Worsening Heart Failure Events in HFpEF: Underlying Biology Not Treatment Location

As part of the patient journey, individuals living with heart failure (HF) may experience episodes of worsening symptoms of congestion and/or low output with varying tempo and frequency. These worsening heart failure (WHF) events contribute to a considerable burden for patients given the association with worse quality of life, re-hospitalizations, and mortality(1). Historically, WHF was synonymous with hospitalization for HF exacerbation. However, these hospitalization events may be preceded by a more gradual and/or less severe worsening in symptoms resulting in one or more outpatient encounters for either intravenous diuretics or escalated oral diuretic dosing. Alternatively, symptomatic worsening may be managed exclusively in the outpatient setting without ever requiring a hospital stay. As such, the definition for WHF has evolved over the years to include events regardless of the location of care (i.e., outpatient or inpatient), while focusing on the need for acute HF therapies based on a patient’s underlying biology(1). Despite the increased recognition of outpatient WHF as an important clinical entity, its true prevalence and prognostic utility among those with HF with preserved ejection fraction (HFpEF) remains poorly understood. Gaps in understanding of the patterns of WHF and the lack of effective therapies to reduce such events serve as unmet needs in treating patients with HFpEF.

In this issue of JACC: Heart Failure, Vaduganathan et al. provide important insights into different types of incident WHF events and their association with recurrent WHF and mortality among patients with HFpEF in the PARAGON-HF trial (Prospective Comparison of ARNI with ARB Global Outcomes in HF with Preserved Ejection Fraction)(2). The authors also assessed the efficacy of sacubitril/valsartan vs. valsartan in reducing a pre-specified expanded composite of cardiovascular death and WHF events, including hospitalized HF and urgent HF visits. Among 884 first WHF events in nearly 4,800 patients over a median follow-up of 35 months, 66 events were urgent HF visits (7.5%). Participants with a first event as an urgent HF visit had similar demographics, comorbidities, natriuretic peptide levels, and mortality risk scores to those who had a first episode of HF hospitalization(2). Vaduganathan et al. demonstrated that the occurrence of either type of WHF episode was associated with increased mortality compared to participants without WHF (albeit, with a significantly higher risk in the hospitalized group compared with outpatient WHF). Both types of incident WHF episodes were also associated with exceedingly high recurrent WHF events or cardiovascular death (>50 per 100 patient-years). These data highlight the prognostic significance of WHF regardless of the location of care. With the inclusion of urgent HF into an expanded composite endpoint of total WHF events (first and recurrent WHF regardless of treatment location) and cardiovascular death (which added 95 events to the original 1,903 for a total of 1,998 – an increase of ~5%), sacubitril/valsartan reduced the expanded composite endpoint by 14% compared to valsartan and met the nominal level of statistical significance (p=0.040)(2).

The authors should be commended on this timely and relevant analysis, which provides (1) important data underscoring the unfavorable prognosis associated with WHF events in HFpEF and (2) additional supportive data to characterize the treatment benefit with sacubitril/valsartan. Importantly, these data inform clinical care. During routine clinical encounters, providers commonly ask patients whether they have had a recent hospitalization for WHF to aid in risk assessment. However, the present data highlight the importance of also asking patients whether they have experienced an outpatient worsening event to better inform management decisions and follow-up cadence. In December of 2020, the Food and Drug Administration’s Cardiovascular and Renal Drugs Advisory Committee voted 12–1 in favor of an expanded indication of sacubitril/valsartan in HFpEF based, in part, on these results of PARAGON-HF towards reducing total HF hospitalizations and urgent HF visits. As in the primary trial results, the present analysis characterized heterogeneity in treatment effect, with particularly favorable treatment effects among women and those with left ventricular ejection fraction ≤ 57% (2). We await the FDA’s formal decision on any label modification and whether use will be recommended in specific HFpEF phenotypes.

