Abstract 3255: Referral Heart function Assessment for sudden death Prevention: a Study Of Determining ICD eligibilitY (RHAPSODY) - repeat heart function assessment using echocardiography and nuclear ventriculography
Background: Guidelines for ICD implant rely heavily on measures of left ventricular ejection fraction (LVEF). Studies suggest that nuclear ventriculography (MUGA) measures are accurate and reproducible. However, echocardiographic measures are used more often. We performed a prospective cohort study to compare echocardiography and MUGA scans in patients referred for ICD implant.
Methods: Between January 2005 - March 2006, patients referred for consideration of ICD implant with an echocardiogram (echo) LVEF ≤30% were prospectively screened for enrollment. Baseline clinical data was obtained including the referral echo. All patients had a repeat assessment of LVEF with an echo and MUGA performed at a tertiary care centre. Eligibility for an ICD was strictly defined as an LVEF ≤30% for the purpose of this study.
Results: Seventy-one patients were included (mean age 62±9 years, 61 (86%) male). Mean referral LVEF was 22±5%; 66 (93%) were for primary prevention and 36 (51%) had ischemic cardiomyopathy. The median time interval was 190 days between the referral and repeated imaging studies, with 23 days between MUGA and the repeated echo. The mean LVEF was 23±8% for the repeat echo (p=0.31 compared to referral echocardiograms) and 28±13% for the MUGA (p<0.001 compared to referral echo, p<0.001 compared to repeat echo). The correlation coefficient for the LVEFs measured by referral and repeat imaging was 0.50 (repeat echo) and 0.39 (MUGA), while the correlation coefficient for the repeat echo and MUGA was 0.65. Only 47 patients (66%) met criteria for an ICD on repeat imaging despite apparent referral eligibility; 8 (11%) were excluded by both the repeat echo and MUGA, and an additional 16 (23%) excluded on the basis of the MUGA alone. There were no cases where a repeat echo found a patient ineligible for an ICD while the MUGA scan result demonstrated an LVEF ≤30%.
Conclusions: Repeat LVEF assessment identifies a substantial proportion of referred ICD candidates who do not meet LVEF guidelines. Nuclear ventriculography is more effective than echocardiography at identifying patients ineligible for ICD implant. Previous trials may have underestimated ICD efficacy by relying on an imprecise measure of ejection fraction.