Abstract :
[en] Language impairment such as receptive aphasia might lead to an underestimation of the level of consciousness in post-comatose patients (Schnakers et al., 2014). Yet, only a few studies focused on the behavioral assessment of residual language function, as based on the Coma Recovery Scale-Revised (CRS-R) and other tools, in line with their neural correlates. As previously recommended by Majerus et al. (2009), we here combined such behavioral and neuroimaging assessments to better explore language functions in patients with disorders of consciousness.
In the Experimental part I, we investigated the neural correlates of the clinical sub-categorization of minimally conscious state (MCS), based on the presence (i.e., MCS+) or absence (i.e., MCS-) of language-related signs of consciousness. As expected, language areas were shown to be more activated in MCS+ compared to MCS-, either using resting state FDG-PET or fMRI. These areas mainly encompassed the left-sided temporal lobule, angular gyrus, inferior/middle frontal cortex, caudate and temporo-occipital fusiform cortex. We also observed increased connectivity in the left frontoparietal network in MCS+ compared to MCS- patients, which was previously associated to language function. This result was particularly consistent across studies, either according to FDG-PET or fMRI measurements. Hence, MCS patients and healthy subjects could be placed along a continuum, from severe left frontoparietal network dysfunction, possibly associated to severely impaired language processing in MCS- patients, to preserved network connectivity in healthy subjects, with MCS+ patients being situated between these two groups. By contrast, increase of grey matter volume as assessed by voxel-based morphometry analyses was not significantly associated to the recovery of command-following, intelligible verbalization and/or intentional communication, unless at the individual-level. Finally, with regard to consciousness areas, a lower impairment of default mode network areas in MCS+ compared to MCS- was only revealed at the single subject-level. The MCS sub-categorization would thus reflect distinct levels of language impairments more than different levels of consciousness
The Experimental part II presents new bedside language assessments to complement the CRS-R and neuroimaging examinations. The Cognitive Assessment by Visual Election (CAVE) comprises tasks involving semantic and reading abilities (i.e., recognition of objects, pictures, letters, numbers, written words and colors), whereas the Brief Evaluation of Receptive Aphasia (BERA) is composed of three subscales examining different domains of language: phonology, semantics and morphosyntax. Good psychometric properties (reliability, validity and sensitivity) were shown using the BERA in aphasic conscious patients.The use of both tools in addition to the CRS-R in patients with disorders of consciousness allowed refining their behavioral language profiles, which were in line with FDG-PET and structural MRI results.
