Repeatability of ejection fraction and left ventricular volume measurement from gated SPECT : using a solid-state CZT gamma camera

Abstract

Background: Many patients undergoing cardiotoxic chemotherapy or with cardiac diseases require re-assessment of left ventricular ejection fraction, and it is important that the test used shows good repeatability. Objectives: The objective of this study was to assess the short-term repeatability of serial re-acquisitions of ejection fraction and volumes from gated single photon emission computed tomography using Quantitative Gated SPECT software and a Discovery NM 530c solid-state gamma camera. Research Design: This study involved retrospective data analysis from 200 patients who had undergone routine repeat Tc-99m tetrofosmin gated SPECT. The patient sample consisted of an equal number of male and female patients and excluded studies with low count rates and small end-diastolic volume (<50 ml). Two successive gated perfusion images had been acquired at rest in the supine position at 45-minutes post injection on the solid-state gamma camera with 10-30 minutes between studies. The standard injected activity was 550 MBq. Results: For the whole population group, there was no clinically significant difference for the ejection fraction (8.9%) between initial and repeat measurements (M = 63%, SD = 12% and M = 59.5%, SD = 13%). The absolute mean difference was 3.8% (SD = 2.83%) for a range of 15 - 88%. The differences in ejection fraction and volumes are similar to previous studies. No significant differences were found between the ejection fraction differences of repeat measurements between groups of age, body mass index, sex and normal/reduced ejection fraction. Conclusions: Results demonstrate comparable repeatability to that from radionuclide ventriculography, indicating that gated single photon emission computed tomography can be used for serial assessment of ejection fraction as an alternative to radionuclide ventriculography. Recommendations: It is recommended that radiographers be properly trained for manual processing when reconstructing the studies in order to minimize errors. As for future work it is suggested that long-term re-acquisitions are compared using a larger sample.