Decreased VMAT2 in the pancreas of humans with type 2 diabetes mellitus measured in vivo by PET imaging

Gary W. Cline, Mika Naganawa, Laigao Chen, Kristin Chidsey, Santos Carvajal-Gonzalez, Sylvester Pawlak, Michelle Rossulek, Yanwei Zhang, Jason Bini, Timothy J. McCarthy, Richard E. Carson, Roberto A. Calle

The progressive loss of beta cell function is part of the natural history of type 2 diabetes. Autopsy studies suggest that this is, in part, due to loss of beta cell mass (BCM), but this has not been confirmed in vivo. Non-invasive methods to quantify BCM may contribute to a better understanding of type 2 diabetes pathophysiology and the development of therapeutic strategies. In humans, the localisation of vesicular monoamine transporter type 2 (VMAT2) in beta cells and pancreatic polypeptide cells, with minimal expression in other exocrine or endocrine pancreatic cells, has led to its development as a measure of BCM. We used the VMAT2 tracer [18F]fluoropropyl-(+)-dihydrotetrabenazine to quantify BCM in humans with impaired glucose tolerance (prediabetes) or type 2 diabetes, and in healthy obese volunteers (HOV). Dynamic positron emission tomography (PET) data were obtained for 4 h with metabolite-corrected arterial blood measurement in 16 HOV, five prediabetic and 17 type 2 diabetic participants. Eleven participants (six HOV and five with type 2 diabetes) underwent two abdominal PET/computed tomography (CT) scans for the assessment of test-retest variability. Standardised uptake value ratio (SUVR) was calculated in pancreatic subregions (head, body and tail), with the spleen as a reference region to determine non-specific tracer uptake at 3-4 h. The outcome measure SUVR minus 1 (SUVR-1) accounts for non-specific tracer uptake. Functional beta cell capacity was assessed by C-peptide release following standard (arginine stimulus test [AST]) and acute insulin response to the glucose-enhanced AST (AIRargMAX). Pearson correlation analysis was performed between the binding variables and the C-peptide AUC post-AST and post-AIRargMAX. Absolute test-retest variability (aTRV) was ≤15% for all regions. Variability and overlap of SUVR-1 was measured in all groups; HOV and participants with prediabetes and with type 2 diabetes. SUVR-1 showed significant positive correlations with AIRargMAX (all groups) in all pancreas subregions (whole pancreas p = 0.009 and pancreas head p = 0.009; body p = 0.019 and tail p = 0.023). SUVR-1 inversely correlated with HbA1c (all groups) in the whole pancreas (p = 0.033) and pancreas head (p = 0.008). SUVR-1 also inversely correlated with years since diagnosis of type 2 diabetes in the pancreas head (p = 0.049) and pancreas tail (p = 0.035). The observed correlations of VMAT2 density in the pancreas and pancreas regions with years since diagnosis of type 2 diabetes, glycaemic control and beta cell function suggest that loss of BCM contributes to deficient insulin secretion in humans with type 2 diabetes.