Prognostic significance of longitudinal strain in dilated cardiomyopathy with recovered ejection fraction

Abstract

Objective: Patients with non-ischaemic dilated cardiomyopathy (NICM) may experience a normalisation in left ventricular ejection fraction (LVEF). Although this correlates with improved prognosis, it does not correspond to a normalisation in the risk of death during follow-up. Currently, there are no tools to risk stratify this population. We tested the hypothesis that absolute global longitudinal strain (aGLS) is associated with mortality in patients with NICM and recovered ejection fraction (LVEF).

Methods: We designed a retrospective, international, longitudinal cohort study enrolling patients with NICM with LVEF <40% improved to the normal range (>50%). We studied the relationship between aGLS measured at the time of the first recording of a normalised LVEF and all-cause mortality during follow-up. We considered aGLS >18% as normal and aGLS ≥16% as of potential prognostic value.

Results: 206 patients met inclusion criteria. Median age was 53.5 years (IQR 44.3-62.8) and 56.6% were males. LVEF at diagnosis was 32.0% (IQR 24.0-38.8). LVEF at the time of recovery was 55.0% (IQR 51.7-60.0). aGLS at the time of LVEF recovery was 13.6%±3.9%. 166 (80%) and 141 (68%) patients had aGLS ≤18% and <16%, respectively. During a follow-up of 5.5±2.8 years, 35 patients (17%) died. aGLS at the time of first recording of a recovered LVEF correlated with mortality during follow-up (HR 0.90, 95% CI 0.91 to 0.99, p=0.048 in adjusted Cox model). No deaths were observed in patients with normal aGLS (>18%). In unadjusted Kaplan-Meier survival analysis, aGLS <16% was associated with higher mortality during follow-up (31 deaths (22%) in patients with GLS <16% vs 4 deaths (6.2%) in patients with GLS ≥16%, HR 3.2, 95% CI 1.1 to 9, p=0.03).

Conclusions: In patients with NICM and normalised LVEF, an impaired aGLS at the time of LVEF recovery is frequent and associated with worse outcomes.

Keywords: cardiomyopathy; dilated; echocardiography; heart failure; systolic.

Figure 1 -. GLS correlates with mortality in patients with NICM and recovered LVEF.

Kaplan Mayer survival curves for the primary outcome (i.e., all-cause mortality). Survival curve for patients with aGLS ≥16% is reported in blue. Survival curve for patients with aGLS <16% is depicted in red. The number of patients at risk at each timepoint within each group is reported below the x-axis. Log-rank p= 0.02

Figure 2 –. GLS correlates with cardiac deaths and MACEs in patients with NICM and recovered LVEF.

Panel A: Kaplan Meier Curves showing survival free of cardiac death. Patients with aGLS <16% at time of LVEF recovery are represented in red, patients with aGLS ≥16% are represented in blue. GLS= global longitudinal strain. Log-rank p= 0.01. Panel B: Kaplan Meier Curves showing survival free of the composite secondary endpoint (Cardiovascular death/LVAD/ICD intervention). Patients with aGLS <16% at time of LVEF recovery are represented by a red line, patients with aGLS ≥16% are represented by a blue line. GLS= global longitudinal strain; ICD= implanted cardioverter defibrillator; LVAD=durable left ventricular assist device. Log-rank p= 0.02.

Figure 3 –. Boxplots showing distribution of aGLS in patients who experienced adverse events during follow up and patients who did not.

Panel A: death from any cause. Panel B: death from any cause or durable left ventricular assist device (LVAD) implant. Panel C: death from any cause, LVAD implant or Implantable Cardioverter Defibrillator (ICD) activation. Panel D: Cardiovascular death. Adverse cardiovascular events during follow up are consistently associated with a lower mean aGLS.

Figure 4 –. Iconographic example of echocardiographic images from a patient with recovered Ejection Fraction (LVEF = 51%) and impaired aGLS (9.2%).

Panel A: Echocardiographic 4 chamber view showing LV in end-diastole. Panel B: Echocardiographic 4 chamber view showing LV in end-systole. Panel C: aGLS bull’s eye plot showing impaired aGLS (9.2%).