Clinical interpretation of quality-of-life outcomes: an investigation of data from the randomized trial of gemcitabine plus paclitaxel compared with paclitaxel alone for advanced breast cancer

Abstract

Quality-of-Life (QL) data are increasingly relevant to trial results, but are often difficult to interpret and apply clinically; methods to address this problem are needed. Exploratory analyses were conducted using QL data from a recent phase III trial comparing front line gemcitabine plus paclitaxel (GT) versus paclitaxel (T) alone in patients with advanced breast cancer. The most troublesome symptom and toxicity items were identified. For each QL domain, associations and change over time were analyzed using clinically relevant data. The impact of tumor response and duration of therapy on QL outcomes was assessed using pooled data from both treatment arms. Baseline QL data were available from 336 patients, out of 266 patients enrolled on the GT arm and 263 patients enrolled on the T arm. The prevalence of disease-related symptoms was low; 7 of the 10 most troublesome items were psychological. Clinical levels of psychological distress were significantly associated with global QL and reduced significantly over time in both arms. Improved QL was significantly associated with reduced functional impairment, tumor response and completing more cycles of therapy. In this patient population, QL may be closely associated with the clinical efficacy of chemotherapy, functional status and psychological well-being, rather than outcomes such as symptom palliation and toxicity.