Whose Opinion Counts?

This issue of the Journal of Clinical Oncology publishes the results of two studies examining the role of palifermin for the prevention of chemotherapy-induced oral mucositis. Stiff et al report on the results from a randomized trial comparing palifermin with placebo for reducing mucositis in patients receiving high-dose chemotherapy. While the primary results of this trial were recently reported, the Stiff et al article is an extension of the previous report, whereby the authors provide data from patientcompleted questionnaires, comparing them with “clinician’s assessment of oral mucositis using the objective scales.” The results reported in this article are in concert with the recent US Food and Drug Administration approval of this drug. This report supports that palifermin is useful for preventing mucositis, to a clinically important degree, in patients that are receiving high-dose chemotherapy with a propensity to cause substantial mucositis, such as the chemotherapy regimen used in this trial. This is a move forward from the viewpoint of a 2003 editorial in the Journal entitled “Keratinocyte growth factor: Not yet ready for prime time.” The second article provides a balanced report of a relatively small randomized, double-blinded, placebo-controlled trial of the same drug, trying to prevent mucositis induced by bolus fluorouracil plus leucovorin. This study did demonstrate biologic activity of palifermin, with decreased mucositis in the treated group. Nonetheless, the results from this individual study are not likely to change clinical practice for several reasons, including the small study size, the decreased use of the bolus fluorouracil plus leucovorin regimen in clinical practice, the availability of oral cryotherapy, not used in this trial, and the costs of the therapy. The remainder of this editorial will address two issues raised in the Stiff et al article: utilization of patient-completed questionnaire data and the propensity to blame assessment measures and clinical trial methodology for negative results when it is more likely to be related to the lack of efficacy of the agent being tested.