ENZARAD (ANZUP 1303)
ForHigh-risk localized/locally-advanced prostate, EBRT candidates, 2yr ADT
HR 0.88
95% CI 0.67-1.15, p=0.34; 8yr 74% vs 72%, did not meet 1° EP
TL;DR8yr MFS 74% vs 72%, HR 0.88 (0.67-1.15), p=0.34: enzalutamide added to 2yr ADT + RT missed its primary endpoint.
Reported via UroToday →
The signal tracks the pelvic field, not the drug: MFS HR 0.47 (0.29-0.76) with pelvic RT planned and 0.43 (0.20-0.92) in cN1, both declared before randomization. RT was 78Gy or 46Gy plus brachy boost with 46Gy elective nodes for cN1, so this transfers directly. It moves the intensification decision toward pts you were already covering nodally.
In cN1 or node-covered high-risk localized prostate going to 2yr ADT plus definitive EBRT, this supports adding enzalutamide; in cN0 prostate-only fields, including 'very high-risk' by Gleason and PSA, it does not.
The benefit tracks the field, not the risk label: MFS HR 0.47 (0.29-0.76) where pelvic RT was planned, versus 0.85 (0.64-1.13) in 'very high-risk' disease. RT was 78Gy or 46Gy plus brachy boost, with 46Gy elective nodes plus gross nodal boost for cN1, so it maps onto standard practice. It ties the intensification call to your nodal coverage decision.
Enzalutamide x24mo was tested against an active NSAA x6mo, not ADT alone, so the flat all-comer MFS HR 0.88 argues against routine ARSI substitution rather than against intensification in general. The cN1 MFS HR 0.43 (0.20-0.92) is where the drug earns its place. Sequencing and biomarker selection remain open.
11 details
International investigator-initiated phase 3, N=802 from 8 countries, accrued March 2014 to June 2018, median follow-up 8 years. Powered at 80% for an HR 0.67. Primary was changed from OS to MFS during the trial because deaths ran below projection.
High-risk clinically localized or locally-advanced disease suitable for EBRT: 90% Gleason 8-10, 36% PSA >20 ng/ml, 12% cN1 by conventional imaging. 40% planned for pelvic RT and 8% for a brachytherapy boost.
Experimental: enzalutamide 160 mg daily x 24 months plus LHRH agonist x24 months. Control: conventional NSAA x6 months plus LHRH agonist x24 months. The control is an active antiandrogen, not ADT alone.
Prostate to 78Gy, or 46Gy plus brachytherapy boost, starting 16 weeks after hormonal therapy. Pelvic nodal RT required for cN1 (46Gy elective nodes plus boost to gross nodes) and optional for cN0 but declared before randomization. QA was unusually tight: credentialing, benchmarking, real-time review of the first 5 plans per site, then 20% random sampling.
Primary: MFS. Secondary: OS, CSS, PSA PFS, clinical PFS, castration resistance, HRQoL, adverse events, cost-effectiveness. Main effects by unstratified log-rank at alpha=0.05, Cox HRs, all p-values nominal without multiplicity adjustment, five prespecified subgroups tested by interaction.
Primary MFS not met. PFS favored enzalutamide at a nominal p=0.044 and OS was flat; the subgroup detail sits in the table above.
| Subgroup | MFS HR (95% CI) | OS HR (95% CI) |
|---|---|---|
| cN1 (regional nodes) | 0.43 (0.20-0.92) | 0.46 (0.17-1.26) |
| Pelvic field RT planned | 0.47 (0.29-0.76) | 0.53 (0.30-0.95) |
| 'Very high-risk' | 0.85 (0.64-1.13) | 0.81 (0.57-1.13) |
STAMPEDE's abiraterone-based intensification in non-metastatic high-risk disease produced an MFS HR 0.53 against ENZARAD's 0.88, and the ENZARAD cN1 subgroup HR mirrors the STAMPEDE result. The gap plausibly reflects population, not drug: cN1 11% vs 39%, median PSA 14 vs 35 ng/ml, cT3-4 47% vs 92%.
