The Misperception of ‘Stable’ Heart Failure

‘There is nothing so stable as change’ Bob Dylan. Heart failure (HF) is a complex syndrome, and the different types of HF are distinguished using various terms and definitions. The American College of Cardiology (ACC) and the American Heart Association (AHA) promoted the use of HF stages (A, B, C, and D), which have served to reinforce the progressive nature of the disease. Physicians who specialize in HF routinely use other common terms like ‘chronic’, ‘acute’, ‘decompensated’, and ‘advanced’, all of which are defined in guidelines and position documents.1-3 However, mirroring the terminology applied to coronary artery disease, we often classify HF as ‘unstable’ and ‘stable’. In HF, the term ‘unstable’ refers to acute, decompensated, or advanced HF. However, ‘stable’ is an orphan concept in chronic HF, even though it is one of the most often used terms in medical charts to describe patient status. Currently, ‘stable HF’ lacks a consensus definition and, in clinical practice, it reflects the physician's perception of the patient's functional situation and prior evolution, or the ambulatory location of care. Physician: “How do you feel?” Patient: “I feel well, the same as at my previous visit.” This exchange involves two participants, the physician and the patient. Responding to the question, “How do you feel?” is not easy for the patient, and the interpretation of the response is difficult for the physician. Indeed, the concordance for the functional class [New York Heart Association (NYHA)] as assessed by the physician and the patient is below 50% and is even worse for low NYHA functional classes.4 In patients classified as NYHA class I, an apparently good NYHA class, the physician often misinterprets the patient's response.4 This lack of concordance accounts for the subjective nature of the perceptions of both the patient and physician, which integrate different and changing views and expectations of wellness, as well as other non-quantitative confounders. Indeed, some patients with chronic NYHA class III over the long term may be accustomed to this condition and consider it their ‘stability’ status, whereas NYHA class III would be perceived as destabilization by the majority of patients. NYHA class I, as defined by the physician, shows the greatest variability in the 6-min walk distances, and both meters walked and peak oxygen consumption overlap substantially between NYHA classes I and II.5, 6 We would say that NYHA class may more practically reflect the physician's ‘eyeball test’ (i.e. do they look ‘great’, ‘ok’, ‘bad’,...), rather than the textbook definition of activity level, which would explain the maintained prognostic value of this qualitative categorization. It is reasonable to ask how a patient maintains a ‘stable’ condition. It seems clear that optimized treatment with neurohormonal agents is the key, and a stable dose of diuretics may also promote stability. However, the reality is not so simple: If a patient is receiving optimized therapy with stable doses of diuretics, is this real stability? Dini et al.7 classified patients with HF and reduced ejection fraction (HFrEF) into stable vs. acute decompensated groups. For stable patients, mortality differed significantly according to diuretic doses, ranging from less than 2% to almost 15% per year.7 In patients with no or mild symptoms (the so-called ‘stable’), furosemide prescription was associated with a higher risk of mortality, especially in patients with optimized treatment and apparently better prognosis.8 Therefore, stability is misinterpreted in patients using diuretics and diuretic requirements are a marker of high risk.8, 9 A clinician should not consider mild symptoms and stability to indicate a good prognosis. Even patients who are less symptomatic are at significant risk. Patients with symptomatic HF at any time (stage C using the ACC/AHA classification) but without symptoms and signs at the clinic visit are at significant risk of death (22% mortality at 5 years).10 This is lower than for patients who are symptomatic (40% mortality at 5 years), but significantly higher than for patients who have never had symptoms nor signs (stage B; 4% mortality at 5 years).10 In a network meta-analysis of 57 randomized clinical trials (phase II–III) in patients with HFrEF, among the 51 949 patient-years included and followed, only 219 patient-years were in NYHA class I at inclusion (0.4%), and the majority were from the PARADIGM-HF trial.11 The 30 years of evidence on the efficacy of neurohormonal blockade in HFrEF with mild-to-moderate symptoms should not mislead us to assume absence of risk. The main evidence that there is no such thing as a truly stable HF patient