ABSTRACT
METHODS AND MATERIALS: A retrospective study was conducted to (1) determine the relationship between baseline prostate-specific antigen (PSA) levels and initial treatment decisions for prostate cancer (surgery, hormone therapy, radiation, or watchful waiting) and (2) estimate the impact of PSA progression (doubling or three consecutive rises) on subsequent treatment decisions. Patient records (n=1116) from three community urology practices and a large academic health system were reviewed. Multivariate models were fitted to assess the relationship between initial treatment and baseline PSA, Gleason score, race, number of comorbid conditions and age and between PSA progression and time to subsequent therapy (adjusted for other factors). RESULTS: Baseline PSA was a significant predictor of initial treatment among men with localized disease with the likelihood of hormone therapy increasing with higher PSA levels and the likelihood of surgery decreasing steadily with higher PSA levels. PSA was the strongest predictor of hormone therapy as first choice followed by age. Age followed by PSA was the strongest predictor of surgery as first treatment as well as radiation therapy. Initial PSA levels did not predict the choice of watchful waiting. Patients with PSA progression were eight times (95% CI: 5.3-12.1) more likely to initiate a subsequent therapy than patients who did not have PSA progression when controlling for other predictors. CONCLUSIONS: In clinical practice, PSA significantly impacts the urologist's primary therapy choice and determines when they introduce subsequent treatments.
