Abstract
1454
Objectives: Neurodegenerative disorders are known to demonstrate typical lobar and regional patterns of parenchymal volume loss with corresponding decreased glucose metabolism. Currently, no data exists to compare semi-quantitative metabolic [F18] fluorodeoxyglucose (FDG) uptake on brain imaging in patients with suspected dementia using age-matched controls versus non-age matched control brain atlas. In this retrospective study, we aim to compare Z-scores provided by MIM Software in hybrid PET/MRI imaging of patients with cognitive impairment and suspected dementia utilizing age-matched versus non-age matched brain atlas as reference comparison. Materials and Methods: 33 patients (23 female and 10 male, average age 70) with cognitive impairment and suspected underlying neurodegenerative disorder underwent [F18]-FDG PET/MRI brain imaging. Patients were categorized by dementia subtype as follows: Alzheimer’s disease (10), Frontotemporal dementia (13), Lewy Body Dementia (7), and Corticobasal degeneration (3). MIM software was utilized to provide semi-quantitative Z-score values of abnormal decreased FDG uptake as compared to normal controls. The Z score subset was obtained both in comparison with age-matched controls (minimum of 5 controls +/- 5 years of age) and a non-age matched control brain atlas which included a total of 43 individuals (19 female, 24 male, mean age 63.8 +/- 10 years).
Results: Ten patients with suspected AD (9 female, 1 male, average age 67) had mean parietal lobe Z-score values of -2.65 and -2.67, when compared to age-matched (AMC) and non-age matched controls (NAMC) respectively (p = 0.98); temporal lobe Z scores when compared to AMC and NAMC were -0.87 and -1.68, respectively (p = 0.40). Thirteen patients with suspected FTD (7 female, 6 male, average age 73) had mean Z-score values in the frontal lobes of -0.96 and -1.05 when compared to AMC and NAMC, respectively (p = .895); temporal lobe Z scores were -0.39 and -1.23 when compared to AMC and NAMC (p = 0.112). Seven patients with suspected LBD (4 female, 3 male, average age 67) had mean parietal, occipital and temporal Z scores of -1.67, -2.15, and -0.77 for AMC and -1.93, -2.2, and -1.02 for NAMC, respectively [p-values for AMC versus NAMC in the parietal, occipital, and temporal lobe were 0.724, 0.948, and 0.708, respectively]. Three patients with suspected CBD (all female, average age 72) had a superior parietal lobule Z score of -0.63 and -0.71 for AMC and NAMC, respectively (p = 0.937).
Conclusions: Our semi-quantitative approach to recognizing lobar-specific patterns of brain hypometabolism in patients with cognitive impairment demonstrates no statistically significant difference when comparing these patients to age-matched versus non-age matched controls. Without the need for age-matched controls in a clinical setting, such a semi-quantitative approach can be more easily applied in the assessment of patients with cognitive impairment and underlying neurodegenerative disease.