18F-FDG PET/CT Brain Glucose Metabolism as a Marker of Different Types of Depression Comorbidity in Chronic Heart Failure Patients with Impaired Systolic Function
Vassiliki Lyra, John Parissis, Maria Kallergi, Emmanouil� Rizos, Gerasimos Filippatos, Dimitrios Th. Kremastinos, Sofia Chatziioannou
Abstract
Abstract Aims Depression is an important issue in heart failure (HF). The study investigated whole-brain and regional brain glucose metabolism in HF patients and its association with depression comorbidity. Methods and results Twenty-nine hospitalized patients with symptomatic systolic HF (left ventricular ejection fraction <40%), New York Heart Association (NYHA) class II–IV and mean age of 55.5 ± 12.0 years, had psychometric questionnaires before discharge and an 18F-FDG PET/CT brain scan after discharge. Semi-automated image analysis was performed on all cases and 30 matched controls. The metabolic parameter mean standardized uptake value (SUVmean) was calculated for the whole brain and three brain regions implicated in depression pathogenesis. A standardized SUVmean was also estimated by dividing regional brain SUVmean with whole-brain SUVmean. Cases had lower average whole-brain SUVmean (3.90 ± 1.49 vs. 5.10 ± 1.35, P = 0.001) and average regional brain SUVmean (4.57 ± 2.31 vs. 9.96 ± 3.58, P < 0.001) compared to controls. Whole-brain SUVmean had a significant correlation with patient age, NYHA class, diabetes, creatinine levels, depression, and cognitive impairment. Regional brain SUVmean was correlated with whole-brain SUVmean and depression. The standardized SUVmean, in particular, was found to be a robust index that could differentiate HF patients with ‘epiphenomenal’ (>0.93) or ‘real’ (≤0.93) depression. Conclusion Heart failure patients with more severe disease showed whole-brain and regional brain hypometabolism in 18F-FDG PET/CT. Depressed HF patients (Beck Depression Inventory score >13) exhibited different metabolic patterns that could be used to differentiate between ‘epiphenomenal’ and ‘real’ depression. Namely, presence of whole-brain hypometabolism suggested ‘epiphenomenal’ depression, whereas absence suggested ‘real’ depression. Presence of significant relative regional brain hypometabolism enhanced the likelihood of ‘real’ depression diagnosis.