Relationships among quality of life and survival in anemic patients with advanced cancer undergoing chemotherapy

Abstract

6027 Background: There are many tools available for measuring cancer patient quality of life (QOL). Baseline QOL data, derived from multiple measurement tools have been identified as prognostic variables for patients with advanced cancer. The purpose of this study was to compare four different ways of assessing QOL in patients with advanced cancer and to describe the relationship between QOL changes and survival in this population. Methods: A double-blind study involved 344 patients with advanced cancer randomized to receive epoetin alfa (40,000 units subcutaneously weekly with escalation to 60,000 units) versus placebo (Witzig, JCO, December, 2004). QOL was assessed monthly by a single-item Uniscale, the Symptom Distress Scale (SDS), the FACT-Anemia scale, and FACT-Fatigue scale. QOL data were translated to a 0–100 point scale. There were no significant differences between treatments in any of the four QOL variables so we undertook a comparative analysis of the four QOL measures across the entire study group. Results: Mean (SD) QOL scores among the 4 QOL tools were significantly different at baseline (p=0.0001) with values of 66 (24) for Uniscale, 74 (13) for SDS, 64 (14) for FACT-An and 51 (22) for FACT-Fatigue. Correlations among the four QOL scales at baseline indicated moderately strong relationships to pseudo-redundancy with correlation coefficients ranging from 0.46 to 0.89. Bland-Altman analysis indicated that there were significant differences between pairs of scales except the Uniscale and the FACT-Fatigue scale (p=0.21). Patients who reported a 10-point improvement in any of the QOL scores survived longer than those who did not (Uniscale: 13 months versus 10 months, p=0.08; SDS 17 months versus 10 months, p=0.0001; FACT-An 16 versus 10 months, p=0.028; and FACT-Fatigue, 15 months versus 9 months, p=0.0001). The variability of the single-item Uniscale was greater than that of the other three scales. Conclusions: Each of the assessments focused on different aspects of patient QOL, resulting in moderate correlation, but not redundancy. While measuring different aspects of patient QOL, changes in scores for the scales indicated an ability to predict survival in this advanced cancer population. No significant financial relationships to disclose.