Recent advance in the treatment of patients with disorders of consciousness: a review of transcranial direct current stimulation efficacy

• ¹ Coma Science Group, University of Liège, Belgium

Introduction: Severe brain injured patients with disorders of consciousness (DOC) represent a challenging population regarding their diagnosis and treatment. Patients in unresponsive wakefulness syndrome (UWS) can open their eyes but present only reflexive behaviours and thus, are clinically unconscious. On the other hand, patients in minimally conscious state (MCS) present fluctuant but consistent cognitively-mediated behaviours (for example through visual pursuit or response to simple commands), and thus, are considered as minimally conscious (1). However, patients with DOC are, by definition, unable to communicate and therefore, cannot actively participate to conventional rehabilitation programs. Regarding diagnosis, the Coma Recovery Scale-Revised (CRS-R) was developed by Giacino et al. in 2004 to overcome the issue of misdiagnosis by allowing differential diagnosis and is now considered to be the gold standard for the behavioural evaluation of these patients (2). Regarding treatment, there are still few options available. So far, research has focused on medication-based interventions (3, 4) and non-invasive brain stimulation techniques, such as transcranial direct-current stimulation (tDCS). TDCS is a non-invasive brain stimulation technique able to modify resting state potential of neurons, thereby increasing their spontaneous excitability (5). As it is safe, inexpensive and does not require active participation from the patient, it represents a promising tool to stimulate the recovery of DOC patients. Several studies have shown that tDCS can transiently improve the level of consciousness of DOC patients, as measured with the CRS-R. However, it is still unclear which proportion of patients can benefit from the technique, how long the effects can last and which brain region represent the best area to be targeted. In an attempt to answer some of these questions, we here reviewed tDCS-based clinical trials on DOC, taking into account published, submitted and accepted articles, in order to shed light on the current state of the art of tDCS interventions in DOC patients. Methods: We reviewed clinical trials on tDCS in patients with DOC, comparing effect sizes and numbers of responders. The studies were selected according to the following inclusion criteria: randomized, double-blind, sham-controlled trials, tDCS as a therapeutic intervention, behavioural evaluation with the CRS-R, population diagnosed with DOC, tDCS applied according to safety parameters, articles written in English. To calculate effect sizes, we used independent samples Cohen’s d. This index was determined by calculating the mean differences between sham and active tDCS effects (CRS-R total scores post- minus pre-stimulation) and then dividing the results by pooled standard deviations. Moreover, to compare the proportion of number of responders in single session studies and in repeated session studies we performed a two-sample proportion test. Results and discussion: Eight articles met our criteria (see Table 1): 3 articles evaluated the effects of a single stimulation session (1 published and 2 submitted) and 5 articles assessed the effects of repeated stimulation sessions (5 published). The areas targeted by the stimulation in the articles included the following regions: five studies targeted the left dorsolateral prefrontal cortex (left DLPFC – F3), one study targeted the posterior parietal cortex / precuneus (Pz), one targeted the primary motor cortex (M1 - C3 or C4) and one targeted the frontoparietal network bilaterally (F3, F4, CP5 and CP6). For what pertains single session studies, Thibaut et al. (6) showed that stimulating the left DLPFC could improve the level of consciousness in 43% of MCS patients, in this study of 2014, 15 out of 55 patients responded positively to tDCS and the observed effect size was 0.38. In one submitted single session study Martens et al. (7) found that stimulating the primary motor cortex didn’t result in a significant improvement in the CRS-R, and only one patient out of 10 responded to tDCS; in another study (8) the same group found that stimulating bilaterally the frontoparietal network, which is part of the External Awareness Network (9), did not improve significantly the behavioural outcome of DOC patients and only 6 out of 46 patients showed an improvement. Regarding repeated sessions studies, on the other hand, the area that was targeted the most frequently was the left DLPFC (5 out of 8 studies), this area being thought to be involved in many high cognitive processes such as