European Review for Medical and Pharmacological Sciences 2017; 21: 4896-4906 Assessment of and response in patients with disorders of consciousness

X.-Y. WANG1, H.-Y. WU2, H.-T. LU1, T.-T. HUANG3, H. ZHANG1, T. ZHANG1

1Rehabilitation Medicine College of Capital Medical University, Chinese Rehabilitation Research Center, Beijing Boai Hospital, Beijing, China 2Key Laboratory of Behavioral Science, Institute of , Chinese Academy of Sciences, Beijng, China 3Central Hospital of Taian City, Taian City, Shandong Province, China

Abstract. – OBJECTIVE: To evaluate changes MMN: Mismatch Negativity; CRS-R: Coma Recovery in mismatch negativity (MMN) and P300 response Scale-Revised; BAEP: Brainstem Auditory Evoked Po- in vegetative state (VS) and minimally conscious tentials; N/A, not applied. state (MCS) patients before and after treatment, and their value for prediction of prognosis. PATIENTS AND METHODS: Event-related po- tentials (ERPs), performed on 11 patients classi- Introduction fied as VS (n = 6) or MCS (n = 5), and five healthy participants (i.e., control group). We performed Disorders of consciousness are common symp- a six months telephone follow-up to monitor toms after injury and among them vege- changes in consciousness recovery. tative state (VS) and minimally conscious state RESULTS: Comparison of the three groups showed significantly higher MMN latency elic- (MCS) are frequently observed. Patients with ited by salient stimuli and P300 elicited by the MCS have severely impaired consciousness con- subject’s own name for the VS group, as well as tent and significantly decreased consciousness significant difference in amplitudes of MMN elic- clarity, but still show minimal, yet specific, pres- ited by frequent stimuli and P300 elicited by oth- ervation of self or environmental awareness, as er first names for this group. The source of MMN well as voluntary eye-opening and sleep-wake and P300 responses was the frontal lobe for the cycles. Although conscious behavioral activities control group, and temporal lobe for the VS and are not continuous, these are reproducible or can MCS groups. CONCLUSIONS: be maintained long enough to be differentiat- The sudden increase in MMN 1,2 amplitude and latency shortening may indicate ed from primitive reflexive behaviors . MCS is an improvement in the state of consciousness. significantly different from VS; although the pa- Neurophysiological evaluations suggest that pa- tient’s consciousness level is far from being stable tients with vegetative state (VS) and minimally or allow the capacity for mutual communication, conscious state (MCS) may preserve patterns there is robust evidence for behaviors showing of higher-order cerebral processing similar to self or environmental awareness. The next stage those observed in conscious patients. of recovery for MCS patients is to be awakened. Key Words: However, VS or MCS can last a very long time, Consciousness disorder, Coma Recovery Scale Re- even until the patient’s death3. Currently, eval- vised, Event-related potentials, Minimally conscious uation of consciousness disorders mainly relies state, Vegetative state. on clinical manifestations of patients, as well as in various assessment scales. The Coma Recov- Abbreviations ery Scale-Revised (CRS-R) is an internationally recognized VS assessment scale4, which can be VS: vegetative state; MCS: minimally conscious state; used to evaluate multiple aspects of patients in a UWS: unresponsive wakefulness syndrome; ERPs: vegetative state, including auditory, visual, per- event-related potentials; DOC: disorder of consciousness; ception, movement, communication, as well as

