In Vivo Evidence of Glial Activation in Clinically Suspected Autoimmune Epilepsy Using TSPO PET Imaging
Abstract number :
3.327
Submission category :
5. Neuro Imaging / 5B. Functional Imaging
Year :
2025
Submission ID :
446
Source :
www.aesnet.org
Presentation date :
12/8/2025 12:00:00 AM
Published date :
Authors :
Presenting Author: Julia Panchenko, BS – NYU
Zanetta Kovbasyuk, MS – NYU
Ayelet Rosenberg, MSc – New York University Langone Health
Patryk Filipiak, PhD – NYU
Eden Tefera, BS – New York University Langone Health
Maria Pleshkevich, MS – NYU
Binita Shah, MD – NYU
Eytan Raz, MD – NYU
Kara Melmed, MD – NYU
Benjamin Brush, MD – NYU
Jakub Mroz, PharmD – NYU
Patrick Carberry, PhD – NYU
Ricksang Jachung, MHA, R.T. – NYU
Kaleigh Timmins, BS, RTN – NYU
Timothy Shepherd, MD – NYU
Seena Dehkharghani, MD – NYU
W. Gordon Frankle, MD – NYU
Steven H. Baete, PhD – NYU
Claude Steriade, MD – Department of Neurology, NYU Langone Medical Center, New York City
Rationale: About 5% of focal epilepsy of unknown cause harbor neural autoantibodies and may respond to immunotherapy (IT), but the diagnosis of autoimmune associated epilepsy (AAE) remains restricted to neural autoantibody testing, which is limited by its uncertain sensitivity for the detection of IT responsive epilepsy. PET radioligands that target TSPO have been used to quantify and visualize glial activation. In this study, we used the [11C]-ER176 TSPO ligand to measure TSPO uptake in the brain of people with clinically suspected but neural autoantibody negative AAE and compared them to GAD65 AAE, structural temporal lobe epilepsy (sTLE), and healthy controls (HC).
Methods: In vivo dynamic TSPO PET (15 mCi of 11C-ER176, 60-90 min acquisition) with concurrent MRI and arterial blood sampling was acquired from 7 HC, 5 sTLE patients, 4 suspected AAE, and 1 GAD65 antibody AAE. Suspected AAE had an antibodies contributing to epilepsy signs (ACES) score (≥3), no known cause, and were neural autoantibody negative. Arterial blood was sampled every 10s for the first 2min, every minute up until 5min, every 5min until 20min and then every 10min afterwards. Metabolite analysis was conducted on 4 arterial samples (5, 15, 30, 40min). Volume of distribution (VT) was determined using the metabolite-corrected arterial input function by fitting time activity curves (TAC) to an unconstrained two tissue compartment model (2TCM), using the extended Hill function with constrained parameters. VT values derived from suspected AAE and GAD65 AAE were compared to averaged VT of HC and sTLE patients across brain regions using z-scores.
Results: Patients with suspected AAE demonstrated on average greater VT values across 10 brain regions compared to both HCs and sTLE. The mean z-score across all brain regions was +1.75 ± 3.82 (vs HC) and +1.04 ± 2.58 (vs sTLE) (VT in specific brain ROIs shown in Fig. 1). The hippocampus (HIPP), amygdala (AMYG), and lateral temporal cortex (LTC) showed the most marked increases, with mean z-scores vs HCs of +5.73 ± 9.91, +4.54 ± 9.38, and +4.54 ± 8.47, respectively, and mean z-scores vs sTLE of +2.84 ± 5.46, +2.83 ± 6.34, and +3.19 ± 6.21 - these changes were largely driven by one individual with significantly elevated z scores, whose last seizure was 40 days prior to the scan. Two of the four clinically suspected AAE had z-score >1.5 in at least one region compared to HC (HIPP), and one had significantly elevated z-scores ( >2) compared to sTLE across multiple limbic regions (HIPP, AMYG, LTC, insula, thalamus). The patient with GAD65 AAE also demonstrated elevated TSPO uptake in limbic regions (HIPP +1.70, AMYG +1.76, and LTC +1.26 compared to HC).
Conclusions: A proportion of patients with clinically suspected but antibody negative AAE exhibit greater TSPO uptake compared to both HC and sTLE, particularly in limbic structures. This increased uptake mirrors and, in one case, exceeds that of a patient with GAD65 AAE. These findings suggest a pattern of elevated neuroinflammation among a subset of patients with suspected autoimmune epilepsy, even in the absence of detectable neural antibodies, who may benefit from IT.
Funding: This work was supported by NINDS (R01NS126156)
Neuro Imaging