Neuronale Langzeitveränderungen im Sprachnetzwerk bei Patienten mit Tumorrezidiven sprachkritischen Regionen

Glial brain tumors in language areas may trigger functional reorganization in language networks on a local and global level. The existence, the nature, and underlying processes of those neuroplastic processes in the language network are still under debate. The aim of this fMRI study was to investigate if there are functional global and local long-term changes due to recurrent brain tumors in language critical areas.

23 patients with glial brain tumors in the language critical areas in the language dominant hemisphere were included in this analysis (mean age 42,83y; 11 females; tumor location: 9 frontal, parietal 3, temporal 11). Patients underwent presurgical fMRI sessions at initial tumor appearance and at recurrent tumor manifestation (mean time between fMRI scans was 919,57 days). During fMRI patients performed covertly a verb, an antonym, and a sentence generation task. fMRI data analysis was done by using SPM12. To assess possible long-term changes in the language system, Dice coefficients of T-maps were calculated. Further cluster sizes in the left and right frontal, temporal, parietal and occipital cortex were compared across the two time points. At least % signal-changes in the left pars opercularis and triangularis of the inferior frontal gyrus, the angular gyrus, the supramarginal gyrus, superior and middle posterior temporal gyrus were calculated and compared across the two time points.

Dice indices of language activation T-maps were .56 for the verb task, .56 for the anonym task and .63 for the syntactic task suggesting medium accordance between activation pattern of the 2 fMRI sessions. Cluster size across the 2 fMRI sessions showed significantly increased cluster sizes during the fMRI session at onset of disease compared to the session of tumor recurrence in the left (p=.022) and trendwise in the right frontal lobe (p=.061) but only for the verb generation task. Finally, % signal changes were significantly higher in all language associated ROIs in the initial fMRI-session compared to the session of tumor recurrence in all language paradigms (p<.05).

Data of this study suggest medium long-term local effects of brain tumors in the language network in the bilateral frontal lobe language by showing increased cluster size before further tumor treatment. On a global level, fMRI-signal intensity is reduced in the whole language network which could possibly be related to continuous tumor growth or tumor treatment.