Abstracts

Rationale: Autoimmune epilepsy and encephalitis is an underrecognized and challenging diagnosis because of heterogenous clinical presentations and workup is often non-specific or unrevealing. It has been suggested that PET may be a more sensitive marker than brain MRI for limbic encephalitis (Pillai et al., 2010). A recent article (Solnes et al., 2017) reported 21 of 23 patients with parietal > occipital lobar hypometabolism in the acute setting of antibody-mediated autoimmune encephalitis of which half had normal MRIs. Another study found hyper and hypometabolism in the temporal regions in 6 out of 8 patients with VGKC/LGI-1 limbic encephalitis (Irani et al., 2011). Thus, the metabolic patterns can be variable and a disease-specific pattern has yet to emerge. We therefore desire to find a specific pattern in a cohort of patients with autoimmune epilepsy and encephalitis. Methods: We retrospectively reviewed 18 patients with a clinical diagnosis of autoimmune epilepsy and encephalitis. At least one positive autoimmune antibodies (VGKC, high GAD65, TPO with other antibodies, LGI-1, NMDA, or CASPR) was detected in serum or CSF in 15 subjects. A total of 20 PET studies were interpreted by the visual inspection of board-certified nuclear medicine physicians (GD, AI, JK). Three patients received treatment (steroids, IVIG, chemotherapy) prior to the PET study. Results: Of the 20 PET studies, 8 were dedicated brain FDG PET studies. The remainder were whole body FDG PET scans primarily for malignancy screening. The time of acquisition of the PET scans was variable: studies were completed both in the acute and chronic settings. Eight out of 20 (40%) scans demonstrated abnormal FDG uptake . Hypometabolic findings occurred in 7 (35%) while 2 (10%) had hypermetabolic findings. One patient had hypermetabolism followed by hypometabolism on serial FDG PET scans. Of the 8 positive scans, abnormalities were localized to the temporal (n=6), parietal (n=1), and cingulate, frontal, occipital, basal ganglia (n=1) regions. Dedicated brain FDG PET scans were abnormal in 6/8 (75%). Whole body FDG PET had abnormal brain uptake in 2/12 (17%). MRI brain scans were abnormal in 90% of the total cohort. There was 1 case of an abnormal PET in the setting of a normal MRI. Conclusions: Overall, 40% of brain and whole body FDG PET scans were abnormal in cases of autoimmune epilepsy and encephalitis. However, 75% of dedicated brain FDG PET scans were abnormal. Thus, whole body FDG PET scans should not be a substitute for brain FDG PET scans. In contrast to prior studies, the majority of abnormalities were localized to the temporal lobe and no specific autoimmune epilepsy and encephalitis pattern has been discovered. This study is limited because scans were done in the acute, chronic, ictal, and interictal settings and includes heterogenous cases. The PET findings were conducted through visual inspection alone. In the future, we may repeat this study with statistical software in comparison with FDG PET scans of normal brains. Funding: None