Left ventricular aneurysm in the scope of gated perfusion SPECT: accuracy of detection and ejection fraction calculation

Abstract

Purpose The objective of this trial was to investigate the capacity of gated perfusion SPECT (GPS) to detect left ventricular aneurysm (ANV) by comparing QGS and 4D-MSPECT (4DM) algorithms with radionuclide ventriculography (RVG). Secondarily, the comparison of GPS ejection fraction (EF) measurements with those of contrast left ventriculography (LVG) and RVG was aimed. Methods Twenty-five patients with ANV confirmed by LVG were studied. The patients underwent RVG and rest Tc-99m-tetrofosmin GPS 1 week after LVG. A 9-segment model was used both in RVG and GPS evaluation. Aneurysm was defined by scoring the wall motion (WM) and phase analysis in RVG; perfusion, wall thickening and WM in GPS. Results The detection rate of ANV was 96%, 84% and 52% for RVG, QGS and 4DM, respectively. The LVG mean EF (43.52% +/- 16.93%) was significantly higher (P < 0.01) than those of RVG n (29.40% +/- 10.90), QGS (30.04% +/- 13.25%) and 4DM (34.92% +/- 13.01%). Moderate to high EF correlation values were obtained between LVG and GPS (r = 0.71-0.79) and GPS-RVG (r = 0.69). There was no significant EF difference between the radionuclide methods except between 4DM- EF and RVG-EF (5.52%, P < 0.05). Wide Bland-Altman limits were observed between the radionuclide methods in EF comparisons (range: 30.5-38.5%). Conclusion GPS seems to have a role in the non-invasive investigation of ANV. QGS-GPS proved to be more reliable (84%) than 4DM- GPS (52%)in the ANV detection. The localization and the extent of the aneurysm itself as well as perfusion and function of adjacent segments may affect aneurysm diagnosis by means of GPS. RVG, QGS-GPS and 4DM-GPS seem not to be interchangeable for routine EF calculation in ANV patients.