Journal de la maladie d'Alzheimer et du parkinsonisme

Abstrait

Occipital and Cingulate Hypometabolism are Significantly Under-Reported on 18-Fluorodeoxyglucose Positron Emission Tomography Scans of Patients with Lewy Body Dementia

Moath Hamed, Frank Schraml, Jeffrey Wilson, James Galvin and Marwan N Sabbagh

Objective: To determine whether occipital and cingulate hypometabolism is being under-reported or missed on 18-fluorodeoxyglucose positron emission tomography (FDG-PET) CT scans in patients with Dementia with Lewy Bodies (DLB).
Background: Recent studies have reported higher sensitivity and specificity for occipital and cingulate hypometabolism on FDG-PET of DLB patients.
Methods: This retrospective chart review looked at regions of interest (ROI’s) in FDG-PET CT scan reports in 35 consecutive patients with a clinical diagnosis of probable, possible, or definite DLB as defined by the latest DLB Consortium Report. ROI’s consisting of glucose hypometabolism in frontal, parietal, temporal, occipital, and cingulate areas were tabulated and charted separately by the authors from the reports. A blinded Nuclear medicine physician read the images independently and marked ROI’s separately. A Cohen’s Kappa coefficient statistic was calculated to determine agreement between the reports and the blinded reads.
Results: On the radiology reports, 25.71% and 17.14% of patients reported occipital and cingulate hypometabolism respectively. Independent reads demonstrated significant disagreement with the proportion of occipital and cingulate hypometabolism being reported on initial reads: 91.43% and 85.71% respectively. Cohen’s Kappa statistic determinations demonstrated significant agreement only with parietal hypometabolism (p<0.05).
Conclusion: Occipital and cingulate hypometabolism is under-reported and missed frequently on clinical interpretations of FDG-PET scans of patients with DLB, but the frequency of hypometabolism is even higher than previously reported. Further studies with more statistical power and receiver operating characteristic analyses are needed to delineate the sensitivity and specificity of these in vivo biomarkers.