Cardiac resynchronisation therapy (CRT) is a well-established treatment in patients with severe and drug-refractory heart failure (New York Heart Association (NYHA) class III or IV) with severe left ventricular (LV) systolic dysfunction, dilated left ventricle and, lastly, LV dyssynchrony defined by a QRS duration ⩾120 ms on surface ECG.1 2 In patients selected on the basis of these criteria, CRT significantly improves symptoms, exercise tolerance and quality of life and also reduces morbidity and mortality.3^–6 Despite a better comprehension of the physiopathology of cardiac dyssynchrony and technical improvements, especially in LV lead positioning in the tributary veins of the coronary sinus, a non-acceptable and stable rate of “non-responders” remains the Achilles’ heel of CRT. However, the real rate of non-responders remains difficult to evaluate, particularly because of the heterogeneity of definitions of non-responders. In the MIRACLE trial, the rate of non-responders defined by a composite definition including death, worsening of heart failure or of global assessment and discontinuation of treatment and lack of improvement in NYHA class was 30%.4 The rate of responders defined as patients alive with stable or improved NYHA class without increase in diuretic use in the latest published trial, the CARE-HF study, was 64%.6 7 There are several reasons to explain the lack of efficacy of CRT in patients who are non-responders:

• An inappropriate or non-optimal selection of the patients on the basis of electrical criterion only (QRS width ⩾120 ms on 12-lead surface ECG) as a marker of cardiac dyssynchrony. Previous echocardiographic and magnetic resonance imaging studies have shown that there is not always a strong correlation between electrical and mechanical dyssynchronies, suggesting that patients with a wide QRS might not have mechanical dyssynchrony within the left ventricle and that by contrast a significant …