Yeong Chul Yun1,2,3, Sabine Wolf2,3, Freya Garhöfer2,3, Katharina Holz2,3, Philipp Vollmuth2, Martin Bendszus2, Heinz-Peter Schlemmer1, Sabine Heiland2, Wolfgang Wick4, Varun Venkataramani4,5, and Felix T. Kurz1,2
1Radiology, German Cancer Research Center, Heidelberg, Germany, 2Neuroradiology, Heidelberg University Hospital, Heidelberg, Germany, 3Faculty of Medicine, Heidelberg University, Heidelberg, Germany, 4Neurology, Heidelberg University Hospital, Heidelberg, Germany, 5Functional Neuroanatomy, Heidelberg University, Heidelberg, Germany
Synopsis
Keywords: Tumors (Pre-Treatment), Brain
Motivation: Identifying brain regions where glioblastoma is associated with epilepsy or neurological deficits could help recognize patients with a higher risk of developing neurological symptoms based on MRI.
Goal(s): We aim to correlate clinical and radiological findings to improve diagnostic evaluation of pre-treatment glioblastoma patients.
Approach: MRIs from 557 patients with de-novo glioblastoma were analyzed retrospectively. We used NANO scale to report neurological deficits and analyzed lesion-frequency-maps for identifying deficits-associated regions.
Results: There was a significant correlation between the NANO scale and lesion volume. For each investigated domain with the NANO scale, radiologically correlated brain regions could be identified.
Impact: We showed that MRI examinations of
pre-treatment glioblastoma patients can provide clinicians and patients
valuable information regarding risk of developing certain neurological deficits
and symptoms. Furthermore, NANO scale and epilepsy-status can provide
information regarding the characteristics of the tumor lesion.
INTRODUCTION
Neurologic Assessment in Neuro-Oncology (NANO)
evaluates neurological deficits in nine domains based on simple clinical
examination1. We aim to correlate clinical and radiological findings
to improve diagnostic evaluation of pre-treatment glioblastoma patients. Identifying
brain regions where glioblastoma is associated with epilepsy, or any other neurological
deficits and characterizing
brain tumor characteristics could help recognize patients with developing a
higher risk of seizures or symptoms based on magnetic resonance imaging (MRI).METHODS
We evaluated 3 Tesla MRI with Gadolinium
contrast administration from 557 patients with pre-treatment IDH-wild-type glioblastoma retrospectively. Neurological
deficits were assessed according to the NANO scale. Furthermore, 145 patients
had a history of epilepsy before receiving the first-line treatment and 412
patients did not have any epileptic seizure before treatment.
Contrast-enhancing tumor (CE) was segmented semi-automatically from
postcontrast T1-weighted MRIs using HD-GLIO2,3. Lesion volumes from
patients without any NANO scale relevant deficits or epilepsy were compared
with symptomatic patients in at least one domain using Mann-Whitney U-tests.
Images were registered to MNI-152-space. Voxel-wise Fisher-exact-test followed
by random permutation analysis (ADIFFI) for multiple comparison correction4
was performed to identify regions with higher occurrence of tumors associated
with the deficits. ADIFFI was performed for epilepsy and each domain of NANO:
gait, strength, ataxia, sensation, visual fields, facial strength, language,
level of consciousness and behavior.RESULTS
There was a significant correlation between NANO scale
and CE lesion volume (Spearman r = 0.3098, p<0.0001, see figure 1). The
volume of CE associated without any deficits (n=106, 6.93 ml ± 0.63 ml) was
significantly smaller (p<0.0001) than lesions with deficits (n=373, 22.8 ml
± 0.9 ml), see figure 1. However, the CE lesion volume associated with epilepsy was
smaller than lesions without epilepsy (6.57 ml ± 0.53 ml vs. 18.04 ml ± 0.90
ml; p<0.0001, see figure 2). For each investigated domain with the NANO scale,
radiologically correlated brain regions could be identified with ADIFFI (example
NANO domain: language – see figure 4). Regions, where epilepsy-related lesions were
significantly more frequent, could not be found with ADIFFI. However, clusters
were found in subcortical regions and in white matter (WM) around the thalamus,
where occurrence of lesions without epilepsy was higher, suggesting a
preference for grey matter (GM) in epilepsy-related lesions (see figure 3).DISCUSSION
The heterogeneous distribution of
epilepsy-associated lesions and preference of GM is in accordance with previous
studies5. However, the volume of symptomatic lesions was smaller,
signifying that location and/or fraction of grey matter and white matter in tumor composition may
be more important than tumor size. Furthermore, we found a clear association
between NANO scale and glioblastoma lesion size and location, in line with other works
using symptom lesion mapping6.CONCLUSION
Glioblastoma lesion located in subcortical regions are less associated with epilepsy. In
cortical regions, the association of lesions with epilepsy was indeterminate
according to ADIFFI analysis. NANO scale may provide an estimate of glioblastoma lesion
volume. In addition, for certain domains of the NANO scale, neurological
deficits are well associated with glioblastoma location. These findings suggest further
investigations into the correlation between NANO and RANO criteria over time
and their usefulness in the clinical management of glioblastoma.Acknowledgements
No acknowledgement found.References
-
Nayak, L., et al., The Neurologic
Assessment in Neuro-Oncology (NANO) scale: a tool to assess neurologic function
for integration into the Response Assessment in Neuro-Oncology (RANO) criteria.
Neuro-Oncology, 2017. 19(5): p. 625-635.
- Isensee,
F., et al., nnU-Net: a self-configuring method for deep learning-based
biomedical image segmentation. Nature Methods, 2021. 18(2): p. 203-211.
- Kickingereder,
P., et al., Automated quantitative tumour response assessment of MRI in
neuro-oncology with artificial neural networks: a multicentre, retrospective
study. The Lancet Oncology, 2019. 20(5): p. 728-740.
- Ellingson,
B.M., et al., Anatomic localization of O6-methylguanine DNA methyltransferase
(MGMT) promoter methylated and unmethylated tumors: A radiographic study in 358
de novo human glioblastomas. NeuroImage, 2012. 59(2): p. 908-916.
- Schaper,
F.L.W.V.J., et al., Mapping Lesion-Related Epilepsy to a Human Brain Network.
JAMA Neurology, 2023. 80(9): p. 891-902.
6. Schiemanck, S.K., et al., Relationship
between Ischemic Lesion Volume and Functional Status in the 2nd Week after
Middle Cerebral Artery Stroke. Neurorehabilitation and Neural Repair, 2005.
19(2): p. 133-138
-
Schiemanck, S.K., et al., Relationship
between Ischemic Lesion Volume and Functional Status in the 2nd Week after
Middle Cerebral Artery Stroke. Neurorehabilitation and Neural Repair, 2005.
19(2): p. 133-138.