Talazoparib plus enzalutamide shows no clear added benefit without DNA repair mutations

Adding the PARP inhibitor talazoparib to enzalutamide delayed disease progression in men with high-volume metastatic hormone-naïve prostate cancer, but the trial did not show a clear benefit over enzalutamide alone in patients whose tumours lacked actionable mutations. The combination also carried greater toxicity. The ZZ-FIRST results, from the Vall d'Hebron Institute of Oncology (VHIO), were published in Nature Medicine.

Trial design and outcomes

ZZ-FIRST was a prospective, multicentre, randomised phase 2 academic trial led by VHIO's Prostate Cancer Group. It enrolled 54 patients with high-volume metastatic hormone-naïve disease. They were randomised 1:2 to enzalutamide plus talazoparib (37 patients) or enzalutamide alone (17 patients).

Median radiographic progression-free survival was 45.3 months with the combination and 31.1 months with enzalutamide alone. Time to PSA progression and to castration resistance was also longer in the combination arm. Even so, the authors concluded that the combination was clinically active but showed no clear additional benefit in patients without actionable mutations, and that it was linked to more toxicity. The small size of the trial, and the uneven allocation between arms, mean the figures should be read with caution.

Why the biological hypothesis was tested in patients

The study addressed a question that has run through a decade of work on PARP inhibitors in prostate cancer. In 2015, a trial led by Joaquin Mateo showed that olaparib was effective in metastatic castration-resistant disease with alterations in DNA repair genes such as BRCA1, BRCA2 and ATM. That result became an early example of a biomarker predicting drug response in this disease and contributed to approval of the drug class in that setting.

Laboratory models had suggested a further possibility. Blocking the androgen receptor pathway might weaken DNA repair in tumour cells, making them vulnerable to PARP inhibition whatever their genetic profile. If true, the benefit could extend to patients without repair-gene mutations.

Spatial biopsies show what happens to DNA repair

Unlike earlier trials, ZZ-FIRST included a period of hormone treatment before randomisation. This allowed tumour biopsies to be collected both before and during androgen blockade. The team combined these longitudinal samples with spatially resolved transcriptomic analysis, tracking tumour adaptation directly in patients rather than relying on preclinical models or single-timepoint clinical data.

Treatment produced marked reprogramming of gene activity, including reduced cell division and proliferation. This did not, however, amount to a functional loss of DNA repair capacity. According to Mateo, the corresponding author, repair genes appeared less active mainly because the cells were proliferating less, not because they had acquired a true deficiency. That finding undermines the mechanistic rationale for using PARP inhibitors in tumours without underlying repair defects.

Implications for biomarker testing

The results are consistent with emerging data from a larger phase 3 study, which indicates that ARPI and PARPi combinations help most in tumours with alterations in DNA repair genes such as BRCA1, BRCA2, CDK12, PALB2 or ATM.

The authors argue that this strengthens the case for biomarker testing to select patients, particularly in earlier disease stages where benefit has to be weighed against long-term toxicity. For researchers, the work also offers a template for embedding serial biopsies and spatial profiling within randomised trials to study resistance mechanisms in situ. Prostate cancer is the second most common cancer worldwide, with an estimated 1.5 million new cases in 2022 and cases projected to double by 2040, so refining patient selection carries wide relevance.

Source: Prostate cancer drug combination adds toxicity without clear benefit for tumors lacking targetable mutations (Medical Xpress, 11 November 2026).

Image: AI-generated illustration.