Although we acknowledge our trial's limitations, we are concerned by the suggestion in the accompanying Comment by Robert Morris and Bradley Monk (Oct 2, p 1120) that patients enrolled in OV05/55955 received ineffective treatment. We believe that the statement is unfounded.

Specific regimens were not mandated in the protocol, but sites used best practice at the time. Clinicians chose the most effective treatment available for their patients. Contrary to Morris and Monk's comment, pegylated liposomal doxorubicin, both as a single agent and in combination with carboplatin, was given to patients within the trial. Bevacizumab was not available to patients but has only very recently been shown to improve progression-free survival modestly in the first-line treatment of ovarian cancer, with data on overall survival and progression-free survival in the relapsed setting awaited.

Our trial shows that the current classification of platinum resistance is open to different interpretations. The concept of platinum resistance was developed at a time when patients were rarely treated on the basis of a rising CA125 concentration alone. Thus a patient considered platinum-resistant by one investigator starting treatment on the basis of CA125 alone could be considered platinum-sensitive by another investigator who only restarts chemotherapy after symptomatic relapse some months later. Figure 3 in our paper shows that there is no difference in overall survival between early and delayed treatment based on the time from end of first-line chemotherapy until randomisation.

OV05/55955 clearly shows that treating patients 4·8 months earlier than is necessary, on the basis of a rise in CA125 concentration alone, does not improve overall survival nor quality of life.