Impact of attenuation correction on quantitative hybrid CZT SPECT/CT imaging in relation to BMI

Objectives 64-slice CT has been routinely used for attenuation correction (AC) with PET, although now is being applied with hybrid CZT SPECT/CT systems for quantification of myocardial defect size. This X-ray CT AC approach has not been validated for SPECT imaging in a large patient population. This study explores the effects of CT-based AC on SPECT defect quantification in relation to BMI.

Methods We conducted a single-center study with 522 consecutive patients (mean age 62 +/-13 yr, BMI 29.3+/-5.9, 60% male) having stress and rest Tc-99m tetrofosmin SPECT images acquired on a hybrid CZT SPECT/64-slice CT system. The integrated myocardial perfusion defect (IMPD) was calculated using standard circumferential profile analyses and gender-matched low disease risk databases applying the WLCQ software on stress images with and without AC. Segments were grouped into coronary territories (LAD/LCx/RCA). A paired two-tail T-test was used to compare the AC and non-AC studies for stress. Patients were further segregated based on the BMI; < 30 kg/m2 (279 patients) or ≥ 30 kg/m2 (243 patients).

Results In patients with BMI <30, only the LAD territory showed a significant difference between the AC and non-AC IMPD, (0.67+/-1.95), p<0.0001 with a 95% CI (0.43 to 0.89). In patients with BMI ≥ 30, both the LAD and RCA territories showed a significant difference in quantitative IMPD (LAD: 0.38+/-1.66%, p<0.0001 with a 95% CI (0.17 to 0.59); RCA: 0.2+/-0.7%, p<0.0001 with a 95% CI (0.11 to 0.29)). Additionally, LAD defects were also more likely to normalize with AC compared to RCA defects in all 522 patients (27% vs. 7%).

Conclusions There is a small but significant reduction in the IMPD with AC in this large cohort of consecutive patients imaged with a hybrid CZT SPECT/CT system, especially in patients with a larger BMI. LAD defects were also more likely to normalize with AC compared to RCA defects in this cohort.