Very low-activity stress/high-activity rest, single-day myocardial perfusion SPECT with a conventional sodium iodide camera and wide beam reconstruction processing

Abstract

Background

A stress (S)/rest (R) 1-day Tc-99m sestamibi protocol is logistically advantageous and facilitates stress-only imaging. However, with conventional 370 MBq (10 mCi) S activity and subsequent 1,110-1,295 MBq (30-35 mCi) R activity there is a risk of S-to-R “shine-through” and underestimation of defect reversibility. New software methods cope with lower counting statistics and should allow for both a reduced S activity and also less likelihood of S-to-R “shine-through.”

Methods

102 prospective patients [49 men, 53 women; mean weight 178 ± 41 lbs (range 98-265 lbs); chest 41.5″ ± 4.0″ (range 32″-52″)] received 192.4 + 18.5 MBq (5.2 ± 0.5 mCi) Tc-99m sestamibi S (25 exercise, 77 regadenoson) activity followed in 30-40 minutes by “full-time” (12 minutes) two-headed NaI camera S SPECT. Immediately thereafter, a 16-minute S SPECT acquisition was also performed in 37/102 patients. Then at 60-80 minute post-S all patients received 1328.3 + 129.5 MBq (35.9 ± 3.5 mCi) Tc-99m sestamibi, and “half-time” (7.5 minutes) R SPECT was acquired. All tomograms were processed with wide beam reconstruction (WBR, UltraSPECT Ltd.) software. A time-adjusted R/S myocardial count density ratio (MCDR) was calculated using automated software. S SPECT quality was visually graded (poor, fair, good, excellent) based upon myocardial definition, cavity contrast, RV visualization, and noise. For comparison, the S/R MCDR was calculated in 581 consecutive patients undergoing a conventional 370 MBq R/1110 MBq S (10 mCi R/30 mCi S) protocol.

Results

S SPECT was normal in 44 patients (43%). Image quality was good-excellent in 93 (91%) patients with 12-minute S SPECT. Also in 37 (98%) patients with 16-minute S SPECT, quality was good-excellent. In patients with >42″ chests 12-minute S SPECT quality worsened with increasing chest circumference, manifested by myocardial “blurring.” Image quality improved by ≥1 grade in the 12/37 patients (32%) also undergoing 16-minute S SPECT. The time- and decay-corrected 12-minute mean R/S MCDR was 5.78, a ratio adequate to minimize S-to-R shine-through, as verified in phantom experiments, and significantly better than a 3.79 S/R ratio achieved in the 581 patients undergoing a conventional R/S protocol.

Conclusions

An approximately 185 MBq (5 mCi S) Tc-99m SPECT processed with WBR provides adequate image quality. For larger patients prolonging image acquisition to 16 minutes is beneficial. For patients with normal S SPECT, a S-only protocol is feasible, affording them a very low (approximately 1.4 mSv) radiation dose. If subsequent R SPECT is necessary, it can be performed with approximately 1,332 MBq (36 mCi) with minimal S-R “shine-through.”

Introduction

A stress/rest single-day Tc-99m sestamibi protocol is logistically advantageous and facilitates stress-only imaging.1, 2, 3 However, with a conventional 370 MBq (10 mCi) stress-injected activity and a subsequent 1,110 MBq (30 mCi) rest activity, there is a risk of stress-to-rest “shine-through” and underestimation of defect reversibility. For example, in a patient with stress-induced ischemia if the stress perfusion distribution is abnormal and if the resting perfusion distribution is entirely normal, the stress defect may be partially apparent in the resting scan if the rest/stress count density is not sufficiently great enough to minimize contribution, i.e., “shine-through,” from the initial stress perfusion scan.

In this study, we report a single-day stress/rest protocol using approximately 185 MBq (5 mCi) stress activity and approximately 6845 MBq (35 mCi) rest activity. We postulate that with this new protocol post-stress perfusion and gated functional image quality will be preserved despite the lower injected activity by means of wide beam reconstruction (WBR) processing, employing iterative reconstruction, resolution recovery, and noise reduction software.4, 5, 6, 7 Moreover, we postulate that the rest-to-stress count density ratio will be sufficiently high to minimize the likelihood of stress-to-rest “shine-through.”