Abstracts

A Pediatric Fmri Memory Task and Tailored Temporal Surgeries Optimize Outcomes

Abstract number : 2.193
Submission category : 5. Neuro Imaging / 5B. Functional Imaging
Year : 2022
Submission ID : 2204586
Source : www.aesnet.org
Presentation date : 12/4/2022 12:00:00 PM
Published date : Nov 22, 2022, 05:25 AM

Authors :
Hannah Goldstein, MD – Seattle Children's Hospital; Dwight Barry, PhD – Seattle Children's Hospital; Jason Hauptman, MD, PhD – Seattle Children's Hospital; Ahmad Marashly, MD – Johns Hopkins University; Edward Novotny, MD – Seattle Children's Hospital; Jeffrey Ojemann, MD – Seattle Children's Hospital; Andrew Poliakov, PhD – Seattle Children's Hospital; Russell Saneto, DO, PhD – Seattle Children's Hospital; Dennis Shaw, MD – Seattle Children's Hospital; Hillary Shurtleff, PhD – Seattle Children's Hospital; Kieu Tran, MD – Seattle Children's Hospital; Molly Warner, PhD – Seattle Children's Hospital; Jason Wright, MD – Seattle Children's Hospital

Rationale: The goals of epilepsy surgery are seizure cessation, maximal preservation of function, and quality of life improvement. In pediatric temporal lobe (TL) cases, lack of fMRI tasks that effectively activate mesial temporal structures hampers preoperative memory risk assessment. We evaluated the effectiveness of memory fMRI tasks we developed by quantifying individual activation in a clinical pediatric setting, analyzing patterns of activation relative to the side of TL pathology, and comparing fMRI and Wada test results. We then evaluated pediatric TL surgery outcome optimization associated with tailored surgical resection informed by our fMRI memory task.

Methods: We retrospectively identified all pediatric patients with TL pathology (seizure focus and/or tumor) who had attempted our memory fMRI tasks during 2009-2018. We calculated the percentage of successfully scanned individual cases, compared activation masks in left vs. right cases, and compared fMRI with Wada tests when available. In a subset of cases with (1) TL resections for focal onset epilepsy; (2) viable fMRI memory scans; and (3) pre- and postoperative neuropsychological (NP) evaluations, we retrospectively evaluated presurgical fMRI memory scans, pre/postoperative NP scores, postoperative MRI scans, and postsurgical Engels. To assess fMRI memory task outcome prediction, we (1) overlayed preoperative fMRI activation onto postoperative structural images, (2) classified patients as having “overlap” or “no overlap” of activation and resection cavities, and (3) compared these findings with memory improvement, stability, or decline, based on Reliable Change Index calculations.

Results: A total of 72 patients, aged 7.6 to 20.9 years, attempted memory fMRI tasks. Of these, 60 (83%), successfully performed the memory fMRI tasks; 12 (17%) failed. Successful fMRI memory studies showed bilateral mesial temporal activation and quantitatively demonstrated: (1) left activation (L-ACT) less than right activation (R-ACT) in left TL cases, (2) nonsignificant R-ACT less than L-ACT in right TL cases, and (3) lower L-ACT plus R-ACT for left TL vs. right TL cases. Patients with viable fMRI and Wada tests had generally concordant results.Twenty of the 72 patients met inclusion criteria for outcomes assessment. Engel scores, at 2.1 (median) years post-surgery, included 16 1A and one each 1B, 1D, 2A, 2D. fMRI activation was linked to NP memory outcome in 19/20 (95%) of cases.

Conclusions: This study demonstrates evidence of an fMRI memory task paradigm that elicits reliable activation at the individual level and can be applied in a real-world clinical setting. Furthermore, fMRI memory task activation effectively predicted individual NP outcomes in the context of tailored TL resections. Patients had excellent seizure and overall good NP outcomes. Findings support individualized surgical intervention using fMRI memory activation.

Funding: Not applicable
Neuro Imaging