Letters Section Editor: Robert M. Golub, MD, Senior Editor.
In Reply: The letters by Dr Lu-Yao and colleagues, Dr Scholz and colleagues, and Dr Mark raise the important issue that prostate cancer-specific mortality was low in both groups, and that the absolute difference (0.6%) may not outweigh the potential long-term adverse effects caused by the treatment. However, we believe that the assessment of prostate cancer-specific mortality has limitations and should be interpreted with caution. SEER reports cause of death information derived from state death certificate data, which have poor correlation with autopsy findings.1,2 The high prevalence of comorbidities in this elderly patient population can make attribution of cause of death challenging. For example, initial treatment course may influence the cause of death that is ultimately listed on death certificates,3 and this “cause of death attribution bias” may have an effect on prostate cancer-specific mortality.4 We believe that since overall survival (our primary endpoint) is more consistently reported, the data regarding prostate cancer-specific mortality should be considered exploratory and investigated further in clinical trials in which investigators have access to detailed medical records.
Wong Y, Mitra N, Hudes G, Schwartz JS, Wan F, Montagnet C, Armstrong K. Outcomes of Treatment vs Observation of Localized Prostate Cancer in Elderly Men—Reply. JAMA. 2007;297(15):1651-1653. doi:10.1001/jama.297.15.1652-c