Does 5-alpha Reductase Inhibitor Use Improve The Efficacy of Intravesical Bacille Calmette-Guérin (BCG) for Non-muscle Invasive Bladder Cancer?

INTRODUCTION

Despite major advancements in systemic therapies and utilization of multimodal approaches for muscle invasive bladder cancer (MIBC), non-muscle invasive bladder cancer (NMIBC) management remains relatively unchanged for decades, with BCG installation as a primary therapy which decreases risk of disease recurrence and progression [1].

Epidemiologic data strongly supports the idea that men are at an increased risk of developing bladder cancer compared to women, although women are more likely to present with advanced disease [2– 4]. While environmental or lifestyle factors, such as industrial chemicals and cigarette smoke exposure may contribute to these differences, recent studies demonstrate that even after adjusting for these factors, sex-related differences in incidence persist [5]. Biologic differences related to sex hormones and sex hormone receptor activity remain a principal explanation [6]. In particular, androgens and the androgen receptor have been implicated in the embryologic origin and development of the bladder as well as bladder carcinogenesis in animal models [6, 7].

Pre-clinical models have suggested surgical castration may lower the incidence of bladder cancer induced by a chemical carcinogen in male mice. More recently, retrospective clinical studies have suggested that 5-alpha reductase inhibitors (5-ARIs) or androgen deprivation therapy may decrease the incidence and progression of bladder cancer [8, 9]. Notably, a study by Shiota et al. indicated that use of 5-ARIs or other androgen suppressive medication may decrease the recurrence rate of NMIBC, with a HR of 0.36 [8].

To our knowledge, the influence of a 5-ARI inhibitor on the efficacy of intravesical instillations of BCG has not been studied. Intravesical BCG remains an integral part of NMIBC treatment for over 40 years [1]. Our objective was therefore to identify if the combination of BCG with 5-ARI use may suppress bladder cancer recurrence compared to BCG alone.

MATERIALS & METHODS

RESULTS

A total of 206 males met the inclusion criteria, including 39 (18.9%) 5-ARI users and 167 (81.1 %) non-users. Thirty-one (79.5%) patients received dutasteride 0.5 mg daily and 8 (20.5%) patients finasteride 5 mg daily. No patients received other androgen suppressive therapy. The median age of men in the entire patient cohort was 70 (interquartile range (IQR): 68.6– 71.6). However, patients who received 5-ARI were generally older than those who did not receive the medications. Median age among 5-ARI recipients was 74 years, which was 6 years greater than the age of controls who did not receive 5-ARIs (p = 0.01). All other baseline characteristics between the two groups were comparable (Table 1). In the entire patient cohort, 121 (58.7%) had a smoking history, 177 (86%) had high-risk NMIBC, 55 (26.7 %) had second look TURBT and 150 (72, 8%) received BCG maintenance therapy (BCG). Median follow up was 3.31 (IQR 3.4– 3.9) years, with only 6 patients dying or being lost to follow-up in the first year. Generally, there was no statistically significant difference in the proportion of all cause between 5-ARI users and non-users. Notably, all patients in the 5-ARI cohort had high-risk disease on initial TUR [3, 16].

Overall, intravesical recurrence was observed during follow-up in 16 (41%) of patients who received 5-ARIs and 74 (44.3%) of patients who did not (p = 0.72).There was no statistical difference in 5-year recurrence free survival between 5-ARI users versus non-users (59% vs 55.7%; p = 0.72). Cox proportional hazards models assessed characteristics associated with disease recurrence in our cohort (Table 2). On multivariate analysis, smoking (HR 1.75, 95% CI = 1.09– 2.83, p = 0.02) was associated with a higher risk of recurrence. Furthermore, a Poisson regression model evaluates the impact of 5-ARI use on the total number of recurrences, which was not significantly different between patients who received 5ARIs and those who did not when adjusted for other baseline risk factors. Disease progression was observed in only one (2.6%) 5-ARI user and in three (1.8%) non-users.

A pre-planned subgroup analysis subsequently evaluated the effect of 5-ARI use according to the completeness of BCG therapy. There a lower prevalence of 5-ARI users among those receiving only induction BCG versus those who also received BCG maintenance (groups I vs II and III) which was above the convention for statistical significance (10.7% vs 22%, p = 0.066). On Kaplan-Meier analysis, there was no significant differences in recurrence-free survival by 5-ARI use among the subgroups of BCG treatment received (Fig. 1). However, the proportion of patients who completed prescribed BCG installations was higher among 5-ARI users (Table 3). Finally, we examined the effect of 5-ARI on tumor recurrence at one year from induction BCG treatment, finding a non-significant decreased recurrence rate among 5-ARI users (33.3% vs 58.3%, p = 0.61).

Fig. 1

Kaplan-Meier analysis of recurrence-free survival with and without 5-ARI stratified by amount of BCG received: (A) Group I < 6 instillations (B) Group II < 1 year maintenance (C) Group III at least 1 year maintenance.

