Between Scylla and Charybdis: Navigating heart failure management in complex older adults

Heart failure is a prevalent and high-risk clinical syndrome.^{1, 2} Among individuals aged 80 year and older, it affects at least 10% of the population and on geriatric hospital wards, up to 30%–35% have heart failure.^{3, 4} Multiple therapies are available to prevent and manage heart failure.² Pursuing the now-standard quadruple therapy, consisting of renin-angiotensin-aldosterone system inhibitor (RAAS-I), beta blocker, sodium glucose cotransporter-2 antagonist and mineralocorticoid receptor inhibitor, in heart failure with reduced ejection fraction (HFrEF) has been estimated to at least halve all-cause mortality.⁵ As a result, this approach has earned a class IA recommendation in current guidelines.²

Importantly, while guideline-directed medical therapies improve hard clinical outcomes, this benefit is primarily seen in ambulatory, chronic HFrEF patients, similar to those enrolled in the landmark trials. As patients deviate from this profile, the less clear the net benefit of such medical therapies on clinical outcomes becomes. Moreover, even despite the availability of trial-tested pharmacotherapy, outcomes remain suboptimal, characterized by a high residual burden, particularly following acute heart failure episodes. For instance, the 90-day mortality and hospitalization rates in the aftermath of an acute heart failure event are on average 10%–15% and 20%–25%.⁶ Even among stable patients exhibiting ‘favourable’ heart failure phenotypes (EF >40%), such as those observed in the FINEARTS HF study, outcomes were far from satisfactory, revealing a 16.4% to 17.4% rate of all-cause mortality during a median follow-up of 32 months.⁷

This brings us to an as yet unresolved paradox. On one hand, older adults with heart failure experience higher event rates, that is, they are at higher risk for poor outcomes. Assuming the relative efficacy of guideline-directed medical therapies holds across age groups, older adults should consequently derive larger absolute benefits in reducing mortality and heart failure hospitalizations. For instance, if the hazard ratio for heart failure hospitalizations is similar across age groups, the absolute risk reduction should be more substantial when baseline event rates are higher. On the other hand, we must acknowledge that managing heart failure in older adults presents unique challenges. Indeed, the concept of less guideline-directed medical therapies in complex older adults is not new; a lower use of ACE inhibitors in older adults with heart failure has been reported as early as 1992.⁸ Their higher complexity and multimorbidity—including frailty—complicate the optimal use of such medical therapies and may even shift the risk–benefit balance. In this regard, the risks of symptomatic hypotension and falls are particularly worrisome.

Several issues merit further consideration. First, older adults have often been underrepresented in clinical trials.⁹ Ironically, the archetypical heart failure patient is typically an older adult.¹⁰ This raises the question how to address this disparity. Subgroup analyses from landmark trials that focus on factors such as frailty, age, multimorbidity and polypharmacy provide some guidance for developing strategies tailored to this population.³ However, such analyses do not equate to robust trial evidence. One could also question the validity of the approaches that were used to determine the level of frailty in such analyses.³ This caution should also extend to other observational data, which have shown that older adults with heart failure are often undertreated, and that this underuse of guideline-directed medical therapies is associated with worse outcomes.^{11, 12} Association does not imply causation.¹³

Addressing the paradox of heart failure care in older adults requires a recognition of their multimorbidity and frailty.¹⁴ This may involve an age-adjusted comparison of the number-needed-to-treat versus the number-needed-to-harm, for example, capped within a year. Unfortunately, data specific to these comparison are lacking for complex older adults with heart failure, and even if available, they were still collected during the aforementioned landmark clinical trials with only limited external validity. It is important however in this regard to consider the net benefit of guideline-directed medical therapies across heart failure phenotypes. Table 1 presents a selection of heart failure trials, depicting a low number-need-to-treat versus a high number-needed-to-harm. In general, most adults derive greater benefits from treatment than that they experience adverse events severe enough to necessitate discontinuation of HF treatment.

Two questions emerge in the context of heart failure management for complex older adults. First, how aggressively should we continue to implement guidelines for the initiation and uptitration of heart failure therapies? It is clear that efforts should be made to adhere to evidence-based medicine, even amidst the challenges of translating the positive findings from studies like STRONG-HF to older populations.²¹ A few consecutive principles may be beneficial, based on the current totality of data and informed by previous expert guidelines on the management of multimorbidity in older adults^{3, 22, 23}: (1) any guideline-directed medical therapy is preferable to none, if tolerated; (2) preference should be given to evidence-based therapies (e.g., bisoprolol instead of propranolol); (3) using more guideline-directed medical therapies is better than using fewer; (4) again, if tolerated, utilizing higher doses should be pursued, particularly for RAAS-I and beta-blockers.

