Myocardial perfusion SPECT shows speedier acquisition possibilities

GE Hawkeye SPECT system. Image courtesy of GE Healthcare.

Given the widespread prevalence of heart disease in the United States, methodology that allows a decrease in the time needed to perform the roughly 9 million annual SPECT myocardial perfusion imaging (MPI) exams would be warmly received by clinicians. A study delivered at the recent Radiological Society of North America annual scientific assembly suggests that some reconstruction algorithms can allow MPI acquisition in a shorter timeframe.

“Recently introduced myocardial perfusion SPECT reconstruction algorithms such as wide-beam reconstruction [WBR, from Haifa, Israel-based UltraSPECT] allow acquisition of MPI in a reduced amount of time,” said Emily Siegal, MD, who presented the results of a study conducted on the protocol at the Lahey Clinic in Burlington, Mass.

Because left ventricular ejection fraction (LVEF), end diastolic volume (EDV), and end systolic volume (ESV) estimates derived from MPI have patient management implications, Siegal said that it is critical to ensure that the estimates with half-time WBR acquisition are as reliable as routine full-time acquisition (FTA).

The group’s study cohort consisted of 41 patients referred for MPI, 29 men and 12 women with a mean age of 62.8 years. All patients were imaged on a GE Healthcare Millennium VG Hawkeye SPECT gamma camera using an FTA protocol of 22 minutes and a WBR acquisition protocol of 10 minutes, and the projection data were processed using routinely accepted methods, according to Siegal.

The researchers used three different programs for estimating LVEF, EDV, and ESV for both the WBR and FTA protocols: the Emory Toolbox (ETB); Cedars-Sinai’s quantitative gated SPECT (QGS) software, and the 4DMspect application. Siegal said that the two protocols were then compared with paired t-test.

“The comparison of LVEF by FTA and WBR protocols by paired t-test showed no significant difference in LVEF between routine and half-time acquisitions with ETB and QGS, with p values of 0.91 and 0.35 respectively,” she said.

Siegal said that this suggests that the estimates from the two protocols were comparable and probably interchangeable. She noted, however, there was a significant difference with 4Dmspect in the LVEF comparison, which had a p value of 0.004.

Also, she reported that there also was more variation in EDV and ESV data between the two protocols, with statistically different values when ETB and QGS were used. In contrast, p values were not significantly different with 4DMspect in regards to the EDV and ESV data in the protocols.

“The results of our small study suggest that WBR could potentially replace FTA for LVEF estimates,” she said. “However, larger studies are necessary to clarify the variability in EDV and ESV estimates. Every lab should compare the two protocols and obtain normal thresholds prior to the routine use of WBR.”