The choice of loop diuretic for decongestion in patients hospitalized with heart failure (HF) may make little difference to survival or readmission risk over the next year, at least when deciding between furosemide or torsemide, a randomized trial suggests.
Both drugs are old and widely used, but differences between the two loop diuretics in bioavailability, effects on potassium levels, and other features have led some clinicians to sometimes prefer torsemide. Until now, however, no randomized HF trials have compared the two drugs.
The new findings suggest clinicians can continue starting such patients with HF on either agent, at their discretion, without concern that the choice may compromise outcomes, say researchers from the TRANSFORM-HF trial, which compared furosemide-first and torsemide-first diuretic strategies in a diverse population of patients with HF.
Given that the two strategies were similarly effective for survival and rehospitalization, clinicians caring for patients with HF can focus more on “getting patients on the right dose for their loop diuretic, and prioritizing those therapies proven to improve clinical outcomes,” said Robert J. Mentz, MD, Duke University Clinical Research Institute, Durham, North Carolina.
Mentz, a TRANSFORM-HF principal investigator, presented the primary results November 5 at the American Heart Association (AHA) Scientific Sessions 2022, held in Chicago and virtually.
The trial had randomly assigned 2859 patients hospitalized with HF and with a plan for oral loop diuretic therapy to initiate treatment with furosemide or torsemide. Clinicians were encouraged to maintain patients on the assigned diuretic, but crossovers to the other drug or other diuretic changes were allowed.
Rates of death from any cause, the primary endpoint, were about 26% in both groups over a median 17-month follow-up, regardless of ejection fraction (EF).
The composite rates of all-cause death or hospitalization at 12 months were also not significantly different, about 49% for those started on furosemide and about 47% for patients initially prescribed torsemide.
As a pragmatic comparative effectiveness trial, TRANSFORM-HF entered diverse patients with HF, broadly representative of actual clinical practice, who were managed according to routine practice and a streamlined study protocol at more than 60 U.S. centers, Mentz observed.
One of the pragmatic design’s advantages, he told theheart.org | Medscape Cardiology, was “how efficient it was” as a randomized comparison of treatment strategies for clinical outcomes. It was “relatively low cost” and recruited patients quickly compared to conventional randomized trials, “and we answered the question clearly.” The trial’s results, Mentz said, reflect “what happens in the real world.”
When Might Torsemide Have the Edge?
Although furosemide is the most commonly used loop diuretic in HF, and there are others besides it and torsemide, the latter has both known and theoretical advantages that set it apart. Torsemide is more than twice as potent as furosemide and more bioavailable, and its treatment effect lasts longer, the TRANSFORM-HF investigators have noted.
In addition, preclinical and small clinical studies suggest torsemide may have pleiotropic effects that might be theoretical advantages for patients with HF. For example, it appears to downregulate the renin-angiotensin-aldosterone system (RAAS) and reduce myocardial fibrosis and promote reverse ventricular remodeling, the group writes.
In practice, therefore, torsemide may be preferred in patients with furosemide resistance or “challenges with bioavailability, especially those with very advanced heart failure with congestion who may have gut edema, where oral furosemide and other loop diuretics are not effectively absorbed,” Biykem Bozkurt, MD, PhD, Baylor College of Medicine, Houston, Texas, told theheart.org | Medscape Cardiology.
In such patients, she said, torsemide “is considered to be a better choice for individuals who have diuretic resistance with advanced congestion.”
The drug’s apparent pleiotropic effects, such as RAAS inhibition, may have less relevance to the TRANSFORM-HF primary endpoint of all-cause mortality than to clinical outcomes more likely associated with successful decongestion, such as HF hospitalization, Bozkurt proposed.
The trial’s pragmatic design, however, made it more feasible to focus on all-cause mortality and less practical to use measures of successful decongestion, such as volume loss or reduction in natriuretic peptide levels, she observed. Those are endpoints of special interest when diuretics are compared, “especially for the subgroup of patients who are diuretic-resistant.”