Second, how should we approach the discontinuation or tapering of heart failure therapies? Deprescribing is an increasingly relevant concept in the management of polypharmacy, mainly in older adults, to mitigate the risk of avoidable medication harm.²⁴ Older adults are likely to experience an elevated risk of adverse drug reactions from heart failure therapies, even though robust data are limited also in this regard.²⁵

The findings from the current systematic review and meta-analysis from Duong et al. further highlight the complexity of this issue and underscore the need for careful, tailored decision-making in older adults with heart failure.²⁶ The authors included a total of 33 studies, encompassing 797 690 participants, of which six were randomized controlled trials (RCTs) involving 532 patients, and 27 were observational studies with a combined total of 797 158 participants. A substantial portion of the data was derived from European (17 studies) and North American (12 studies) study cohorts, with an important representation of patients with HFrEF (12 studies, 407 492 participants). The age distribution in the RCTs ranged from 61 to 78 years, while observational studies reported ages from 64 to 84 years.

The review evaluated the feasibility and risks associated with medication discontinuation or tapering based on drug class and frailty status, yielding three key conclusions. First, more data are urgently needed to inform decision-making in this area. For instance, the authors were unable to conduct a meta-analysis for ‘hard’ clinical outcomes such as mortality or hospitalization due to insufficient RCT data. Second, in patients with chronic kidney disease, it was feasible to taper RAAS-I (i.e., reduce the dose by 50%) without any increase in harm. This evidence was primarily derived from a very small trial (n = 47) that focused specifically on CKD patients.²⁷ This study had an open design, was powered to assess CKD outcomes, but lacked sufficient power for mortality. Despite its limitations, it yielded interesting outcomes that warrant further investigation. Third, in patients with heart failure with preserved ejection fraction, discontinuation of beta-blockers was shown to be feasible. Overall, these conclusions are consistent with current clinical guidelines, some of which have recently been emphasized in a position paper focusing on patient profiling.¹⁴

Collectively, the available evidence at this point suggests at least that some pharmacotherapy principles may also apply to frail older adults with heart failure. In practice, this means adhering to RCT-proven strategies as closely as possible, as recommended in current guidelines. For HFrEF, clinicians should be weary of the hidden opportunity cost of deprescribing potentially beneficial agents, particularly in terms of preventing hospitalizations. Conversely, in HF presentations other than HFrEF, it is reasonable to avoid aggressively pursuing typical HFrEF therapies and instead focus on agents with proven efficacy. While the current (largely observational) data support some conclusions regarding outcomes such as hospitalizations and mortality, significant gaps remain regarding patient-reported outcomes, including NYHA scores and quality of life.

The findings of Duong et al. also raise questions on the level of evidence required to modify our prior beliefs regarding the initiation, continuation and restarting of potentially disease-modifying therapies, that is, GDMT, in older adults with heart failure, particularly those considered frail. We should resist succumbing to the same ‘nihilism’ that continues to overshadow chronic heart failure management at the opposite end of the age spectrum.²⁸

The necessity for comprehensive trials like DANTON or FRAIL-AF is evident.^{29, 30} We urgently need studies comparing holistic, patient-centred approaches with guideline-directed therapies in older adults with HFrEF, particularly those over 75 years old who are also frail. A non-inferiority design would be preferable to identify acceptable trade-offs between hard outcomes, for example, heart failure hospitalizations, and drug-related harm. Based on the following rough estimates (mortality 35%, allowing for a 15% non-inferiority margin; total medication harm: 45% with an expected reduction to 40%), approximately 770 patients would be needed per arm with a 12-month follow-up, not taking into account loss due to attrition.

In conclusion, the challenge of managing heart failure in older adults is inherently complex and multifaceted.³ While awaiting robust trial data to better inform deprescribing strategies in older adults with heart failure, particularly those who are frail, clinical decisions should be guided by a patient-centred, reasoned and pragmatic framework. This approach requires synthesizing the available evidence, accounting for pathophysiological mechanisms and evaluating the benefit–risk profile for each patient. Only through such an informed, patient-centred strategy can we aim to optimize outcomes in this vulnerable population.

All authors contributed to the manuscript. The first draft was written by Lorenz Van der Linden, and Ross Tsuyuki reviewed and edited the draft. Both authors read and approved the final manuscript.

The authors declare no relevant conflicts of interest for this editorial.