HERO
ForAdvanced prostate cancer, completing 48wk ADT, no ongoing ADT planned
TL;DR90-day testosterone recovery to ≥280 ng/dL 53.9% with relugolix vs 3.2% with leuprolide after 48wk ADT.
Reported via UroToday →
For intermediate-risk prostate cancer treated with RT plus short-course ADT, the recovery kinetic is the deliverable: median 86.0 days to normal T on relugolix vs 2 of 47 recovering on leuprolide by day 90. That gates the choice of agent when ADT is a fixed 4 to 6 month adjunct to RT and the goal is off-treatment hypogonadism time, not depth of suppression.
In men finishing a defined short-course ADT with RT and no plan for ongoing suppression, this informs agent choice on recovery speed; it does not apply to men continuing ADT indefinitely or to those where prolonged suppression is the therapeutic aim.
When ADT is a defined 4 to 6 month adjunct to definitive RT, agent choice can be made on recovery speed rather than suppression depth. Median 86.0 days to normal testosterone (95% CI 65.0, 92.0) on relugolix is a concrete planning figure for counselling on off-treatment hypogonadism.
Relevant only where ADT is time-limited by design. The parent trial's depth advantage (96.7% vs 88.8% sustained castration) and this recovery signal point the same way, but neither is tied to an oncologic endpoint, so sequencing and continuous-suppression decisions are untouched.
8 details
Subset analysis of the phase 3 HERO trial, which randomized 934 men with advanced prostate cancer 2:1 to relugolix or leuprolide for 48 weeks. This report covers a 90-day recovery period after treatment discontinuation.
184 men who completed 48 weeks of assigned therapy and had no plan to start alternative ADT within the next 12 weeks (24 weeks after a last leuprolide 3-month depot). 137 had received relugolix, 47 leuprolide.
Relugolix 120 mg orally once daily after a single 360 mg loading dose on Day 1, versus leuprolide injections every 12 weeks, each for 48 weeks.
Time to testosterone recovery to ≥280 ng/dL by Kaplan-Meier, PSA during the recovery phase, and adverse events during recovery. No oncologic efficacy endpoint in this analysis.
During recovery, 96% of men had at least one adverse event and 15% had a grade ≥3 event, similar in both groups.
The 90-day window sits inside the pharmacologic tail of a 3-month leuprolide depot, so part of the 53.9% vs 3.2% gap is measurement timing rather than a durable biologic difference. The leuprolide median of 112.0 days derives from 2 recovery events.
Consistent in direction with the parent HERO result, where relugolix gave deeper sustained castration (96.7% vs 88.8%, difference 7.9%, 95% CI 4.1 to 11.8, P<0.001). The trade the two readings define is depth of suppression on-treatment against speed of recovery off-treatment, both favoring the oral antagonist.
This settles a kinetic question, not a clinical one: no recurrence, quality-of-life, or cardiometabolic endpoint is tied to the faster recovery here. Whether restoring normal testosterone months earlier translates into measurable benefit, or into a PSA that reflects recovering physiology rather than disease, remains untested.
| Metric | Relugolix | Leuprolide |
|---|---|---|
| n in recovery subset | 137 | 47 |
| Baseline T entering recovery (mean±SD) | 427±142 ng/dL | 404±127 ng/dL |
| Recovered T | 74 | 2 |
| Median time to recovery | 86.0 d (95% CI 65.0, 92.0) | 112.0 d (95% CI 112.0, NE) |
| Median PSA at day 90 | 0.39 ng/mL (0 to 233.1) | 0.06 ng/mL (0 to 14.0) |
Post-hoc subset of a phase 3 trial, non-randomised entry criterion, 137 vs 47 arms; recovery kinetics are a pharmacologic expectation, not a new clinical outcome.
- Does faster testosterone recovery change QoL or cardiometabolic outcomes
- Is higher day-90 PSA on relugolix physiologic or disease-driven
- Recovery kinetics after short-course ADT with definitive radiotherapy
📚 Sources · 📄 1 paper
Abstract
The longer read
The value of this analysis is not the 53.9% versus 3.2% figure, which is close to what the pharmacology predicts, but what it does and does not license a clinician to conclude. Relugolix is an oral non-peptide GnRH antagonist with a short half-life; leuprolide is a depot agonist dosed every twelve weeks. A recovery window opened at the moment of last dose is therefore not a symmetric comparison, because a man in the leuprolide arm may still be within the release profile of his final injection for much of the ninety days being measured. The eligibility rule acknowledges this by requiring that leuprolide patients have no planned alternative ADT for twenty-four weeks rather than twelve, but the recovery assessment itself still stops at day 90. That two of forty-seven leuprolide patients recovered, with a median of 112.0 days and an upper confidence bound that is not estimable, is a statement about where the observation window ended as much as about the drug.
What survives that objection is the relugolix number itself, which does not depend on the comparator. A median of 86.0 days to testosterone above 280 ng/dL, with a confidence interval of 65.0 to 92.0, is a usable planning figure. For a radiation oncologist prescribing a fixed short course of androgen suppression alongside definitive treatment, the relevant question is rarely which agent suppresses more deeply; the HERO parent result already answered that, and the margin there was 96.7 percent against 88.8 percent, a difference of 7.9 percent. The question is how long the patient carries the metabolic, sexual, and musculoskeletal cost after the course ends. This analysis puts a number on that for one agent and leaves the other unquantified within the window studied.
The methodological weakness that should move confidence most is not the post-hoc framing but the entry criterion. Inclusion required that a patient had completed forty-eight weeks and that no further suppression was planned. That is a clinical judgment made by treating physicians with knowledge of each patient's disease trajectory, applied after randomization, and it selects differently across arms in ways the reported baseline testosterone values cannot rule out. The arms entered recovery at 427 and 404 ng/dL respectively, which is reassuring for comparability of the starting point but says nothing about why these particular 184 of 934 men were the ones whose clinicians were willing to stop.
The PSA finding deserves more skepticism than the write-up gives it. Median PSA at day 90 was higher on relugolix, 0.39 against 0.06 ng/mL, with a relugolix range extending to 233.1. That upper value cannot be explained by testosterone recovery alone in every patient, and no analysis separates PSA rise attributable to returning androgen physiology from PSA rise attributable to disease. Framing faster recovery as a benefit while reporting a higher PSA without that separation is the part of this presentation a skeptic would push on hardest, and it is the part a future analysis would need to resolve before recovery speed can be treated as an unqualified advantage rather than a trade.