Auditory localization should be considered as a sign of minimally conscious state based on multimodal findings

Abstract Auditory localization (i.e. turning the head and/or the eyes towards an auditory stimulus) is often part of the clinical evaluation of patients recovering from coma. The objective of this study is to determine whether auditory localization could be considered as a new sign of minimally conscious state, using a multimodal approach. The presence of auditory localization and the clinical outcome at 2 years of follow-up were evaluated in 186 patients with severe brain injury, including 64 with unresponsive wakefulness syndrome, 28 in minimally conscious state minus, 71 in minimally conscious state plus and 23 who emerged from the minimally conscious state. Brain metabolism, functional connectivity and graph theory measures were investigated by means of 18F-fluorodeoxyglucose positron emission tomography, functional MRI and high-density electroencephalography in two subgroups of unresponsive patients, with and without auditory localization. These two subgroups were also compared to a subgroup of patients in minimally conscious state minus. Auditory localization was observed in 13% of unresponsive patients, 46% of patients in minimally conscious state minus, 62% of patients in minimally conscious state plus and 78% of patients who emerged from the minimally conscious state. The probability to observe an auditory localization increased along with the level of consciousness, and the presence of auditory localization could predict the level of consciousness. Patients with auditory localization had higher survival rates (at 2-year follow-up) than those without localization. Differences in brain function were found between unresponsive patients with and without auditory localization. Higher connectivity in unresponsive patients with auditory localization was measured between the fronto-parietal network and secondary visual areas, and in the alpha band electroencephalography network. Moreover, patients in minimally conscious state minus significantly differed from unresponsive patients without auditory localization in terms of brain metabolism and alpha network centrality, whereas no difference was found with unresponsive patients who presented auditory localization. Our multimodal findings suggest differences in brain function between unresponsive patients with and without auditory localization, which support our hypothesis that auditory localization should be considered as a new sign of minimally conscious state. Unresponsive patients showing auditory localization should therefore no longer be considered unresponsive but minimally conscious. This would have crucial consequences on these patients’ lives as it would directly impact the therapeutic orientation or end-of-life decisions usually taken based on the diagnosis.


IV. Behavioral acquisitions using the Coma Recovery Scale-Revised.
For each CRS-R evaluation, experienced clinicians assessed all the items of the CRS-R, which ensured to evaluate the auditory localisation every time for each patient. The auditory localisation was assessed as indicated in the CRS-R guidelines: an auditory stimulus was presented for five seconds by the assessor standing behind and out of view of the patient, for a total of four trials, two on each side. Auditory localisation was considered present when an orientation of the head or the eyes towards the stimulus could be observed on both trials for at least one side. Importantly, the CRS-R was systematically performed on the day of the PET, MRI and EEG examinations. The own name was used as it has been shown to elicit more responses than neutral sound. (Cheng et al., 2013) V. 18 F-FDG-PET, MRI and EEG data acquisition parameters, preprocessing and analysis.
FDG-PET: The scan started 30 minutes after an intravenous injection of the tracer (approximately 150-300 MBq of FDG) and lasted 12 minutes. 18 FDG-PET images were reconstructed with standard 2x2x2 mm 3 voxels using iterative list mode time-offlight algorithm and corrections for attenuation, dead-time, random and scatter events were applied. Images for each subject were manually reoriented using Statistical Parametric Mapping (SPM 12, www.fil.ion.ucl.ac.uk/spm). They were then preprocessed including spatial normalization, smoothing (with an isotropic 14 mm FWHM Gaussian filter) and proportional scaling.
Some patients required light sedation during the scanning to prevent excessive movements but this does not affect the results.
For the normalization procedure we used a study-specific template created with DARTEL obtained from patients and HCS. (Ashburner, 2007) , (Di Perri et al., 2013) , (Peelle et al., 2012) This template was used to minimize normalization difficulty as it decreases the degree of warping necessary for patient brains in the normalization step and reduces the likelihood of misclassification and normalization errors that can occur during the voxel-based morphometry process. For BOLD noise reduction, we used the anatomical component-based noise correction method (Behzadi et al., 2007) as implemented in the CONN functional connectivity toolbox(Whitfield-Gabrieli and Nieto-Castanon, 2012). A temporal band-pass filter of 0,008-0,09 Hz was applied on the time series as classically performed in seed-correlation analysis (Greicius et al., 2003;Fox et al., 2005). Regarding motion correction, we used the artefact detection toolbox (ART; http://nitrc.org/projects/artifact_detect) as described elsewhere (Aubinet et al., 2018), using a composite motion measure. With this approach, a volume was defined as an outlier (artifact) if the largest voxel movement detected was above the specified thresholds. Specifically, an image was defined as an outlier (artifact) image if the head displacement in x, y or z direction was greater than 0.5 mm from the previous frame, or if the rotational displacement was greater than .02 rad from the previous frame, or if the global mean intensity in the image was greater than 3 SD from the mean image intensity for the entire resting scan. Outliers in the global mean signal intensity and motion were subsequently included as nuisance regressors (i.e., one regressor per outlier within the first-level general linear model). In doing so, the temporal structure of the data was not disrupted.
For some patients, a light sedation was required to reduce the severity of movement artefact during the fMRI data acquisition (6/8 LOCA patients and 9/25 NO LOCA patients).

VI. Seeds coordinates used for fMRI analyses, for the auditory, default mode and fronto-parietal networks.
Networks ( IX. fMRI results. Comparaison between LOCA and HCS (on the left) and between NO-LOCA and HCS (on the right) of the correlation between the auditory (first row), default mode (second row) and fronto-parietal (third row) networks and the time series from all other brain voxels. The blue spots show significantly decreased functional connectivity between patients and HCS. Statistical maps are thresholded at p<0.05 false discovery rate corrected at non-parametric cluster-mass with clusters made of voxels surviving a p<0.001 (whole-brain level).