No Longer from Pillar to Post: The First Effective Step in Treating Heart Failure with Preserved Ejection Fraction

This article refers to ‘Health status across major subgroups of patients with heart failure and preserved ejection fraction’ by T.J. Siddiqi et al., published in this issue on pages 1623–1631. The classification of heart failure using the New York Heart Association (NYHA) functional grouping began in 1921, prior to the advent of imaging or natriuretic peptides, and well before the grouping of heart failure into reduced or preserved ejection fraction. This system based on the clinician assessment of patient symptoms, provided physicians with a method of communicating prognosis and risk of mortality, highlighting the intertwined link between quality of life and cardiovascular mortality. Over time however, it has become increasingly recognized that the tool lacks precision1; dependence on the clinician's view rather than the patient's experience, significant subjectivity and poor reproducibility between the mid-range classes with frequent transitions, and limited prognostic value. Consequently, patient-reported outcome measures have been developed, such as the Kansas City Cardiomyopathy Questionnaire (KCCQ) score, which outperforms the NYHA classification with regard to association with mortality or hospitalization.2 The KCCQ has also demonstrated superiority over other scores for prognostication and is more strongly associated with baseline functional status.3 Heart failure with preserved ejection fraction (HFpEF), which has emerged as the prevalent form of heart failure within Western societies, continues to pose a significant therapeutic challenge due to the dearth of effective treatment options. Over the past decade, various therapies that began as heart failure with reduced ejection fraction (HFrEF) trials have come and gone, with failures from three of the four pillars of heart failure. The DELIVER4 and EMPEROR-Preserved5 trials were the first large scale pharmacologic trials to demonstrate benefit of a therapy across the spectrum of heart failure. In this issue of the Journal, Siddiqi and colleagues present a substudy of the EMPEROR-Preserved trial, focused on the changes in quality of life scores measured following 12 months.6 The complete EMPEROR-Preserved trial has been published previously.5 To summarize, the trial was a randomized controlled trial of patients with predominantly NYHA class II heart failure with mildly reduced ejection fraction (HFmrEF) and HFpEF, run across 23 countries and including a total of 5988 patients (2979 with KCCQ measurements included in this sub-analysis). The overall trial primary outcome was composite of cardiovascular death or hospitalization for heart failure; there was a significant reduction in the group assigned to empagliflozin, with a hazard ratio of 0.79 (95% confidence interval [CI] 0.69–0.90), with benefits regardless of diabetes status. The trial measured quality of life using the well validated KCCQ score, with data captured at baseline, 12, 32 and 52 weeks post-randomization. The results of this trial are in line with the findings from EMPULSE,7 which evaluated the impact of empagliflozin on quality of life in patients admitted with decompensated heart failure, regardless of ejection fraction. The EMPERIAL-Preserved trial evaluated treatment with empagliflozin in patients with HFmrEF/HFpEF, however was significantly smaller (n = 315) than the EMPEROR-Preserved and for 12 weeks only. In this trial, there was a similar median KCCQ score, age, body mass index, rate of atrial fibrillation/flutter and median ejection fraction, treated for 12 weeks. The overall primary outcome of 6-min walk test distance was neutral, with little change in the overall KCCQ overall summary score (+0.27 points [95% CI −3.09 to 3.64]). The DELIVER trial was a similar international double blind randomized controlled trial of dapagliflozin in patients with HFmrEF/HFpEF. A dedicated substudy of health status has been published, evaluating 5795 patients with a median KCCQ of 72.9, slightly higher than this trial.8 Dapagliflozin also improved KCCQ values early, and to a similar extent to that seen in EMPEROR-Preserved, implying a class effect. PRESERVED-HF was a smaller randomized trial in NYHA class II–IV patients with an ejection fraction >45%.9 In this trial, there was a smaller proportion of NYHA class II (57% vs. 81%) and a lower median KCCQ (∼63 points). This group of patients had a greater improvement in the overall KCCQ score, increasing by 4.5 points at 3 months, with a numerically greater improvement in those with a higher body mass index. Importantly, this is not directly comparable to the present study, in view of the much shorter study duration. Quality of life scores have been variably captured across previous studies of HFpEF, predominantly either using the KCCQ or the