Darren Crowley 1, Jack Gleeson 2,3, Harriet Bennett-Lenane 1, Brendan Griffin 1, Maria Donovan 1
1 School of Pharmacy, University College Cork (Cork, Ireland), 2 Cancer Research @UCC, College of Medicine and Health, University College Cork (Cork, Ireland), 3 CUH/UCC Cancer Centre, Cork University Hospital (Cork, Ireland)
Objectives
Glioblastoma multiforme (GBM) is a poor prognosis cancer, with studies suggesting that only 4% of patients survive beyond five years¹. Temozolomide (TMZ) is the current standard of care for systemic treatment for GBM, which has shown a survival benefit for these patients². However, dose-limiting haematological toxicities necessitate treatment interruption or discontinuation for some patients, which may impact the benefit of TMZ³. TMZ is administered to patients in two phases, namely, the concomitant phase and adjuvant phase. During the concomitant phase, patients receive focal radiotherapy and daily TMZ for 42 days. Four weeks post the concomitant phase, patients enter the adjuvant phase, where they receive up to six cycles of a higher dose of TMZ³.
In this study, we aimed to perform a population pharmacokinetic (PK) analysis of TMZ to evaluate the impact of covariates on key PK parameters such as absorption rate (Ka) volume of distribution (Vd) and clearance (Cl). We also wanted to assess whether there was a connection between individuals with these covariates and differing TMZ concentrations.
Methods
Adult patients diagnosed with a high-grade glioma (WHO grade 3 or grade 4 astrocytoma, oligodendroglioma or glioblastoma) being treated with TMZ in Cork University Hospital, Cork, Ireland, were enrolled. TMZ plasma concentrations were quantitatively measured using a validated HPLC method in University College Cork. TMZ was administered orally at a dose of 75mg/m² daily over a six-week period during the concomitant phase of treatment. Blood sampling was performed up to seven hours post dose at three different timepoints over the period of treatment. Nonlinear mixed-effects (NLME) was used to perform the popPK analysis, with the PK model and covariate search being conducted using Phoenix NLME modelling software (version 8.6.1.6; Certara L.P., St. Louis, MO, USA). Covariates were chosen to be included in the stepwise covariate search by visual examination of the population covariate box plots and population covariate plots. Covariates tested included age, sex, body surface area, weight, height, labarotory parameters, tumour side, type and grade. Using the base structural model as a baseline, an increase in the accuracy of the model after covariate addition was determined by assessment of the negative value of two times the log-likelihood (-2LL). P-values were used to determine the -2LL values at which a covariate should be retained or removed. In this study, the P-value for adding a covariate was set at 0.05 and the P-value for removing a covariate was set at 0.01.
Results
The PK analysis dataset included 27 plasma concentration versus time data points obtained from 11 patients. A one compartment model with first order elimination best described the observed data, with observed versus predicted concentration plots lying near the line of unity. On examination of 21 potential covariates that could influence TMZ PK parameters, the effect of age was found to significantly influence TMZ clearance (p < 0.05). The cone-shaped pattern of the conditional weighted residual (CWRes) versus predicted concentration plot flattened after covariate inclusion, showing improved consistency in residuals and model fit. Upon reviewing Eta versus covariate plots, increasing age was found to reduce the estimated clearance value for individuals. The final model estimated a typical absorption rate of 1.21/hr, volume of distribution of 32.1L and clearance of 11.7L/hr. Other covariates that were indicated to influence PK parameters but not to a significant level were: White blood cell count on Ka, height, creatinine clearance, glioma type and tumour side on Vd, and neutrophil count on Cl.
Conclusions
Age was found to be a covariate that significantly influenced the clearance of TMZ. Older patients were predicted to have reduced predicted clearance values. Future studies are required to see if other potential covariates influence PK parameters of TMZ and if they lead to a clinically significant difference in TMZ concentrations and toxicity rate.
References:
[1] National Cancer Registry Ireland. Cancer Trends - Brain. Cancer Trends. 2015
[2] Stupp R, Mason WP, Van Den Bent MJ, Weller M, Fisher B, Taphoorn MJB, et al. Radiotherapy plus Concomitant and Adjuvant Temozolomide for Glioblastoma [Internet]. 2005. Available from: www.nejm.org​
[3] CHMP. Temodal Summary of Product Characteristics. 1999
Reference: PAGE 34 (2026) Abstr 11971 [www.page-meeting.org/?abstract=11971]
Poster: Methodology - Covariate/Variability Models