We thank Dr. Nguyen and his colleagues for their letter about our article “Population Pharmacokinetics of Busulfan in Children: Increased Evidence for Body Surface Area and Allometric Body Weight Dosing of Busulfan in Children” (1). We would like to respond to the issues raised in his letter with a few comments.
The European Medicines Agency (EMA) recommends an intravenous busulfan dosing regimen according to the patient's body weight, with 5 dosing groups defined according to Nguyen and colleagues (2) as follows: 1.0 mg/kg for <9 kg; 1.2 mg/kg for 9 to <16 kg; 1.1 mg/kg for 16–23 kg; 0.95 mg/kg for >23–34 kg; and 0.80 mg/kg for >34 kg. In contrast, studies in children have shown that similar area under curve (AUC) values to those observed in adults can be achieved when dosing busulfan according to body surface area (BSA; refs. 3–5).
In our analysis, we propose 2 new individualized i.v. busulfan dosing regimens in children, either based on BSA or on allometric body weight and not based on just body weight as pointed out in Nguyen's letter. Using just body weight compared with allometric body weight is a substantial difference in drug dosing. In allometric body weight dosing, the body weight is raised to a power of 0.75, resulting in a nonlinear relationship between clearance (CL) and body size (6).
Nguyen and colleagues have applied our new dosing regimens and the EMA dosing regimen to a larger cohort of i.v. busulfan pediatric patients which has previously been described in the literature. It is hard to follow up on the cited literature as the article by Paci and colleagues (7) had not yet been published at the time of submission of this Response and the second citation (8) leads to a conference abstract by Daher Abdi and colleagues.
Both studies are from the same group and therefore include most likely the same patients. As mentioned in the abstract from Daher Abdi and colleagues (8), 90 patients were reused from 2 prospective clinical trials. These two trials contributed to the EMA dosing recommendation of i.v. busulfan. The trial by Nguyen and colleagues (2) was the dose-finding study and the study by Vassal and colleagues (9) was the validation of the new dosing proposal by Nguyen and colleagues. By reusing patient data which were used for dose regimen findings once, the analysis is biased toward the already implemented dosing recommendation.
In our analysis, completely new data were used which have not been used in a prior dose-finding analysis. In addition, our model was evaluated by using an external data set which was not used during model building and originates from a different dosing schedule.
The dose-finding study by Nguyen and colleagues (2) included just 4 children <9 kg of body weight, which was fewer than in our analysis with 5 children having a body weight <9 kg. In Nguyen's letter, this was claimed to be too few. In addition, we had 5 more children with a body weight <9 kg in our external evaluation data set. We agree with the fact that our data are unbalanced with respect to the body weight groups but we have not seen any literature on busulfan dosing out yet where the groups were not unbalanced. We further disagree with the statement in Nguyen's letter that the small group of patients in this extreme body weight group of <9 kg might have markedly influenced the shape of the model as all other dosing groups included far more patients which would therefore rather influence the shape.
In the newly suggested dosing recommendation by Daher Abdi and colleagues (8), the body weight is raised to the same power for allometric body weight dosing in the children >9 kg of body weight as in our dosing recommendation. Therefore, it would have been interesting to see the comparison between Daher Abdi's dosing regimen and our dosing in addition to the comparison made by Nguyen (Fig. 1 in Nguyen's letter). We do know that our clearance values are lower in most cases, which we stated clearly in our article. Our calculated clearance values derived from new patient data. One has to keep in mind that the patient population in pediatric oncology is very heterogeneous and that comedications which may influence the clearance of busulfan can contribute to the differing results.
Furthermore, by looking at our prediction of corrected visual predictive checks in the published article (1), one can see that the model describes the i.v. data sufficiently as well as the external data set which consists of just i.v. data. This confirms that the oral data included in the data set did not lead to a bias in the model. The AUC simulation plots for our dosing recommendation divided into the different body weight dosing groups show no over- or underprediction in either one of the body weight groups.
Furthermore, the dosing recommendation which is proposed in the ongoing European Group for Blood and Marrow Transplantation clinical protocol is in the version of 2011 and not as cited by Nguyen in the version from 2010. In addition, the article by Bartelink and colleagues is not available yet. The only source available is a conference abstract (10).
Addressing the last point of Nguyen's letter “our Bu concentration measurement techniques are not cross-validated which could lead to the possibility of unreliable estimates of CL values by the model,” we mentioned in our article that plasma samples were either measured via a validated high-pressure liquid chromatography with UV detector or by a validated liquid chromatography/mass spectrometry method. Second, we have tested different error models for each dosing or laboratory method during the model building step before using one error model for all data to describe the residual unexplained variability.
As we already stated, our dosing scheme is based on simulations and has to be tested prospectively before giving a general recommendation for the adequate dosing of busulfan in children.
See the original Letter to the Editor, p. 2715
Disclosure of Potential Conflicts of Interest
No potential conflicts of interests were disclosed.