Heart Failure Clinical Trials: From Beta-Blockers to AI

Heart failure clinical trials have, over the past three decades, transformed a condition once considered a death sentence into a treatable chronic disease. A network meta-analysis of drug trials spanning 30 years found that combining three classes of medication reduced mortality by roughly 56% compared with placebo, and newer four-drug regimens push that number even further.¹ The story of how those gains accumulated, where the remaining gaps lie, and what the next generation of trials looks like is worth knowing if you or someone you know lives with heart failure.

How Beta-Blockers Changed the Conversation

Before the 1990s, the prevailing wisdom actually warned against giving beta-blockers to people with heart failure. The logic seemed sound on paper: heart failure means a weakened pump, and beta-blockers slow the heart, so why weaken it further? Several landmark trials flipped that thinking. CIBIS-II, COPERNICUS, and MERIT-HF each showed that specific beta-blockers (bisoprolol, carvedilol, and extended-release metoprolol) reduced mortality in patients with advanced heart failure.² A systematic overview of 24 randomized trials involving over 3,000 patients found a 31% reduction in the odds of death with beta-blockers, dropping the average annual death rate from about 10% to about 7.5%.³ A separate meta-analysis pooling 17 reports confirmed the benefit held regardless of whether heart failure was caused by coronary artery disease or another condition.⁴

These results did more than add a drug to the toolkit. They demonstrated that the body’s stress-hormone response to a failing heart, while helpful in the short term, actually accelerates damage over time. Blocking part of that response lets the heart recover. This concept of neurohormonal inhibition became the organizing principle for the next two decades of heart failure drug development.

Stacking Therapies and the Rise of ARNI

Once beta-blockers proved their worth, trials began testing what happens when you layer multiple drugs that each target a different arm of the body’s stress response. ACE inhibitors (which block a hormone called angiotensin) had already been established. Mineralocorticoid receptor antagonists (which block aldosterone) added another layer. The 30-year network meta-analysis estimated that combining an ACE inhibitor, a beta-blocker, and a mineralocorticoid receptor antagonist together produced a 56% reduction in mortality versus placebo.⁵

Then came the PARADIGM-HF trial, one of the most consequential heart failure studies of the past decade. It tested sacubitril/valsartan, a drug that combines an angiotensin blocker with a neprilysin inhibitor (a molecule that enhances the body’s own protective hormones). Compared to enalapril, an established ACE inhibitor, sacubitril/valsartan reduced death from any cause by about 16% and cut hospitalizations for heart failure by 21%.⁶ That trial was stopped early because the benefit was so clear. The same network meta-analysis found that replacing the ACE inhibitor with sacubitril/valsartan in the triple-drug stack produced the greatest mortality reduction of any combination tested: a 63% reduction versus placebo.⁷ Subgroup analyses confirmed the benefit held across different background therapies.⁸

SGLT2 Inhibitors and the Fourth Pillar

The most surprising recent chapter in heart failure trials came from diabetes medicine. SGLT2 inhibitors were developed to lower blood sugar by making the kidneys excrete more glucose. But in cardiovascular safety trials required by regulators, researchers noticed something unexpected: patients taking these drugs had fewer heart failure hospitalizations, whether they had diabetes or not. That observation launched dedicated heart failure trials.

The DAPA-HF trial randomized over 4,700 patients with heart failure and reduced pumping function to dapagliflozin or placebo on top of standard therapy. Dapagliflozin cut the combined risk of worsening heart failure or cardiovascular death by 26%. Death from any cause dropped by about 17%.⁹ When pooled with the parallel EMPEROR-Reduced trial of empagliflozin, the combined data across nearly 8,500 patients showed a 13% drop in all-cause death and a 26% reduction in the composite of cardiovascular death or first heart failure hospitalization.¹⁰ Guidelines quickly adopted SGLT2 inhibitors as the fourth foundational drug class, establishing what clinicians now call “quadruple therapy.”

The Preserved Ejection Fraction Problem

For years, roughly half of all heart failure patients were left behind by clinical trial successes. These patients have heart failure with preserved ejection fraction, meaning their heart squeezes normally but fills poorly and stiffens. Every major drug trial that worked for reduced ejection fraction failed when tested in this group. It became one of the most frustrating dead ends in cardiology.

