TMPRSS2-ERG (TE) results in androgen-driven overexpression of ERG, which is involved in resistance to taxanes in preclinical models. In prior work we showed that TE expression in blood correlated with taxanes resistance in metastatic castration-resistant prostate cancer (mCPRC). Here, we studied if the detection of TE in primary tumors predicts taxanes activity in CPRC. We also explored the impact of prior abiraterone or enzalutamide (A/E) in blood TE detection and in TE predictive value.
mCRPC patients (pts) treated with taxanes in a multicenter biomarker study were included. Formalin-fixed paraffin-embedded (FFPE) tumors and peripheral blood mononuclear cells (PBMCs) fraction were tested for TE presence by RT-qPCR. FFPE were retrospectively obtained. PBMCs were prospectively collected prior to taxane initiation. PSA-PFS was evaluated by Kaplan-Meier analysis using log-rank test. Univariate analysis of TE status (+ vs-) was performed with Cox regression.
124 pts were included: 111 (89.5%) received docetaxel (Dx), 13 (10.5%) cabazitaxel (Cz) and 27 (21.8%) both. Fifty-seven (45.9%) tumors were TE+. Overall, no correlation between tumor TE expression and taxane benefit was observed in the whole population, or in the Dx or Cz group separately. However, in Dx-treated pts without prior A/E (N = 80, 72.1%), tumor TE+ correlated with lower PSA-PFS (median 8.6 vs 13.6 months; HR 1.7, p ≤ 0.05). No differences were observed in Dx treated pts with prior A/E (N = 31, 27.9%) according to tumor TE expression. In 44 pts, matched tumor and PBMC samples were available. Concordance between tumor an blood was 92.8% and 63.3% for pts with and without prior A/E, respectively. TE in blood was + in 1 (7%) pts with prior A/E and in 7 (23.3%) pts without prior A/E. As observed in FFPE samples, in patients without prior A/E to Dx (N = 28; 63.6%), blood TE+ correlated with lower PSA response (0% vs 61.9%, p ≤ 0.01) and reduced median PSA-PFS (3.34 vs 8.2 mM; HR 4.1 p ≤ 0.01).
The predictive value of TE in taxane resistance may be different depending on prior exposure to A/E. This is being tested in a multicenter prospective study.
Clinical trial identification
Legal entity responsible for the study
Hospital Clínic of Barcelona/Institut d'Investigacions Biomèdiques Agust Pi i Sunyer
A. González del Alba: Advisory boards: Sanofi, Janssen, Astellas, Bayer Travel expenses: Astellas, Sanofi, Janssen. All other authors have declared no conflicts of interest.