Effect of cardiac disease history on gated blood pool SPECT LV measurement accuracy

Abstract

Background: Recently software has become available for the routine analysis of left ventricular (LV) end-diastolic volume (EDV) and ejection fraction (EF) from gated blood pool SPECT (GBPS) data. This investigation examined whether accuracy of GBPS measurements, using cardiac magnetic resonance (CMR) imaging as the standard, are influenced by unusual LV shapes, as would be expected for patients with prior MI. Methods: LV functional parameters were measured for 28 subjects (age=58±12 years; 85% male, 13 with prior MI (MI) and 15 with non-ischemic cardiomyopathy (CM) by GBPS, planar-GBP and by true-FISP CMR. GBPS calculations from a count-threshold algorithm (BP-SPECT, Columbia University, New York, NY), and a surface gradient algorithm (QBS, Cedars-Sinai Medical Center, Los Angeles, CA) were evaluated. Results: By linear regression analysis BP-SPECT, QBS and planar-GBP EF's correlated similarly with CMR (r=0.92, 0.88 and 0.90, respectively) and both GBPS methods correlated similarly versus CMR for EDV (r=0.88 and 0.89, respectively). By the paired t-test, CMR EF's (38%±17%) were not different from BP-SPECT, QBS or planar-GBP (44%±18%, 36%±19%, and 38%±16%, respectively) for MI and CM patients. BP-SPECT and QBS EDV's were found to be significantly lower than CMR for MI patients (159±51 ml and 140±69 ml versus 207±70 ml) but not for CM patients (223±108 ml and 206±111 ml versus 274±109 ml). Conclusion: GBPS LV function measurements closely parallel those of CMR and planar-GBP, but volume accuracy is influenced by presence or absence of prior myocardial infarction.