The TOPCAT trial of spironolactone is a case study in why preserved-ejection-fraction trials are so hard to run. The overall result was negative. But a post hoc analysis revealed a roughly fourfold difference in event rates between patients enrolled in the Americas versus those enrolled in Russia and Georgia. Spironolactone showed clear benefit in the Americas cohort, with significant reductions in cardiovascular death and heart failure hospitalization, while in the Russia/Georgia cohort there was no detectable drug effect at all.¹¹ Subsequent analysis suggested the regional disparity likely reflected differences in patient selection and possibly adherence, raising hard questions about how multinational trials verify that enrolled patients actually have the disease being studied.¹²

SGLT2 inhibitors finally cracked the preserved-ejection-fraction barrier. The EMPEROR-Preserved trial showed that empagliflozin reduced the combined risk of cardiovascular death or heart failure hospitalization by 21%, driven mainly by fewer hospitalizations.¹³ A meta-analysis of randomized trials confirmed the signal: SGLT2 inhibitors cut heart failure hospitalizations by 26% in preserved-ejection-fraction patients. However, the same pooled data showed no significant reduction in all-cause or cardiovascular death in this group, a meaningful difference from the reduced-ejection-fraction results.¹⁴ The drugs keep patients out of the hospital, but whether they extend life in this population remains an open question.

Device Trials That Reshaped Treatment

Not all heart failure trials test drugs. Implantable devices have their own evidence base, and some of the most dramatic results in the field have come from device trials.

Cardiac resynchronization therapy (CRT) uses a specialized pacemaker to coordinate the timing of contractions in a failing heart. The MADIT-CRT trial showed that adding CRT to an implantable defibrillator reduced heart failure events by 41% in patients with wide electrical conduction delays on their ECG, over an average follow-up of about two and a half years.¹⁵ Real-world registry data confirmed that these device benefits hold across men and women and across racial groups, with no significant differences in treatment effect.¹⁶ ¹⁷

More recently, the COAPT trial tested a catheter-based clip that reduces mitral valve leakage, a condition where a leaky valve worsens heart failure in a vicious cycle. Patients who received the clip on top of standard medical therapy had their rate of heart failure hospitalizations nearly halved, and death from any cause at two years dropped from about 46% to 29%.¹⁸ The degree of valve repair mattered: patients who achieved only a mild reduction in leakage had substantially worse outcomes than those whose valves were fixed well, underscoring that the technical quality of the procedure is inseparable from the clinical result.¹⁹ A post-approval study of over 5,000 patients in real-world practice confirmed durable results and large improvements in quality of life.²⁰

Mechanical Hearts for End-Stage Disease

At the far end of the severity spectrum, trials have tested left ventricular assist devices (LVADs), mechanical pumps implanted inside the chest to take over the heart’s workload. The original REMATCH trial demonstrated that LVADs improved survival and quality of life compared with medical therapy in patients too sick for transplant.²¹ Newer continuous-flow devices dramatically improved on those early results: two-year survival climbed from 29% with the original device to 77% with the next generation, and adverse events dropped by about 38%.²² LVADs remain a major surgery with real risks, including infection and bleeding, but for patients with no other options, the trial evidence supports a meaningful extension and improvement of life.

Newer Drug Targets on the Horizon

The current wave of trials is exploring mechanisms beyond neurohormonal blockade. Vericiguat, a drug that enhances the nitric oxide signaling pathway to help blood vessels relax and reduce the heart’s workload, was tested in the VICTORIA trial. It modestly reduced the composite of cardiovascular death or heart failure hospitalization, with the primary outcome occurring in about 36% of treated patients versus 39% on placebo.²³ The benefit was consistent across subgroups, including patients whose qualifying event was a recent hospitalization versus an outpatient visit.²⁴ The effect size was smaller than what beta-blockers or SGLT2 inhibitors achieved, which has generated some debate about where vericiguat fits in the treatment hierarchy, but it offers something for patients who remain symptomatic despite standard quadruple therapy.

Omecamtiv mecarbil takes an entirely different approach: instead of blocking harmful hormones, it directly makes the heart muscle contract more forcefully by activating cardiac myosin, the molecular motor inside heart cells. The GALACTIC-HF trial enrolled over 8,200 patients with heart failure and reduced ejection fraction, and omecamtiv mecarbil reduced the primary composite endpoint by 8%.²⁵ That modest headline number masks a potentially important detail: the drug appeared to benefit patients with the most severely weakened hearts more than those with milder dysfunction, suggesting it could fill a niche for people who need more contractile support. Regulatory decisions on its future are still unfolding.

Trials for Disease-Specific Causes

Heart failure is not a single disease but a final common pathway for dozens of conditions. An increasingly important frontier is running trials targeted at specific causes. Transthyretin amyloid cardiomyopathy, a condition where misfolded proteins deposit in the heart and stiffen it, illustrates this well. It was long considered untreatable and often went undiagnosed. The ATTR-ACT trial tested tafamidis, a drug that stabilizes the misfolding protein. Over 30 months, tafamidis cut all-cause mortality (about 30% on drug versus 43% on placebo) and reduced cardiovascular hospitalizations by about a third.²⁶ Post hoc echocardiographic analysis showed the drug also slowed the decline in heart function over time.²⁷ This was one of the first trials to show that identifying and treating the specific cause of heart failure, rather than just managing symptoms, can meaningfully change the trajectory.

