Functional brain imaging using 18F-fluorodeoxyglucose positron emission tomography/computerized tomography in 138 patients with Kleine-Levin syndrome: an early marker?

Abstract

Kleine-Levin syndrome is a rare disorder characterized by relapsing-remitting episodes of severe hypersomnia, cognitive impairment, apathy, derealization and behavioural disturbances. Between episodes, most patients experience normal sleep, mood and behaviour, but they may have some residual abnormalities in brain functional imaging; however, the frequency, localization and significance of abnormal imaging are unknown, as brain functional imaging have been scarce and heterogenous [including scintigraphy 18F-fluorodeoxyglucose positron emission tomography/computerized tomography (FDG-PET/CT) and functional MRI during resting state and cognitive effort] and based on case reports or on group analysis in small groups. Using visual individual analysis of 18F-fluorodeoxyglucose positron emission tomography/computerized tomography at the time of Kleine-Levin syndrome diagnosis, we examined the frequency, localization and clinical determinants of hypo- and hypermetabolism in a cross-sectional study. Among 179 patients with Kleine-Levin syndrome who underwent 18F-fluorodeoxyglucose positron emission tomography/computerized tomography, the visual analysis was restricted to the 138 untreated patients studied during asymptomatic periods. As many as 70% of patients had hypometabolism, mostly affecting the posterior associative cortex and the hippocampus. Hypometabolism was associated with younger age, recent (<3 years) disease course and a higher number of episodes during the preceding year. The hypometabolism was more extensive (from the left temporo-occipital junction to the entire homolateral and then the bilateral posterior associative cortex) at the beginning of the disorder. In contrast, there was hypermetabolism in the prefrontal dorsolateral cortex in half of the patients (almost all having concomitant hypometabolism in the posterior areas), which was also associated with younger age and shorter disease course. The cognitive performances (including episodic memory) were similar in patients with versus without hippocampus hypometabolism. In conclusion, hypometabolism is frequently observed upon individual visual analysis of 18F-fluorodeoxyglucose positron emission tomography/computerized tomography during asymptomatic Kleine-Levin syndrome periods; it is mostly affecting the posterior associative cortex and the hippocampus and is mostly in young patients with recent-onset disease. Hypometabolism provides a trait marker during the first years of Kleine-Levin syndrome, which could help clinicians during the diagnosis process.

Keywords: FDG-PET/CT; Kleine–Levin syndrome; adolescence; cognition; functional imaging.

Graphical Abstract

Figure 1

Flow chart of patients with KLS and frequency of abnormal (with hypometabolism) FDG-PET/CT.

Figure 2

Associations between areas with FDG-PET/CT hypometabolism in patients with KLS.

Figure 3

Example of brain FDG-PET/CT images in three patients with KLS. The images are displayed in axial and coronal views in radiological convention [right (R) is on the left (L)] with the French colour scale. Patient A, 18 y.o., had two episodes with hypersomnia, derealization, apathy and mood disturbances before the FDG-PET/CT. The cortical and subcortical metabolism was normal. Patient B, 17 y.o., had six episodes with hypersomnia, derealization, apathy, behavioural disturbances, hallucinations and psychiatric symptoms before the FDG-PET/CT scan. There was mild left temporo-occipital (white arrow) and hippocampal hypometabolisms (white dotted arrow). Patient C, 36 y.o., had 21 episodes with hypersomnia, derealization, apathy, behavioural disturbances and psychiatric symptoms. There was a bilateral associative parieto-temporo-occipital (white arrow) and hippocampal hypometabolism (white dotted arrow).

Figure 4

Patients with FDG-PET/CT hypometabolism according to disease course. Percentage of patients with FDG-PET/CT hypometabolism in the posterior associative area (upper panel) and mesotemporal structure (lower panel) according to disease course. Patients with imprecise disease course are omitted.