Targeted androgen pathway suppression in localized prostate cancer: a pilot study

Abstract

Purpose: Ligand-mediated activation of the androgen receptor (AR) is critical for prostate cancer (PCa) survival and proliferation. The failure to completely ablate tissue androgens may limit suppression of PCa growth. We evaluated combinations of CYP17A and 5-α-reductase inhibitors for reducing prostate androgen levels, AR signaling, and PCa volumes.

Patients and methods: Thirty-five men with intermediate/high-risk clinically localized PCa were randomly assigned to goserelin combined with dutasteride (ZD), bicalutamide and dutasteride (ZBD), or bicalutamide, dutasteride, and ketoconazole (ZBDK) for 3 months before prostatectomy. Controls included patients receiving combined androgen blockade with luteinizing hormone-releasing hormone agonist and bicalutamide. The primary outcome measure was tissue dihydrotestosterone (DHT) concentration.

Results: Prostate DHT levels were substantially lower in all experimental arms (0.02 to 0.04 ng/g v 0.92 ng/g in controls; P < .001). The ZBDK group demonstrated the greatest percentage decline in serum testosterone, androsterone, and dehydroepiandrosterone sulfate (P < .05 for all). Staining for AR and the androgen-regulated genes prostate-specific antigen and TMPRSS2 was strongly suppressed in benign glands and moderately in malignant glands (P < .05 for all). Two patients had pathologic complete response, and nine had ≤ 0.2 cm(3) of residual tumor (defined as a near-complete response), with the largest numbers of complete and near-complete responses in the ZBDK group.

Conclusion: Addition of androgen synthesis inhibitors lowers prostate androgens below that achieved with standard therapy, but significant AR signaling remains. Tissue-based analysis of steroids and AR signaling is critical to informing the search for optimal local and systemic control of high-risk prostate cancer.

Trial registration: ClinicalTrials.gov NCT00298155.

Conflict of interest statement

Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.

Figures

Fig 1.

CONSORT diagram. ZBD, goserelin, bicalutamide, and dutasteride; ZBDK, goserelin, bicalutamide, dutasteride, and ketoconazole; ZD, goserelin and dutasteride.

Fig 2.

Prostate androgen levels after 3 months of multitargeted neoadjuvant androgen suppression. Tissue androgens from prostatectomy specimens after no treatment (No Rx), or 3 months of combined androgen blockade with a luteinizing hormone-releasing hormone agonist (goserelin) and bicalutamide (ZB); goserelin combined with the SRD5A inhibitor, dutasteride (ZD); goserelin combined with bicalutamide and dutasteride (ZBD); and all three of these agents combined with the CYP17A inhibitor, ketoconazole (ZBDK). Levels of (A) dehydrotestosterone (DHT), (B) testosterone, (C) androstenedione (AED), and (D) dehydroepiandrosterone (DHEA) were measured by mass spectroscopy. The difference in tissue androgen levels between each treatment cohort and the control (ZB) group was evaluated by linear regression. Statistically significant P values (P < .05) or those trending toward significance (P < .10) are presented in italics below the relevant treatment group. Absolute values of the mean and standard deviations are presented in Table 2. (E) Residual prostate testosterone levels are shown on an expanded x-axis (corresponding to the region denoted with a dark vertical bar in panel B). (*) P < .05. (†) P < .005.

Fig 3.

Change in serum androgen levels after 3 months of multitargeted neoadjuvant androgen suppression. The percentage change in androgen levels from baseline are depicted for each treatment group after 3 months of combined androgen blockade with goserelin and bicalutamide (ZB), goserelin and dutasteride (ZD), goserelin with bicalutamide and dutasteride (ZBD), and all three agents combined with ketoconazole (ZBDK). Levels of (A) dehydrotestosterone (DHT), (B) testosterone, (C) dehydroepiandrosterone sulfate (DHEA-S), and (D) androsterone were measured by mass spectroscopy in serum samples obtained before starting treatment and on the morning of prostatectomy. Differences between each treatment cohort and the control (ZB) group were evaluated by linear regression. Statistically significant P values (P < .05) or those trending toward significance (P < .10) are presented in italics below the relevant treatment group. No comparisons for DHEA or androstenedione were significant (data not shown).

Fig 4.

Immunohistochemical expression of androgen receptor (AR) and androgen-regulated genes after 3 months of multitargeted neoadjuvant androgen suppression. A tissue microarray comprising cores of benign and cancer tissue from each patient was analyzed for (A, B) nuclear expression of AR, and (C, D) cytoplasmic expression of prostate-specific antigen (PSA) and (E, F) TMPRSS2. Immunostaining was scored separately in benign (A, C, E) and cancer glands (B, D, F) using a compositional method based on the percentage of cells at each intensity level (0 for none, 1 for faint, and 2 for intense). The stacked bar graphs represent the proportion of cores in each treatment group that stain at the indicated intensity level (light gray, none; medium gray, faint; dark gray, intense). Differences between the indicated cohort and the untreated tissues were evaluated by unpaired t tests with Welch's correction. Statistically significant P values (P < .05) are presented in italics below the relevant treatment group. Rx, treatment; ZB, goserelin and bicalutamide; ZBD, goserelin, bicalutamide, and dutasteride; ZBDK, goserelin, bicalutamide, dutasteride, and ketoconazole; ZD, goserelin and dutasteride.

Fig 5.

Distribution of tumor volume and prostate-specific antigen (PSA) nadir by treatment group. Tumor volume at (A) prostatectomy and (B) nadir PSA are depicted for each treatment group after 3 months of combined androgen blockade with goserelin and bicalutamide (ZB), goserelin and dutasteride (ZD), goserelin with bicalutamide and dutasteride (ZBD), and all three agents combined with ketoconazole (ZBDK). Dotted lines represent near complete response less than 0.2 cm3 (A) and nadir PSA less than 0.2 ng/dL (B). P values for differences in treatment groups were assessed by two-sample tests of proportions.

Fig A1.

Prostatic expression of androgen receptor (AR) and prostate-specific antigen (PSA) after 3 months of multitargeted neoadjuvant androgen suppression. A tissue microarray comprising six cores of benign and cancer tissue from each patient was analyzed (A) for nuclear expression of AR and (B) for cytoplasmic expression of PSA in benign and malignant prostate epithelium. Representative examples of high- and low-intensity staining from different patients in the untreated and multitargeted treatment arms are shown for both benign and cancer cores. ZBDK, goserelin, bicalutamide, dutasteride, and ketoconazole .

Fig A2.

Association of prostate-specific antigen (PSA) nadir with tumor volume, serum androgens, and prostatic immunohistochemistry (IHC). Nadir PSA values were used to dichotomize all study patients into those with PSA nadir less than or more than 0.2 ng/dL to evaluate differences in (A) tumor volume, (B) serum androgen levels, and prostate IHC expression of androgen receptor (AR), PSA, and TMPRSS2 in (C) benign and (D) cancer tissue. Differences were evaluated by linear regression. Statistically significant P values (P < .05) or those trending toward significance (P < .10) are presented in italics below the relevant treatment group. AED, androstenedione; DHEA, dehydroepiandrosterone; DHT, dihydrotestosterone.