is that HF is almost always a progressive disease; even when a ‘stable’ patient does not show overt signs of progression, the biomarker levels reflect progression. In the Val-HeFT study, which studied chronic HF in stable patients, levels of high-sensitivity troponin T (hs-TnT) as low as 7.2 pg/mL were associated with a higher risk of death,12 as were N-terminal pro-B-type natriuretic peptide (NT-proBNP) levels of 305 pg/mL.13 In these apparently stable patients with optimized treatment, small changes in the levels of biomarkers that reflect myocardial stretch and ongoing injury were associated with an impaired prognosis.12, 14 Even in chronic HF patients, the persistence of severe left ventricular systolic dysfunction despite optimal treatment represents another biomarker of disease progression.15, 16 We must acknowledge that left ventricular recovery is found in about 25% of patients with HFrEF and, as in those who normalized injury and stretch biomarkers, it confers a better prognosis. This potential to recovery should stimulate both physician and patient to actively look for and to keep on the best possible treatment all the time.17 In addition, sudden cardiac death must not be overlooked given that patients at lower NYHA classes exhibit the highest risk of sudden cardiac death. In this context, only the National Institute for Health and Care Excellence (NICE) guidelines include a quality statement about how to review people with stable chronic HF.18 This statement defines adults with stable HF as ‘adults diagnosed with chronic HF whose clinical condition has not deteriorated, whose heart medication has not been changed and who have not been admitted to hospital because of HF.’ The guidelines propose that such patients be reviewed at least every 6 months to assess whether their condition has deteriorated, whether their medications should be changed, or whether other procedures should be considered. The critical part of this statement is the need to check the patient's condition (e.g. using biomarkers) and specially to check whether medications should be changed, and compliance reinforced. It accounts for the main risk in the ‘stable’ patient, where physician misperception is that no drug changes are needed, and patient misperception is that compliance with life-saving medications is not so important. Stable HF is thus an oxymoron and should not be used as a reason to do or change nothing. Indeed, among patients with mild symptoms in the PARADIGM-HF study, a significant percentage (31%) deteriorated symptoms at 8 months and 11% of patients in the control group exhibited any adverse event at 1 year, being death the first manifestation in 33% of these patients.19 Recently developed risk calculators take into account state-of-the art treatment and biomarker levels and reveal that there is no such thing as truly stable HF. Lupon et al.20 developed a novel HF risk calculator that included multiple markers that reflect different pathophysiological pathways in HF (NT-proBNP, soluble ST2, and hs-TnT). The Barcelona Bio-HF calculator (BCN Bio-HF calculator; www.bcnbiohfcalculator.org)20 is a web-based calculator that refines risk stratification and allows quick and easy interactive calculations of prognosis and life expectancy for individual patients. Figure 1 illustrates its use to calculate the risk of mortality and hospital readmission in one such patient that NICE guidelines would consider stable. For many practicing clinicians, signs and symptoms, together with left ventricular ejection fraction as determined by transthoracic echocardiography, are the main reasons for starting therapy and for stratifying the risk of an individual patient. Companion biomarkers, some of them already incorporated in clinical guidelines, may provide valuable molecular information about heart health. Further, modern imaging techniques are increasingly utilized, particularly cardiac magnetic resonance imaging, and they reveal the anatomy and function of the heart with almost anatomical precision and provide detailed information about accumulating fibrosis in cardiac walls. The more widespread use of biomarkers and molecular imaging techniques will give us a more refined understanding of the multi-faceted phenotypes present in HF and provide a more precise understanding of HF risk.21 In the meantime, the term ‘stable’ HF should be progressively removed from medical charts and guidelines to avoid the misperception that stability carries good prognosis. Instead, the term ‘optimized vs. non-optimized’ treatment should be introduced in medical charts to keep us alert that optimization of guideline-directed medical treatment is the only way to slow or halt disease progression. Conflict of interest: D.