Abstract
721
Objectives: The vesicular monoamine transporter type 2 (VMAT2) in CNS monoaminergic synaptic vesicles is also expressed by pancreatic β cell mass (BCM). A VMAT2 ligand, [11C]DTBZ (DTBZ), has been proposed as a potential biomarker of BCM in rodent models of diabetes. We attempted PET quantification of BCM in patients with long-standing (> 20 y) type 1 diabetes (T1D), which results from immune-mediated BCM destruction.
Methods: Six T1Ds (35 ± 2 y) and 9 controls (30 ± 3 y) had dynamic abdominal DTBZ PET for 90 min (550 MBq bolus), and MRI. VMAT2 binding potential BPND was mapped voxel-wise by a reference tissue model (MRTM2) using renal cortex as reference tissue. 3D pancreatic regions of interest (ROIs) were identified on the BPND map by k-cluster analysis-based PET tissue classification algorithms and MRI. VMAT2 binding was compared between the two groups by pancreatic BPND in three ways: mean ROI values; voxel frequency distributions; and the functional binding capacity (FBC, defined as the sum of all voxels with BPND above a threshold times the voxel volume).
Results: In T1D, the ROI mean BPND was decreased by 14% (1.86 ± 0.05 vs. 2.14 ± 0.08; p = 0.01). The voxel BPND frequency distribution showed that high BPND signal was markedly decreased than was low signal compared with controls. FBC with BPND ≥ 2.5 was decreased by 70% in TID (17 ± 5 mL vs. 59 ± 7 mL; p<0.002,) and this FBC correlated with stimulated C-peptide secretion measurements (r =0.82, p < 0.002).
Conclusions: DTBZ PET may allow quantitative evaluation of pancreatic BCM noninvasively. The cause of the low BPND signal in TID is unclear, but may be due to underestimation of nonspecific binding in the pancreas.
Research Support: The PHS, NIH, NIDDK, 1 RO1 DK077493 and 5RO1 DK63567 (PEH), and funds provided by the Naomi Berrie Diabetes Center.
- Society of Nuclear Medicine, Inc.