- Design
- Post hoc landmark analysis of a randomised placebo-controlled trial
- Population
- 1,648 patients with HFpEF in TOPCAT Americas
- Primary outcome
- CV death, HF hospitalisation or aborted cardiac arrest by early eGFR dip
- Effect
- Dip 33% vs 20% (OR 1.97); treatment HR 0.75 with dip vs 0.80 without (P interaction 0.81)
A post hoc analysis of 1,648 patients from TOPCAT Americas examined an early eGFR dip, defined as a fall of 15% or more between baseline and week 4, after starting spironolactone or placebo in heart failure with preserved ejection fraction.
A dip occurred in 33% on spironolactone and 20% on placebo (OR 1.97). A dip was associated with worse subsequent cardiovascular outcomes in both groups. But spironolactone's effect on cardiovascular death, heart failure admission or aborted cardiac arrest was similar with a dip (HR 0.75) and without one (HR 0.80), with no interaction, and at every size of eGFR fall, risk was lower with spironolactone than placebo.
An early creatinine rise after starting an MRA often prompts nephrologists and physicians to stop it. This analysis suggests that reflex is wrong in HFpEF: the dip is a marker of risk, not evidence the drug is harming the patient. Potassium, not creatinine alone, should decide whether to continue.
- Recheck creatinine and potassium within 1 to 4 weeks of starting spironolactone
- Do not stop spironolactone for a modest early eGFR dip alone
- Stop or reduce for hyperkalaemia or a large, progressive fall in eGFR
- Look for volume depletion and adjust loop diuretics
- Treat a dip as a marker of higher risk and follow the patient more closely
Why it matters
Stopping an MRA for an expected early creatinine rise may give up its cardiovascular benefit.
Don't overread it
This is a post hoc analysis in which the subgroup hazard ratios individually cross 1; the finding is a lack of interaction, not proof of benefit in each group.
The statistics, in plain English
The hazard ratios of 0.75 (0.53 to 1.08) with a dip and 0.80 (0.64 to 1.00) without are similar, and the interaction P value of 0.81 means there is no evidence the drug worked differently. Each subgroup interval crosses or touches 1 because the subgroups are smaller than the whole trial. Using the Americas cohort avoids TOPCAT's known enrolment problems in Russia and Georgia.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for nephrology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free