ORIGINAL ARTICLE Interassay reproducibility of myocardial perfusion gated SPECT in patients with atrial fibrillation Santiago Aguade ´-Bruix, MD, a Guillermo Romero-Farina, MD, PhD, FESC, b Gemma Cuberas-Borros, PhD, b Marta Mila `-Lopez, MD, a Virginia Pubul-Nu ´n ˜ez, MD, c Rosa Siurana-Escuer, MD, PhD, d David Garcı ´a-Dorado, MD, PhD, FESC, b and Jaume Candell-Riera, MD, PhD, FESC b Aim. The aim of this study was to assess interassay reproducibility of myocardial perfusion gated-SPECT for calculation of end-diastolic volume (EDV), end-systolic volume (ESV), and left ventricular ejection fraction (LVEF) in patients with atrial fibrillation (AF). Methods. One hundred and fifteen consecutive patients with AF from three participating hospitals (mean age 68.9 years, 39 women) were included in the study. All patients underwent two image gated acquisitions at rest with a 30 minute interval between them. Quantitative data were obtained using the QGS and ECT software algorithms. Results. Heart rate was similar in both studies: 74.94 ± 15.2 vs 73.03 ± 15.57. QGS yielded an LVEF of 54.4%/53.8%, an EDV of 100 mL/101.5 mL, and an ESV of 51 mL/52.3 mL; and ECT showed an LVEF of 63.6%/62.9%, an EDV of 125.8 mL/127.4 mL and ESV of 54.1 mL/ 56.3 mL. Correlation between the two acquisitions was high (>0.948) for both methods for LVEF, EDV and ESV. Regression and Bland-Altman graphics showed a good agreement between all parameters. Interassay variation coefficients for each method (QGS/ECT) were 5.29% vs 4.83% for LVEF, 4.94% vs 5.17% for EDV, and 9.94% vs 12.78% for ESV. Conclusions. Interassay reproducibility of LVEF and EDV with gated-SPECT in patients with AF is good, whereas for ESV it is suboptimal, particularly when ESV is small. (J Nucl Cardiol 2010;17:450–8.) Key Words: Atrial fibrillation Æ gated-SPECT Æ interassay reproducibility Æ myocardial perfusion imaging INTRODUCTION Single photon emission computed tomography R-wave gating, also known as myocardial perfusion gated SPECT (Single Photon Emission Computed Tomography), is a technique used to assess left ven- tricular volume (LVV) and ejection fraction (LVEF) with high reproducibility and agreement with other techniques such as isotopic equilibrium ventriculogra- phy, 1 magnetic resonance imaging 2 and contrast ventriculography 3 in patients in sinus rhythm. Atrial fibrillation (AF), one of the most prevalent arrhythmias in the population studied with gated SPECT, produces changes in cardiac cycle length, meaning that image acquisition is actually a sum of cycles of different duration. If very strict selection cri- teria (heart rate or image duration) are adopted, the result is very poor quality images due to count incon- sistencies at different projections (Figure 1). When signals from detector to computer are disrupted, back- projection slice reconstruction can result in pronounced perfusion artifacts. 4 Furthermore, since gated SPECT geometry-based algorithms are based on definition of myocardial shapes from perfusion images, inconsistent From the Department of Nuclear Medicine, a Hospital Universitari Vall d’Hebron, Universitat Auto `noma de Barcelona, Barcelona, Spain; Cardiology Department, b Hospital Universitari Vall d’Hebron, Universitat Auto `noma de Barcelona, Barcelona, Spain; Department of Nuclear Medicine, c Complejo Universitario Hospitalario, San- tiago de Compostela, Spain; Department of Nuclear Medicine, d Hospital Universitari Joan XXIII, Tarragona, Spain. This study was conducted for the RECAVA Investigator’s Group RECAVA (Cardiovascular Disease Research Network). Received for publication Jun 26, 2009; final revision accepted Dec 11, 2009. Reprint requests: Santiago Aguade ´-Bruix, MD, Department of Nuclear Medicine, Hospital Universitari Vall d’Hebron, Universitat Auto `- noma de Barcelona, Passeig Vall d’Hebron, 119-129, 08035 Barcelona, Spain; saguade@vhebron.net. 1071-3581/$34.00 Copyright Ó 2009 by the American Society of Nuclear Cardiology. doi:10.1007/s12350-009-9186-6 450