Abstract 696: Role of Magnetic Resonance Imaging to Predict Response to Cardiac Resynchronization Therapy: Assessment of Left Ventricular Dyssynchrony and Scar Extent
To assess the value of a novel measure of left ventricular (LV) dyssynchrony derived from cardiac magnetic resonance imaging (CMR) to predict response to cardiac resynchronization therapy (CRT). Furthermore, delayed enhancement CMR was applied to evaluate the impact of the extent and location of scar tissue on CRT efficacy. A total of 35 consecutive heart failure patients scheduled for CRT were included. CMR was performed before CRT to assess LV dyssynchrony and the extent/location of scar tissue. Using a standard 16 segment model, LV dyssynchrony was defined as the standard deviation of time-to-maximum radial wall thickness (SDt-16) obtained from a cine-set of short-axis slices. Standard echocardiography was performed before and 6 months after implantation to determine response to CRT, defined as a reduction ≥ 15% in LV end-systolic volume. Twenty healthy subjects were included as normal controls. At 6 months follow-up, 21 patients (60%) were classified as responders. LV dyssynchrony (SDt-16) was significantly higher in patients [88 ms (67–99)] as compared to controls [26 ms (22–31), p <0.001] and in responders as compared to non-responders [97 ms (90–106) vs. 60 ms (47–71), p <0.001]. Furthermore, SDt-16 was strongly associated with response to CRT (OR=3.8, 95% CI 1.9–7.2, p <0.05). ROC curve analysis revealed that a cut-off value of SDt-16 ≥78 ms yielded a sensitivity of 82% and a specificity of 80% to predict echocardiographic response to CRT (AUC 0.9). In addition, the total amount of scar (TOTscar) was significantly greater in non-responders as compared to responders [35% (12–48) vs. 3% (0–8), p<0.001], while no difference was found between the 2 groups for the scar location. The TOTscar was significantly associated with response to CRT (OR=0.87, 95% CI 0.73– 0.90, p<0.05) and ROC curve analysis revealed that a cut-off value of 13% yielded a sensitivity of 85% and a specificity of 79% to predict response to CRT (AUC 0.85). Of note, a match between the LV lead position, assessed by fluoroscopy, and a transmural scar was found in 57% of non-responders and in 14% of responders (p<0.05). CMR offers the unique opportunity to assess in a single setting both LV dyssynchrony and viability and to predict echocardiographic response to CRT.