1. ¹Division of Urology, Mount Sinai Hospital and Women's College Hospital, University of Toronto, Toronto, Ontario, Canada
2. ²The Scott Department of Urology, Baylor College of Medicine, Houston, Texas, USA

Correspondence: Dr E Grober, Murray Koffler Urologic Wellness Centre, Mount Sinai Hospital, 60 Murray Street, 6th Floor, Box 19, Toronto, Ontario, Canada M5T 3L9. E-mail: egrober@mtsinai.on.ca

Received 9 June 2008; Revised 3 September 2008; Accepted 13 September 2008; Published online 9 October 2008.

Abstract

The primary objective of this study was to correlate simultaneous measures of prostate-specific antigen (PSA) and serum testosterone among large samples of eugonadal, untreated hypogonadal and hypogonadal men treated with testosterone replacement therapy (TRT). From 2001 to 2007, laboratory records were reviewed to identify men who underwent simultaneous measurement of PSA and serum testosterone levels. The data were stratified based on three groups of men: group 1 consisted of eugonadal men (T>300 ng per 100 ml) evaluated for BPH, reproductive failure or sexual dysfunction; group 2 consisted of untreated hypogonadal men (T<300 ng per 100 ml); and group 3 comprised symptomatic hypogonadal men receiving TRT. Correlations were found between PSA (total and free fractions) and total serum testosterone levels among the three groups. Group 1: eugonadal men (n=385 patients), mean PSA and serum testosterone were 1.60 ng ml^-1 and 484 ng 100 ml, respectively. There was no significant correlation between PSA and total serum testosterone levels (r=-0.01, P=0.8). Group 2: untreated hypogonadal men (n=229 patients), mean PSA and serum testosterone were 1.49 ng ml^-1 and 269 ng per 100 ml, respectively. There was no significant correlation between PSA and total serum testosterone levels (r=0.03, P=0.6). Group 3: hypogonadal men on TRT (n=229 patients and 994 individual samples analyzed) mean PSA and serum testosterone were 1.50 ng ml^-1 and 555 ng per 100 ml, respectively. There was no significant correlation between PSA and serum testosterone levels (r=-0.005, P=0.9). Mean total serum testosterone levels were increased significantly (P<0.001) following treatment. Mean PSA levels did not increase in a statistically or clinically significant manner following TRT (mean PSA increase from baseline 0.05 ng ml^-1, P=0.6). In conclusion, TRT does not appear to significantly influence serum PSA expression and no significant correlation was identified between PSA and serum testosterone among eugonadal, untreated hypogonadal and hypogonadal men receiving TRT.