Search

Aims: Carboplatin dosage is calculated by using the estimated glomerular filtration rate (GFR) to achieve a target plasma area under the plasma concentration–time curve (AUC). The aims of the present study were to investigate factors that influence the pharmacokinetics of carboplatin in children with high‐risk neuroblastoma, and whether target exposures for carboplatin were achieved using current treatment protocols. Methods: Data on children receiving high‐dose carboplatin, etoposide and melphalan for neuroblastoma were obtained from two study sites [European International Society for Paediatric Oncology (SIOP) Neuroblastoma study, Children's Hospital at Westmead; n = 51]. A population pharmacokinetic model was built for carboplatin to evaluate various dosing formulas. The pharmacokinetics of etoposide and melphalan was also investigated. The final model was used to simulate whether target carboplatin AUC (16.4 mg ml–1·min) would be achieved using the paediatric Newell formula, modified Calvert formula and weight‐based dosing. Results: Allometric weight was the only significant, independent covariate for the pharmacokinetic parameters of carboplatin, etoposide and melphalan. The paediatric Newell formula and modified Calvert formula were suitable for achieving the target AUC of carboplatin for children with a GFR <100 ml min–1 1.73 m–2 but not for those with a GFR ≥100 ml min–1 1.73 m–2. A weight‐based dosing regimen of 50 mg kg–1 achieved the target AUC more consistently than the other formulas, regardless of renal function. Conclusions: GFR did not appear to influence the pharmacokinetics of carboplatin after adjusting pharmacokinetic parameters for weight. This model‐based approach validates the use of weight‐based dosing as an appropriate alternative for carboplatin in children with either mild renal impairment or normal renal function. [ABSTRACT FROM AUTHOR]

Estimation of glomerular filtration rate (GFR) using the clearance of chromium 51 EDTA (51Cr-EDTA) (or other radiolabelled isotopes) is reliable, but invasive and not always practicable. Mathematical models have been devised for estimating GFR using readily obtainable patient characteristics. Unfortunately, these models were developed using various patient populations and may not provide the optimal prediction of GFR in children with cancer. The current study uses population pharmacokinetics to determine the relationship between 51Cr-EDTA clearance, and patient covariates in 50 paediatric cancer patients. These models were validated using a separate group of 43 children and were compared with previously published models of renal function. Body size was the major determinant of 51Cr-EDTA clearance and inclusion of weight or surface area reduced the residual variability between individuals (coefficient of variation) from 61 to 32%. Serum creatinine was the only other parameter that significantly improved the model. Mean percentage error values of -5.0 and -1.1% were observed for models including weight alone or weight and creatinine, respectively, with precision estimates of 21.7 and 20.0%. These simple additive models provide a more rationale approach than the use of complex formulae, involving additional parameters, to predict renal function.British Journal of Cancer (2004) 90, 60-64. doi:10.1038/sj.bjc.6601484 www.bjcancer.com [ABSTRACT FROM AUTHOR]