Tackling overtreatment of prostate cancer: the role of comorbidity.

The management of localized prostate cancer continues to be an area of tremendous controversy. In 1 of the only clinical trials supporting an aggressive approach, an overall survival benefit for radical prostatectomy versus watchful waiting was limited to men aged <65 years at diagnosis and could be demonstrated only after 10 years of follow-up. In addition, a small minority of the men were diagnosed by prostate-specific antigen (PSA) screening, making generalizability of those results to contemporary prostate cancer cohorts questionable. Moreover, older men or men with low-risk tumors are unlikely to die from their prostate cancer even without any aggressive intervention, such as radical prostatectomy or radiation therapy. Therefore, although a wide variety of treatment options exists, the main challenge has been, and continues to be, how to identify the men for whom it would be more appropriate to adopt a conservative approach. In this issue of Cancer, Daskivich et al add to the literature in this area. By using data from a cohort of nearly 1500 men with prostate cancer from 2 Veterans Affairs (VA) Medical Centers in California, they examined the impact of comorbidity on nonprostate cancer mortality with a competing risks analysis in groups stratified according to the tumor risk criteria described by D’Amico et al. Daskivich et al reasoned that the outcome of nonprostate cancer mortality was of particular interest, because it would help discern which men would be unlikely to benefit from aggressive treatment. In a model adjusted for age, race, D’Amico et al tumor risk, and treatment received, an increasing score on the Charlson comorbidity index predicted higher nonprostate cancer mortality. Men who had a Charlson score 3, for whom the nonprostate cancer mortality rate was 74% at 10 years, had a greater than 8-fold increase in the hazard of death from causes other than prostate cancer compared with men who had a Charlson score of 0. In general, those men with higher Charlson scores and/or low-risk to intermediate-risk tumors were far more likely to die from causes other than their prostate cancer. The concept that comorbidity should be incorporated into decision-making about treatment for prostate cancer is not new, with some prior reports dating back more than a decade. The advantages of this study over others include the use of competing risk analyses to account for both nonprostate and cancer-specific mortality in estimating the hazard of death associated with each, coupled with the heterogeneous mix of patients with respect to tumor characteristics and treatment received. In addition, comorbidity was assessed through review of medical records, which may be more accurate than the assessments in many previous studies based on administrative claims data. However, generalizability is likely to be limited given the relatively small sample of patients, the fact that the cohorts were derived from only 2 centers, and the nature of the veteran population, with its particular socioeconomic and comorbidity issues. Previous studies have attempted to derive nomograms to predict 10-year life expectancy and have presented information on predictive accuracy, neither of which were done in this study. Thus, the practical utility of the study findings are limited to the generality that men with high degrees of comorbidity and low-risk to intermediate-risk tumors probably should receive conservative management. Ultimately, we believe that the primary value of this particular investigation is that it continues the dialogue about how best to identify the subset of patients who are unlikely to benefit from definitive treatment. The sobering reality is that, despite calls for more conservative management of localized prostate cancer in both the published literature and in practice guidelines, overtreatment remains rampant. An earlier publication by the same