Pcn380 - The Prognostic Significance of Health-Related Quality of Life From Cancer Randomized Controlled Trials: A Systematic Evaluation of the Statistical Methods

Abstract

OBJECTIVES:PROs from open label trials are traditionally regarded as being more biased than PROs from double blinded studies, as the patient's knowledge of their treatment might influence their perception of their symptoms, functioning and HRQoL. We explored this hypothesis by comparing PROs from the QLQ-C30 which was used in patients treated with dabrafenib plus trametinib (D+T) for BRAF+ metastatic melanoma with the same data collection schedule in two separate studies e COMBI-D (double blinded vs dabrafenib, n¼211 in D+T arm) and COMBI-V (open label vs vemurafenib, n¼352 in D+T arm). METHODS: We qualitatively and quantitatively compared responses to the QLQ-C30 in the active dabrafenib+trametinib arms. Mean values and standard deviations for each QLQ-C30 domain were compared between trials at each time point until <20% of patients remained on trial. RESULTS: Qualitatively, there were no differences with graphs visually overlaying. This was consistent across all domains, including those showing clinically relevant improvements (HRQoL, pain) and those with no change. A small visual trend was observed in the COMBI-V open label study for constipation (worse in COMBI-V by mean of 2.0 points) and emotional functioning (better in COMBI-V by mean of 3.4 points) but not appetite loss, pain, dyspnea, fatigue, financial difficulties, insomnia, nausea and vomiting, physical functioning, role functioning, cognitive functioning or global health status/HRQoL. Quantitatively, in only 1 out of 98 timepoints, a difference larger than the MID of 5 was seen (emotional functioning, week 32, D5.4pts). CONCLUSIONS: This analysis demonstrates that there was little influence on the PRO results for the open label study COMBI-V versus COMBI-D a blinded study. The results were remarkably consistent qualitatively and numerically across these two independent studies. Future research is needed to validate this result across therapies and diseases to give decision makers confidence in interpreting open label PRO data.