The impact of diurnal variation of PSA on timing of measurement in prostate biopsy

• Published in: The Prostate Vol. 79; no. 14; pp. 1666 - 1672
• Main Authors: Uzun, Hakkı, Arpa, Medeni
• Format: Journal Article
• Language: English
• Published: United States Wiley Subscription Services, Inc 01.10.2019
• Subjects: 17β-Estradiol; Aged; Aged, 80 and over; Biopsy; Circadian Rhythm; Diurnal; diurnal variation; Humans; Lower Urinary Tract Symptoms; Male; Middle Aged; prostate; Prostate - pathology; Prostate cancer; Prostate-Specific Antigen - blood; Prostatic Neoplasms - blood; Prostatic Neoplasms - diagnosis; Prostatic Neoplasms - pathology; PSA; Statistical analysis; Testosterone; Testosterone - blood; Time Factors; Ultrasound; Urinary tract; Variation; prostate; testosterone; diurnal variation; PSA
• ISSN: 0270-4137; 1097-0045
• DOI: 10.1002/pros.23891

Background Prostate‐specific antigen (PSA) synthesis is related to testosterone, which has a diurnal rhythm. PSA might have a diurnal variation and the timing of measurement could change the clinical practice for prostate biopsy. Methods Male patients complaining of lower urinary tract symptoms (group 1) and diagnosed with prostate cancer (group 2) were recruited into the study. Morning fasting blood samples were withdrawn between 9.00 and 11.00 am for the determination of biochemical parameters, PSA (PSA1), total testosterone (T1), and estradiol (E1) levels. In the afternoon, between 15.00 and 15.30 pm, blood samples were again obtained from the same participants at the same day and the serum concentration of PSA (PSA2), total testosterone (T2), and estradiol (E2) were measured. Results A total of 160 and 30 patients were enrolled in groups 1 and 2, respectively. One hundred forty (87.5%) and 26 (86.6%) patients had a decrease in the PSA levels when measured in the afternoon. The Wilcoxon signed‐rank test determined a statistically significant difference between the PSA levels measured in the morning and in the afternoon in each group. An analysis of covariance test revealed no statistically significant difference in PSA concentration between the groups after adjustment for baseline concentration (F(1.187) = 0.203, P = .653). There was a weak positive correlation between PSA1/PSA2 and T1/T2, rs (160) = 0.163, P = .034. An extra unit increase in PSA1 concentration leads to a 0.805 (95% confidence interval [CI], 0.781‐0.830) and 0.828 (95% CI, 0.807‐0.849) ng/mL increase in PSA2 concentration in groups 1 and 2, respectively, that is, patients with and without prostate cancer had a similar decrease in the PSA levels. When measured in the afternoon, 66.6% and 50% patients with a morning PSA level over 3 or 4 ng/mL had a PSA drop below these levels, respectively. Conclusions PSA has a diurnal variation and the timing of measurement may alter the decision of the clinician for transrectal ultrasound prostate biopsy.