Changes in the levels of testosterone profile over time in relation to clinical parameters in a cohort of patients with prostate cancer managed by active surveillance.

Abstract

To characterize testosterone profile changes over time in a cohort of prostate cancer (PCa) patients managed with active surveillance (AS) and to assess its correlation with the initial disease characteristics and further progression. We conducted retrospective chart review of PCa patients managed with AS. Patients were followed with PSA, total, free and bioavailable testosterone measurements, physical examination, and by repeat biopsies or periodic magnetic resonance imaging. Disease progression was identified by follow-up biopsy changes or by imaging. A Cox proportional hazard regression models were used to assess the association between testosterone profile at baseline and the risk of progression. For the 122 patients included in analyses, the mean age at diagnosis was 65.8years; the mean follow-up time was 7.8years. At baseline, 108 (88.5%) patients had a Gleason score of ≤ 6. In all, 45 (36.8%) patients had disease progression, with a mean time to progression of 4.6years. During follow-up, PSA levels showed a rising trend, while testosterone profile levels showed a trend of decrease over time. There was no significant correlation between PSA and testosterone profile (total, free, and bioavailable) level changes over time (ρ = -0.14, -0.11 and -0.16, P = 0.16, 0.34, and 0.20, respectively). In addition, multivariable analysis showed that serum-free testosterone was an independent predictor of disease progression (HR 0.93, 95% CI 0.88-0.99, P = 0.029). Our study results showed that testosterone profile measurements tended to decrease over time in PCa patients managed with AS. Free testosterone was a significant independent variable of disease progression.