The pelvic-RT subgroup is not a clean randomized comparison of pelvic coverage: it carried 28% N1 and 62% Gleason 9/10 against 0% N1 and 49% Gleason 9/10 in the no-pelvic-RT group, so risk enrichment and biological interaction are entangled. The presenter named both explanations and could separate neither.
Two prespecified subgroups moving together on MFS and OS, with OS HRs tracking MFS HRs as ICECaP surrogacy predicts, is more internally consistent than a lone subgroup blip. It still does not establish that enzalutamide works only when the pelvis is treated.
Adequately powered phase 3 missed its primary MFS endpoint against an active NSAA control; benefit rests on two prespecified subgroups with overlapping risk composition.
- Which pelvic-RT subgroups drive the enzalutamide benefit
- Biomarkers identifying who needs ARSI intensification
- Whether pooled ARPI trial data confirms the nodal signal
📚 Sources · 📄 1 paper
Abstract
The longer read
A negative primary endpoint here is less surprising than it first reads, because the control arm was not ADT alone. Both arms received two years of LHRH agonist and the comparator carried six months of a conventional non-steroidal antiandrogen, so the question ENZARAD asked was whether swapping a modern ARSI for an older antiandrogen buys anything on top of an already-intensified backbone with high-quality radiation. Against that comparator, an all-comer MFS HR of 0.88 with a confidence interval crossing 1 is a fair answer: for most men with high-risk localized disease treated to 78Gy with two years of ADT, enzalutamide is not adding a detectable amount.
The more interesting question is why the STAMPEDE platform's non-metastatic high-risk result sits so far from this one, at an MFS HR of 0.53. Population is the obvious candidate and the numbers support it. ENZARAD enrolled 11% cN1 against 39%, a median PSA of 14 against 35 ng/ml, and cT3-4 in 47% against 92%. A trial that enrolls closer to the boundary of high risk has fewer events to prevent and a smaller absolute reservoir of occult micrometastatic disease for systemic intensification to act on. Eight-year MFS of 72% in the control arm is the tell: this population does well on radiation and two years of ADT, which compresses the room any additional agent has to work in.
The two subgroups that moved deserve more weight than a typical subgroup finding, without being mistaken for a randomized result. Both cN1 and planned pelvic RT were prespecified among five, and the pelvic-RT declaration was made before randomization rather than reconstructed afterward, which removes the worst version of the selection problem. The MFS and OS hazard ratios move together in both, which is what ICECaP surrogacy would predict if the effect were real rather than noise. Against that, five subgroups tested with nominal unadjusted p-values will occasionally produce a pair like this by chance, and the pelvic-RT and cN1 groups overlap substantially, so they are not two independent confirmations.
The confounding is the part that should temper confidence most. Pts selected for pelvic RT were 28% node-positive with 62% Gleason 9/10, against 0% node-positive and 49% Gleason 9/10 in those treated to the prostate alone. That is a materially higher-risk group, and higher-risk pts are precisely where an ARSI would be expected to show benefit regardless of what field was treated. The alternative reading, that sterilizing pelvic nodes with radiation is a precondition for enzalutamide to prevent distant metastases, is biologically coherent but no more supported by these data than the simpler risk-enrichment explanation. Both may hold. Neither is testable within this trial.
What this changes: for a node-negative man with Gleason 9 disease and a normal-appearing pelvis being treated prostate-only, ENZARAD is reassurance that skipping the ARSI is defensible, and the null result in the 'very high-risk' subgroup (HR 0.85) blunts the argument that Gleason and PSA alone should trigger intensification. For cN1 disease going to elective nodal coverage, it points the other way. The honest position is that ENZARAD relocates the intensification question from 'which high-risk pts' to 'which nodal-risk pts', and a pooled analysis across the ARSI trials in this setting is what would settle it.