Acute heart failure treatment: a light at the end of the tunnel?

Acute heart failure (AHF) research and management have been fraught with failure for decades with every study attempting to introduce a new intervention in this field showing neutral results and sometimes even harm.1, 2 Twenty years ago, in the pages of the European Journal of Heart Failure,3 it was suggested that the treatment of AHF be divided into three periods1: (1) initial stabilization, (2) first day(s) in the hospital, and (3) pre- and post-discharge therapy aimed at reducing heart failure (HF) (re)admissions and death. Since then, no new therapy or intervention has been approved and implemented throughout the world in any of these three stages of AHF therapy.1, 2 Patients with AHF are commonly treated pre- and post-discharge with low doses of renin–angiotensin–aldosterone system (RAAS) inhibitors and beta-blockers. This is done even though most AHF patients have a left ventricular ejection fraction (LVEF) >40% and these therapies have been demonstrated to be effective only in stabilized chronic HF patients with reduced LVEF (i.e. none were shown to be effective during and immediately after an AHF admission),4 although some retrospective analyses have suggested that they may be of benefit in all AHF patients.5, 6 These failures have led some experts to doubt whether AHF is indeed a specific entity.7 However, a new wave of studies and analyses have shown promising results, culminating with the publication of the SOLOIST (Study to Evaluate the Effects of Sotagliflozin on Clinical Outcomes in Hemodynamically Stable Patients with Type 2 Diabetes Post Worsening Heart Failure) study demonstrating a very significant effect for sotagliflozin, a combined sodium–glucose co-transporter (SGLT) 1 + 2 blocker, in treating patients with a recent AHF admission (third phase of AHF therapy) and type 2 diabetes.8 The first signs for a possible light at the end of the tunnel were seen during the last 2 years. During the last two decades attempts to develop therapies have focused mostly on drugs that were administered intravenously for stage 1 – initial stabilization, such as natriuretic peptides, levosimendan, tezosentan and rolofyline.1, 2 All these drugs failed to show benefits on longer term outcome. However, the first glimmer of hope came with serelaxin, a naturally occurring hormone, that is administered intravenously starting at the first hours of an AHF admission. In a meta-analysis of a series of studies,9 intravenous (IV) serelaxin was shown to have some benefits on worsening HF and death, without a benefit on HF admissions. These findings, however, did not lead to approval and commercialization, mainly because the largest serelaxin study in AHF (Relaxin in Acute Heart Failure 2 – RELAX-AHF-2) did not demonstrate significant effects on either of its primary endpoints.10 Beyond the initial stabilization, during the rest of admission for AHF (stage 2 – first days in hospital and stage 3 – pre-discharge) the goals of therapy shift towards assessing general and specific cardiovascular triggers for the event, continued decongestion and initiation of disease-modifying therapies intended to improve outcomes over the long term. In these stages, a second study signaled that outcomes of AHF patients can indeed be improved. In the PIONEER-HF study (Comparison of Sacubitril–Valsartan versus Enalapril on Effect on NT-proBNP in Patients Stabilized from an Acute Heart Failure Episode), 881 patients were randomized to 8 weeks of sacubitril–valsartan or enalapril. The study drug was started between 1 to 10 days after an AHF admission while the patient was still hospitalized. The study demonstrated a faster improvement in natriuretic peptides in patients randomized to sacubitril–valsartan.11 But also, despite not showing effects on a key secondary endpoint of clinical events, the study drug was associated in a post hoc analysis with a nominally significant reduction in cardiovascular death or HF rehospitalization with event rates of 9.2% vs. 15.2% (P = 0.007).12 This was followed in recent months by the simultaneous presentation and publication of three positive or nearly positive studies examining new interventions administered to patients with AHF during admission or immediately post-discharge (third phase). All three studies were presented in the 2020 American Heart Association (virtual) meeting, and subsequently published. In the GALACTIC-HF (Global Approach to Lowering Adverse Cardiac Outcomes Through Improving Contractility in Heart Failure) study, patients with significant systolic HF (LVEF ≤35%), and systolic blood pressure <140 mmHg were randomized to receive the myosin activator omecamtiv mecarbil or placebo.13 The study met its primary endpoint with a reduction in the time to cardiovascular death or HF admission [hazard ratio (HR) 0.92, 95% confidence interval (CI) 0.86–0.99, P = 0.03] in the omecamtiv mecarbil group. Of note, the study was designed to recruit 25% of patients during an AHF admission. In this AHF subgroup, omecamtiv's effect on the primary endpoint was HR 0.89 (95% CI 0.78–1.01); driven by the effect in AHF patients with higher levels of natriuretic peptides (HR 0.75, 95% CI 0.61–0.92). These results extend to quality of life that was measured by the Kansas City Cardiomyopathy Questionnaire (KCCQ) Clinical Summary score. As with the outcome endpoints, those recruited as inpatients (AHF third phase) had a significant improvement (mean treatment difference 2.5, 95% CI 0.5–4.5, points) while no effect was observed in outpatients with chronic HF. At the same time another study – AFFIRM-AHF (A Randomised, Double-Blind