GnRH Antagonist or Agonist in the Man With Established Cardiovascular Disease
A man with metastatic prostate cancer and a prior MI needs androgen deprivation started now. HERO says one GnRH class is safer for his heart; PRONOUNCE, the trial built to actually test that question, says it can't tell the difference.
P.A., a 68-year-old man, retired two years ago after four decades running a family-owned print shop, and now spends most weekday mornings at the same diner counter with two friends from his old bowling league. Eighteen months ago he had an anterior STEMI, treated with primary PCI to his LAD; he has been on dual antiplatelet therapy, a statin, and a beta-blocker since, with a most recent LDL of 62 and no further cardiac events. Last month, a routine PSA of 38 led to a bone scan showing three vertebral metastases and a diagnosis of metastatic hormone-sensitive prostate cancer — disease that will need androgen deprivation therapy started promptly, in a man whose cardiovascular history is not incidental background but the central complicating fact of the decision in front of him.
The choice between a GnRH antagonist and a GnRH agonist for ADT has real cardiovascular stakes that didn't used to get this much scrutiny. HERO, the pivotal trial for oral relugolix, reported a fifty-four percent relative reduction in major adverse cardiovascular events compared with leuprolide — two point nine percent versus six point two percent at forty-eight weeks — and in the subgroup of patients with a documented prior cardiovascular event, the gap widened sharply, three point six percent versus seventeen point eight percent. That subgroup describes P.A. almost exactly. But HERO's cardiovascular comparison was a secondary endpoint, not the trial's primary design, and the one trial actually built and powered to test cardiovascular safety head-to-head — PRONOUNCE, degarelix versus leuprolide in men with known atherosclerotic disease — found no significant difference between the drug classes at twelve months, terminated early with fewer events than projected. Two real trials, asking overlapping but not identical questions, and neither one alone settles what to start him on.
In clinic, choosing a GnRH class after a prior MI
I'd start relugolix. HERO's overall cardiovascular finding was already striking — a fifty-four percent relative reduction in MACE — but the subgroup that actually matters here is the one with a prior cardiovascular event: three point six percent MACE on relugolix versus seventeen point eight percent on leuprolide. That's not the trial's average patient, that's a man who looks like P.A. specifically. Avoiding a testosterone flare is a smaller, secondary reason on top of it — not the main argument, but not nothing in a man with recent coronary disease either.
I want to be careful about which trial we're actually leaning on. HERO's cardiovascular result came from a prespecified safety analysis, not a powered primary endpoint — the trial was designed and sized to demonstrate sustained testosterone suppression, not to test MACE. PRONOUNCE was the trial actually built for this question: men with known atherosclerotic cardiovascular disease, randomized specifically to compare cardiovascular safety between a GnRH antagonist and agonist. It found no significant difference at twelve months, five and a half percent versus four point one percent, and that's the population that actually looks like P.A. — established ASCVD, not just advanced age.
The prior-event subgroup you're citing from HERO is real, but it's a subgroup analysis within a trial that wasn't cardiovascular-outcome-powered to begin with. I'd trust the properly designed trial's null result over an underpowered trial's secondary subgroup finding, even a striking one.
I don't think these two trials are actually in conflict once you look past the headline numbers. HERO's overall MACE finding carries a fragility index of three — meaning as few as three additional events, redistributed, would erase its statistical significance, which is a real caution about how much weight to put on it, subgroup or no subgroup. PRONOUNCE, meanwhile, stopped at 545 of its planned 900 patients and accrued 26 events against the 66 it was powered for, because the leuprolide event rate it assumed — ten point two percent — never materialized, which is its own kind of fragility, just in the direction of failing to detect a real difference rather than manufacturing one. Neither trial, read honestly, gives P.A. a clean cardiovascular answer.
What I'd actually weigh instead: relugolix avoids the testosterone flare a GnRH agonist causes in the first weeks of treatment, a real, mechanistically clear risk in a man with recent coronary disease whose myocardium doesn't need an androgen surge on top of whatever ADT itself does hemodynamically. That argument doesn't depend on either contested MACE trial being right — it's a smaller, more mechanistically grounded reason that happens to point the same direction as the urologic oncologist's read, for a different and more defensible reason.
Started on relugolix, loading dose followed by the standard oral maintenance regimen, with continued cardiology follow-up on his existing DAPT and statin regimen unchanged. The cardiologist's caution about HERO's own statistical fragility was accepted as valid, but the pharmacologist's flare-avoidance argument — independent of either contested MACE trial — gave the team a reason to prefer relugolix that didn't require resolving which trial to trust more.
Not agreed: whether the cardiologist's preference for PRONOUNCE's null result as the more trustworthy cardiovascular evidence should generalize to future patients with less clear-cut secondary indications for avoiding a testosterone flare. The cardiologist remains unconvinced that HERO's subgroup finding should carry real clinical weight for anyone; the urologic oncologist still reads the subgroup gap as too large to dismiss as noise, even granting the pharmacologist's fragility-index caveat.