APCCC 2026: High-risk Localised, Locally Advanced, and Biochemically Recurrent PC
TL;DRConsensus (≥75%) reached on 44 of 134 questions (32.8%); fractionation for high-risk localised PC reached no consensus despite 88% favoring some hypofractionation.
HEATENZARADSTAMPEDEHYPO-RT-PCSHARPATLASPEACE-7THUNDER
Fractionation drew 88% for some form of hypofractionation but no consensus on which form, and 44 panellists abstained or declared themselves unqualified. Moderate hypofractionation (2.5–3.4 Gy/fx) is framed as the established option. SBRT for high-risk disease rests on non-randomised SHARP pooling, with 5-yr G2 GU toxicity of 23%, so it stays a trial question.
In very high-risk cN0 PC planned for RT + long-term ADT, panel opinion (65–71%) favors adding 2 yr abiraterone over enzalutamide. This does not settle the choice between ultrahypofractionation and moderate hypofractionation for high-risk disease.
Fractionation reached no consensus: 88% favored some hypofractionation, and 44 panellists abstained. Moderate hypofractionation (2.5–3.4 Gy/fx) is the established option, while SBRT for high-risk disease rests on non-randomised SHARP pooling. DIL focal boost up to 95 Gy over 77 Gy/35 fx is presented as validated dose escalation.
In very high-risk cN0 disease, 65–71% backed 2 yr abiraterone with RT + long-term ADT, and very few backed enzalutamide, tracking ENZARAD's null MFS (HR 0.88). A combined 83% would stop ADT after 2 yr with abiraterone. ATLAS, PEACE-7, THUNDER and DASL-HiCaP will test other ARPIs.
32–38% favored surgery as part of a multimodal approach for high-risk cN0 disease. The panel reached consensus against perioperative ADT + ARPI before RP, but voted before PROTEUS (pCR/MRD 8.9% vs 1.0%) was presented, so that position may shift.
9 details 3 trials watching
Modified Delphi process: 144 international experts drafted 313 multiple-choice questions, 134 on high-risk localised / locally advanced PC, PSA persistence and recurrence. 133 panellists voted by web survey before the conference.
Consensus defined a priori as 75% agreement, strong consensus 90%. Consensus reached on 44 of 134 questions (32.8%), seven of them strong.
No consensus on fractionation (Q19) despite a combined 88% favoring some hypofractionation. Moderate hypofractionation (2.5–3.4 Gy per fraction) is named as an option. MRI-defined DIL focal boost up to 95 Gy over 77 Gy in 35 fractions is described as validated in a randomised trial.
Very high-risk disease: RT + long-term ADT (2–3 yr) with the option of abiraterone for 2 yr. A combined 83% voted ADT can be stopped after 2 yr with abiraterone in most or selected pts (Q7).
The intensification votes follow STAMPEDE over ENZARAD, whose null MFS (HR 0.88, 95% CI 0.67–1.15) came in a lower-risk population. ATLAS, PEACE-7, THUNDER and DASL-HiCaP are testing other ARPIs in this setting.
44 panellists abstained or declared themselves unqualified on fractionation, which shrinks the informed denominator. The Q6 vote against perioperative ADT + ARPI before RP predates PROTEUS, which was presented and published after the conference.
The panel accepts RT and multimodal surgery as defensible for high-risk cN0 disease without settling the comparison. Its abiraterone-not-enzalutamide vote is drug-specific, not a class effect, and the ongoing ARPI trials will decide whether that distinction holds.
| Question | Setting | Panel vote |
|---|---|---|
| Q1–Q2 | High-risk cN0 M0 (conventional or next-gen imaging) | RT + long-term ADT ± ARPI 62–68%; surgery as multimodal 32–38% |
| Q3–Q4 | Very high-risk cN0 M0 | RT + long-term ADT + abiraterone 2 yr 65–71%; very few for enzalutamide |
| Q5 | Neoadjuvant systemic therapy before RT | Selected or most pts, combined 83%+ (87% combined) |
| Q6 | Perioperative ADT + ARPI before RP | Consensus against |
| Q7 | Stop ADT after 2 yr with abiraterone | Most or selected pts, combined 83% |
| Q19 | Fractionation, high-risk localised / locally advanced | No consensus; combined 88% some hypofractionation; 44 abstained / unqualified |
- Can other ARPIs replace abiraterone in very high-risk localised PC? recruiting Relugolix and Enzalutamide in Combination With Radiation Therapy for the Treatment of Very High Risk Prostate Cancer, OPTIMAL Trial Phase 2n=50 · primary completion 2030-11 · ph2 enzalutamide + relugolix + RT, NCCN very-high-risk
- Role of SBRT in high-risk localised PC n=146 · primary completion 2024-02 · randomised ph2 SBRT ± ADT, NCCN IR/HR, PET N0M0n=90 · primary completion 2028-08 · ph2 SBRT ± brachy boost + 6mo ADT in NCCN high-risk
- Does PROTEUS change the stance on perioperative ARPI before RP?
📚 Sources · 📄 1 paper
The longer read
For a radiation oncologist, this report is most useful where the panel split, not where it agreed. Only 44 of 134 questions crossed the prespecified 75% threshold. The questions closest to daily RT practice, treatment selection for high-risk cN0 disease and fractionation, both fell short of it. The 62–68% vote for RT plus long-term ADT ± ARPI over a multimodal surgical approach describes a field that treats both pathways as defensible, not one that has resolved the comparison. None of the evidence the authors cite tests RT against surgery directly in this population.
The intensification votes follow the trial evidence closely. A panel majority backs 2 yr abiraterone in very high-risk disease, and very few panellists voted for enzalutamide. That mirrors the split between STAMPEDE and ENZARAD. The authors point out that ENZARAD enrolled a lower-risk population (cN1 11% vs 39%, cT3–T4 47% vs 94%). Its null MFS result (HR 0.88, 95% CI 0.67–1.15) therefore does not cleanly test ARPI substitution in the population STAMPEDE treated. The panel is voting on drug-specific evidence, not a class effect, and that is defensible. ATLAS, PEACE-7, THUNDER and DASL-HiCaP will show whether the distinction holds, especially in pts without pelvic nodal disease. The ENZARAD signal in cN1 disease, or when pelvic RT is given for another reason, is preplanned but still a subgroup. It should not yet move practice on its own.
Fractionation is where a quick read of the vote misleads most. A combined 88% for some form of hypofractionation sounds decisive. But no single schedule reached consensus, and 44 panellists abstained or called themselves unqualified, so informed opinion is a smaller group than the headline suggests. The cited evidence supports caution on ultrahypofractionation for high-risk disease. HYPO-RT-PC had few high-risk pts, and the SHARP data are pooled from non-randomised prospective series. Moderate hypofractionation is presented as the established option. SBRT for high-risk localised disease remains a question for trials.
The focal boost discussion is the most transferable RT content. The authors describe a randomised trial in which an MRI-defined boost of up to 95 Gy on top of 77 Gy in 35 fractions improved bPFS and local failure, with no significant difference in late GU or GI toxicity. This makes dose escalation to the dominant lesion an option alongside conventional fractionation. How boost combines with moderate hypofractionation or SBRT is not addressed in the material provided.
The document generates no new outcome data. The authors say there was no formal literature review or meta-analysis, and the votes are distributions of expert opinion. One timing problem deserves attention. The consensus against perioperative ADT + ARPI before RP was reached before PROTEUS was presented and published. That trial reported a higher pCR or minimal residual disease rate and an MFS difference driven by PSMA-PET. So the Q6 vote reflects pre-PROTEUS opinion and may move when the question returns at the 2027 localised consensus conference.