Despite this study’s strengths, there are several notable limitations. The authors acknowledge that relatively few urgent HF visits occurred in PARAGON-HF based on the protocol-specified definition requiring intravenous therapy (66 first events; 95 total events)(2). Participant recruitment for PARAGON-HF started in 2014, when outpatient intravenous diuretic clinics were not widely available in many world regions(3). In addition, the study population was from a large, international clinical trial including ~12% recruitment in North America (N=559) with specific eligibility criteria that may have resulted in population differences from individuals seen in routine practice. Therefore, there is some uncertainty regarding current generalizability given both regional variation and temporal changes in contemporary HF management practices. Furthermore, outpatient WHF events in PARAGON-HF required the use of intravenous decongestive therapy. Recent data have shown that outpatient oral diuretic dose intensification is also associated with increased risk for adverse events(4). It is likely that the degree of congestion combined with the severity of renal dysfunction are the true prognostic markers to emphasize rather than the route of diuretic administration. Thus, the specific components of the optimal definition of outpatient WHF are not yet clear. Broadening the WHF criteria to include oral diuretic intensification could represent an advantage in terms of patient-centered trial design (i.e., better characterizing the patient experience) and also support increased event accrual and study power with the potential to shorten trial duration. However, this may be countered by lower specificity in discriminating true (and modifiable) clinical worsening which could create more noise in endpoint assessment.

Comparing WHF events across clinical trials is challenging due to variations in how clinical events are defined and evaluated. A key strength of the present study was the use of a clinical endpoint committee (CEC) with charter definitions for blinded reviewers to independently apply pre-specified criteria when assessing each WHF event. This approach is a rigorous method to reduce subjectivity by applying pre-specified criteria such as one HF-related symptom, two WHF signs (by either exam, natriuretic peptide levels, and/or imaging), and treatment administration in order to positively adjudicate a WHF event. Yet, CECs add incremental costs and the committee may be confronted with challenges due to insufficient clinical data for evaluation (e.g., the natriuretic peptide criteria cannot be evaluated if the biomarker was not assessed or data are not available). Particularly relevant to patients with HFpEF, who commonly have multiple comorbidities (e.g., lung and kidney disease) with signs and symptoms that overlap with HF exacerbation, it can be difficult for clinicians (and CEC reviewers) to tease out whether HFpEF or another disease state is driving the worsening event. On one hand, CEC reviewers may review source documents and feel that the event “looks like WHF” but it may not meet all of the pre-specified criteria (i.e., “HF light”). On the other hand, a patient seen in the emergency room with mixed COPD and HFpEF exacerbation treated with steroids and IV diuretics could meet the pre-specified WHF definition yet the pathophysiology may be more consistent with a pulmonary event. Thus, CEC approaches continue to evolve to account for new insights such as those from the present study.

Since study power is driven, in part, by the number of clinical events, the use of investigator-reported events may offer advantages in blinded, randomized trials. In the case of PARAGON-HF, use of investigated-reported events as compared to adjudicated events increased the number of HF hospitalizations and urgent HF visits by approximately 500 and 200, respectively. This was balanced by approximately 100 fewer cardiovascular deaths by investigator-reported event assessments. Nonetheless, a post-hoc analysis of the expanded primary composite with investigator-reported WHF and cardiovascular death events met the nominal level of significance with a rate ratio of 0.83, p=0.006 (vs. 0.86 with adjudicated events, p=0.040)(5). Thus, the extra events only modestly affected the estimate of treatment effect and these additional analyses demonstrate the overall consistency of the evidence supporting benefits of sacubitril/valsartan on reducing WHF.

Outpatient WHF is increasingly recognized as a relevant endpoint for disease progression in HFpEF. There exists an unmet need for developing effective evidence-based therapies to mitigate WHF occurring both in the hospitalized and outpatient settings. The present study results and the recent FDA Advisory Committee’s recommendation to expand labeling of sacubitril/valsartan provide promise for HFpEF treatment options. Creating and consistently applying uniform definitions of WHF may help to elucidate its true burden and support future efforts to identify new effective treatment strategies in HFpEF populations.