4896 Corresponding Author: Hao Zhang, MD and Tong Zhang, MD; e-mail: [email protected] MMN and P300 response in disorders of consciousness to reflect symptom changes in patients in a more found that increased MMN amplitude indicated sensitive manner. Therefore, this is the main improved consciousness, while ERP assessments method used to diagnose, track, and evaluate showed that patients with a high amplitude and MCS in clinical research. However, such clinical short latency of MMN usually recovered to a scales are highly subjective, and have difficulties relatively higher level of consciousness. Cavinato in reflecting subtle changes in consciousness. On et al14 compared the latency and amplitude of N1 the other hand, nerve electrophysiology assess- and P3 waves in different groups of patients, in- ments possess certain advantages for the eval- cluding 6 patients with PVS and 11 with MCS as uation of patients with consciousness disorders well as 10 healthy volunteers, by exposing them because of their strong objectivity and high tem- to sinusoidal sound waves, their own names, and poral resolution. In recent years, many research- sinusoidal sound waves with the names of others. ers have used event-related potential (ERP) to Results showed that P3 latency slowly became provide prognosis for patients with consciousness longer in all the healthy volunteers and patients disorders5. Evidence-based medical analysis of with MCS as the complexity of auditory stimuli auditory ERP components (i.e., , mismatch increased, whereas there was no change in pa- negativity and P300) has shown that mismatch tients with PVS. This indicated that prolongation negativity (MMN) and P300 are relatively better of P3 latency in response to complexity of sound predictors for patients with low responsiveness stimuli may represent activity in high-level pro- in comatose or other conditions6. Earlier studies7 cessing and integration centers, which predicted had focused on the use of electrophysiology to the transition from VS to MCS in non-conscious predict consciousness recovery in VS patients af- patients without corresponding changes in their ter traumatic brain injury. Fischer et al8 found that clinical symptoms. Cavinato et al7 conducted fol- MMN, which is a component of ERP, can be re- low-up visits, measurements with the disability corded in patients with consciousness disorders. rating scale, as well as electroencephalograms A single salient auditory stimulus interspersed (EEG), brainstem auditory evoked potentials among a series of frequent regular auditory stim- (BAEP), somatosensory evoked potentials, and uli triggers MMN in the , and it P3assessments of 34 patients with VS within 2-3 can predict the consciousness recovery in coma months after injury. They found that 26 patients patients. Schnakers et al9 showed that they were (76%) recovered consciousness, and P3 was the able to record P300 in 6 out of 11 VS patients, only parameter contributing significantly to pre- which is consistent with previous studies showing diction of conscious recovery. Fischer et al8 found that P300 can be elicited in 38% of VS patients. that MMN can be recorded in some patients with At the same time, it was observed that in MCS PVS or MCS, suggesting that these patients still patients, P300 latency generated by the patient’s retained some conscious activity, with the possi- own names was longer than in VS patients. It is bility of further recovery. However, these authors believed that such replicable ERP wave forms, did not conduct follow-up visits in their study, which are related to recognition and processing, thus there was no indication of outcomes for these show that patients with traumatic brain injury patients. Lew et al13 conducted an auditory ERP can execute a certain type of cognitive functions, study in 22 patients with consciousness disorders and that the longer latency of P300 can be used after traumatic brain injury, and found that the to distinguish VS from MCS. Previous studies occurrence of P300 in these patients was linked showed that the occurrence of MMN and P300 to a better outcome. Fischer et al15 conducted an can help determine the prognosis of patients auditory ERP study in 50 coma patients (includ- with consciousness disorders8, 10-13 since dynamics ing patients with brain injury, stroke, and hypox- changes in MMN seem to reflect consciousness ic-ischemic encephalopathy), and found that the recovery in patients with consciousness disor- occurrence of P300, but not MMN, was relevant ders. As patients recovered a certain level of for the determination of patient awakening. Tak- consciousness, their capability to distinguish au- en together, the results of the analysis of the two ditory stimuli slowly improved, and the abrupt above-mentioned component waves show that increase in MMN amplitude showed that patients the occurrence of MMN and P300 is valuable to were able to communicate with their environ- predict patient awakening. ment8,10-13. Wijnen et al12 used P3 and MMN to Because of the large variations in electro- predict functional recovery in 10 patients who physiology results generated using various para- were in permanent vegetative state (PVS), and digms, an increasing number of researchers have

