Neurokognitive Auswirkungen von rechts- versus linkshemisphärischen T2/FLAIR-Läsionen des Tractus Arcuatus bei Patienten mit erstdiagnostiziertem Glioblastom

There has been increasing evidence that not only the distinct segments of the dominant (usually left-hemispheric) arcuate fasciculus (AF), but also its right-hemispheric counterparts are crucial to communication functions, including the recognition of facial expressions. However, the non-dominant AF is rarely included in invidual surgical plannings for brain tumour resections. In this study, we set out to investigate the correlation between cognitive functions and tumour infiltration of bihemispheric AF segments.

121 patients (79 males, 43 females, 116 right-handed, median age 62 years [36-83 years]) with a newly diagnosed unilateral glioblastoma underwent a preoperative cognitive assessment consisting of 11 subtests. Cognitive performance was expressed in percentile ranks (PR), normalised regarding age, gender and education. The PR and data-driven cluster analysis were used to build 7 different cluster scores, each describing a specific neurocognitive domain. Cranial FLAIR/T2-weighted MR images were acquired on a Phillips 3T scanner within 5 days. The respective tumour/oedema volumes were segmented using BrainLab Elements. A NatBrainLab atlas template was used to calculate the overlap of the lesion and the 3 segments of the AF in both hemispheres. Kendall-tau rank correlation coefficients between the cluster scores and the overlap were calculated in R and corrected for multiple comparisons.

Correlations were to be found between right-hemispheric lesions of the long AF segment and visuospatial judgment ( τ= -0,17; p= 0,046), of the posterior segment and visuospatial memory (τ= -0,18; p= 0,038) as well as visuospatial construction abilities (τ= -0,21, p= 0,013).
Lesions in all segments of the left AF correlated with working memory, albeit most pronounced in the long segment (τ= -0,26; p= 0,004). Similarly, lesions in all left AF segments correlated with lexical fluency and verbal memory, especially regarding the long AF segment (τ= -0,29; p= 0,001).

Our findings confirm the strong involvement of the left AF and, particularly, its long segment, in cognitive language functions such as verbal fluency and working memory, but also emphasise the importance of the right AF for visuospatial functions, which are for instance involved in emotional communication aspects. This stimulates the discussion whether appropriate mapping and monitoring tasks should be integrated in the awake surgery setting and calls for further analysis of the patients" follow-up outcomes.