In our opinion, the excellent results in terms of antileukemic efficacy reported by Lengfelder et al do not support some of the original paper's statements, which they repeat in their extended reply. To deal in specific detail with all of the nine points argued by Lengfelder et al would be too exhaustive, so we will address some selected issues. For example, when examining the data for toxicity and compliance, the AMLCG study showed that, in four of the 45 patients (9%) consolidation was omitted due to severe infections during induction. Moreover, 13 patients (29%) did not receive maintenance therapy (including the four patients that had not received consolidation). In 11 additional cases (24%), maintenance therapy was terminated early. This significant lack of compliance was mainly due to toxicity, leading to refusal by the patients in 31%; physicians’ decision after previous severe infections in 26%; prolonged neutropenia between courses in 31%; chronic hepatitis in 6%; and vasculitis in a further 6%. In our opinion, only the enthusiasm generated by the antileukemic efficacy of the AMLCG protocol, but not by such telling data on toxicity and compliance, could lead Lengfelder et al to claim ‘the treatment was feasible with an acceptable toxicity and a good compliance’.
Our counts for the cumulative dose of anthracyclines administered in the AMLCG study, in comparison with other protocols, are slightly different to those counted by Lengfelder et al. It is true that the cumulative DNR equivalent dose given during induction and consolidation in the AMLCG study was 485 mg/m2, but maintenance therapy was scheduled to reach a cumulative DNR dose of 540 mg/m2. However, the cumulative DNR equivalent dose given in the GIMEMA and PETHEMA studies was clearly lower (440 mg/m2). How can Lengfelder et al rule out the impact of this and other factors on their results and state it was essentially due to ‘an ara-C effect’?
This is a preview of subscription content, access via your institution