4897 X.-Y. Wang, H.-Y. Wu, H.-T. Lu, T.-T. Huang, H. Zhang, T. Zhang begun to think about the concept of “conscious- lung, liver, kidney, or other important organs; ness”. Consciousness may not be a single con- (6) Patients whose families refused to sign the cept; instead, it contains multiple functions that informed consent forms. are correlated with each other16. We designed this study based on this idea; therefore, electro- Control Group physiological method was used to study patients Healthy volunteers between 18 and 65 years of with consciousness disorders. In addition, clin- age were selected. The group included four males ical follow-ups were conducted for 6 months, in and one female. Age, gender, and education back- order to investigate the difference between VS ground were matched to the patient groups. and MCS, as well as to assess the correlation between electrophysiological and clinical evalu- ation results, with the hope of further elucidat- Clinical Data Collection ing the remaining brain function in patients with consciousness disorders. General Data Medical history of all participants, including course of disease, was collected by physicians. Patients and Methods In addition, we collected relevant demographic and individual characteristics including gender, Patients age, education background, and left/right hand- All patients were hospitalized and diagnosed edness. with consciousness disorders at the Neural Re- habilitation Department in Beijing Boai Hospital Physical Examination of Neural System between March 2013 and April 2014. Patients The neural systems of all participants were selected for this study included six patients with assessed in detail by physicians to record positive VS (four males and two females; age 26-60 years physical characteristics. old) and five patients with MCS (four males and one female; age 18-54 years old). Brain Imaging Data and This study was approved by the appropri- Electrophysiological Examination ate Institutional Review Board, and was in All patients underwent computed tomogra- accordance with the Helsinki Declaration on phy head scans (GE Healthcare, Little Chalfont, research with human subjects (1975, 2000). Buckinghamshire, UK) at the start of the study Written informed consent was obtained for all in order to determine the injury site. In addition, participants or their guardians prior to the start brain stem auditory evoked potentials (BAEP) of the study. examination was conducted to exclude organic Patients had to meet the following criteria to be damage to the ear or cochlear nerve, and to en- included in the study: (1) A clear history of brain sure that participants were able to hear auditory injury; (2) Patients with severe consciousness stimuli. disorders were required to meet the diagnostic standards of VS or MCS; (3) Disease duration Treatment was longer than 3 months, and age was between All patients with consciousness disorders re- 18 and 65 years; (4) Patients with impaired audi- ceived comprehensive rehabilitative treatment tory nerves were excluded using auditory evoked including physical therapy for limb and body potentials; (5) No use of drugs (e.g. sedatives) that function, hyperbaric oxygen therapy, and acu- could affect the study and our observations for at puncture. least one week before the study. Patient presenting the following characteristics Clinical Evaluation were excluded from the study: (1) Patients with The CRS-R (Coma Recovery Scale-Revised) hypersomnia, locked-in syndrome, or mental dis- was used to evaluate patients with consciousness orders that should be distinguished from VS or disorders in six aspects, including communica- MCS; (2) Age less than 18 or more than 65 years; tion, arousal, as well as auditory, visual, motor, (3) Patients who were unable to complete the as- and verbal functions according to the previous sessments included in our study; (4) Patients who study17. The criteria used for the inclusion of had a large area of the skull damaged or missing; patients in the VS or MCS group are presented (5) Patients who had severe diseases of the heart, in Table I.

4898 MMN and P300 response in disorders of consciousness

Table I. Diagnostic standard for vegetative state and ville, TN, USA). E-prime software (Grand Island, minimally conscious state patients. NY, USA) was used to present stimuli, a Net- CRS-R Vegetative Minimally Station system (San Diego, CA, USA) was used function state conscious state to record evoked potentials, and NetStationv4.3 analysis software (San Diego, CA, USA) was Auditory ≤ 2 3-4 used to analyze ERP data. Visual ≤ 1 2-5 Motor ≤ 2 3-5 Verbal ≤ 2 2 Event-Related Potentials Stimulus Array Communication ≤ 0 1 Participants sat in an electrically shielded and Arousal N/A N/A quiet testing room, keeping their body relaxed. In order to complete a task, participants heard a series of continuous sounds interrupted by spe- The CRS-R comprises six functional catego- cific stimuli through double-channel earphones. ries including auditory (score 0-4), visual (score (1) Pure tone stimuli: frequent stimuli were 1000 0-5), motor (score 0-6), verbal (score 0-3), com- Hz, and salient stimuli were 1500 Hz. Both were munication (score 0-2), and arousal (score 0-3), presented in an with an 80:20% with a total score ranging from 0-23. The first five ratio. Sounds were presented 300 times in total; scales were used to distinguish vegetative state with duration of 75 ms, bandwidth of 400 Hz, (VS) vs. minimally conscious state (MCS) pa- and sound intensity of 90 dB. (2) Name stimuli: tients, while arousal was not used for diagnosis. subject’s own name (SON) and three other first Note that to classify a patient as VS all five-mea- names (OFNs) formed by characters with a sim- surement score criteria were met, while for MCS ilar word frequency to the patient’s own name the presence of any of the measurement criteria from the Chinese Han character word frequency was enough to consider the patient as MCS. Table were used. The OFNs could not be the same Since error rates for clinical diagnosis in pa- or similar to the names of the patient’s relatives. tients with severe consciousness disorders is as Four names were presented in an oddball para- high as 40%, two medical practitioners trained digm, with four arrays, 80 words per array (each for scale evaluation diagnoses the patients in this name was 25%), and a 1300-1400 ms interval study, and patients who had consistent diagnoses between each stimulus. All names were present- were ultimately recruited. Activities of patients ed using the same female voice with a moderate with severe consciousness disorders were affect- tone, and the voice intensity was 90 dB. ed by time; therefore, scale evaluation was con- ducted in patients at four different time points. Electrode Placement and Event-Related Patients were evaluated once every 15 days, and Potentials Recording ERP examination was conducted if there was any Patients wore an electrode cap with 128 chan- change in the evaluation results at any time. An nels (EGI). Each patient had tasks from two ERP reexamination was conducted if evaluation stimuli arrays. MMN was recorded from the pure results were unchanged for two months. tone stimuli array, and P300 was recorded from the name stimuli array.

