ENZAMET: Decipher score may refine triplet therapy use in mHSPC

The Decipher Prostate Metastatic Classifier (DPMC) score predicted which patients with metastatic hormone-sensitive prostate cancer (mHSPC) would benefit from the addition of docetaxel to androgen deprivation therapy (ADT) plus enzalutamide (Xtandi), according to a prespecified biomarker analysis of the ENZAMET trial (NCT02446405) presented at the 2026 American Society of Clinical Oncology Annual Meeting.¹

Christopher Sweeney, MBBS, DHS, a medical oncologist and director of the immunoGENomics Cancer Institute at the University of Adelaide in Australia, presented the findings. He characterized the results as level 1B evidence consistent with prior findings from the CHAARTED (NCT00309985) and STAMPEDE trials (NCT00268476), identifying patients more likely to benefit from docetaxel.

Background

Since 2014, the treatment landscape of mHSPC has been transformed by the sequential demonstration that adding docetaxel to ADT alone improves survival in patients with poorer prognosis disease, that androgen receptor pathway inhibitors (ARPIs) added to ADT improve outcomes across all major prognostic groups, and more recently that triplet therapy regimens incorporating abiraterone acetate (Zytiga), darolutamide (Nubeqa), or enzalutamide added to ADT plus docetaxel further prolong survival.^1,2 With approximately 50% of patients alive at 8 years in the ENZAMET study,¹ a key unresolved question is which patients specifically benefit from the addition of docetaxel to an ARPI-based backbone, as no randomized data comparing ADT plus ARPI with or without docetaxel exist.

Prior work in CHAARTED and STAMPEDE demonstrated that the Decipher Prostate Genomic Classifier—a 22-gene expression panel scored from 0 to 1.0 using a GLMNET algorithm on formalin-fixed paraffin-embedded tissue—predicts OS benefit from adding docetaxel to ADT, with higher Decipher scores associating with greater docetaxel benefit.^1,3 That finding reached level 1B evidence and informed the hypothesis tested in ENZAMET.

Study hypothesis and design

ENZAMET is an investigator-sponsored phase 3 trial that randomly assigned 1125 patients with mHSPC to ADT plus a standard nonsteroidal antiandrogen (arm A, control) or ADT plus enzalutamide 160 mg/day (arm B). After 88 patients had accrued, the protocol was amended following the 2014 CHAARTED results to allow planned early docetaxel; docetaxel use was added as a stratification factor rather than a randomization arm, reflecting recognition that docetaxel is not standard of care for all patients. Accrual to ADT plus enzalutamide plus docetaxel was contemporaneous with accrual to ADT plus enzalutamide alone.

A prespecified statistical analysis plan—written prior to any analysis of ENZAMET specimens, locked and stored with the University of Sydney Clinical Trials Centre in October 2025—hypothesized that patients with tumors having a DPMC score greater than 0.85 (the approximate median score from the combined CHAARTED and STAMPEDE analysis using Decipher version 1.2) treated with docetaxel added to ADT plus enzalutamide would have a survival benefit compared with those receiving ADT plus enzalutamide alone, independent of known prognostic clinical variables. Conversely, patients with tumors with a DPMC score of 0.85 or lower were hypothesized not to benefit from adding docetaxel to the doublet. A secondary hypothesis was that patients with tumors with a DPMC score greater than 0.85 treated with ADT plus enzalutamide would have a worse prognosis than those with a score of 0.85 or lower on the same treatment.

Methods

Of 1125 enrolled participants, 1071 consented for biomarker analyses. Participant samples were centralized at the Chris O'Brien Lifehouse biobank (N=809) with centralized hematoxylin and eosin review. Gene expression profiling was performed at a Clinical Laboratory Improvement Amendments-certified commercial laboratory (Veracyte) on 764 samples; the final biomarker cohort consisted of 634 patients with reliable RNA profiles. The ADT plus enzalutamide plus or minus docetaxel cohort used for the primary analysis comprised 320 patients.

The statistical approach proceeded in 3 steps: univariable analyses, followed by multivariable analysis with individual patient weighting and propensity scoring using inverse probability of treatment (docetaxel) weights (IPTW)—estimated via LASSO-penalized logistic regression—to address the lack of randomization to docetaxel, and finally a test for interaction to assess prediction. All analyses were designed and performed independently by 2 groups of statisticians.

Prognostic value of DPMC in the full biomarker cohort

In the full biomarker cohort (N=634), a higher DPMC score was significantly associated with worse OS (HR, 1.53; 95% CI, 1.19–1.97; P < .001). Patients with tumors with a DPMC score greater than 0.85 were more likely to be younger, have synchronous metastases, and have a higher Gleason score (Gleason 8–10: 82% vs 59%; P < .001), and more likely to have synchronous vs metachronous metastases (75% vs 62%; P = .001).

Patients treated with docetaxel in the overall cohort had greater rates of multiple poor prognostic features—including higher DPMC score, higher Gleason score, synchronous and high-volume metastases, and younger age—and demonstrated worse unadjusted OS compared with patients not treated with docetaxel (HR, 1.30; 95% CI, 1.02–1.66; P = .03), underscoring the substantial confounding inherent to nonrandomized docetaxel assignment.

Results in the ADT plus enzalutamide plus or minus docetaxel cohort

In the ADT plus enzalutamide plus or minus docetaxel cohort (N=320), patients treated with the triplet similarly had a greater rate of adverse prognostic features compared with those receiving the doublet. In unadjusted analyses by DPMC cutoff, patients with a DPMC score of 0.85 or lower who received the triplet had significantly worse OS than those receiving the doublet (HR, 3.02; 95% CI, 1.59–5.76; P < .001), despite the doublet group having more favorable prognostic characteristics. In contrast, patients with a DPMC score greater than 0.85 treated with the triplet had essentially the same OS as those treated with the doublet (HR, 1.08; 95% CI, 0.69–1.71; P = .73), even though the triplet group in this stratum carried a greater burden of poor prognostic features, including higher Gleason score, synchronous metastases, and higher ACE-27 comorbidity score.

