Quality of life (QOL) and pain in advanced stage prostate cancer: impact of missing data on evaluating palliation in SWOG 9916

Abstract

4579 Background: Both regimens, estramustine plus docetaxel (Arm 1) and mitoxantrone plus prednisone (Arm 2), are known to palliate pain in hormone refractory metastatic prostate cancer. In SWOG 9916, Arm 1 was hypothesized to have greater clinical efficacy as well as equivalent or better palliation of disease-related symptoms. Methods: Men with stage D1 or D2 prostate cancer were randomized to one of the two regimens. We report preliminary data for the Present Pain Intensity (PPI) scale of the McGill Pain Questionnaire and the Global QOL (GQOL) scale score from the EORTC QLQ-C30 at baseline, week 10, months 6, and 12. Results: 333 eligible patients were registered to each arm. Arm 1 had statistically better survival (72% vs. 62% at year 1). The table presents cross-sectional means and sample size for patients by treatment arm and time point. Arm 1 has better submission rates than Arm 2, consistent with Arm 1's longer survival. Data not shown suggest that patients who submitted fewer forms over time had worse scores at baseline and worsening QOL at time of drop-out. Across time and treatment arms, standard errors were on the order of .08 for the PPI but ranged from 1.38 to 2.58 for GQOL. Conclusions: QOL/pain data are particularly important given superior Arm 1 clinical efficacy. However, since reliance on cross-sectional comparisons can be misleading, we will compare treatment arms using 1) pattern mixture model sensitivity analyses for the GQOL outcome and 2) a composite measure (pain reduction and duration, analgesic use) to identify pain palliation responders to examine the influence of informative missing data. Mean QOL Scores by Assessment Time (n) Author Disclosure Employment or Leadership Consultant or Advisory Stock Ownership Honoraria Research Funding Expert Testimony Other Remuneration Aventis Aventis