Examination of Event-Related Potentials Data Processing ad Analysis Brain electrical signals were continuously Preparation Before Examination collected, with a sample collection rate of 1024 Participants were asked to wash their hair be- Hz, band-pass filter wave of 0-30 Hz, notch filter fore examination, and use of any hair oil or con- of 50 Hz, and scalp resistance maintained at less ditioner was avoided to prevent wave disappear- than 5 KΩ. When recording, bilateral mastoids ance or disturbance. Examination was conducted were used as reference electrodes, and during 1 or 2 hours after a meal to avoid the influence of data analysis their readings were set as the mean low blood sugar on test results. Researchers tried value for readings from all electrodes. A level to ensure that patients were quiet and relaxed. of 100 ms before stimuli presentation was set as baseline, eye movement was set at an amplitude Equipment and Software ≥ 60 μV, eye blinking was set at an amplitude We used an EEG system manufactured by ≥ 200 μV, and other electromyography arti- Electrical Geodesics Incorporated (EGI, Nash- facts were removed when then mean was taken.

4899 X.-Y. Wang, H.-Y. Wu, H.-T. Lu, T.-T. Huang, H. Zhang, T. Zhang

MMN was defined as the maximum negative wave 100-250 ms after stimuli presentation, VS VS VS MCS MCS MCS MCS and P300 was defined as the maximum positive MCS MCS Awake Awake

wave 300-600 ms after stimuli presentation. The interview Follow-up electrode location with the clearest waveform was set as the point of analysis: Fz point was Y Y Y Y Y N Y N Y selected for MMN, and Cz point was selected for Y Y P300

P300. The main parameters for analysis were the Elicited

amplitude and peak value latency period of the various waveforms. Y Y Y Y Y Y Y N Y Source localization analysis: The Standardized Y Y MMV Low-Resolution Electromagnetic Tomography Elicited (sLORETA) method was used for the source localization of evoked potentials on the scalp,

VS VS VS VS which is included in the EGI analysis software VS VS MCS MCS MCS MCS MCS package. VSor MCS

Statistical Analysis 7 9 11 11 13 9 6 7 6 6 Data were analyzed using SPSS v16.0 soft- 7 ware (SPSS Inc., Chicago, IL, USA). Quantita- CRS-R tive results were presented as mean ± standard errors, and comparison between groups was done using One-way ANOVA test followed by Least 8 11 10 12 12 7 6 5 4 3 Significant Difference (LSD). In addition, quali- 8

tative results were examined using the modified (months) small-sample χ2-test. Differences were consid- ered significant when p < 0.05. duration Disease

8 7 9 11 11 16 Results 12 (years) Education We summarize in Table II the medical history data collected for patients included in the study, 31 31 8 41 45 43 45 7 50 9 26 48 46 60 5 overall CRS-R scores, as well as the results of as- Age sessing ERPs and diagnosis determination based on follow-up interviews.

F F F M M M M M M M There were no significant differences t( -test, p M

> 0.05) in age of disease onset (41 ± 8.33 vs. 44 Gender ± 5.2), or years of education (9.33 ± 2.78 vs. 9.4 ± 1.68) between the vegetative state (VS) and min- imally conscious state (MCS) groups.

