Search

BackgroundTreosulfan may offer advantages over busulfan for haematopoietic stem cell transplant (HSCT) conditioning, but it is unlicensed in children and the only published dosing information came from a 7-patient first-in-child pharmacokinetic study. We aimed to investigate treosulfan’s pharmacokinetics, efficacy and safety in a GOSHCC-funded trial.MethodsPatients were recruited from GOSH and Newcastle. All received fludarabine +/-serotherapy and treosulfan dosing was based on scaling of adult exposure and expected pharmacokinetic maturation (14 g/m2 in children, 12 g/m2 in infants aged less than 1 year, 10 g/m2 in children aged 3 months or below). Pharmacokinetic sampling was performed during the first and third doses, and population pharmacokinetic modelling performed. Short-term toxicity and long-term (at least 1 year post transplant) outcomes were measured. Successful myeloid engraftment (chimerism greater than 20% at last follow-up) death were the primary outcome measures. Multivariate modelling was used to investigate predictors of these outcomes, and a quadratic model in a generalised linear modelling framework used to define a therapeutic index.Results84 patients were recruited (34 in a pilot phase, 50 in the main study) and provided pharmacokinetic data. A two-compartment model with allometric scaling for body size gave an adequate fit. Age (younger patients had lower per m2 clearance) and renal function (higher creatinine correlated with lower clearance) were significant covariates. After excluding 10 patients from the pilot study who did not meet the main study inclusion criteria, multivariate analysis on the remaining 74 patients (8 of whom had poor engraftmnet, 9 of whom died) showed low treosulfan area under the curve (AUC) was associated with poor engraftment (p=0.047), and high AUC with probability of death (p=0.006). CD34 dose (p=0.02) and mis-matched haplotype donor (p=0.02) were the only other variables associated with success and death respectively. AUC was similar across age and weight. A cumulative AUC of 4703 mg h/L maximised the probability of success, whereas >80% probability of success was associated with AUCs between 3500 and 6319 mg h/L. Only 65% of our study patients were in this target range.ConclusionsWe have shown poor engraftment and mortality are both associated with treosulfan AUC, and the therapeutic index seems narrow enough to warrant personalised dosing.