Minnesota Living with Heart Failure score, however it is equally reduced in HFpEF and HFrEF.10 The baseline quality of life score in this trial was higher at baseline compared with other studies of HFpEF. The participants of the TOPCAT study, evaluating spironolactone versus placebo, had a median KCCQ overall summary score of 54, lower than that estimated in CHARM-Preserved or I-PRESERVE,11 with an improvement of only 1.5 points. Patients in REDUCE LAP-HF-II, evaluating the role of an interatrial shunt in NYHA class II–IV ambulatory HFpEF alone, had a markedly lower KCCQ, at ∼40 points.12 PARAGON-HF, assessing sacubitril/valsartan in HFmrEF/HFpEF, had a baseline of 71 points with a change of 1 point at 8 months.13 Quality of life has been previously analysed in depth by Reddy et al.14 in the NEAT-HFpEF and INDIE-HFpEF groups, trials of isosorbide mononitrate and inorganic nitrites, respectively. In those trials, patients with the worst quality of life were more likely to be obese, diabetic with low physical activity levels. Interestingly, the worst tertile was younger (KCCQ <45, median 66 years) whereas older age was noted in this substudy (KCCQ <62.5, median age 73 years), albeit with different KCCQ cutoffs for each tertile. Siddiqi and colleagues elegantly demonstrate the association between KCCQ scores over time with the development of cardiovascular outcomes in patients with HFpEF,6 similar to what has been show in HFrEF previously.13, 15 Although outcomes such as cardiovascular death and heart failure hospitalization are no doubt of importance to patients, the generally older and more comorbid patients with HFpEF are also concerned with day to day quality of life; the majority of heart failure patients are willing to accept a decrease in longevity for an improved health status.16 Quality of life in HFpEF is influenced by numerous elements, and with the escalating significance of managing comorbidities, it is critical to acknowledge that their effective management not only ameliorates the underlying physiological complexities associated with the disease, but also improves overall quality of life. There are several limitations to note of this study however; due caution must be exercised while interpreting results from any substudy, due to the inherent reduction in statistical power, albeit recognizing the large initial size of the trial. The magnitude of improvement in KCCQ is small, relative to typical definitions.16 Two separate groups of patients are combined in this study – those with HFmrEF and HFpEF – and mechanistic data clearly separate these as different types of patients. There was notable regional variation in outcomes across the trial; North American trial participants seemed to derive very little benefit, however this analysis may be underpowered as only made up 12% of the trial population. The participants in this trial had significantly higher median KCCQ scores than most other trials of HFpEF or HFrEF, and the magnitude of benefit was smaller than that seen in patients with KCCQ scores of 60 or below. A common question in heart failure research pertains to what constitutes a meaningful change in KCCQ scores, particularly in the heterogeneous population of those with HFpEF. Presently, a numerical change of 5 in the score is accepted as the minimal clinically important difference,16 although this remains an arbitrary designation. Taken together, the research presented by Siddiqi et al.6 represents an important step forward in the management of HFpEF. Firstly, validation of the effectiveness of sodium–glucose cotransporter 2 (SGLT2) inhibitors to improve not only cardiovascular death and hospitalization, but positively impact the quality of life in the short to medium term in patients with heart failure with preserved ejection fraction. Secondly, the recognition that obese patients gained significant benefit, a group previously highlighted as having the worse quality of life and one representing a significant phenotype of HFpEF across Western nations. Finally, identification of the association between early changes in quality of life as a predictor to the development of harder outcomes at 12 months, and longer term data is awaited to assess the disease modifying abilities of SGLT2 inhibitors in these patients with less severe HFpEF. The era of therapeutic nihilism is finally coming to an end in the management of HFpEF. This trial, taken with the recent body of data around SGLT2 inhibitors, demonstrate efficacy beginning at hospital discharge and extending beyond 12 months. These agents appear to offer benefits across the quality of life spectrum, however certain groups seem to derive greater benefit than others; the key trials of the future will be based on the identification of anatomical and physiologic phenotypes, to better target pharmacologic and device-based therapies. Conflict of interest: none declared.