Comparison of Mismatched Negativity and P300 Waves in the Three Groups

MMN was elicited in five out of six patients of the VS group, while P300 was elicited in four out of six patients. Both MMN and P300 were Disease elicited in all five patients with MCS, as well as in all five healthy volunteers of the control Encephalorrhagia Hypoxic-ischemic encephalopathy Encephalorrhagia Cerebral infarction Traumatic brain injury brain Traumatic Encephalorrhagia Brainstem infarction Brainstem Traumatic brain injury brain Traumatic Hypoxic-ischemic encephalopathy group. The superimposed averaged waveforms Encephalorrhagia

of all groups are shown in Figure 1. The compar- Patients’ medical history, CRS-R scores, event-related potentials, and diagnosis after interview. follow-up ison of the amplitudes and latency periods for 8 MMN and P300 for the three groups is shown 2 Encephalorrhagia11 7 9 10 6 5 4 3

Patient 1 in Table III. II. Table

4900 MMN and P300 response in disorders of consciousness

Figure 1. Mismatch negativity and P300 response graphs. Panels on the left present mismatch negativity (MMN) response graphs for the control (upper panel), vegetative state (VS; middle panel), and minimally conscious state (MCS; lower panel) groups. Panels on the right present P300 response graphs for the control (upper panel), VS (middle panel), and MCS (lower panel) groups.

Source Localization Analysis sity of 0.039062 nA, at Brodmann area 20 (upper In the Control Group (Figure 2, upper panel), inferior temporal gyrus of the temporal pole). MMN elicited by salient stimuli was located at In the MCS Group (Figure 2, lower panel), -3, 52, -6, with an intensity of 0.010099 nA, at MMN elicited by salient stimuli was located at Brodmann area 10 (i.e., middle frontal gyrus of 46, 10, -34, with an intensity of 0.070311 nA, at the frontal pole). MMN elicited by frequent stim- Brodmann area 21 (i.e., middle superior temporal uli was located at -3, 52, -6, with an intensity of gyrus of the occipital pole). MMN elicited by 0.011403 nA, at Brodmann area 10. frequent stimuli was located at -3, -81, 1, with an In the VS Group (Figure 2, middle panel), intensity of 0.066694 nA, at Brodmann area 18 MMN elicited by salient stimuli was located at (i.e., lingual gyrus). -52, -67, 6, with an intensity of 0.024283 nA, at In the Control Group (Figure 3, upper panel), Brodmann area 19 (i.e., middle temporal gyrus P300 elicited by SON was located at -3, 52, -6, of the occipital pole). MMN elicited by frequent with an intensity of 0.037858 nA, at Brodmann stimuli was located at -59, -39, -20, with an inten- area 11(i.e., middle frontal gyrus of the frontal

4901 X.-Y. Wang, H.-Y. Wu, H.-T. Lu, T.-T. Huang, H. Zhang, T. Zhang

pole). P300 elicited by OFNs was located at -3, 52, -6, with an intensity of 0.019716nA, at Brod- mann area 10. ± 1.12 ± 1.56 ± 1.47 ± 1.12 # # # # (μv) In the VS Group (Figure 3, middle panel), 2.45 ± 1.78 2.45 Amplitude 1.57 1.49 1.82 1.68 P300 elicited by SON was located at -52, -67, 1, with an intensity of 0.093389nA, at Brodmann

area 37 (i.e., middle temporal gyrus). P300 elic- ited by OFNs was located at -52, -67, 1, with an

(ms) intensity of 0.071744nA, at Brodmann area 37.

Latency In the MCS Group (Figure 3, lower panel), ± 68.78 264.76 ± 95.76 235.94 ± 112.65 212.34 ± 43.25 258.24 244.00 ± 21.00 P300 elicited by SON was located at 46, -67, -6,

with an intensity of 0.110801nA, at Brodmann ar- ea 37 (i.e., lower superior temporal gyrus). P300

Other first names elicited by OFNs was located at 4, -81, 1, with an

(μv) intensity of 0.063377nA, at Brodmann area 18 ± 1.78 3.25 ± 2.15 3.65 ± 1.87 3.39 ± 1.28 3.26 ± 1.82 3.445 Amplitude (i.e., lingual gyrus).

Discussion * ± 72.68 * ± 67.83 #, #,

P300 (Cz) P300 Consciousness disorders include three differ- 18 ± 89.46 168.78 ± 76.85 176.65 ± 75.65 182.56 ent states: comatose, VS, and MCS . In order to 289.75 312.62 improve the diagnostic accuracy of VS and MCS, this study evaluated patients with consciousness Latency disorders using the CRS-R scale at four different

Indicates a significant difference compared to the control group ( p < 0.05). *Indicates group control the to compared difference significant a Indicates time points. Patients were included in the VS or # Subjects’ own name

