Serum Testosterone Levels in Prostate Cancer Patients Undergoing Luteinizing Hormone-Releasing Hormone Agonist Therapy

© 2017 Elsevier Inc. Prostate cancer guidelines have recommended serum testosterone measurement during androgen deprivation therapy to assess its efficacy and diagnose castration resistance. The present study compared the testosterone levels from a widely used chemiluminescent assay (CLIA) and liquid chromatography tandem mass spectrometry method, which is the recommended method, in prostate cancer patients undergoing luteinizing hormone-releasing hormone agonist therapy. The CLIA overestimated the testosterone levels and suggested, incorrectly, the presence of inadequate castration in ≤ 15% of patients. Background: Serum testosterone measurement is recommended to assess the efficacy of androgen deprivation therapy (ADT) and to diagnose castration resistance in patients with prostate cancer (PCa). Currently, the accepted castrate level of serum testosterone is 50 ng/dL. Liquid chromatography and tandem mass spectrometry (LC MSMS) is the appropriate method to measure testosterone, especially at low levels. However, worldwide, chemiluminescent assays (CLIAs) are used in clinical laboratories, despite their lack of accuracy and reproducibility, because they are automatable, fast, sensitive, and inexpensive. Materials and Methods: We compared serum testosterone levels measured using LC MSMS and CLIAs in 126 patients with PCa undergoing luteinizing hormone-releasing hormone (LHRH) agonist therapy. Results: The median serum testosterone level was 14.0 ng/dL (range, 2.0-67.0 ng/dL) with LC MSMS and 31.9 ng/dL (range, 10.0-91.6 ng/dL) with CLIA (P <.001). The serum testosterone levels, measured using LC MSMS, were < 20 ng/dL in 83 patients (65.9%), 20 to 50 ng/dL in 40 (31.7%), and > 50 ng/dL in 3 patients (2.4%). These ranges were found in 34 (27%), 72 (57.1%), and 20 (15.9%) patients when testosterone was measured using CLIA (P <.001). The castrate level of serum testosterone using LC MSMS and CLIA was 39.8 ng/dL (95% confidence interval [CI], 37.1-43.4 ng/dL) and 66.5 ng/dL (95% CI, 62.3-71.2 ng/dL), respectively. Conclusion: We found that CLIA overestimated the testosterone levels in PCa patients undergoing LHRH agonist therapy. Thus, the castration level was incorrectly considered inadequate with CLIA in almost 15% of patients. The true castration level of serum testosterone using an appropriate method is < 50 ng/dL.