Determinants of fatigue improvement in outpatient oncology according to baseline categories of fatigue severity.

Abstract

9112 Background: Understanding the determinants of fatigue response may help distinguish between different fatigue phenotypes and inform trial designs. Methods: Patients with invasive cancer of the breast, prostate, colon/rectum, or lung were enrolled from multiple sites. At baseline and 4-5 weeks later patients rated their symptoms on a 0-10 numerical rating scale. A 2-point change was considered clinically significant for fatigue change. Logistic regression models and ordinal logistic regression models were used to examine the effects of demographic and clinical factors on fatigue change. Separate models were constructed for fatigue deterioration in patients with mild fatigue at baseline and lack of improvement for those with severe fatigue at baseline or any significant change for those with moderate baseline fatigue. Results: 3,106 pts were enrolled at baseline and 3,032 were analyzable for fatigue change. At baseline, 23% had no fatigue, 35% mild, 25% moderate, and 17% severe. Improvement varied by baseline fatigue score with response in 54% with severe fatigue, 31% with moderate, and 9% with mild. Overall, 13 different parameters were significant in these 3 models of fatigue change, but no single parameter was common to all 3 categories of fatigue. Exposure to anxiolytics and duration of current treatment were significant in models of mild and severe fatigue; antidepressant exposure was significant for mild and moderate fatigue, and baseline fatigue level was a significant parameter in moderate and severe fatigue. In colorectal and lung cancer patients, gender was not significant for fatigue change in patients with mild or moderate fatigue, but gender was the strongest predictor fatigue improvement in patients with severe fatigue with men most likely to significantly improve (OR=2.49, 95% CI 1.15-5.41, p=0.021). Conclusions: Clinically significant improvement in fatigue varies between 9% and 54%. Predictors of fatigue change vary depending on categories of baseline fatigue severity. Gender is the strongest predictor for fatigue improvement among lung and colorectal outpatients with severe baseline fatigue. Future trial designs should account for gender and baseline fatigue severity.