(μv) (ms) MCS group depending on the results obtained ± 1.02 1.15 ± 1.67 1.72 ± 1.28 1.39 ± 0.77 1.34 ± 1.15 1.44 using the CRS-R scales. In addition, a control Amplitude group composed of healthy volunteers was also included. Pure tone stimuli were used to elicit

MMN, while name stimuli were used to elicit P300. MMN occurred in five out of six patients

(ms) with VS, P300 occurred in four patients, while Latency both MMN and P300 occurred in all patients ± 32.76 52.63 ± 24.32 61.29 56.24 ± 27.68 ± 28.93 48.86 66.88 ± 26.36 with MCS. In spite of this, there were no signif- icant differences between the two groups. Pre- vious studies have reported variable results on Frequent

p < 0.05). the occurrence rate of MMN and P300. Similar 19 * ± 1.26 * ± 1.05 to our study, Perrin et al used name stimuli #, #, ± 1.78 3.74 3.42 ± 2.36 3.42 ± 1.15 3.34 (1 SON, 7 OFNs) for their assessment of P300, 1.63 1.89 they found that P300 was elicited in three out of five patients with VS, and in all six patients with 8 Amplitude MCS. In 2010, Fischer et al used pure tone (with varied durations) and name (presented by differ- MISMATCH negativity (Fz) * ± 15.45 * ± 21.78 #, #,

(μv) (ms) ent voices) stimuli to investigate MMN and P300. ± 15.36 42.50 46.50 ± 7.36 48.00 ± 9.56 Latency They found that MMN and P300 were elicited in three out of 11 patients with MCS, and MMN was elicited in two; also, P300 was elicited in 66.13 62.25 Salient

2 1 2 1 N three out of 16 patients with VS. Possible reasons for such variability in experimental results in-

Amplitudes and latency period mismatched of negativity and P300 responses. clude different evaluation methods for conscious- ness disorders, different course of disease, varied

Group stimuli presentation methods, and varied ERP MCS Control

VS

Table III. Table state. VS: vegetative of times; N: number state; conscious MCS: minimally Abbreviations: a significant difference compared to the MCS group ( a significant difference data analysis methods.

4902 MMN and P300 response in disorders of consciousness

Comparison of Mismatched Negativity cantly lower than in the control group, indicating and P300 Waves in the Three Groups that there was a lower awareness for unfamiliar In the VS group, MMN amplitude elicited by names stemming from the remaining brain func- salient stimuli was significantly lower than in the tion in patients with VS and MCS. Comparison MCS and control groups, while MMN latency of the MCS and control groups showed that there was significantly longer than in the MCS and was a significant difference in P300 amplitude control groups. The abrupt increase in MMN am- elicited by OFNs stimuli, but no difference for plitude and shortened latency by salient stimuli other stimuli. The results above showed that the may suggest the recovery of the conscious state, amplitude and latency of ERP may be valuable to which is consistent with the study by Wijnen et predict the difference between VS and MCS, and al12. P300 latency elicited by SON stimuli in the to determine whether VS can transition to MCS. VS group was significantly longer than in the MCS group, suggesting that to a certain extent Source Localization Analysis VS patients were aware of their own names. On MMN responses were located in the frontal the other hand, P300 amplitude elicited by OFNs lobe (Brodmann area 10) in the control group, stimuli in the VS and MCS groups was signifi- while they were located in the temporal lobe

Figure 2. Mismatch negativity brain images. The upper, middle, and lower panels present brain images for the control, vegetative state (VS), and minimally conscious state (MCS) groups, respectively. For each group, sagittal, coronal and axial images are presented (left, mid, and right, respectively).

4903 X.-Y. Wang, H.-Y. Wu, H.-T. Lu, T.-T. Huang, H. Zhang, T. Zhang

Figure 3. P300 response brain images. The upper, middle, and lower panels present brain images for the control, vegetative state (VS), and minimally conscious state (MCS) groups, respectively. For each group, sagittal, coronal and axial images are presented (left, mid, and right, respectively). in the VS (Brodmann area 19 or 20) and MCS and inferior temporal gyri. Despite there was no groups (Brodmann area 18 or 21). P300 in the significant difference in the amplitude and laten- control group was mainly located in the frontal cy between MCS and control groups, the source lobe (Brodmann area 10 or 11), while in VS and location in the MCS group (i.e., the temporal MCS groups it was located in the temporal lobe lobe) was clearly different from the control group, (Brodmann area 18 or 37). The source for MCS which is likely the reason why MCS patients was mainly located in the superior temporal and have better clinical responses than VS patients. middle temporal gyri (i.e., closer to the high- Although they were still not fully awake, and er-level cortices of the frontal lobe), while the remained as MCS clinically. Therefore, although source for VS was mainly located in the middle VS and MCS had poorer responses to the envi-

