Residency · Residency · Nuclear Medicine
Brain FDG PET in Dementia Evaluation
Introduction
Fluorodeoxyglucose (FDG) PET imaging of the brain is a powerful tool for evaluating neurodegenerative dementias. By mapping regional cerebral glucose metabolism, FDG PET reveals characteristic hypometabolic patterns that correspond to specific dementia subtypes.
Physiological Basis
Glucose Metabolism and Neuronal Activity
The brain consumes approximately 20 to 25% of total body glucose despite comprising only 2% of body weight. FDG is a glucose analog transported into neurons via GLUT-1 and GLUT-3 transporters. After phosphorylation by hexokinase to FDG-6-phosphate, it becomes metabolically trapped within the cell. Regional FDG uptake therefore reflects synaptic activity and neuronal metabolic demand. Neurodegeneration leads to synaptic loss and hypometabolism before significant structural atrophy develops, which is why FDG PET can detect changes earlier than structural imaging.
Patient Preparation and Imaging Protocol
Pre-Scan Preparation
The patient should fast for 4 to 6 hours prior to injection. Blood glucose must be checked and ideally should be below 150 to 200 mg/dL. The patient should rest in a quiet, dimly lit room for 20 to 30 minutes before and after injection to minimize non-specific cortical activation from sensory stimulation. Sedation should be avoided prior to injection because it alters metabolic patterns.
Acquisition Parameters
The dose is 5 to 10 mCi (185 to 370 MBq) of F-18 FDG administered intravenously. The uptake period is 30 to 45 minutes post-injection. Imaging consists of a 10 to 20 minute acquisition on a PET/CT scanner, with iterative reconstruction and attenuation correction.
Interpretation Approach
Normal FDG Distribution
The highest uptake is seen in gray matter, particularly the cortex, basal ganglia, thalami, and cerebellum. White matter demonstrates significantly lower metabolism. Uptake is expected to be symmetric between hemispheres. The visual cortex often shows the highest cortical metabolism.
Systematic Review Strategy
Cortical metabolism should be compared to the cerebellum and sensorimotor cortex, which are often preserved in neurodegenerative disease. The interpreter evaluates for asymmetry between hemispheres. Quantitative tools such as statistical parametric mapping (SPM) or 3D-SSP (Neurostat) compare the patient's metabolism to normative databases. Findings should always be correlated with clinical history, neuropsychological testing, and structural imaging.
Dementia-Specific Metabolic Patterns
Alzheimer Disease (AD)
Alzheimer disease produces bilateral temporoparietal and posterior cingulate/precuneus hypometabolism. Frontal involvement appears in advanced disease. The sensorimotor cortex, visual cortex, basal ganglia, thalami, and cerebellum are typically spared. The sensitivity for moderate to severe AD is 90 to 95%.
Frontotemporal Dementia (FTD)
Frontotemporal dementia shows frontal and anterior temporal hypometabolism. The behavioral variant demonstrates bilateral frontal predominance. The semantic variant shows left anterior temporal predominance. The non-fluent/agrammatic variant produces left frontal and insular hypometabolism.
Dementia with Lewy Bodies (DLB)
The pattern in Lewy body dementia is similar to AD with the addition of occipital hypometabolism. The "cingulate island sign" refers to relative preservation of the posterior cingulate compared to the precuneus and occipital cortex. Occipital involvement is the key feature distinguishing DLB from AD.
Corticobasal Degeneration (CBD)
Corticobasal degeneration produces asymmetric frontoparietal hypometabolism, contralateral to the clinically more affected side. Basal ganglia and thalamic asymmetry may also be present.
Vascular Dementia
Vascular dementia produces scattered, asymmetric areas of hypometabolism, often corresponding to vascular territories. Both cortical and subcortical structures may be involved.
| Dementia Subtype | FDG PET Hypometabolic Pattern | Preserved Regions | Key Distinguishing Feature |
|---|---|---|---|
| Alzheimer disease | Bilateral temporoparietal, posterior cingulate/precuneus | Sensorimotor, visual cortex, basal ganglia, cerebellum | Posterior cingulate involved early |
| Frontotemporal dementia (bvFTD) | Bilateral frontal and anterior temporal | Posterior parietal, occipital | Frontal predominance |
| FTD (semantic variant) | Left anterior temporal | Right hemisphere, posterior | Asymmetric left temporal |
| Dementia with Lewy bodies | Temporoparietal + occipital | Posterior cingulate ("island sign") | Occipital hypometabolism |
| Corticobasal degeneration | Asymmetric frontoparietal (contralateral to symptoms) | Opposite hemisphere relatively preserved | Marked asymmetry |
| Vascular dementia | Scattered, asymmetric, multi-territory | Variable | Corresponds to vascular territories |
Quantitative Analysis Tools
Several quantitative tools aid interpretation. 3D-SSP (Neurostat) generates surface projection maps compared to age-matched controls. Statistical Parametric Mapping (SPM) provides voxel-based statistical comparison. The PALZ score is a quantitative index for AD-pattern hypometabolism. SUVr (standardized uptake value ratio) is a semi-quantitative metric using the cerebellum or pons as a reference region.
Clinical Utility and Indications
FDG PET is particularly valuable for differentiating AD from FTD when the clinical presentation is ambiguous, evaluating atypical dementia presentations, and assessing patients with early-onset dementia (age under 65). CMS approves FDG PET for differentiating AD from FTD. FDG PET is complementary to amyloid PET, which assesses pathology rather than function.
Pitfalls and Limitations
Medications including sedatives, antiepileptic drugs, and corticosteroids can alter metabolic patterns. Psychiatric conditions such as depression and anxiety may produce frontal hypometabolism. Uncontrolled diabetes may reduce cortical FDG uptake globally. Normal aging produces mild frontal hypometabolism that should not be confused with pathology. Early disease may show subtle changes that require quantitative analysis for detection.
Clinical Pearls
The posterior cingulate and precuneus are among the earliest regions affected in Alzheimer disease and should always be carefully evaluated.
Occipital hypometabolism in a dementia patient strongly suggests Lewy body dementia rather than Alzheimer disease.
Always correlate FDG PET findings with the clinical phenotype, as overlap exists between dementia subtypes. Quantitative analysis tools significantly improve diagnostic confidence.
References
- Defined Indications and Reporting Standards for Brain FDG PET. Journal of Nuclear Medicine. 2022;63(5):e1-e15.
- Defined Clinical Indications for Neuroimaging with FDG PET/CT. European Journal of Nuclear Medicine and Molecular Imaging. 2021;48:1075-1096.
- Defined Patterns of FDG PET in Neurodegenerative Dementias. Seminars in Nuclear Medicine. 2020;50(4):281-293.
- Defined Role of Quantitative Analysis in Brain FDG PET Interpretation. Clinical Nuclear Medicine. 2021;46(9):e456-e463.