Abstract
BACKGROUND
Treatment recommendations for localized prostate carcinoma are based on the patient's remaining life expectancy (RLE), which is influenced by age, comorbidity, and tumor grade. Previous studies have evaluated the influence of age and comorbidity, but to the authors' knowledge not RLE, on actual treatment decisions.
METHODS
An ageβstratified random sample of 347 patients was generated from a cohort of all patients with newly diagnosed prostate carcinoma in the Ontario Cancer Registry between May 1, 1995 and April 30, 1996 (n = 5192). Chart review was performed to obtain detailed tumor, comorbidity, and treatment information. RLE was estimated from a published model derived from a cohort of 451 men with untreated prostate carcinoma who were followed for 15 years. Multivariable logistic regression was performed to evaluate predictors of treatment, such as radical prostatectomy (RP), radiotherapy (RT), or potentially curative therapy (RP or RT), in relation to patient age, comorbidity, tumor characteristics, and RLE.
RESULTS
RP was provided within 6 months of diagnosis to 58.7%, 32.1%, 2.6%, and 0% of patients of ages < 60 years, 60β69 years, 70β79 years, and 80+ years, respectively. The results for RT were 6.4%, 30.9%, 23.4%, and 3.3%, respectively. Increasing comorbidity decreased rates of RP but did not affect use of RT. After controlling for comorbidity and tumor characteristics, older men were found to be treated with RP less often than younger men with similar RLE, whereas RLE did not appear to influence receipt of RT.
CONCLUSIONS
Although different mechanisms may account for these results, an age bias may be present among urologists and radiation oncologists treating men with localized prostate carcinoma. Cancer 2004;100:72β81. Β© 2003 American Cancer Society.