When examining OS by DPMC score and disease volume in the ADT plus enzalutamide arm, patients with high-volume disease and DPMC greater than 0.85 had a 5-year OS of 41% vs 81% for those with DPMC of 0.85 or lower (HR, 2.83; 95% CI, 1.07–7.49). In the ADT plus enzalutamide plus docetaxel arm, this prognostic difference was attenuated, with 5-year OS of 48% for DPMC greater than 0.85 vs 58% for DPMC of 0.85 or lower (HR, 1.07; 95% CI, 0.60–1.92), suggesting that in high-volume disease, the triplet may largely abrogate the negative prognosis conferred by high DPMC. A similar but more attenuated pattern was observed in low-volume disease.

Propensity score–weighted multivariable analysis

Standardized mean difference (SMD) analysis demonstrated substantial baseline imbalance in unfavorable risk factors in patients treated with the triplet vs the doublet. After IPTW propensity scoring using factors known at the time of randomization, most covariates moved within the predefined acceptable balance boundary (SMD of 0.2 or lower); however, residual imbalance remained for tumor volume, Gleason score, and timing of metastases in the triplet arm. Sweeney noted that propensity scoring did not fully adjust for all factors that drove the decision to prescribe docetaxel, including undefined clinical and/or biological factors.

In the IPTW-weighted propensity score multivariable analysis of OS, a statistically significant interaction between docetaxel and DPMC cutoff was observed (interaction P = .04). For patients with a DPMC score of 0.85 or lower, the weighted HR for OS with docetaxel added to ADT plus enzalutamide vs ADT plus enzalutamide alone was 1.94 (95% CI, 0.95–3.96; P = .07), suggesting no benefit and possible harm. For patients with a DPMC score greater than 0.85, the weighted HR was 0.75 (95% CI, 0.43–1.33; P = .33), providing initial evidence of potential OS benefit from the addition of docetaxel in this biomarker-selected group.

"This is the first time we're starting to see some evidence that the triplet does work for a biomarker-selected patient compared with the hormonal therapy alone approach," Sweeney said.

The DPMC interaction finding was consistent across secondary end points. For progression-free survival, the interaction P was .03 in the weighted analysis, with a weighted HR of 1.83 (95% CI, 0.98–3.42) for DPMC of 0.85 or lower vs 0.74 (95% CI, 0.44–1.25) for DPMC greater than 0.85. For prostate-specific antigen progression-free survival, the interaction P was .03, with weighted HRs of 1.72 (95% CI, 0.93–3.19) and 0.70 (95% CI, 0.43–1.14), respectively. For prostate cancer–specific survival, the interaction P was .01, with weighted subdistribution HRs of 3.49 (95% CI, 1.46–8.33) and 0.88 (95% CI, 0.47–1.64), respectively.

Limitations

Sweeney outlined 4 key limitations. First, the lack of randomization to docetaxel is the primary limitation; imbalance was partially addressed with IPTW and propensity scoring, but residual confounding from undefined clinical and/or biological factors could not be excluded. Second, subgroup sample sizes were small, and results should be interpreted with appropriate caution. Third, the analysis did not include patients receiving radiation to the prostate or stereotactic body radiation therapy to metastatic disease, and the clinical decision to use docetaxel vs radiation in those with DPMC greater than 0.85 remains at the clinician's discretion. Fourth, global access to the Decipher test remains a practical barrier to implementation; health technology assessments are being planned to evaluate the health care utilization and quality-of-life implications of more effective cancer control with docetaxel in the DPMC greater than 0.85 group and avoidance of docetaxel in those with DPMC of 0.85 or lower.

Conclusions

Sweeney concluded that a DPMC score greater than 0.85 is prognostic for worse OS on ADT plus enzalutamide and is associated with evidence of OS benefit from adding docetaxel to the doublet. Conversely, patients with a DPMC score of 0.85 or lower showed less evidence of benefit and may be spared docetaxel toxicity. He characterized the findings as level 1B evidence consistent with CHAARTED and STAMPEDE, representing a step toward identifying patients more likely to benefit from docetaxel in the modern ARPI-based treatment era.

"There is some evidence here, supported by our statistical analysis plan, that the classifier predicted the postulated benefits of adding docetaxel to ADT and enzalutamide," Sweeney said.

REFERENCES

1. Sweeney C, Proudfoot JA, Kaur S, et al. Assessment of the ability of Decipher Prostate Genomic Classifier (DGC) >0.85 to identify patients who benefit from adding docetaxel (DOC) to androgen deprivation therapy (ADT) plus enzalutamide (ENZ): level 1B evidence from the ENZAMET study. J Clin Oncol. 2026;44(suppl 16). Abstract 5001). doi:10.1200/JCO.2026.44.16_suppl.5001

2. Zhang AY, Davis ID, Thomas H, et al; Australian and New Zealand Urogenital and Prostate Cancer Trials Group (ANZUP). 8-year outcomes of enzalutamide (ENZA) versus a non-steroidal anti-androgen (NSAA) for metastatic, hormone-sensitive prostate cancer (ENZAMET; ANZUP 1304). J Clin Oncol. 2025;43(suppl_16):5090. doi:10.1200/JCO.2025.43.16_suppl.5090

3. Grist E, Dutey-Magni P, Parry MA, et al. Tumor transcriptome-wide expression classifiers predict treatment sensitivity in advanced prostate cancers. Cell. 2025;188(20):5717-5734.e10. doi:10.1016/j.cell.2025.07.042