Spironolactone and hyperkalemia in congestive heart failure
In 1999 Pitt and colleagues published the results of the RALES trial, an important study showing that the addition of a relatively small dose of the aldosterone antagonist spironolactone to a regimen that included angiotensin-converting enzyme (ACE) inhibitors for patients with severe congestive heart failure (NYHA Class III-IV) had a striking benefit on mortality.1 The rationale for this therapy was the documented “escape” from converting enzyme blockade, allowing for the conversion of angiotensin I to angiotensin II by other mechanisms. The resulting increase in aldosterone release is recognized to have effects not only on renal sodium-potassium transport, but also importantly on myocardial remodeling and fibrosis. More recently, the newer aldosterone antagonist eplerenone has been shown, in the EPHESUS trial, to be beneficial in patients with recent myocardial infarction complicated by heart failure.2
Spironolactone, used in larger doses as a diuretic, is generally well tolerated, the most common side effect being gynecomastia in men. Eplerenone is more selective and is not associated with gynecomastia. The major problem with both these drugs is hyperkalemia, since they block the tubular reabsorption of sodium and the concomitant secretion of potassium. Because congestive heart failure (CHF) patients will be on ACE inhibitors, which may well induce some degree of hyperkalemia via their inhibition of aldosterone release, the addition of the aldosterone antagonist may well increase the degree of potassium retention. Hence there is need for caution, both in assessing renal function prior to prescribing these agents and subsequently in following renal function and potassium concentrations.
As in most good clinical trials, subjects in the trials mentioned were screened to assure adequate renal function and normal potassium levels at initiation, and their potassium levels were carefully and frequently monitored, per protocol. It is by no means assured that the use of these drugs in routine clinical care will be accompanied by such careful monitoring.
In this issue of Cardiology Review, Shah and Gottlieb describe a retrospective study done within the Veterans Health Administration (VHA) system in Maryland. Electronic records in the VHA allow for a quite thorough review of orders, lab data, and diagnoses. They examined the records of patients prescribed spironolactone for heart failure, recording initial renal function and serum potassium levels as well as demographics, background medication, and ejection fraction.
Noteworthy was their finding of a much higher rate of hyperkalemia (≥ 5.5 mEq/L) and severe hyperkalemia (≥ 6.0 mEq/L) than that reported in the RALES trial. As they note, the cohort that they examined tended to be older and had more underlying renal insufficiency than the RALES cohort. Their cohort also included a group of patients whose baseline renal function (with creatinine levels ≥ 2.5 mg/dL or with baseline serum potassium concentration of ≥ 5.0 mEq/L) would have excluded them from the RALES study. But perhaps the clinical finding of most importance was that these patients were less closely followed after initiation of spironolactone.
How should we use spironolactone in CHF? There is little question that this drug reduces mortality—at least in advanced CHF, for which it is indicated. Clearly, however, the risk of hyperkalemia must be considered, and patients with baseline elevated levels of creatinine or of serum potassium must be evaluated with great care, both at the start of treatment and frequently thereafter. Within the VHA computerized patient record, it is in fact quite easy to post lab requirements for use of a drug, and to post reminders to check on selected lab values. In addition, it would be prudent to follow the RALES protocol in not using this drug when serum creatinine is much over 2.0 mg/dL or potassium is ≥ 5.0 mEq/L. And in patients on ACE inhibitor and aldosterone antagonist, any further interference with the renin-angiotensin-aldosterone system, as for example with an angiotensin-receptor blocker, must be approached with great caution. Primum non nocere.