Prostate cancer is the most common malignancy in Western societies and the second most common cause of male cancerrelated death in the UK and USA, accounting for≈ 12% of all male cancer-related deaths [1, 2]. When confined to the prostate gland, the disease is curable with local therapy (radical prostatectomy, external beam radiotherapy, brachytherapy or cryotherapy). However, despite the use of PSA for screening, in 15–33% of men local therapy fails and they develop incurable metastatic disease [3, 4]. Activation of the androgen receptor (AR) by androgenic steroids including testosterone and its more potent 5α-reduced metabolite, 5α-dihydrotestosterone (DHT), regulates the transcription of a diverse range of target genes involved in prostate cell proliferation, differentiation, and apoptosis [5]. Androgen deprivation by medical or surgical castration remains the mainstay of treatment and> 90% of men with metastatic prostate cancer initially respond rapidly and often dramatically to castration, with improvements in bone pain, regression of soft tissue metastasis, and steep declines in PSA level. However, the duration of response is frequently short (12–33 months) and in almost all patients is followed by the emergence of castration-resistant prostate cancer (CRPC) that, untreated, is invariably fatal within 9–12 months [3]. Systemic chemotherapy with docetaxel for patients with CRPC confers a modest survival advantage (2–3 months) and is effective for palliating symptoms [6], but the duration and rate of response is limited. Other chemotherapy, eg mitoxantrone, has not been shown to improve survival but might help with symptom control [6]. There is an urgent need for new agents that provide palliation and improve survival.

Continued androgen-dependent activation of a ‘hypersensitive’AR in castrated patients secondary to AR gene amplification [7], mutations in the AR gene [5], increased AR expression [8], or alterations in AR corepressor/co-activator function [9] appear to be important mechanisms of castration resistance. These could be reversed by suppressing circulating androgens, inhibiting the binding of biologically active androgenic steroids to the ARs or disrupting AR