4904 MMN and P300 response in disorders of consciousness ronment, they still presented various levels of nitive processes, which are mainly located in awareness responses, which were mostly located the frontal lobes. While MMN and P300 in VS at the level of the auditory cortex. are mainly located in the sensory cortex of the temporal lobes, which suggests that processing Patient’s Follow-up and Effect of of auditory stimuli in patients was primarily fo- Treatment on Mismatched Negativity cused in the auditory cortex, further processing and P300 Responses of the information was limited, which may be In this work, no significant differences were the primary reason why these patients failed to found between MCS and VS groups pre-and respond well to the external environment. The post-treatment, which was likely related to the results of the four ERP examinations performed fact that the selected patients were mostly in supported this view. MCS patients had better recovery, and the time of follow-up visit was rel- clinical responses than VS patients, and their atively short. source localization was closer to higher-level Furthermore, we interviewed all patients (or cortices in the frontal lobes. their families) over the phone after six months. Among the four patients in the VS group who pre- sented MMN and P300 responses, two were still Conclusions VS, one was awake, and the other transitioned to MCS. Noteworthy, the patient who presented We believe that there is limited value for MMN responses (but not P300) had transitioned MMN and P300 responses as predictors of the to MCS, while the patient who did not respond to recovery of conscious state, although the abrupt MMN or P300 was still VS. In the MCS group, changes of amplitude and latency may have where all five patients presented MMN and P300 some relevance for the change of conscious responses, two were awake at the time of fol- state. Source localization may be more persua- low-up interview, while the other three were still sive evidence to determine conscious state, and MCS. One of the patients awoke 1 month after may be valuable to distinguishing and diagnose leaving the hospital, and after reviewing the ERP VS and MCS. Our study had a relatively small results of this patient, we found that the MMN sample size, and the follow-up interview period before leaving the hospital also showed a trend was relatively short, while patients had a rel- with high amplitude and short latency, consis- atively long course of disease of at least three tent with the above-mentioned conclusion. The months, and their recovery was relatively slow. other patient awoke four months after leaving Therefore, clinical studies with larger sample the hospital, although no similar MMN could be sizes and long-term clinical follow-ups will be observed in this patient. This may be because a needed to verify this view, which will be our long time passed between ERP and the time the next research focus. Moreover, the recovery patient awoke; therefore, this result was not re- from consciousness disorders is related to ma- corded. This suggests that there may be a specific ny factors in the patients, including cause of time window for the change in ERP results. More disease, age, degree of pathological changes, continuous clinical monitoring will be needed to course of disease, which should be considered further understand similar cases. in future research. Based on the results of ERP amplitudes and latency of this study, the abrupt changes of am- plitude and latency in patients may be valuable Funding Sources to predict changes in consciousness state; how- This study was supported by the National Natural Science ever, there was no enhanced ERP amplitude or Foundation of China (grant number: 81171852). prolonged latency as the CRS score improved, suggesting that there is limited value for the amplitude and latency of MMN and P300 to Authors’ Contributions predict or distinguish between VS and MCS. ZH and ZT participated in the design of the study, WXY Variations in the source localization of MMN participated in the design of the study, carried out experi- may show that although MMN can be elicited ments, wrote the manuscript; WHY analyzed experimental results; LHT and HTT conceived of the study, and partici- when patients were not paying , MMN pated in its design and coordination and helped to draft the generated from healthy volunteers requires an manuscript. All authors read and approved the final man- attentive state, as well as other high-level cog- uscript.