DISCUSSION

The treatment landscape of bladder cancer continues to evolve. The clinical relevance of the AR and AR inhibition in bladder cancer remains unclear, but the accessibility and familiarity to urologists or AR-targeting medications present excellent potential for combination therapies. Intravesical BCG is the only treatment recognized to decrease both recurrence pathologic progression rates of NMIBC [1]. However, BCG therapy still has limitations: approximately 30% of patients treated with intravesical BCG failed to respond, and treatment may cause bladder and systemic toxicity which is responsible for treatment discontinuation. Moreover, recurrent BCG shortages in certain regions in the last several years have been problematic. Therefore, novel therapeutic strategies are needed to improve the effectiveness of BCG for NMIBC.

In preclinical models the androgen receptor (AR) has been implicated in the development and progression of bladder cancer [6, 7, 11]. More recently, several clinical studies have suggested that men who receive 5-ARIs may have better bladder cancer outcomes [8, 9, 12, 13]. Shiota et al. demonstrated a reduced bladder cancer recurrence rates among patients who received 5-ARIs or androgen deprivation therapy in a cohort of 228 Japanese men [8]. Wu et al. similarly reported in a cohort of 274 men that the use of androgen suppression therapy was associated with a lower risk of bladder tumor recurrence [9]. In a population-based cohort of 10,720 Finnish men diagnosed with bladder cancer, those who received 5-ARIs had a 25% reduction in disease specific survival compared to the control group who received α-blockers [12]. Finally, an evaluation of AR expression in bladder tumors from patients who also received androgen deprivation therapy suggests that these beneficial effects may be due to an AR-mediated effect of the loss of androgens [14].

To our knowledge this is the first study to specifically assess whether 5-ARIs may impart a benefit in men with NIMBC receiving BCG therapy. BCG therapy relies on the interaction between the bacillus and immunocompetent cells which stimulate the secretion of inflammatory cytokines such as interleukin (IL)-1 IL-2, IL-6 and interferon). Prior pre-clinical research has suggested that targeting AR signaling may enhance BCG immunotherapeutic effects via cytokine alterations [15].

Our results contrast previous reports which found a statistically significant association of androgen suppression therapy (AST) with disease progression for NIMBC [8, 9]. AST in these studies was defined as 5-ARI, GnRH agonist or antagonist, or anti-androgen therapy; our study evaluated only 5-ARIs. The time period in our study was shorter and more recent, though the follow-up time and number of patients is similar between studies. The most important difference appears that our study includes a more homogeneous cohort of patients, where only patients who received induction BCG were evaluated. Thus, the proportion of high-risk patients differences widely between our study and prior studies. With our inclusion criteria, 92% of patients in our study were high-risk, compared to 51% in the recent study by Wu [9]. Thus, 20% of high-risk patients in our study received 5-ARIs, compared to 11% in the study by Wu et al. No significant improvement in recurrences or progression with AST in their high-risk subgroup, with all NMIBC patients who progressed under AST initially showing high-risk disease [9]. With a highly homogeneous cohort, our study thus suggests 5-ARIs are ineffective to slow the progression of high-risk NMIBC.

The corollary is that the previously observed benefit of 5-ARIs may be most pronounced in patients with low grade, AR-expressing tumors. Prior immunohistochemistry studies have suggested that AR expression is more frequent in low grade tumors. Ide et al. conducted a meta-analysis to assess the expression of androgen receptor (AR) and estrogen receptors (ERs) in bladder cancer and found downregulation of AR expression in high-grade bladder tumors [16].

As a novel finding, we found fewer patients receiving 5-ARIs discontinued BCG treatments compared to men who did not receive 5-ARIs. Reasons for discontinuation were presumed due to intolerance, though this information was not available from our chart review. Nonetheless, our results suggest it is worthwhile to further investigate whether 5-ARI use may improve the tolerability of men to BCG-induced lower urinary tract symptoms.

Our study has several strengths and limitations. Accurate assessment of 5-ARIs use was performed routinely at pre-operative pharmacist consultations. Further, the number of BCG treatments received was accurately recorded by our pharmacy department. Finally, we present a relatively large and homogeneous cohort of patients whose care aligned with established guidelines. Limitations include those inherent to single-institution retrospective studies and the lack of AR staining of tumor samples. Further, the proportion of patients who received 5-ARIs was low in our study, but comparable to prior studies.

In summary, our results suggest that 5-ARI use does not significantly decrease the risk of intravesical recurrence of NMIBC patients receiving BCG. In contrast to prior studies, this negative finding may be attributable to our population of NMIBC which was almost exclusively high-risk. Notably, we did however observe that discontinuation of BCG appears less frequent among patients who received 5-ARI compared to controls who did not. Further research is warranted to investigate this finding.