Comparability of heart-to-mediastinum ratio determination from I-123 mIBG SPECT and planar imaging: New analyses from ADMIRE-HF

Catégorie Primary study
JournalJournal of the American College of Cardiology
Year 2011
Loading references information
Background: The purpose of this study was to evaluate a new method of global quantitation of cardiac uptake on I123 mIBG SPECT for distinguishing patients with abnormal and normal cardiac uptake in comparison to standard planar heart/mediastinum ratio (H/M). Methods: The study included 1016 subjects (923 heart failure (HF), 93 controls) from the ADMIRE-HF trial. SPECT images at 4 hours post injection were reconstructed by three methods filtered backprojection (FBP); ordered-subsets expectation maximization (OSEM); and OSEM with deconvolution of septal penetration (DSP). OSEM with DSP used camera-specific point spread functions for compensation for septal penetration. Two volumes of interest (VOI) were created in the transaxial images, one including the entire left ventricle, the other in the mediastinum above the level of the heart base, and a SPECT H/M was calculated by dividing the mean counts in the heart VOI by the mean counts in the mediastinum VOI. Receiver operating characteristic (ROC) analysis was used to assess the capability of each SPECT and the planar H/M (as determined by consensus of three blinded readers in ADMIRE-HF) for distinguishing the HF subjects and controls. Results: The ROC analysis yielded optimal cutoff values for FBP, OSEM, and DSP H/M of 2.9, 2.7, and 4.5 respectively. The sensitivity and specificity for distinguishing the HF subjects and controls using these cutoff values were 0.77 and 0.70 for FBP, 0.81 and 0.78 for OSEM, and 0.76 and 0.74 for DSP, respectively. By comparison, the optimal cutoff for the planar H/M was 1.58, yielding sensitivity and specificity of 0.78 and 0.79 respectively. The areas under ROC curves (AUC) were 0.82, 0.86, and 0.82 for FBP, OSEM, and DSP, respectively, which were not statistically different from the AUC of planar H/M (0.85). Conclusions: H/M obtained from I123 mIBG SPECT was equivalent to the planar H/M for distinguishing abnormal and normal cardiac uptake. The present results indicate that global quantification of cardiac I123 mIBG SPECT may represent a viable alternative to the planar H/M, which could have implications for users of small field-of-view dual-head cardiac cameras and new SPECT-only systems.
Epistemonikos ID: 4b7eccc47a36b81ce0e619d2bc04ea1b200fc73b
First added on: Feb 04, 2025