Author + information
- Jort Merken, MD,
- Hans-Peter Brunner-La Rocca, MD,
- Jerremy Weerts,
- Job Verdonschot, MD,
- Mark Hazebroek, MD, PhD,
- Georg Schummers,
- Marcus Schreckenberg,
- Joost Lumens, PhD,
- Stephane Heymans, MD, PhD and
- Christian Knackstedt, MD, PhD∗ (, )@MaastrichtUMC@maasuniversity
- ↵∗Maastricht University Medical Center, Department of Cardiology, PO Box 5800, 6202 AZ Maastricht, the Netherlands
Recently, the term heart failure (HF) with recovered ejection fraction (HFrecovEF) was introduced for patients with a normalization of the left ventricular ejection fraction (LVEF) (1). The question remains whether this means true recovery of systolic left ventricular function or simply normalization of LVEF as many patients show remaining abnormalities in biomarker profiles and high hospitalization rate (1). Therefore, this research aims to assess the additional value of global longitudinal strain (GLS) in patients with HFrecovEF as physicians are left with uncertainties regarding the therapeutic approach (e.g., halt or continue HF medication).
Patients were identified via the electronic hospital health record system (EHR) based on the documented, routine LVEF measurement. In total, 212 patients met the inclusion criteria of an initial LVEF <35% with improvement to ≥55%, and had a complete follow-up. Patient characteristics were collected from the EHR. The reason for developing HF was rated according to the current ESC guidelines (2). GLS was measured on routine echocardiographic images at the point of time that LVEF was normalized using automated software (AutoLV module; part of the TomTec-Arena 2.20.10; TomTec Imaging Systems, Unterschleissheim, Germany). GLS was considered normal if ≤−21.5% and abnormal if >−21.5% (3). The primary endpoint was death or hospitalization due to HF during a follow-up of 56 ± 21 months (range 12 to 90 months) analyzed from normalization of LVEF onwards.
The reason for HF in the HFrecovEF patients was classified as: nonischemic (n = 41), due to arrhythmias (n = 70), abnormal loading conditions (n = 34), and “other” (e.g., Takotsubo) (n = 38). Patients with an ischemic cause (n = 29) showed similar survival and combined primary endpoint results as compared with patients with a nonischemic etiology (n = 183) (p = 0.91 and p = 0.92, respectively). The duration between onset of symptoms and initial echocardiogram was overall 10 ± 7 (1 to 18) weeks, between initial echocardiogram and follow-up echo was 16 ± 8 (5 to 23) months. Overall, image quality was rated as good (91%), moderate in 8%, and poor in 1% (p = 0.26 between GLS groups).
Seventy-nine percent (n = 168) of patients still had an abnormal GLS despite normalization of LVEF. No significant differences regarding symptoms, medication, physical findings, further echocardiographic measurements (e.g., mitral annular plane systolic excursion, diastolic function) or underlying etiologies of HF were detected comparing patients with normal and abnormal GLS during follow-up. Also, no significant differences regarding interventions, operations, or new diagnoses were observed between GLS groups. The vast majority of patients (96%) showed a normal diastolic function at the time of normalization of LVEF with no significant difference between patients with normal/abnormal GLS (p = 0.11). Importantly, an abnormal GLS was associated with a significantly worse outcome of all-cause death and hospitalization due to HF (Figure 1) compared with patients having a normal GLS. There were 22 deaths in the group with abnormal GLS and only 1 in the other group (p = 0.05).
Thus, HFrecovEF is purely defined on improvement/normalization of LVEF. Still, normalization of geometry and/or LVEF does not necessarily mean full myocardial recovery (4). This difference is clinically important as decisions are taken based on dimensions and function rather than on tissue characteristics. Therefore, the mere depiction of LVEF, that is, volumetric changes, does not seem to be a sufficiently comprehensive marker of systolic function. GLS seems to provide additional characterization of systolic function as a significantly better death and HF hospitalization-free survival was seen in patients with normal GLS. These findings may have 2 direct implications for physicians caring for patients with HFrecovEF: 1) because systolic function remains altered in most patients, medication should not be ceased in patients with HFrecovEF; and 2) because many physicians rely on the measurement of LVEF, it calls for implementation of modern imaging tools (e.g., GLS or stress echocardiography).
In conclusion, the majority of patients with HFrecovEF still have an abnormal systolic function, as measured by GLS. Patients with an abnormal GLS, despite normalization of LVEF, showed a significantly worse outcome. Further studies should evaluate therapeutic consequences.
Please note: Dr. Lumens is supported by research grants from the Dutch Heart Foundation (NHS-2015T082) and the Netherlands Organization of Scientific Research (NWO- 016.176.340). Mr. Schummers and Mr. Schreckenberg are employees of TOMTEC Imaging Systems. Dr. Lumens has received research software and hardware from TOMTEC Imaging Systems. Dr. Knackstedt has received travel grants, research software, and hardware from TOMTEC Imaging Systems. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2018 American College of Cardiology Foundation
- Basuray A.,
- French B.,
- Ky B.,
- et al.
- Ponikowski P.,
- Voors A.A.,
- Anker S.D.,
- et al.
- Cohn J.N.,
- Ferrari R.,
- Sharpe N.,
- on Behalf of an International Forum on Cardiac Remodeling