Phase II proof-of-concept study of atorvastatin in castration-resistant prostate cancer

Abstract

Objectives: To test for evidence of statin-mediated effects in patients with castration-resistant prostate cancer (CRPC) as post-diagnosis use of statins in patients with prostate cancer is associated with favourable survival outcome.

Patients and methods: The SPECTRE trial was a 6-weeks-long proof-of-concept single-arm Phase II treatment trial, combining atorvastatin and androgen deprivation therapy in patients with CRPC (regardless of metastatic status), designed to test for evidence of statin-mediated effects in patients with CRPC. The primary study endpoint was the proportion of patients achieving a ≥50% drop from baseline in prostate-specific antigen (PSA) levels at any time over the 6-week period of atorvastatin medication (PSA response). Exploratory endpoints include PSA velocity and serum metabolites identified by mass spectrometry .

Results: At the scheduled interim analysis, one of 12 patients experienced a ≥50% drop in PSA levels (primary endpoint), with ≥2 patients satisfying the primary endpoint required for further recruitment. All 12 patients experienced substantial falls in serum cholesterol levels following statin treatment. While all patients had comparable pre-study PSA velocities, six of 12 patients showed decreased PSA velocities after statin treatment, suggestive of disease stabilization. Unbiased metabolomics analysis on serial weekly blood samples identified tryptophan to be the dominant metabolite associated with patient response to statin.

Conclusions: Data from the SPECTRE study provide the first evidence of statin-mediated effects on CRPC and early sign of disease stabilization. Our data also highlight the possibility of altered tryptophan metabolism being associated with tumour response.

Keywords: #PCSM; #ProstateCancer; #uroonc; atorvastatin; castration resistant prostate cancer; cholesterol; prostate specific antigen; statins.

Fig. 1

Schematic of trial workflow. Left panel: Overall study design along with details of study parameters (before, during and after atorvastatin treatment) and study endpoints. Right panel: The SPECTRE study was designed (at 90% power, 10% 1‐sided significance level) to distinguish an ‘ineffective’ PSA response rate of ≤10% from an ‘effective’ PSA response rate of ≥30%. This required 12 patients to be recruited at the first stage. If ≤1 of these 12 patients responded, recruitment would not proceed to the second stage, which would require a total of 35 evaluable patients (with 40 as the recruitment target; Appendix S1).

Fig. 2

(A) Cholesterol levels as indicated for each informative patient (n = 12; ***P < 0.001, ns = not significant; two‐way ANOVA with Sidak's test; mean values ± SD are shown). Free fatty acid (B; n = 11) and triglyceride (C; n = 12) levels at the start and end of treatment for each patient (*P < 0.01, **P < 0.001; Wilcoxon signed rank test; mean values ± SD are shown). (D) Waterfall plot showing the maximal percentage change in PSA levels for each patient that occurs at any point after treatment start (n = 12). (E) PSA (ng/mL) levels as indicated at the start and end of treatment for each patient (n = 12; ns = not significant; Wilcoxon signed rank test; mean values ± SD are shown).

Fig. 3

(A) PSA velocity for each patient before and during treatment (n = 12; ns = not significant; Wilcoxon signed rank test; mean values ± SD are shown). (B) Patients stratified into groups of those with increasing or decreasing PSA velocity comparing before and during treatment (n = 6 per group; *P < 0.05, **P < 0.005; Mann–Whitney test; mean values ± SD are shown). (C) Waterfall plot showing the change in PSA velocity comparing before and during treatment (n = 12). The dotted line separates the patients with increasing velocity (left) and those with decreasing velocity (right).