Spironolactone or Eplerenone: The Fourth Drug He Stopped Without Saying Why
He already tried the fourth medication in his heart-failure regimen once, and quietly stopped taking it himself without saying why. Restarting it means deciding whether to give him the same drug at a lower dose, or the costlier alternative built specifically to avoid the reason he stopped.
S.T., a 64-year-old man, spent thirty years teaching industrial arts before retiring, and still spends most weekends in a church-basement workshop building wheelchair ramps and raised garden beds for neighbors who can't manage the lumber themselves. Hypertension had shadowed him for fifteen years before a work-up for progressive breathlessness three years ago found a dilated, poorly contracting left ventricle — an ejection fraction of 28 percent, no significant disease on his coronary angiogram, no clear trigger anyone could name. Over the following two years his heart failure team built him up, drug by drug, to the four medications proven to extend life in this disease: sacubitril/valsartan, carvedilol, dapagliflozin, and, added last, spironolactone. His ejection fraction climbed to 38 percent on the first three. The fourth never really got the chance to help.
About four months into taking spironolactone, he noticed his chest had started to feel tender under the skin, then visibly fuller on both sides. He didn't mention it to anyone — not his wife, not his cardiologist — and about six weeks before today's visit, he simply stopped filling the prescription. It was a pharmacy adherence flag, not anything he volunteered, that brought the subject up today. Off the drug, the tenderness has resolved and the swelling is mostly gone. His potassium and kidney function, checked at the time he was on it and again today, were never the problem — his labs are exactly why the discontinuation went unnoticed by anyone but him. The question the team is now working through is not whether he needs a mineralocorticoid receptor antagonist — the case for the fourth pillar of his regimen doesn't weaken because the first attempt at it went badly — but which one he can actually be expected to stay on this time, and why the two drugs in the class produced such a different experience for identical intent.
In clinic, six weeks after he quietly stopped it
Restart the same drug and you are testing exactly the mechanism that already burned him once. Spironolactone doesn't only block the mineralocorticoid receptor — it binds androgen receptors with meaningful affinity, increases peripheral conversion of testosterone to estradiol, and displaces estradiol from sex hormone binding globulin. That's not incidental toxicity; it's why RALES itself reported gynecomastia or breast pain in ten percent of the men who took it, against one percent on placebo, with two percent of that trial's spironolactone arm stopping the drug specifically because of it. Eplerenone was built to solve exactly this problem — selective for the mineralocorticoid receptor, minimal androgen or progesterone cross-reactivity, and gynecomastia rates in its own pivotal trial that never separated from placebo. He isn't a hypothetical patient who might react this way. He already did, reproducibly, at a standard dose.
That doesn't mean eplerenone carries zero risk of this — isolated cases exist even on the selective agent. What it doesn't carry is the specific off-target receptor mechanism that explains what actually happened to him.
I don't dispute the receptor pharmacology. What I'd push back on is treating this as though the outcomes case for the two drugs is equivalent, just with a side-effect tiebreaker. Spironolactone has RALES behind it — the largest, longest-standing mortality trial any aldosterone antagonist has in a patient this sick, at this ejection fraction, with decades of real practice built on top of it. No trial has ever randomized spironolactone against eplerenone head to head. The pooled comparison everyone cites for eplerenone's advantage is observational, and when the most recent systematic review split its own results by study design, the randomized-only data showed no significant difference in either treatment withdrawal or gynecomastia between the two drugs — the entire advantage came from the observational cohorts.
That's exactly what you'd expect if healthier or better-resourced patients are the ones who end up switched to the pricier alternative, not necessarily what you'd expect if the drug itself is doing something different. I'm not confident which explanation is correct — I just don't think the population-level number should carry as much weight as it's being given here. And I'll grant the other side of my own argument: he was NYHA III when we added this drug, but he's class II now. Strictly, EMPHASIS-HF is the better-matched trial for who he is today, not RALES — but EMPHASIS-HF was never run against spironolactone, only against placebo on top of triple therapy, so it doesn't actually settle which agent is right for him either.
Whichever way this goes, the actual failure we're trying to prevent already happened once, and it wasn't a pharmacology failure — it was a silent one. He didn't call, didn't flag it at a visit, just stopped filling it. If we hand him a prescription for eplerenone today without knowing what it will cost him at his pharmacy, we risk the identical outcome through a different door. His plan may cover it at a reasonable tier, may require a prior authorization citing his documented reaction, or may leave him paying several times what he's used to for a generic heart-failure drug. I want that answered today, in front of him, before he leaves — not assumed. If it turns out to be a real barrier, the honest fallback isn't no fourth drug at all; it's spironolactone at the lowest effective dose with an explicit plan for him to report anything at two weeks, not six months.
Agreed: don't restart spironolactone. Start eplerenone 25mg daily today, with explicit counseling to report any breast changes immediately rather than waiting, and a potassium and renal-function recheck in one week per routine MRA-initiation monitoring. Before he leaves the clinic, the pharmacist runs a real-time check of his formulary tier and any prior-authorization requirement for eplerenone.
Not agreed, and the reason today's plan carries an explicit branch point rather than a single expectation:
He fills it today and stays on the selective agent — the plan holds as written.
He starts spironolactone 12.5mg instead, with an explicit two-week check-in call rather than waiting for his next scheduled visit to ask how it's going.
Nobody on the team disputed which drug is pharmacologically cleaner for this specific history. What stayed unresolved was whether "cleaner on paper" and "the one he'll actually keep taking" are guaranteed to be the same drug — and that depends on an answer nobody in the room had yet.