Association between prostate‐specific antigen and serum testosterone: A systematic review and meta‐analysis

Abstract

Background Serum testosterone assays are an important tool in the clinical evaluation of a number of endocrine disorders including male hypogonadism. However, serum testosterone has a limited role in real clinical use due to its inaccuracy. We aimed to assess the association between prostate specific antigen (PSA) and testosterone as well as the effects of various types of testosterone replacement therapy (TRT) for PSA level. Methods Two electronic databases were screened: PUBMED (1966 through December 2018) and Cochrane Library (1993 through December 2018). The first strategy compared the overall increase in PSA following testosterone treatment compared with placebo. The second strategy analyzed the overall association between PSA and testosterone among the observational studies. Results In the first strategy, twenty‐two articles were included in the final analysis. In the second strategy, 18 studies were included. Testosterone replacement therapy (TRT) showed a significant change in PSA level compared to that in the placebo group (mean difference [MD]: 0.13, 95% CI: 0.01 to 0.25, P=0.04). Compared to placebo, only intramuscular (IM) TRT shows a significant change in PSA level group (MD: 0.16, 95% CI: 0.01 to 0.30, P=0.04), as neither the oral nor topical type showed a significant change in PSA. In the second strategy analysis, there was no overall correlation found between PSA and testosterone (z=0.04, 95% CI: ‐0.04 to 0.12, P=0.04; r = 0.039). However, in the subgroup of non‐BPH (benign prostate hyperplasia), a significant correlation between PSA and testosterone (z=0.07, 95% CI: 0.01 to 0.13, P=0.009; r = 0.089) was found. Conclusions We found that TRT, particularly IM TRT, significantly changed the PSA level compared with the placebo group. Furthermore, there was a significant correlation between PSA and testosterone in patients with non‐BPH. According to these findings, we suggest the possibility of PSA as a surrogate marker of testosterone.