4905 X.-Y. Wang, H.-Y. Wu, H.-T. Lu, T.-T. Huang, H. Zhang, T. Zhang

Acknowledgements Moonen G, Laureys S. Voluntary brain processing The authors thank all the patients and patients’ families. in disorders of consciousness. Neurology 2008; 71: 1614-1620. 10) Qiao J, Qian LJ, Zhao HF, Gong GH, Geng DQ. Conflict of Interest The relationship between quality of life and clin- No conflict of interest exits in the submission of this man- ical phenotype in patients with treatment resis- uscript, and manuscript is approved by all authors for pub- tant and non-treatment resistant depression. lication. Eur Rev Med Pharmacol Sci 2017; 21: 2432- 2436. 11) Qin P, Di H, Yan X, Yu S, Yu D, Laureys S, Weng X. Mismatch negativity to the patient’s own name References in chronic disorders of consciousness. Neurosci Lett 2008; 448: 24-28. Del CA, Baillet S, Dehaene S 1) . Brain dynamics un- 12) Wijnen VJ, van Boxtel GJ, Eilander HJ, de Gelder B. derlying the nonlinear threshold for access to Mismatch negativity predicts recovery from the consciousness. PLoS Biol 2007; 5: e260. vegetative state. Clin Neurophysiol 2007; 118: 2) Giacino JT, Ashwal S, Childs N, Cranford R, Jennett 597-605. B, Katz DI, Kelly JP, Rosenberg JH, Whyte J, Zafon- 13) Lew HL, Dikmen S, Slimp J, Temkin N, Lee EH, New- te RD, Zasler ND . The minimally conscious state: ell D, Robinson LR. Use of somatosensory-evoked definition and diagnostic criteria. Neurology 2002; potentials and cognitive event-related potentials 58: 349-353. in predicting outcomes of patients with severe 3) Bruno MA, Vanhaudenhuyse A, Thibaut A, Moonen traumatic brain injury. Am J Phys Med Rehabil G, Laureys S. From unresponsive wakefulness to 2003; 82: 53-61, 62-64, 80. minimally conscious PLUS and functional locked- 14) Cavinato M, Volpato C, Silvoni S, Sacchetto M, Meri- in syndromes: recent advances in our under- co A, Piccione F. Event-related brain potential mod- standing of disorders of consciousness. J Neurol ulation in patients with severe brain damage. Clin 2011; 258: 1373-1384. Neurophysiol 2011; 122: 719-724. Schnakers C, Vanhaudenhuyse A, Giacino J, Ventu- 4) 15) Fischer C, Dailler F, Morlet D. Novelty P3 elicited ra M, Boly M, Majerus S, Moonen G, Laureys S . Di- by the subject’s own name in comatose patients. agnostic accuracy of the vegetative and mini- Clin Neurophysiol 2008; 119: 2224-2230. mally conscious state: clinical consensus versus Soros P standardized neurobehavioral assessment. BMC 16) . The neural correlates of consciousness: Neurol 2009; 9: 35. Electrophysiological and neuroimaging evidence for conscious processing in vegetative state. Clin anhaudenhuyse A, Laureys S, Perrin F 5) V . Cognitive Neurophysiol 2011; 122: 641-642. event-related potentials in comatose and post-co- Schnakers C, Majerus S, Giacino J, Vanhaudenhuyse matose states. Neurocrit Care 2008; 8: 262-270. 17) A, Bruno MA, Boly M, Moonen G, Damas P, Lam- Daltrozzo J, Wioland N, Mutschler V, Kotchoubey B. 6) bermont B, Lamy M, Damas F, Ventura M, Laureys S. Predicting coma and other low responsive patients A French validation study of the Coma Recovery outcome using event-related brain potentials: a me- Scale-Revised (CRS-R). Brain Inj 2008; 22: 786- ta-analysis. Clin Neurophysiol 2007; 118: 606-614. 792. Cavinato M, Freo U, Ori C, Zorzi M, Tonin P, Picci- 7) 18) Kolarik J, Waberzinek G, Opavsky J. [Comatose one F, Merico A. Post-acute P300 predicts recov- states in neurology. I. Physiology of conscious- ery of consciousness from traumatic vegetative ness disorders]. Acta Univ Palacki Olomuc Fac state. Brain Inj 2009; 23: 973-980. Med 1980; 94: 65-85. Fischer C, Luaute J, Morlet D 8) . Event-related poten- 19) Perrin F, Schnakers C, Schabus M, Degueldre C, tials (MMN and novelty P3) in permanent vege- Goldman S, Bredart S, Faymonville ME, Lamy M, tative or minimally conscious states. Clin Neuro- Moonen G, Luxen A, Maquet P, Laureys S. Brain re- physiol 2010; 121: 1032-1042. sponse to one’s own name in vegetative state, 9) Schnakers C, Perrin F, Schabus M, Majerus S, Ledoux minimally conscious state, and locked-in syn- D, Damas P, Boly M, Vanhaudenhuyse A, Bruno MA, drome. Arch Neurol 2006; 63: 562-569.

4906