Follicle-Stimulating Hormone and the Pituitary-Testicular-Prostate Axis at the Time of Initial Diagnosis of Prostate Cancer and Subsequent Cluster Selection of the Patient Population Undergoing Standard Radical Prostatectomy

Aim: A preceding exploratory analysis has shown that follicle-stimulating hormone (FSH) was significantly correlated to and predicted by prostate-specific antigen (PSA) in a prostate cancer population. The aim of the study was to evaluate FSH physiopathology along the pituitary-testicular-prostate (PTP) axis at the time of initial diagnosis of prostate cancer in an operated population clustered according to the FSH/PSA ratio. Patients and Methods: The study included 93 patients who underwent standard radical prostatectomy. Age, percentages of positive cores at transrectal ultrasound scan biopsy (TRUSB) (P+), biopsy Gleason score (bGS), pathology Gleason score (pGS), luteinizing hormone (LH), FSH, prolactin hormone (PRL), total testosterone (TT), free testosterone (FT), estradiol (ESR) and PSA were the continuous variables. Category variables were pT and biopsy/pathology Gleason pattern I/II (b/pGPI/II). The population was clustered according to the FSH/PSA ratio which was computed from empirical data and then ranked for clustering the population as groups A (range 0.13 ≤ FSH/PSA ≤ 0.20), B (range 0.20 < FSH/PSA ≤ 0.50), C (range 0.50 < FSH/PSA ≤ 0.75), D (range 0.75 < FSH/PSA ≤ 1.00), E (range 1.00 < FSH/PSA ≤ 1.25), F (range 1.25 < FSH/PSA ≤ 2.00), G (range 2.00 < FSH/PSA ≤ 2.25), H (range 2.25 < FSH/PSA ≤ 6.40) and I (range 6.40 < FSH/ PSA ≤ 19.40). The model was assessed by simple linear regression analysis and differences between the groups were investigated by analysis of variance (ANOVA) for continuous variables and by contingency tables for category variables. Results: FSH was significantly correlated to and predicted by PSA in groups A (p = 0.04), B (p < 0.0001), C (p < 0.0001), D (p < 0.0001), E (p < 0.0001), F (p < 0.0001), G (p < 0.0001), H (p = 0.0001) and I (p = 0.001). Also, clusters (A-I) differed significantly for mean values of FSH (p < 0.0001), LH (p < 0.0001), TT (p = 0.04), PSA (p < 0.0001), bGS (p = 0.005), pGS (p = 0.01) and PSA/FT ratio (p < 0.0001); moreover, the nine groups showed significant different frequency distributions of pGPI (p = 0.02), pGPII (p = 0.0002) and bGPI (p = 0.04). Conclusion: The ranking FSH/PSA ratio significantly clustered, along the PTP axis, an operated population diagnosed with prostate cancer. Also, the ranking FSH/PSA ratio selected prostate cancer clusters expressing different levels of hormonal disorder along the PTP axis and prognostic potential with different risks of progression. As a theory, in the current advancing world, the ranking FSH/PSA model might be considered as an interesting and effective tool for prostate cancer study as well as individualized, risk-adapted approaches of the disease. However, confirmatory studies are needed.