Immediate or Delayed Docetaxel: Does Progression-Free Survival Really Reflect Two Strategies?

Abstract

TO THE EDITOR: We have read the article by Fidias et al with great interest. It made a valuable contribution from a clinical point of view. Their trial was designed to answer one question that clinicians come across frequently in everyday oncological management of patients with advanced non–small-cell lung cancer: should nonprogressing patients after a sufficient amount of first-line therapy proceed with immediate second-line chemotherapy, or should they defer it until documentation of first progression? The authors stated that progression-free survival (PFS) in the immediate docetaxel arm was significantly longer than that in the delayed docetaxel arm (5.7 v 2.7 months; P .0001) and that there was a trend of improved overall survival (OS) in the same arm (12.3 v 9.7 months; P .0853), probably owing to the greater number of patients who were able to receive treatment in the immediate docetaxel arm. OS was the primary end point.WehavesomecommentsonPFS,oneof thesecondaryendpoints. Rather than one PFS, the delayed arm has actually two PFS rates by definition, given that patients should wait until the documentation of progression before receiving chemotherapy: one from random assignment to documentation of first progression just before time of docetaxel institution (PFS1), and one from docetaxel institution to the documentation of progression while still receiving chemotherapy (PFS2). However, there is only one sort of PFS in the immediate docetaxel arm: the one from random assignment to progression on docetaxel chemotherapy. Although it is not clear from the article, the investigators probably reported a comparison of PFS1 in the delayed arm with PFS in the immediate arm. We think that this may be the reason why there was a strong statistical difference (2.7 v 5.7 months; P .0001). Such a comparison does not fully reflect the expected answer to the question of delayed versus immediate docetaxel after first-line therapy. Rather, it gives information concerning the question if docetaxel improves PFS after four cycles of gemcitabin-carboplatin in nonprogressing patients. It is highly probable that no PFS difference between delayed and immediate docetaxel would have been detected if the investigators had calculated PFS from the time of random assignment to progression during docetaxel chemotherapy for both arms.