Reply to T. Maughan et al

Abstract

We thank Maughan et al for their interest and commentary on the OPTIMOX2 (Optimized Leucovorin-Fluorouracil-Oxaliplatin) study, which should be considered along with the editorial by Hochster. Some remarks have already been discussed. Maintenance therapy and chemotherapy-free interval were stopped at progression in most patients. They were stopped in six patients (3%) before (but not until) the tumor reached the baseline measure in case of previous response, or in case of tumor-related symptoms. The COIN (COntinuous chemotherapy versus INtermittent chemotherapy) trial showed that continuous therapy was associated with a trend toward better survival compared with chemotherapy discontinuation, as in the OPTIMOX2 study. The hazard ratio for overall survival in the OPTIMOX2 study was 1.14, with a 95% CI of 0.8259 to 1.574, compared with the hazard ratio of 1.084 in the COIN trial, with a 80% CI of 1.008 to 1.165. We used duration of disease control (DDC) because it has been proven that progression-free survival (PFS) is not a suitably reliable end point to predict overall survival (OS) when patients can receive more than one course of treatment. It has been proposed that DDC can capture the treatment effect in strategic trials. It is defined as the sum of PFS of the first treatment phase and the PFS of the following treatment phases when progression is not observed at their first evaluation. We showed in a pooled analysis that there was a strong association between DDC and OS (r 0.93; P .001), but there was no statistically significant correlation between PFS and OS (r 0.45; P .49). There is no difficulty in using DDC if data postprogression are available. On the basis of the OPTIMOX study, we decided not to plan chemotherapy discontinuation at baseline, but only in selected patients after at least 6 months induction chemotherapy. A normal carcinoembryonic antigen level at 3 months from baseline was found to be a predictive factor for a prolonged chemotherapy-free interval and favorable prognostic factor for overall survival. We agree that additional large data sets will give us new information about patient eligibility for chemotherapy discontinuation. Benoist Chibaudel, Frederique Maindrault-Goebel, Christophe Tournigand, and Aimery de Gramont Hopital Saint-Antoine, Assistance Publique des Hopitaux de Paris, Paris, France