Who Gets Into These Trials and Who Does Not

A persistent concern in heart failure research is that the people enrolled in trials do not look like the people living with the disease. Women make up roughly half of heart failure patients in the general population, yet their representation in clinical trials has hovered around 20 to 30% for four decades with no meaningful improvement.²⁸ A systematic review of contemporary trials found women constituted 27% of participants overall, and no significant upward trend over 15 years.²⁹ Older patients are similarly underrepresented: the average trial participant is younger than the average heart failure patient, particularly in preserved-ejection-fraction studies where real-world patients average 73 years but trial participants average 71.

The picture for racial and ethnic minorities is mixed. An earlier review found only 15% of trial participants were nonwhite, and fewer than half of trials even reported racial data.³⁰ More recent data show improvement: nonwhite enrollment rose from about 13% to 30% between 2001 and 2016, and reporting of race/ethnicity increased, though it still occurs in barely over half of trials.³¹ These gaps matter because heart failure behaves differently across demographic groups, with different leading causes, different prevalence, and potentially different responses to treatment. Trials that enroll a narrow slice of patients leave clinicians guessing about how to apply the results to everyone else.

How Trial Design Is Evolving

The endpoints used in heart failure trials have shifted as the field has matured. Early trials could use all-cause death as their primary measure because mortality was high enough to power a study. As treatments improved and death rates fell, trials needed to capture a broader picture. Today most large trials use composite endpoints that combine cardiovascular death with hospitalizations for heart failure, and many also track quality-of-life scores and repeat hospitalizations.³² ³³

There is also growing interest in using biomarkers as surrogate endpoints to make trials faster and cheaper. A modeling analysis demonstrated that NT-proBNP, a blood protein released when the heart is under stress, can predict the clinical outcomes of a trial after less than 10 weeks of treatment, with confidence intervals comparable to those obtained from years of follow-up.³⁴ If regulators increasingly accept biomarker endpoints, future trials could be shorter, smaller, and more affordable, which could accelerate the pace at which new therapies reach patients.

Decentralized trial designs are another innovation starting to take root. The CHIEF-HF trial of canagliflozin ran entirely without in-person visits: patients consented electronically, received medications by mail, wore Fitbits to monitor activity, and completed quality-of-life questionnaires on a phone app.³⁵ This model could help close the representation gap by reaching patients who live far from academic medical centers or cannot easily travel, precisely the populations currently missing from trial data.³⁶

The Gap Between Trial Evidence and Real-World Practice

Perhaps the most sobering finding in recent heart failure research has nothing to do with a new drug. It is the observation that therapies proven effective in trials are not reaching patients at the doses shown to work. A systematic review and meta-analysis of real-world prescribing data found that while ACE inhibitors or angiotensin receptor blockers were prescribed to about 81% of eligible patients and beta-blockers to about 78%, mineralocorticoid receptor antagonists were used in only about 47%, and CRT devices in under 6%.³⁷ Even when prescribed, drugs are frequently under-dosed below the target levels proven beneficial in trials.

The full quadruple therapy regimen (beta-blocker, renin-angiotensin system inhibitor, mineralocorticoid receptor antagonist, and SGLT2 inhibitor) was being prescribed to fewer than 40% of patients with reduced ejection fraction in a recent international registry, and under 8% of those with preserved ejection fraction.³⁸ Impaired kidney function was the most commonly identified barrier. Delayed initiation, uncertain sequencing of drugs, and clinical inertia are all contributors to this gap.³⁹ The implication is stark: the biggest potential gain in heart failure outcomes right now may not come from the next blockbuster trial, but from getting existing proven therapies into the hands of the people who need them.

AI Monitoring and the Trials Still to Come

Wearable devices and artificial intelligence are starting to intersect with heart failure research in ways that go beyond decentralized trial logistics. The SMART-CARE study, currently enrolling, is testing whether AI algorithms analyzing continuous data from wrist-worn and chest-worn sensors, including oxygen levels, heart rate variability, respiratory rate, and sleep patterns, can detect worsening heart failure early enough to prevent hospital admissions.⁴⁰ If trials like this succeed, they could shift the treatment model from reactive (responding to a hospitalization) to preemptive (intervening days before a patient decompensates). That would represent a different kind of advance from the drug trials discussed earlier, not a new molecule but a new way of deploying existing therapies at the right time. Results are pending, but the trial’s design reflects a growing conviction that the next frontier in heart failure is not only what we give patients but when and how we know they need it.