Spironolactone Near the Threshold: Completing Quadruple Therapy Under Borderline Hyperkalemia Risk
A patient with heart failure has improved on three of the four guideline-recommended drug classes. The fourth — a mineralocorticoid receptor antagonist — carries its own proven mortality benefit, but today's labs sit close enough to the thresholds used to screen it out that starting it is no longer a simple next step.
W.K., a 68-year-old man, spends most of his retirement in the pole barn behind his house restoring antique farm tractors — his son is helping him get a 1948 Farmall ready for next month's county fair. He taught auto shop for thirty-one years before retiring, and the hypertension he's carried since his forties and the type 2 diabetes that followed in his fifties — both poorly controlled through most of his working years, by his own account — have left him with chronic kidney disease he now manages as a matter of routine.
He was diagnosed with heart failure with reduced ejection fraction eight months ago, after a winter of worsening breathlessness he'd blamed on the cold barn air. His echocardiogram then showed an ejection fraction of 25%. Lisinopril was started at diagnosis; three months ago, once his blood pressure and renal function had proven stable on it, his cardiologist switched him to sacubitril/valsartan for its added mortality benefit over an ACE inhibitor alone. Metoprolol succinate has been at target dose for two months, and dapagliflozin was added six weeks ago, his renal function comfortably above the threshold required to start it. He is noticeably better for it — NYHA class III symptoms down to class II, and back working on the tractor most afternoons.
At today's routine follow-up, his ejection fraction has improved to 32%, and he and his cardiologist are ready to discuss the fourth pillar of guideline-directed therapy: a mineralocorticoid receptor antagonist. But today's labs complicate that plan. His potassium, drawn the same morning, is 5.0 mEq/L. His eGFR is 33 mL/min/1.73m², down from 38 three months ago — a decline consistent with the expected early effect of starting an ARNI, but one that has not yet been distinguished from a genuine ongoing decline.
Spironolactone carries a real, trial-proven mortality benefit in HFrEF at his ejection fraction and symptom class. But starting it in a patient whose labs already sit close to the potassium and renal-function boundaries the major MRA trials used to exclude patients is a materially different decision than starting it in a patient comfortably clear of them — and it is a decision his case doesn't currently answer on its own.
At today's follow-up, discussing the fourth pillar
He's already on three of the four GDMT pillars, and RALES and EMPHASIS-HF are unambiguous about what an MRA does in HFrEF — RALES in class III–IV, EMPHASIS-HF in class II, which is where he sits today. This isn't a marginal benefit we're weighing against a marginal risk. A potassium of 5.0 and an eGFR of 33 sit right on the lines rather than past them: both trials excluded a potassium above 5.0, and the guideline recommendation is written for a potassium below 5.0 with an eGFR above 30. Nothing here tells me his renal function is on a one-way trip down rather than settling into a new stable baseline after the ARNI switch. I'd start spironolactone today at 12.5 mg, with a repeat basic metabolic panel in three to five days rather than waiting for it before we start.
I don't disagree about what the drug does — I disagree about what today's numbers actually tell us. An eGFR drop from 38 to 33 over three months is exactly what we'd expect from starting an ARNI, and it's also exactly what an early, progressive decline would look like at this stage. We can't tell those two apart from one data point, and starting a second drug that raises potassium on top of an unresolved renal trend is how a borderline number turns into a real hyperkalemic event.
I'd hold off starting anything today and repeat the labs first — if they're stable, I'm not opposed to spironolactone at a reduced dose; I just don't want to find out which trend we're on after he's already on the drug.
Both of you are treating this as start-now-and-monitor versus wait-and-see, but the potassium-binder trials give us a third option that isn't just splitting the difference — AMBER showed patiromer keeping patients with advanced CKD and resistant hypertension on spironolactone, and DIAMOND showed the same enabling effect in HFrEF specifically. The binder was built to stop hyperkalemia from being the thing that limits RAAS-modulating therapy. I'd start spironolactone at 12.5 mg and patiromer together today, and still repeat the potassium in about a week, the same recheck either of you would want — the difference is he's not losing three to five days of a mortality-benefit drug while we wait to find out.
The real cost isn't clinical, it's practical: he's already managing three heart-failure medications, and a fourth pill with its own dosing schedule is a real adherence burden worth naming, not just a footnote.
Agreed: repeat the basic metabolic panel (potassium, creatinine/eGFR) within about a week, rather than waiting for the next routine visit.
Not agreed, and the reason today's plan carries three different starting points rather than one:
Start spironolactone 12.5 mg today; the mortality benefit is worth acting on now, before the repeat labs return.
Hold off entirely until the repeat labs confirm the renal-function trend has stabilized, not just paused.
Start spironolactone and patiromer together today; the binder converts the "wait to be safe" question into a non-issue.
All three agree on the recheck itself; none of them agree on what happens before it comes back, and the case does not resolve that difference for them.