Euro Radio: Combining [18F]FDG-PET with MRI predicts the likelihood of postoperative seizure recurrence in patients with temporal lobe epilepsy

Temporal lobe epilepsy (TLE) is the most prevalent type of focal epilepsy, with recurrent episodes as the main clinical feature, and surgery is the current first-line option
.
Of TLE patients, about 30-40% are
MRI-negative.
Despite the development of multimodal techniques including magnetoencephalography (MEG), stereoencephalogram (SEEG) and molecular imaging, as well as multidisciplinary team (MDT), 15-20% of patients still have
severe seizure recurrences after surgery.
Recurrence of seizures after surgery is often related
to the extent of the epileptogenic zone (EZ), the extent of resection, and associated brain changes.

In MDT, 18F-fluorodeoxyglucose positron emission tomography ([18F]FDG-PET) is the most commonly used functional imaging modality and is often used
in combination with MRI in preoperative evaluation.
Regarding the quantitative analysis method of PET data, the threshold processing of clusters with normal standard deviation (z-score) is generally considered to be an important method
for locating epilepsy.
In recent studies, in addition to these comprehensive analyses of the area of seizure, metabolic patterns of [18F]FDG-PET have also been used to predict seizure outcomes after surgery, where hypometabolism
outside the surgical area is noted.
In our recent work, in addition to the PET model, the combination of PET and functional MRI also has a predictive effect
on seizure outcomes.
However, the structural basis for the predictive value of seizure outcomes as the basis for these functional images has not been adequately demonstrated
.

With the development of diffusion imaging, diffusion spectroscopy imaging (DSI) has been introduced into clinical applications
.
DSI has been reported to have quantitative properties
of the structure of white matter in the brain.
Unlike traditional diffusion methods that capture and decode the average direction within voxels, DSI can also acquire and analyze numerous diffusion directions
of fibers.
Surface-based analysis techniques can be used to quantify morphological indicators
.

Recently, a study published in the journal European Radiology explored the anatomical basis of white matter structural patterns, combined with the metabolism of [18F]FDG-PET, to further realize the possibility of predicting seizures after surgery before surgery, and provided technical support
for the formulation and evaluation of treatment plans.

This retrospective study enrolled 63 patients with
drug-resistant temporal lobe epilepsy.
Quantitative anisotropy (QA) based on Z-transform, cortical thickness, and diffusion spectral imaging consistent/inconsistent with cortectomy was used to quantify its predictive value
for postoperative seizures.

Of the 63 patients, 47 had a low metabolic zone of PET consistent
with the surgical area.
42 patients were seizure-free
after surgery.
The sensitivity and specificity of PET in predicting epilepsy degrees of freedom were 89.
4% and 68.
8%,
respectively.
Lesions with abnormally overlapping PET, cortical thickness, and QA were completely removed, and 27 patients obtained Engel I, a good predictor of epileptic degrees of freedom, with a probability (OR) of 19.
57 (95% CI 2.
38-161.
25, P = 0.
006).

。 Hypobolism involving multiple lobes (OR = 7.
18, 95% CI 1.
02 to 50.
75, p = 0.
048) and hypometabolic lesions with abnormal QA/cortical thickness outside the surgical area (OR = 14.
72, 95% CI 2.
13 to 101.
56, p = 0.
006) were the two main predictors
of Engel III/IV outcomes 。 The OR values for QA prediction of Engel I and seizure recurrence were 14.
64 (95% CI 2.
90-73.
80, p = 0.
001) and 12.
01 (95% CI 2.
91-49.
65, p = 0.
001),
respectively.

Figure Example of a patient with low PET metabolism outside the surgical area.

At the same time, outside the surgical area, low metabolism is accompanied by abnormalities
in QA or cortical thickness.
Case 1: Female, 24 years old, with intermittent seizures for more than 18 years
.
PET images show new-onset hypometabolism in the frontal, parietal, and occipital lobes of the two cerebral hemispheres outside the temporal surgical area
.
At the same time, on the cortical thickness map, there are also abnormal areas
in the occipital lobe.
The patient was rated Engel III
at one year postoperative follow-up.
Case 2: Female, 55 years old, paroxysmal systemic tics for 46 years
.
The patient has hypometabolism in the frontal, parietal, and occipital lobes of the right cerebral hemisphere outside the surgical area
.
Abnormalities in the temporal lobes and wedges on the cortical thickness map overlap
with hypometabolism outside the surgical area.
After the excision, the patient was rated Engel IV after 1 year

This study suggests that the consistency of [18F]FDG-PET, QA, and cortical thickness has predictive value
for seizures.
Abnormal QA or cortical thickness present in the hypometabolic zones of multiple lobes and outside the surgical area are factors in
predicting a poor prognosis of Engel type III/IV seizures.
These observations will help clinically screen patients who are less likely to benefit from resection surgery and reduce unnecessary surgical harm
.

Original source:

Zhen-Ming Wang,Peng-Hu Wei,Chunxiu Wang,et al.
Combined [ 18 F]FDG-PET with MRI structural patterns in predicting post-surgical seizure outcomes in temporal lobe epilepsy patients.
DOI:10.
1007/s00330-022-08912-2