Spironolactone for HF: Clinical-trial and real-world experiences clash in observational study

August 04, 2004

Toronto, ON - Prescriptions for spironolactone in Ontario rose sharply among older HF patients on ACE inhibitors after the 1999 publication of a major trial that showed the drug can dramatically cut HF-related morbidity and mortality, according to an observational study[1]. The increased spironolactone use paralleled a jump in hyperkalemia-related hospitalization and associated in-hospital death rates, write investigators in the August 5, 2004, issue of the New England Journal of Medicine.

Our findings indicate that spironolactone-related hyperkalemia is a much greater problem in everyday practice than in the setting of a clinical trial.

Dr David N Juurlink

(Source: University of Toronto)

This clash between clinical-trial and real-world experiences may have partly resulted from the application of results from the Randomized Aldactone Evaluation Study (RALES)[2] to patients who didn't necessarily fit its entry criteria, write Dr David N Juurlink (University of Toronto, ON) and colleagues. "Our findings indicate that spironolactone-related hyperkalemia is a much greater problem in everyday practice than in the setting of a clinical trial," they write.

"There's no reason to believe that the same phenomenon wouldn't happen with eplerenone," Juurlink told heartwire , referring to another aldosterone inhibitor that can potentially cause hyperkalemia.

What the Ontario study found

Juurlink and colleagues compared prescription and hospital-admissions records for the more than 1.3 million persons aged >66 years in Ontario for the years 1994 through 2001. Among those treated with ACE inhibitors who had recently been hospitalized for HF, the spironolactone-prescription rate in the years before the RALES publication was at a consistent level. But it began to shoot upward once the trial was published. Increases in hospitalization for hyperkalemia and associated in-hospital mortality followed similar curves, indicating significant sharp increases starting with the trial's release. No significant change was seen in the rate of HF hospitalizations over the same period. RALES was published online July 19, 1999 and in print September 2, 1999.

We're giving spironolactone to patients who would never have been randomized to RALES.

Juurlink sees several main reasons for discrepancies between the RALES and real-world experiences. "The first, and there's evidence to support this from other research[3], is that we're giving spironolactone to patients who would never have been randomized to RALES." For example, he said, the Ontario cohort was older than the RALES patients and thus more likely to have renal dysfunction, diabetes, or other comorbidities that elevate the risk of hyperkalemia.

"I think the other important [reason] is that we're not monitoring patients as closely in clinical practice. In RALES, not only were those patients selected to be at low risk of hyperkalemia to begin with, but they were watched very closely," Juurlink said.

Effects of the RALES publication among patients treated with ACE inhibitors and recently hospitalized for HF, 1994-2001

Parameter Early 1994 Early 1999 Late 2001*
Spironolactone prescriptions (per 1000 patients) 34 30 149
Hospitalizations for hyperkalemia (per 1000 patients) 2.4 4.0 11.0
Hyperkalemia-related in-hospital death (per 1000 patients) 0.3 0.7 2.0
*All differences vs earlier time periods, p<0.001. RALES published online July 19, 1999 and in print September 2, 1999.

The Ontario data are not an indictment of the aldosterone inhibitor, Juurlink cautioned. "Here is a drug with very clear evidence from a well-done trial of impressive benefit in terms of morbidity and mortality. So under no circumstances should we stop writing prescriptions for spironolactone. We should just be more selective about whom we give it to." He also suggested that physicians check patients' potassium levels after the first week instead of at 30 days, as is common"and maybe think twice about using it in a patient with diabetes."

In RALES, 1663 patients with NYHA class 3-4 HF and no history of renal insufficiency or hyperkalemia were randomized to 25-mg spironolactone or placebo daily. Patients were often on high doses of diuretics and typically receiving ACE inhibitors and beta blockers. Patients assigned to active therapy showed relative risk reductions of 30% for all-cause mortality, 29% for sudden death, and 35% for hospitalization for worsening HF over two years.

