Frontiers in Physics (Oct 2023)

Dynamic 18F-FDG-PET kinetic parameters for epileptogenic zone localization in drug-resistant epilepsy

  • Kitiwat Khamwan,
  • Kitiwat Khamwan,
  • Chanan Sukprakun,
  • Chusak Limotai,
  • Chusak Limotai,
  • Suda Jirasakuldej,
  • Suda Jirasakuldej,
  • Attapon Jantarato,
  • Thiravat Hemachudha,
  • Supatporn Tepmongkol,
  • Supatporn Tepmongkol,
  • Supatporn Tepmongkol,
  • Supatporn Tepmongkol

DOI
https://doi.org/10.3389/fphy.2023.1233059
Journal volume & issue
Vol. 11

Abstract

Read online

Objective: Precisely localizing the seizure onset zone remains a challenging task in drug-resistant epilepsy (DRE) patients especially given its critical role in successful surgery and effective management. This study aimed to investigate the kinetic parameters of regional 18F-fluorodeoxyglucose (FDG) uptake in DRE patients, aiming to identify the kinetic parameters best enabling the identification of the epileptogenic region.Methods: Consecutive DRE patients with clinically mandated interictal 18F-FDG PET/CT were recruited from October 2019 to September 2020 for pre-surgical evaluation. Immediately after injecting 18F-FDG of 112–179 MBq, dynamic data were acquired for 90 min. The motion correction and resampling to the Montreal atlas was performed in order to generate a transformation matrix. 116 volume of interests (VOIs) and regional time-activity curves (TACs) were generated by employing the automated anatomical labeling (AAL) template using PMOD software. Kinetic parameters of FDG unidirectional blood-brain clearance (K1), efflux (k2), phosphorylation (k3), and net metabolic flux (Ki) were derived using irreversible 2-tissue-compartment model with an image-derived input function (IDIF). The kinetic parameters values obtained from all regions were ranked and compared with the presumed epileptogenic zone (EZ).Results: Eleven DRE patients (5 males, 6 females, mean age 35.1 ± 10.2 years) were analyzed. We found that the region with the lowest values of Ki provided correct lateralization in 7/7 (100%) of patient with temporal lobe epilepsy (TLE) and the region with the lowest Ki and k3 parameters showed concordance with the EZ in 100% and 71.4% of patients, respectively.Conclusion: The present parametric approach to the evaluation of FDG-PET may be more sensitive than semi-quantitative approaches for the detection of pathophysiology in the EZ of patients with medically unresponsive TLE in addition to the routine clinical investigations.

Keywords