Over the last 20 years or so, “we’ve learned that hospitalized heart failure is a very different disease process with a different natural history,” observed Clyde W. Yancy, MD, MSc, Northwestern University, Chicago, who was not part of the current study.
“So, the idea that something as nuanced as choice of one loop diuretic over the other, in that setting, would be sufficient to change the natural history, may be still a high bar for us,” he told theheart.org | Medscape Cardiology.
“Based on these data, one would have to argue that whichever loop diuretic you select for the hospitalized patient — and a lot of that is driven by market exigencies right now — it turns out that the response is indistinguishable,” Yancy said. “That means if your hospital happens to have furosemide on the formulary, use it. If furosemide is not available but torsemide is available, use it.”
Yancy said he’d like to see a trial similar to TRANSFORM-HF but in outpatients receiving today’s guideline-directed medical therapy, which includes the sodium-glucose cotransporter 2 (SGLT2) inhibitors, drugs that increase the fractional excretion of sodium and have a “diuretic-like” effect.
Such a trial, he said, would explore “the combination of not one, or two, but three agents with a diuretic effect — a loop diuretic, a mineralocorticoid antagonist, and an SGLT2 inhibitor — in ambulatory, optimized patients. It might make a difference.”
HF Regardless of EF
The trial enrolled patients hospitalized with worsening or new-onset HF with a plan for long-term loop diuretic therapy who had either an EF of 40% or lower or, regardless of EF, elevated natriuretic peptide levels when hospitalized.
Of the 2859 participants, whose mean age was about 65 years, about 36% were women and 34% African American. Overall, 1428 were assigned to receive furosemide as their initial oral diuretic and 1431 patients were assigned to the torsemide-first strategy.
The rate of death from any cause in both groups was 17 per 100 patient-years at a median of 17.4 months. The hazard ratio (HR) for torsemide vs furosemide was 1.02 (95% CI, 0.89 – 1.18; P = .77).
The corresponding HR at 12 months for all-cause death or hospitalization was 0.92 (95% CI, 0.83 – 1.02; P = .11). The relative risk for any hospitalization was 0.94 (95% CI, 0.84 – 1.07).
Pragmatic Design: Other Implications
Dosing was left to clinician discretion in the open-label study, as was whether patients maintained their assigned drug or switched over to the other agent. Indeed, 5.4% of patients crossed over to the other loop diuretic and 2.8% went off loop diuretics entirely between in-hospital randomization and discharge, Mentz reported. By day 30, 6.7% had crossed over and 7% had stopped taking loop diuretics.
The diuretic crossovers and discontinuations, Mentz said, likely biased the trial’s outcomes, such that the two strategies performed about equally well. Efforts were made, however, to at least partially overcome that limitation.
“We put measures in place to support adherence — sending letters to their primary doctors, giving them a wallet card so they would know which therapy they were on, having conversations about the importance of trying to stay on the randomized therapy,” Mentz said in an interview. Still, some clinicians saw differences between the two agents that prompted them, at some point, to switch patients from one loop diuretic to the other.
But interestingly, Mentz reported, the two strategies did not significantly differ in all-cause mortality or the composite of all-cause mortality or hospitalization in analysis by intention to treat.
Mentz discloses receiving honoraria from AstraZeneca, Bayer/Merck, Boehringer Ingelheim/Lilly, Cytokinetics, Pharmacosmos, Respicardia, Windtree Therapeutics, and Zoll; and research grants from American Regent and Novartis. Bozkurt discloses receiving honoraria from AstraZeneca, Baxter Health Care, and Sanofi Aventis and having other relationships with Renovacor, Respicardia, Abbott Vascular, Liva Nova, Vifor, and Cardurion. Yancy discloses modest relationship with Abbott.
American Heart Association (AHA) Scientific Sessions 2022. Session LBS.01. Presented November 5.