Placebo Controlled Trial Comparing the Effect of Intravenous Ferric Carboxymaltose on Hospitalisations and Mortality in Iron Deficient Patients Admitted for Acute Heart Failure) – was presented and published.14 In this study patients admitted for AHF with LVEF <50%, high natriuretic peptides and iron deficiency were randomized to receive IV ferric carboxymaltose or placebo. After 1 year the primary endpoint of total hospitalizations for HF and cardiovascular death showed borderline significance (rate ratio 0.79, 95% CI 0.62–1.01, P = 0.059). The effect was more significant for the endpoint of first HF hospitalization or cardiovascular death (HR 0.80, 95% CI 0.66–0.98, P = 0.030). Furthermore, the effects were more pronounced in patients recruited before the COVID-19 pandemic. Noteworthy, AHF patients with iron deficiency represent more than 70% of these patients.15 Finally, the SOLOIST-WHF study was presented and published at the same time.8 In this study, patients with diabetes mellitus and a recent admission for AHF were randomized during the admission or within 72 h of discharge to receive placebo or the SGLT-1 and -2 blocker sotagliflozin. Patients were enrolled without LVEF limitation and only slightly elevated natriuretic peptide levels at baseline as compared to other AHF studies. Hence, these patients had higher LVEF and lower levels of natriuretic peptides and renal failure than in other chronic HF studies, making the patient's risk elevated solely because they were in the third stage of an AHF event. The study was discontinued before its planned sample size was reached. As stated in the methods,8 SOLOIST-WHF was planned as an event-driven trial that was supposed to enroll until 947 adjudicated primary endpoint events (first occurrence of either death from cardiovascular causes or hospitalization for HF) were accrued, resulting in an estimated sample size of approximately 4000 patients. However, the study was terminated after only 1222 patients were enrolled, leading to 600 primary events. Despite this substantially reduced sample size, the SOLOIST-WHF study met both its revised and original endpoints. The rate of cardiovascular death or HF readmissions or urgent visits due to HF (revised primary) was lower in the sotagliflozin group than in the placebo group (51.0 vs. 76.3 events/100 patient-years; rate ratio 0.67, 95% CI 0.52–0.85; P < 0.001). In addition, the original primary endpoint of the trial (the first occurrence of either death from cardiovascular causes or hospitalization for HF) was also significantly affected with a HR of 0.71 (95% CI 0.56–0.89), as was the effect on quality of life as measured by the KCCQ-12 score with a 4.1 (95% CI 1.3–7.0) point difference. How should we interpret these results? In aggregate the three studies (GALACTIC-HF, AFFIRM-AHF and SOLOIST-WHF together with those of PIONEER-HF) demonstrate that patients with a recent AHF admission are at higher risk for adverse outcomes and therefore can benefit most from interventions. Indeed, as depicted in the Table 1, event rates at 1 year were >40% in SOLOIST-WHF and AFFIRM-AHF (the same data are not available for the GALACTIC-HF AHF stratum). These event rates are triple those seen in recent chronic HF studies such as DAPA-HF or PARADIGM-HF,16, 17 and even higher than those in the VICTORIA study enrolling patients with systolic dysfunction and a worsening HF event within 6 months of enrolment.18 Patients with AHF are at higher risk due to more severe HF characteristics and the fact that patients are at the end of an AHF admission which may be caused by different pathophysiological mechanisms19 rather than simply having more severe chronic HF. But despite having more adverse events, the results of some studies such as the EVEREST study20 questioned how much of these post-AHF events may be modifiable (i.e. amenable to therapy). The three recent studies, added to the previous results of the serelaxin meta-analysis and PIONEER-HF, suggest that the outcomes of patients enrolled with very recent AHF can be improved substantially, as in all studies there were significant or close to significant effects on hard outcomes, and in two of the studies improvements in quality of life were also reported. However, the most outstanding of these studies is SOLOIST-WHF. In this study (Table 1) the effect sizes on cardiovascular death or HF admission are to the best of our knowledge the highest of all phase 3 HF studies ever conducted. The number needed to treat (NNT) to prevent an event is less than five, while that observed in other modern HF studies such as PARADIGM-HF,17 DAPA-HF,16 EMPEROR-Reduced,21 and VICTORIA18 was 15–25. The NNT observed in the SOLOIST-WHF study was even lower than that observed in the first wave of HF studies conducted more than 20 years ago when patients' background therapy was mostly diuretics, establishing the importance of RAAS inhibitors and beta-blockers in HF such as SOLVD Treatment and COPERNICUS.22, 23 This puts sotagliflozin at the level of ‘breakthrough’ therapy that enables for the first time to treat and improve the adverse outcomes of AHF. Beyond the large effects observed on hard outcomes in the SOLOIST-WHF study, the study is also notable for enrolling patients with (relatively) low levels of natriuretic peptides and no LVEF restriction. These design features and the lack of interaction seen between the effect size on the primary endpoint and baseline LVEF suggest that patients with AHF may benefit from novel interventions regardless of these characteristics, opening the door to development of new interventions for patients with AHF and HF with preserved and mid-range