aim to inform better prescribing practice for this patient group. Methods: Data were generated from a phase III clinical trial pharmacology sub-study in children with AML, where patients received mitoxantrone at either 12 mg/m²/day or 0.4 mg/kg/day (for children under 12 months old, weighing =10 kg, or with a BSA of <0.5 m²). Plasma concentrations from 44 patients (313 samples) were analysed, with associated patient characteristics, haematology/biochemistry and toxicity data collected. Patients received IV infusions of mitoxantrone (~1 h) each day for 3-4 days and were sampled from 0-24 hours post-first infusion, as well as 48- and 72-hours post-end of final day infusion. Due to the pharmacokinetic profile and extent of late-timepoint sampling, >40% of the data were below the lower limit of quantification (BLLOQ). A population PK model was developed using NONMEM (version 7.5), utilising the M3 method for handling BLLOQ data [7]. Model evaluation included a combination of objective function value assessment, informed assessment of final parameter estimates, goodness-of-fit plots, visual predictive checks, and sampling importance resampling. Normalised prediction distribution error (NPDE) evaluation was extended to BLLOQ data [8]. Individual clearance rates were used to calculate mitoxantrone area under the curve (AUC) for each patient and the relationships between dosing regimen, PK and exposure were assessed. Results: A two-compartment model with inter-individual variability included on clearance (CL) and volume of distribution of the second compartment (V2) best described the data. Allometric scaling was included, with weight effect on clearances and volumes fixed to 0.75 and 1, respectively and scaled to the population median body weight of 27.5 kg. A combined (additive and proportional) residual error model was included, with additive error fixed to 2.5 ng/mL, reflecting 50% of the assay LLOQ. Stochastic approximation expectation maximisation (SAEM) estimation was used, with the M3 method used to handle BLLOQ data. V1 was fixed to an existing literature value (23.2 L) to improve model stability [8]. Final population parameter estimates were CL 37.1 L/h (RSE 11%); inter-compartmental clearance (Q) 29.1 L/h (RSE 14%); V2 118 L (RSE 25%); IIV on CL 64.6% (RSE 27%); IIV on V2 108% (RSE 29%). Patients receiving the reduced mg/kg (n=3) dosing regimen obtained lower AUCs compared to the mg/m² (n=41) group (mg/m² = 332 ± 146 µg/L.h; mg/kg = 160 ± 44 µg/L.h), while CL scaled to body surface area was comparable between groups (mg/m² = 42.4 ± 18.5 L/h/m²; mg/kg = 53.4 ± 12.0 L/h/m²), indicating potential under-dosing of the mg/kg group. Conclusions: Mitoxantrone PK in children is described here using the largest studied paediatric cohort to date, increasing understanding of its properties in this patient group. Younger children on a reduced dosing regimen of 0.4 mg/kg/day had lower mitoxantrone AUCs than those receiving 12 mg/m²/day dosing, potentially placing them at risk of suboptimal drug exposure. These findings support those of previous studies calling for updated, evidence-based chemotherapy dosing guidelines for neonates and infants [2,6]. The influence of mitoxantrone PK and exposure on toxicity and clinical response in this patient cohort is now being investigated.
[1] Zarnegar-Lumley S, Caldwell KJ, Rubnitz JE. Relapsed acute myeloid leukemia in children and adolescents: current treatment options and future strategies. Leukemia. 2022;36(8):1951-1960. [2] Veal GJ, Boddy AV. Chemotherapy in newborns and preterm babies. Semin Fetal Neonatal Med. 2012;17(4):243-248. [3] BC Cancer Drug Manual. Mitoxantrone monograph. Published online May 2019. Accessed November 26, 2024. http://www.bccancer.bc.ca/drug-database-site/Drug%20Index/Mitoxantrone_monograph.pdf [4] Lacayo N, Lum B, Becton D, et al. Pharmacokinetic interactions of cyclosporine with etoposide and mitoxantrone in children with acute myeloid leukemia. Leukemia. 2002;16(5):920-927. [5] O’Brien MM, Lacayo NJ, Lum BL, et al. Phase I study of valspodar (PSC-833) with mitoxantrone and etoposide in refractory and relapsed pediatric acute leukemia: A report from the Children’s Oncology Group. Pediatric Blood Cancer. 2010;54(5):694-702. [6] Nijstad AL, Barnett S, Lalmohamed A, et al. Clinical pharmacology of cytotoxic drugs in neonates and infants: Providing evidence-based dosing guidance. Eur J Cancer. 2022;164:137-154. [7] Bergstrand M, Karlsson MO. Handling Data Below the Limit of Quantification in Mixed Effect Models. AAPS J. 2009;11(2):371-380. [8] Nguyen THT, Comets E, Mentré F. Extension of NPDE for evaluation of nonlinear mixed effect models in presence of data below the quantification limit with applications to HIV dynamic model. Journal of Pharmacokinetics and Pharmacodynamics. 2012;39(5):499-518.
Reference: PAGE 33 (2025) Abstr 11355 [www.page-meeting.org/?abstract=11355]
Poster: Drug/Disease Modelling - Paediatrics