Effects of plasma glucose levels on regional cerebral 18F-fluorodeoxyglucose uptake: Implications for dementia evaluation with brain PET imaging.

PURPOSE: Hyperglycemia affects FDG uptake in the brain, potentially emulating Alzheimer's disease in normal individuals. This study investigates global and regional cerebral FDG uptake as a function of plasma glucose in a cohort of patients. METHODS: 120 consecutive male patients with FDG PET/CT for initial oncologic staging (July-Dec 2015) were reviewed. Patients with dementia, cerebrovascular accident, structural brain lesion, prior oncology treatment or high metabolic tumor burden (recently shown affecting brain FDG uptake) were excluded. 53 (24 nondiabetic) eligible patients (age 65.7 ± 2.8 mean ± SE) were analyzed with parametric computer software, MIMneuro™. Regional Z-scores were evaluated as a function of plasma glucose and age using multi variable linear mixed effects models with false discovery analysis adjusting for multiple comparisons. If the regression slope was significantly (p < 0.05) different than zero, hyperglycemia effect was present. RESULTS: There was a negative inverse relationship (p < 0.001) between global brain FDG uptake and hyperglycemia. No regional hyperglycemia effect on uptake were present when subjects were normalized using pons or cerebellum. However, regional hyperglycemia effects were seen (p < 0.047-0.001) when normalizing by the whole brain. No obvious pattern was seen in the regions affected. Age had a significant effect using whole brain normalization (p < 0.04-0.01). CONCLUSIONS: Cortical variation in FDG uptake were identified when subjects were hyperglycemic. However, these variations didn't fit a particular pattern of dementia and the severity of the affect is not likely to alter clinical interpretation.

Effects of Tumor Burden on Reference Tissue Standardized Uptake for PET Imaging: Modification of PERCIST Criteria.

Purpose To examine the effect metabolic burden (tumor and/or cardiac myocyte uptake) has on fluorine 18 fluorodeoxyglucose (FDG) distribution in organs and tissues of interest. Materials and Methods Positron emission tomographic (PET)/computed tomographic (CT) scans at the Ann Arbor Veterans Affairs hospital from January to July 2015 were reviewed. A total of 107 scans (50 patients; mean age, 64.3 years ± 13.2 [standard deviation]) had metabolic tissue burden assessed by using total lesion glycolysis (TLG) obtained from autosegmentation of the tumor and/or cardiac tissue. Standardized uptake value (SUV) and subsequent normalized SUV uptake in target organs and tissues were compared with 436 FDG PET/CT scans previously reported in 229 patients as a function of TLG to describe the effect(s) that metabolic burden has on reference tissue (blood pool, liver, and brain) FDG uptake. Subsequent regression by using linear mixed-effects models was used. If the slope of the regression was significantly (P < .05) different than zero, then an effect from TLG was present. Results There was a negative inverse relationship (P < .0001) between FDG uptake within reference tissues (blood pool, liver, and brain) and TLG in comparison to the study population at similar blood glucose levels. This TLG effect was no longer statistically significant (P > .05) when FDG uptake was normalized to a reference tissue (eg, blood pool or liver). Conclusion Metabolic tissue burden can have a significant effect on SUV measurements for PET imaging. This effect can be mitigated by normalizing FDG uptake to a reference tissue. © RSNA, 2018.

Effect of hyperglycemia on brain and liver 18F-FDG standardized uptake value (FDG SUV) measured by quantitative positron emission tomography (PET) imaging.

PURPOSE: Blood glucose is routinely measured prior to 18F-fluorodeoxyglucose (FDG) administration in positron emission tomography (PET) imaging to identify hyperglycemia that may affect image quality. In this study we explore the effects of blood glucose levels upon semi-quantitative standardized uptake value (SUV) measurements of target organs and tissues of interest and in particular address the relationship of blood glucose to FDG accumulation in the brain and liver. METHODS: 436 FDG PET/CT consecutive studies performed for oncology staging in 229 patients (226 male) at the Ann Arbor Veterans Administration Healthcare System were reviewed. All patients had blood glucose measured (112.4±34.1mg/dL) prior to injection of 466.2±51.8MBq (12.6±1.4mCi) of FDG. SUV measurements of brain, aortic arch blood-pool, liver, and spleen were obtained at 64.5±10.2min' post-injection. RESULTS: We found a negative inverse relationship of brain SUV with increasing plasma glucose, levels for both absolute and normalized (either to blood-pool or liver) values. Higher blood glucose levels had a mild effect upon liver and blood-pool SUV. By contrast, spleen SUV was independent of blood glucose, but demonstrated the greatest variability (deviation on linear regression). In contrast to other tissues, liver and spleen SUV normalized to blood-pool SUV were not dependent upon blood glucose levels. CONCLUSION: The effects of hyperglycemia upon FDG uptake in brain and liver, over a range of blood glucose values generally considered acceptable for clinical PET imaging, may have measurable effects on semi-quantitative image analysis.