Which echocardiographic definition should be used to define response to cardiac resynchronization therapy?

Introduction: The definition of response to cardiac resynchronization therapy (CRT) remains controversial, with different criteria being used to define a positive response. The PROSPECT trial recently demonstrated that echocardiography is not sufficiently accurate to identify responders to CRT. However, it is possible that the definition used in this study was not the most appropriate.

Objective: To compare different echocardiographic definitions of response to CRT with peak oxygen consumption (VO2), in an attempt to identify the best echocardiographic definition.

Methods: Thirty consecutive patients who underwent echocardiography and cardiopulmonary exercise testing (CPET) before and 6 months after CRT were studied. An improvement of > or =1 NYHA class defined clinical responders; a > or =15% decrease in left ventricular end-systolic volume (LVESV) defined remodeling responders; a > or =25% improvement in left ventricular ejection fraction (LVEF) identified responders according to LVEF; a >25% improvement in left ventricular dP/dt defined responders according to dP/dt; and a ?10% improvement in peak VO2 defined CPET responders.

Results: There were 47% responders according to the reverse remodeling definition, 60% according to LVEF and 67% according to dP/dt; 77% were clinical responders and 40% CPET responders. The only baseline characteristic that differed between CPET responders and non-responders was the sphericity index (57 +/- 12% vs. 72 +/- 16%, p = 0.019), which showed an inverse correlation with CPET response (r = -0.455, p = 0.011). LVEF response showed the best agreement with CPET response (83% positive and 56% negative concordance). Clinical and echocardiographic responses were often discordant: 48% of clinical responders were non-responders according to reverse remodeling, 35% according to LVEF and 39% according to dP/dt. However, of clinical responders who did not respond on echocardiographic criteria, a positive NYHA response paralleled the CPET definition in 35% of cases.

Conclusion: The best agreement between echocardiographic definitions of response and CPET was achieved with LVEF. In 35% of cases of discrepancy between clinical and echocardiographic responses, the clinical response paralleled CPET, which implies a benefit of CRT undetected by echocardiography and not a placebo effect.