Heterogeneity of PET FDG Myocardial Viability Images in Normal Subjects

Submitted May 2, 1996 to the American Heart Association's 69th Scientific Sessions

Authors:

Andrew B Barclay, Irvin H Naylor, Robert L Eisner, Randolph E Patterson

Abstract:

Despite attenuation correction, PET myocardial fluoro-18-deoxyglucose (FDG) images have been reported to show a lateral wall "hot spot" in many patients. Since this could distort FDG myocardial viability images (MVI), we investigated the normal range of FDG MVI spatial heterogeneity in 16 non-diabetic subjects with low clinical probability of disease who also had normal resting PET rubidium-82 (Rb-82) myocardial perfusion images (MPI). Following acquisition of Rb-82 MPI, subjects received a glucose load and 10 mCi of FDG. FDG MVI were acquired 50 minutes later. Polar maps were generated from short axis slices and segmented into 13 regions for quantitative comparisons. The Rb-82 MPI showed little heterogeneity and little deviation from our 50 patient Rb-82 normal file (p=NS in 12 of 13 regions). The FDG MVI showed much larger standard deviations (2-fold) than the Rb-82 MPI in all 13 regions. The most significant feature of the average FDG MVI was a 15% count increase ("hot spot") in the lateral wall relative to other regions. The range of count ratios was large, with 2 patients actually exhibiting a 7% relative decrease in the lateral wall. The table shows count ratios between different myocardial wall segments. In summary, the FDG MVI in normal subjects showed (1) a lateral wall "hot spot", (2) a large range of count ratios between wall segments, and (3) greater variability across subjects than Rb-82 MPI.

Table:

	    Septal/Lateral         Anterior/Inferior
	=====================    =====================
	 avg+-sd      range       avg+-sd      range
	==========  =========    ==========  =========
Rb-82	1.04+-0.05  0.93-1.13    0.97+-0.04  0.87-1.03
FDG	0.89+-0.15  0.68-1.22    1.04+-0.12  0.86-1.34