The clinical fallout

The Ontario group attempted to gauge the clinical impact of how clinicians interpreted the RALES trial. "We estimate that every 1000 additional prescriptions for spironolactone issued after RALES led to 50 additional admissions for hyperkalemia," they write. Extrapolating their Ontario numbers to the US population, the group writes that the clinical fallout after RALES would "correspond to 37000 additional hospitalizations and 4200 additional deaths each year."

Dr Prakash C Deedwania

(Source: UCSF)

Dr Prakash C Deedwania (University of California, San Francisco, and VA Central California Health Care System, Fresno), a RALES investigator, told heartwire he has serious reservations about the magnitude of those numbers. The Ontario study is "flawed " and "filled with speculation," he said, and it failed to include crucial information about its population that might have supported its conclusions. For example, he noted, no specific mention was made of how many in the cohort had renal dysfunctionespecially among the 40% with diabetesor NYHA class 2 vs class 3-4 HF. Limiting the analysis to patients getting ACE inhibitors is in no way the same as including only those with class 3-4 HF, Deedwania said. Nor, he added, were their serum potassium or creatinine levels included in the analysis.

Deedwania also contested the Ontario group's suggestion that increasing beta-blocker use after RALES in patients receiving aldosterone blockade, especially with ACE inhibitors, may have contributed to rising hyperkalemia rates. Contrary evidence emerged from the EPHESUS trial, in which eplerenone was associated with significant clinical benefits in patients with post-MI LV dysfunction or HF, he noted[4].

Many physicians are inappropriately giving spironolactone to patients who do not fit the RALES entry criteriaI agree with that 100%.

However, Deedwania said, "Many physicians are inappropriately giving spironolactone to patients who do not fit the RALES entry criteriaI agree with that 100%." The Ontario study's main message, he said, is that spironolactone should be used only in HF patients who are similar to the RALES population, especially in having HF of NYHA class 3-4. "Keeping in mind that if you give it to patients who are not RALES-like, particularly those with diabetes and renal insufficiency, you are going to expose the patient to significant risk of hyperkalemia, especially if you do not monitor potassium closely."

Why more hyperkalemia?

Juurlink and colleagues speculate that one or more of the following conditions might explain increased risk for hyperkalemia among HF patients receiving spironolactone in the post-RALES era:

  • Potassium levels may not be monitored closely enough in patients receiving spironolactone.

  • "Physicians may extend the RALES findings to patients who, unlike the patients in that study, do not have left ventricular dysfunction," such as those with HF due to diastolic dysfunction or cor pulmonale.

  • Physicians "may neglect baseline attributes that predispose patients to hyperkalemia," such as diabetes.

  • Conditions that develop during therapy, such as renal dysfunction, may be missed.

  • Some patients, such as those on furosemide, may deliberately increase their potassium consumption.

  • Dosages of spironolactone or other drugs that promote hyperkalemia may be inappropriately high.

The subtext

The climb in spironolactone prescriptions after RALES "suggests that a major clinical trial can significantly influence prescription practices in the absence of direct marketing forces from the pharmaceutical industry," Juurlink and his colleagues write. This goes even for a generic drug that is not aggressively marketed by the pharmaceutical industry, he told heartwire . "A well-done trial can have a dramatic and rapid influence on practice, especially when it's published in a major journal and picked up by the media."

Another message of the study is that more attention should be paid to a treatment's potential for adverse effects. When considering whether a trial's findings might apply to their patients, clinicians tend to focus more on benefits, Juurlink observed. "We should probably be more attuned to issues of harm, not just of efficacy."

Other insights

Spironolactone dosages higher than those used in RALES and a greater prevalence of renal dysfunction likely explain the posttrial rise in hyperkalemia, write Drs John JV McMurray and Eileen O'Meara (Western Infirmary, Glasgow, UK) in an editorial accompanying the Ontario study[5]. Its median spironolactone dosage of 25 mg/day was the precise assigned dosage in RALES, suggesting that a large proportion received more than was given in the randomized trial. "Familiarity with a long-established treatment and lack of the usual educational activities related to the safe use of a new drug may have contributed," they write. "Similarly, the index of suspicion and the intensity of surveillance for adverse effects related to a new product would have been higher than those for an established drug."


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.