Recent retrospective cohort studies have supported these pre
Recent retrospective cohort studies have supported these preclinical findings in animals indicating the correlation of AR activation with the induction of inhibitor of apoptosis proteins carcinogenesis. First, men with prostate cancer who underwent androgen deprivation therapy were shown to have a significantly lower risk of subsequent development of bladder cancer [0/266 (0%)], compared with those undergoing surgery alone [5/437 (1.1%)] or radiotherapy [14/631 (2.2%)] (Shiota et al., 2015). Second, in 162 men with a history of both prostate and bladder cancers, androgen deprivation therapy used for the treatment of the former strongly prevented the recurrence of the latter, compared with those without hormonal therapy (5-year recurrence-free survival: 76% vs. 40%, P < 0.001; number of 5-year cumulative recurrence episodes: 0.44 vs. 1.54, P < 0.001) (Izumi et al., 2014b). In this cohort, androgen deprivation therapy (hazard ratio = 0.24, P < 0.001) (Izumi et al., 2014b) as well as AR expression in 72 non-muscle-invasive bladder tumors (hazard ratio = 0.27, P = 0.005) (Izumi et al., 2016) was found to be an independent prognosticator for bladder tumor recurrence. Third, in 228 men with bladder cancer, androgen deprivation therapy (for their prostate cancer) or a 5α-reductase inhibitor dutasteride treatment (for their benign prostatic hyperplasia) significantly reduced the rate of bladder tumor recurrence, compared with control patients without hormonal treatment (hazard ratio = 0.36, P = 0.024) (Shiota et al., 2017). Additionally, in a prospective study involving 72,370 men, a 5α-reductase inhibitor finasteride primarily prescribed for their symptomatic benign prostatic hyperplasia was associated with a significantly reduced risk of bladder cancer development (hazard ratio = 0.634, P = 0.0004) (Morales et al., 2016), while an animal study failed to show a considerable effect of finasteride on the development of BBN-induced bladder tumors as described above (Imada et al., 1997). Androgen-mediated AR signals have been shown to involve the modulation of the activity of certain enzymes responsible for the metabolism of bladder carcinogens. These enzymes include cytochrome P450 (e.g. CYP4B1) and UDP-glucuronosyltransferase (e.g. UGT1A subtypes) that are known to contribute to the activation and detoxification, respectively, of bladder carcinogens, such as aromatic amines. The levels of CYP4B1 gene expression were higher in intact male mouse bladders than in intact female or castrated male mouse bladders (Imaoka et al., 2001). Similarly, the expression levels of mouse Ugt1a subtypes were considerably up-regulated in the bladders from intact female or ARKO male mice, compared with those from intact/control male mice, and orchiectomy in intact male mice resulted in significant decreases in the expression of some Ugt1a subtypes in their bladders. (Izumi et al., 2013). Additionally, in SVHUC human normal urothelial cells stably expressing wild-type full-length AR, DHT treatment was found to reduce the expression of all UGT1A subtypes, and an AR antagonist hydroxyflutamide blocked the DHT effects (Izumi et al., 2013). Moreover, using a transgenic mouse model, castration was shown to reduce sensitivity to a bladder carcinogen 4-aminobiphenyl via modulating Ugt1a3 in the liver (Bhattacharya et al., 2015). Meanwhile, in an immunohistochemical study using human bladder tissue specimens, UGT1A expression was down-regulated in bladder cancers, compared with non-neoplastic urothelial tissues, as well as in high-grade or muscle-invasive tumors, compared with low-grade or non-muscle-invasive tumors, and loss of UGT1A was an independent prognosticator for disease progression in patients with muscle-invasive tumor (Izumi et al., 2014a). GATA3 is a zinc-finger transcription factor and is highly expressed in urothelial cells. As a result, GATA3 immunohistochemistry has been widely used as a promising urothelial marker in diagnostic surgical pathology practice. However, a subset of bladder cancers, especially high-grade and/or muscle-invasive tumors (20–28%), was found to lose GATA3 expression (Miyamoto et al., 2012a), and loss of GATA3 expression in upper urinary tract urothelial tumors was associated with poor prognosis as an independent predictor (Inoue et al., 2017), suggesting its role as a tumor suppressor. Indeed, in SVHUC cells exposed to a chemical carcinogen, GATA3 knockdown resulted in the induction of neoplastic transformation as well as down- or up-regulation of the expression of tumor suppressors (e.g. p53, p21, p27, PTEN, UGT1A) or oncogenic molecules (e.g. c-myc, cyclin D1/D3/E, FGFR3), respectively (Li et al., 2014). AR overexpression or androgen treatment (only in AR-positive cells) was also shown to reduce GATA3 expression in SVHUC sublines (Li et al., 2014). In addition, orchiectomy augmented the expression levels of GATA3 in the mouse bladders (Li et al., 2014). Thus, AR activation appears to correlate with down-regulation of the expression of GATA3 that contributes to the prevention of urothelial tumorigenesis.