attention, planning, decision making (10-11) and others. Four studies (12-15) found significant improvements after tDCS when targeting the left DLPFC with effect sizes varying from 0.30 (moderate) to 2.22 (high). Another study by Huang et al. (16), in which the posterior parietal cortex targeting the precuneus (Pz) was stimulated, also found positive improvements in the CRS-R scores with an effect size of 0.31. Left DLPFC was the most used target, both for single and repeated session studies and has showed to effectively increase patient’s responsiveness in the CRS-R compared to other stimulated regions. Moreover, in one study (Zhang et al. 2017), neurophysiological changes were assessed with EEG. The authors found that stimulating this region resulted in an increased P300 evoked-response potential amplitude. In another study (Estraneo et al. 2017) employing EEG as well, an amelioration of EEG organization and background activity was observed after active tDCS in patients who clinically responded to it. Repeated session studies generally show larger effect size of tDCS treatment and higher number of responders as compared to single session. Furthermore, using a statistical proportion test we found that the number of responders of repeated session studies (39 out of 110 in total) is significantly higher (p=0.0125) than the number of responders of single session studies (22 out of 111 in total). It seems that applying multiple tDCS sessions induces longer behavioural effects and increases the number of responders, this is also shown by effect sizes that appear to be smaller in single session studies with average effect size being 0.20 as opposed to 0.75 for repeated session studies. Conclusions: From this retrospective exploration of tDCS clinical trials, it emerged that the left DLPFC seems to be the most powerful and promising target to improve behavioural responsiveness of patients with DOC (see Figure 1). Still, many other alternatives could be tested, including multi-channels montages targeting specific brain networks. For instance, the cerebellum might be an important area to stimulate since it presents rich afferent and efferent connections (17). Nevertheless, to date, the left DLPFC can be considered as the most effective target whereas the strength and duration of tDCS aftereffects seems to increase with the number of sessions. As compared to pharmacological interventions, it is interesting to notice that the study evaluating the effects of 20 sessions of left DLPFC tDCS (15) showed a bigger effect size (i.e., 0.53) than the one observed in amantadine-based interventions (i.e., between 0.19 and 0.40) (18) However, individual responsiveness to tDCS is still a key issue in therapeutic use of non-invasive brain stimulation techniques. For instance, patients in MCS seem to benefit more from the treatment as compared to patients in UWS. It should also be noted that only a subgroup of MCS patients show these clinical improvements after tDCS, probably underpinning metabolic and atrophic individual differences; as shown in a retrospective study of Thibaut et al. 2015 where they found that tDCS-responders had higher residual brain metabolism and grey matter integrity as compared to non-responders with the same diagnosis (19). In conclusion, the next steps to improve tDCS efficacy in the treatment of DOC patients would be to identify the brain features of tDCS-responders and to develop patient-tailored MRI-guided stimulation montages and protocols, in order to optimize the current field based on the patient’s brain lesions. Table 1. Methods and results of the included studies: In the target column the anodal electrode is between parenthesis according to 10-20 EEG standard. Number of responders is given out of the whole sample with no information about MCS/UWS. Effect size is calculated with Cohen’s d index. Author and year of publication Target area Number of sessions Time since onset Sample size Number of responders Effect size Huang et al., 2017 Posterior Parietal (Pz) 5 5 months N=33 33 MCS 9/33 0.31 Zhang et al., 2017 Left DLPFC (F3) 10 5 months N=26 11 UWS 15 MCS 10/26 2.22 Estraneo et al., 2017 Left DLPFC (F3) 5 3 months N=13 7 UWS 6 MCS 5/13 0.30 Thibaut et al., 2017 Left DLPFC (F3) 5 3 months N=16 16 MCS 9/16 0.43 Martens et al., 2018 Left DLPFC (F3) 20 9 years N=22 22 MCS 6/22 0.53 Thibaut et al., 2014 Left DLPFC (F3) Single session 36 months N=55 25 UWS 30 MCS 15/55 0.38 Martens et al., (Submitted 2018) Primary motor (C3 or C4) Single session 9 months N=10 4 UWS 6 MCS 1/10 0.12 Martens et al., (Submitted 2018) Frontoparietal network (F3, F4, CP5, CP6) Single session 38 months N=46 17 UWS 23 MCS 6/46 0.12 Figure. 1 Visualization of tDCS targets and relative effect size indicated by symbols