Can targeting the androgen receptor in localized prostate cancer provide insights into why men with metastatic castration-resistant prostate cancer die?

Abstract

The improved survival observed with abiraterone acetate or enzalutamide in patients with metastatic castration-resistant prostate cancer (mCRPC) who experience disease progression despite conventional androgen-deprivation therapy (ADT) validates the androgen receptor (AR) as a viable target in this group of patients. Yet many questions remain, particularly in light of the so-called bookends of the published clinical data that lead to the approval of these drugs. On one end of the spectrum, the fairly remarkable success that has been observed with these agents has been paired with the equally sobering realization that the development of resistance to these agents is nearly universal and is often associated with the development of a highly aggressive phenotype with rapidly progressive, fatal disease, for which few viable therapeutic options exist. At the other end of the spectrum, investigators have been pushing the envelope to use these active agents in earlier disease states, where disease burden is lower, patients are less heavily pretreated, and potentially, a broader therapeutic index can be exploited. In the article that accompanies this editorial, Taplin et al report the results of a phase II clinical trial of preprostatectomy (neoadjuvant) abiraterone acetate. Although the results of this trial do not support the use of this approach as a standard of care, they provide important insights into the questions at both ends of the therapeutic spectrum. For many cancer types, the development of systemic therapy for high-risk localized disease has been predicated on a potential impact on both local and micrometastatic disease control. In the case of localized prostate cancer, there is every reason to believe that neoadjuvant ADT should be of utility. First, a cohort of patients with a high risk of recurrence can be readily identified, so enrichment strategies that result in a more robust activity signal can be developed. Second, oncogenesis and disease progression is deeply dependent on a single signaling pathway, the AR, with the vast majority of cancers relying on this pathway for growth and survival. Third, the pathway can be effectively and easily pharmacologically targeted. Fourth, there is ample evidence that inhibition of the specific pathway is sufficient to halt the neoplastic phenotype. Finally, this approach has been extensively tested and validated in patients with advanced disease, for whom ADT is the standard of care, and widely used because of its significant palliative and life-prolonging effects. On the basis of these observations, the utility of neoadjuvant (preprostatectomy) ADT has been extensively explored. Surprisingly, although several series demonstrated a reduction in the rate of positive surgical margins, in no patient has this observation translated into an improvement in prostate-specific antigen (PSA) –progression-free survival, a generally accepted intermediate outcome end point, nor has a survival advantage been demonstrated for this approach. Consequently, neoadjuvant ADT is neither widely used nor recommended in treatment guidelines and algorithms, with National Comprehensive Cancer Network guidelines specifically indicating that, “Neoadjuvant ADT for RP [radical prostatectomy] is strongly discouraged.” Why has neoadjuvant ADT failed to demonstrate clinical benefit in patients with localized prostate cancer? It is possible that localized prostate cancer is simply not as uniformly and exquisitely dependent on androgens as mediators of transcriptional activity as anticipated. Stated another way, there might exist prostate cancer cells, even at this early stage of disease, that are resistant to ADT. Whether such resistance is innate or adaptive in response to the selective pressure of therapy is an important and unresolved question. Alternatively, it has been postulated that the problem lies not in a lack of sensitivity to ADT, but rather in inadequate androgen deprivation itself. These hypotheses can each be tested. The former—that there exist clones of cells that are not sensitive to ADT—is being prospectively tested by the Cancer and Leukemia Group B (CALGB [Alliance]) study 90203, an intergroup trial randomly assigning men with high-risk localized prostate cancer to neoadjuvant ADT plus docetaxel followed by radical prostatectomy versus immediate prostatectomy with no neoadjuvant therapy. The second premise: that conventional ADT is inadequate and results in incomplete tissue suppression of androgens, which in turn results in incomplete tumor killing, is the hypothesis addressed by Taplin et al. This hypothesis was put forth by Mostaghel et al, who had previously demonstrated that so-called standard ADT inadequately suppressed androgen-regulated gene expression, and suggested that ablation or reduction of circulating androgens was not necessarily reflective of androgen ablation within the prostate itself. JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 32 NUMBER 33 NOVEMBER 2