LVEF.24, 25 What is the explanation for the remarkable benefit observed in the SOLOIST-WHF study? As described in a recent review,24 although SOLOIST-WHF enrolled only patients with diabetes and its results should be explored in non-diabetic patients too, the absolute benefit seen in SOLOIST-WHF was three times larger than that observed in studies of SGLT2 inhibitors in chronic HF such as DAPA-HF16 and EMPEROR-Reduced.21 Some of this larger benefit can be attributed to the higher event rate observed in the control arm since patients were enrolled immediately after an AHF admission. However, such high event rates did not immediately translate into higher benefit in the GALACTIC-HF, AFFIRM-AHF and VICTORIA studies where the NNT was in the 15–25 range,13, 14, 18 as compared to less than five in SOLOIST-WHF. This may suggest that the double (SGLT1 + 2) blockage effect of sotagliflozin may have an additional benefit beyond those seen with SGLT2 inhibitors alone. As seen in all SGLT inhibitor studies, the beneficial effects of these drugs appear relatively early – within days of randomization. This may imply that some of the benefit is driven by more effective fluid removal. In this respect, patients recently admitted for AHF may be more susceptible to congestion and, therefore, better decongestion is an important goal of AHF therapy beyond the initial stabilization.26 Hence, SOLOIST-WHF's very large effect, in our opinion, is not a simple result of higher event rates, but of the fact that SGLT blockage is especially effective in AHF. These observations are in line with a recent analysis of the BIOSTAT-CHF database demonstrating that, first, patients admitted with AHF have a worse prognosis than those with worsening HF that does not lead to hospital admission. Second, there are different biomarker signatures present in patients admitted due to AHF and, third, different biomarkers predict adverse outcome in patients with AHF than those with outpatient worsening HF.27 Although not conclusive, this suggests that AHF is indeed a sperate disorder driven by different pathophysiological mechanisms which may lend themselves better to the specific effects of specific therapies. In some cases AHF patients will respond less well to certain therapies and in others, as in the SOLOIST-WHF study, remarkable benefits will be detected. Finally, the results of these three studies confirm our previous suggestion that in HF research bigger is not always better.28 With significant and close to significant results seen in the inpatient (AHF) subset of the GALACTIC-HF study enrolling 25% of the total 8256 patients (i.e. slightly >2000 patients), the 1525 patients enrolled in AFFIRM-AHF, the 1222 patients enrolled in SOLOIST-WHF, and even the 881 patients enrolled for only 8 weeks in PIONEER-HF, the assumption that mega-studies can be substantially downsized while focusing on sicker and more vulnerable patients has been largely confirmed. As suggested previously, smaller studies focusing on better site and patient selection are not less likely and maybe more likely to show benefits in patients who are most in need of such novel interventions.29, 30 From the perspective of regulatory approvals and improving post-AHF patient care, these positive findings in AHF are not certain, regretfully, to translate into new therapeutic options for AHF patients. Both serelaxin and omecamtiv mecarbil development were halted by Novartis and Amgen31, respectively, and the commercialization of sotagliflozin is uncertain at this time., However, it is possible that the findings of the SOLOIST-WHF study will increase the use of all SGLT inhibitors in the treatment of patients post-AHF. Equally we hope that iron therapy will be used more commonly in patients who are found to have lower ferritin or have iron deficiency during an AHF episode. In this respect, we believe that some of the reluctance to develop new therapies for AHF and implement those results relates to the pre-conceived belief that AHF may not be a specific disorder requiring specific therapy.7 Hopefully recently published findings suggesting that AHF is a different disorder27 and the current positive studies would change this perception and help move the needle towards more research and commercialization of therapies specifically targeted at AHF. Equally, those involved in furthering treatment of HF should consider the high event rate and larger effect sizes observed in AHF studies. One should consider whether there is a need to develop more therapies for patients with chronic HF and low LVEF, or rather focus on where the unmet need is – patients with AHF regardless of their LVEF. Conflict of interest: G.C. is an employee of Momentum Research who received research grants from Abbott Laboratories, Amgen, Celyad, Cirius Therapeutics, Roche Diagnostics, Sanofi, and Windtree Therapeutics Inc. B.A.D. is an employee of Momentum Research who received research grants from Abbott Laboratories, Amgen, Celyad, Cirius Therapeutics, Roche Diagnostics, Sanofi, and Windtree Therapeutics Inc. C.E. is an employee of Momentum Research who received research grants from Abbott Laboratories, Amgen, Celyad, Cirius Therapeutics, Roche Diagnostics, Sanofi, and Windtree Therapeutics Inc. K.T. received speaker's honoraria from Otsuka, Sumitomo Dainippon, AstraZeneca, and Bayer and consultancy fees from Terumo. A.C.S. reports personal fees from Novartis, Servier, MSD, AstraZeneca, Vifor Pharma, Amgen, CVRX, Menarini, Roche Diagnostics, BMS, outside the submitted work. A.M. received personal fees and/or research grants from Novartis, Orion, Roche, personal fees from Servier, Adrenomed, Abbott, Sanofi.