gms | German Medical Science

Objective: While cognitive dysfunction is commonly observed in brain tumor patients, metacognitive abilities as such as self-referential processing have as yet only scarcely been investigated. It is however unclear, to what extent and depending on which anatomical structures the tumor lesioned brain is able to sustain an adequate sense of self. We investigated self-referential processing in glioma patients compared to healthy controls under consideration of tumor localization.

Methods: Self-referential processing was investigated preoperatively in 19 glioma patients (mean 44.2 yrs) and 19 matched healthy controls (mean 48.4 yrs), using a trait-judgement (TJ) and a trait recall (TR) task, as well as standardized neuropsychological testing. In the TJ task, subjects had to ascribe 50 adjectives as either self- or non-self-describing. The percentage of chosen positive and negative adjectives served as dependent variables, in order to assess the so-called Self-Serving-Effect (SSE), denominating a normally observed bias for positive traits. In the TR task, subjects had to recall as many of the presented adjectives as possible. The dependent variables to assess the Mnemic-Neglect-Effect (MNE, terming the memory advantage for positive traits) were the percentage of recalled positive and negative self- and non-self-describing traits.

Results: Tumor location varied, but with sparing of cortical midline structures in our cohort. Overall, patients performed worse compared to controls in all cognitive domains, especially with regard to verbal recall (p<.01). However, although patients showed longer reaction times (RT) in the TJ task, a significant effect for valence (ascribing positive traits faster than negative traits to oneself) was preserved in patients, as well as no between-group differences were found with regard to the SSE, i.e. all subject ascribed more positive than negative traits to themselves. Accordingly, no significant between-group difference was found with regard to the MNE, showing a memory advantage for positive as compared to negative self-ascribed traits, although patients recalled overall significantly fewer adjectives than controls, complying with the observed deficits in verbal recall.

Conclusion: Despite relevant cognitive impairment, glioma patients in our study showed preserved metacognitive abilities regarding self-referential processing. Our results seem to comply with current theories of self-referential processing converging neuroanatomically onto cortical midline structures.