ABSTRACT & COMMENTARY
Heart Failure with Recovery of LV Function
By Michael H. Crawford, MD, Editor
SOURCE: Basuray A, et al. Heart failure with recovered ejection fraction: Clinical description, biomarkers, and outcomes. Circulation 2014;129:2380-2387.
The clinical implications of the recovery of left ventricular (LV) function after treatment of patients with initial systolic heart failure is poorly understood. Thus, these investigators from the Penn Heart Failure Study sought to characterize this population, believing them to be different from patients with heart failure with persistently reduced LV function and those with heart failure with preserved LV function. Patients were recruited from three specialized heart failure centers and divided into three groups: heart failure (HF) with reduced LV ejection fraction (EF) < 50% (HFrEF); HF with preserved EF > 50% (HFpEF); and HF with recovered EF (HFrec), meaning it was < 50% and then was > 50% after treatment. Patients with all causes and treatments of HF were included, except those with hypertrophic and infiltrative cardiomyopathies. The final study population included 1821 patients, most of whom had HFrEF (n = 1523); 122 had HFpEF and 176 had HFrec. At baseline the HFrec group was younger and had less comorbidities such as coronary artery or kidney disease. During follow-up for up to 8.9 years (median 3.6), 507 experienced a terminal event (335 death, 129 transplant, 43 ventricular assist device placements). HFrEF had the highest likelihood of reaching the combined endpoint (hazard ratio [HR], 4.1; 95% confidence interval [CI], 2.4-6.8; P < 0.001) vs HFrec. HFpEF patients were also more likely to reach this endpoint vs HFrec (HR, 2.3; 95% CI, 1.2-4.5; P < 0.02). However, after 8 years nearly 20% of HFrec patients had achieved this endpoint. Hospitalization for HF was more common in the HFrEF group, but not different between the other two groups. The authors concluded that HFrec patients have an increased event-free survival compared to HFrEF and HFpEF patients, but experience similar numbers of subsequent hospitalizations as HFpEF patients.
I have several patients who had systolic heart failure for a variety of reasons, who responded to therapy and now have a normal LVEF on medical therapy. They keep asking me if they have to keep taking their HF medications. This paper sheds some light on this issue. These investigators propose that this is a unique group with different characteristics than HFrEF and HFpEF patients, and with a different prognosis. Their data show that these patients have a lower mortality than the other two types of heart failure, but they continue to have HF hospitalizations, so they are not cured. Also, they have continued evidence of neurohumeral activity such as elevated brain natriuretic peptide levels and detectable troponin levels. They suggest these data support the rationale for continued medical therapy. However, we have all seen such patients who have self-discontinued their medications and they have done fine. So some must be cured, but how do we tell whom? Perhaps by biomarkers or sophisticated imaging, but this retrospective observation study doesn’t address this issue. In this study 88% of the HFrec patients were on beta-blockers, 85% on ACEI, and 69% were on diuretics. Clearly the authors are keeping most of their patients on their HF medications indefinitely.
The investigators raise another interesting issue. Perhaps some patients we are categorizing as HFpEF are really HFrec patients, except we don’t have prior echocardiograms or other measures of LV function to determine this. This thought should prompt a more thorough search for outside hospital or other old records in these patients, because the treatment of HFpEF is unknown, but HFrec patients should probably stay on HFrEF therapy indefinitely.
Of course, LVEF is a crude measure of LV performance. Perhaps more sophisticated imaging or hemodynamic studies would discover subtle abnormalities of LV function in the HFrec patients. Also, the use of a 50% LVEF cut point is arbitrary. Some studies have used 45% or 55%. Another limitation of this study is that the deployment of medical therapy and devices was not analyzed further. Also, we don’t have an etiologic breakdown. This could be important. For example, I rarely see a CAD patient with HRrEF return to a normal LVEF after revascularization, but post-chemotherapy patients often do. More detailed clinical data will have to come from prospective studies, which hopefully are in the works.