Patients with severe aortic stenosis (AS) and reduced ejection fraction (EF) should be promptly operated on. Global longitudinal strain (GLS) has been proposed as a subtle subclinical marker of left ventricular (LV) systolic dysfunction and a potential prognostic factor in asymptomatic patients with severe AS and preserved EF. However, the relation between longitudinal strain, EF, AS severity and symptoms has not been fully evaluated.

Methods and results .– This prospective study included 171 patients (73 ± 12 years, 40% female) with at least mild AS (defined by mean pressure gradient > 10 mmHg). All patients had a comprehensive transthoracic echocardiography (TTE). Global and basal longitudinal strain (BLS) was measured offline with a speckle tracking method.

Among the global population, mean aortic valve area 1 cm ² ; mean pressure gradient 44 mmHg and mean EF 62%.

Four subgroups were identified: 55 asymptomatic patients with non-severe AS and normal EF (group 1); 37 patients with asymptomatic severe AS and normal EF (group 2); 60 patients with symptomatic severe AS and normal EF (group 3) and 19 patients with severe AS and LV systolic dysfunction (group 4). GLS was significantly lower in patients with severe AS and reduced EF than in patients with AS and preserved EF (−13.7 ± 5 versus −18.3 ± 3 respectively, P = 0.0006). In univariate analysis, the best correlation was found between GLS and EF ( r = 0.48, P < 0.0001). In multivariate analysis performed after adjustment for AVA and EF; GLS was not independently associated with symptomatic status ( P = 0.15).

Basal longitudinal train decreased gradually according to AS severity, symptomatic status and LV dysfunction: BLS was higher in patients with asymptomatic AS than in patients with symptomatic severe AS or patients with severe AS and LV dysfunction (−16.1 ± 4 versus −12,5 ± 3 versus −9,7 ± 5, P < 0.0001). In univariate analysis, the best correlation was found between BLS and mean pressure gradient ( r = 0.46, P < 0.0001).

In multivariate analysis including BLS, EF and AVA; the independent determinants of symptomatic status were AVA ( P = 0.0003) and BLS ( P = 0.0003).

Conclusions .– In a large population with a wide range of AS severity, EF and symptoms, BLS but not GLS, was independently associated with symptoms in patients with AS, suggesting it may provide interesting prognosis information. However, because of an important overlap for the same values of BLS, a cut-off value of BLS associated to symptomatic status seemed difficult to be applied to the global population.

Further studies looking at the relation between the progressive decrease of BLS for each individual and the apparition of symptoms may help to confirm BLS as a prognostic marker among patients with asymptomatic severe AS and preserved EF.