spironolactone in diastolic heart failure

the current nejm study follows one i sent out last year from jama (see chf preserved EF spironolactone jama 2013 or JAMA. 2013;309(8):781-791), of the Aldo-DHF, a european study looking at whether spironolactone improved diastolic function and exercise tolerance in 422 ambulatory patients with diastolic heart failure, finding an improvement in left ventricular diastolic function but no difference in maximal exercise capacity, patient symptoms or quality of life.

in the current TOPCAT nejm study (see chf preserved EF spironolactone nejm 2014 in dropbox, or DOI: 10.1056/NEJMoa1313731), 3445 pts with symptomatic diastolic heart failure (EF>45%) given low dose spironolactone (15-45mg/d) vs placebo, assessing primary outcome of cardiovasc death, aborted cardiac arrest, or hosp for heart failure. followed 3.3 years. results:

    --baseline characteristics: ave age 69, half female, 90% white, LVEF 56%, 96% NYHA functional class 2 or 3, serum K=4.3, 82% on diuretic, 84% ACE-I/ARB, 78% b-blocker, and either had hospitalization for heart failure in past year (72%) or elevated BNP in previous 60 days.

    --primary outcome in 320 spironolactone patients (18.6%) vs 351 placebo (20.4%), with HR=0.89 (0.77-1.04, nonsignificant). of note, in assessing prespecified subgroups, spironolactone significantly reduced the primary outcome in those patients stratified by BNP levels vs those stratified by prior hospitalizations.

    --of the components of the primary outcome, hospitalization for heart failure significantly improved, with 206 pts on spironolactone (12%) vs 245 on placebo (14.2%): HR 0.83 (0.69-0.99).

    --for deaths from cardiovasc causes: spironolactone assoc with 160, placebo 172 (HR 0.90, 0.73-1.12, nonsignif). insignif # of aborted cardiac arrest.

    --adverse events: overall no diff between groups, but spironolactone assoc with increased hyperkalemia (18.7% vs 9.1%) though less hypokalemia (16.2% vs 22.9%) as well as doubling of creatinine (10.2% vs 7.0%). no diff in serious adverse events with frequent monitoring (eg, no creat>3 or dialysis)

    --from their figure 1, the efficacy of spironolactone was largely evident from 12--48 months after start of trial

so, as the background issue, unlike systolic heart failure, the pharmacologic therapy for diastolic is much less effective. mineralocorticoid-receptor antagonists (eg spironolactone) have been found to decrease overall mortality and hospitalizations in those with systolic dysfunction, as have ACE-I/ARBs and b-blockers. diastolic heart failure, though the general approach to treatment is the same, does not have these important benefits, though there is symptomatic improvement and the CHARM-preserved trial did find a lower hospital admission rate in those on candasartan.  b-blocker efficacy is less clear. to me, this study at least shows statistical benefit in some results of spironolactone use, so i think it is reasonable to first use the basic meds (esp diuretics, ACE/ARB), consider adding b-blockers (esp if afib/fast rate or persistent hypertension, but also may get some improvement of cardiac hemodynamics), but then use spironolactone (carefully monitoring K) if persistent symptoms or high BNP.