Clinical Neurophysiology 126 (2015) 1532–1538
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
The P300 in middle cerebral artery strokes or hemorrhages: Outcome predictions and source localization Mana R. Ehlers a,b,c, Carmen López Herrero a, Andreas Kastrup a, Helmut Hildebrandt a,d,⇑ a
Klinikum Bremen-Ost, Department of Neurology, Züricher Str. 40, 28325 Bremen, Germany University of Bremen, Center for Cognitive Sciences, 28359 Bremen, Germany c University of British Columbia, Department of Psychology, Vancouver, BC V6T 1Z4, Canada d University of Oldenburg, Institute of Psychology, 26111 Oldenburg, Germany b
a r t i c l e
i n f o
Article history: Accepted 26 October 2014 Available online 1 November 2014 Keywords: Stroke Severe impairment Early rehabilitation Outcome prediction P300 Lesion localization
h i g h l i g h t s The P300 amplitude predicts outcome after lesions in the territory of the middle cerebral artery. Impairment of attention may impede successful rehabilitation. The left superior temporal lobe and the left premotor/prefrontal areas are essential for P300
generation.
a b s t r a c t Objectives: There are no reliable outcome predictors for severely impaired patients suffering from large infarctions or hemorrhages within the territory of the middle cerebral artery. This study investigated whether the amplitude of the event-related potential (ERP) component P300 predicts if a patient will be transferred to the next stage of rehabilitation (positive outcome) or to a nursing home (negative outcome). The second goal was to look for lesion locations determining the generation of the P300 amplitude. Methods: Forty-seven patients performed an auditory oddball task to elicit the P300 and were assessed with different scores for activities of daily living (ADL). Patients were divided in two groups according to their outcome. P300 amplitudes were compared between these groups controlling for age and gender. Post-hoc analyses were performed to analyse the relationship between P300 amplitude and neurological outcome scores. In addition, lesion overlaps were created to detect which lesion pattern affects P300 generation. Results: Patients with a positive outcome showed higher P300 amplitudes at frontal electrode sites than those with a negative outcome. P300 amplitude correlated with ADL score difference. Lesions in the superior temporal gyrus, middle and inferior frontal and prefrontal regions led to visibly diminished P300 amplitudes. Conclusions: The findings suggest that an impairment of attention (P300 amplitude reduction) negatively influences successful neurological rehabilitation. Left superior temporal lobe and the left premotor/prefrontal areas are essential brain areas for the generation of the P300. Significance: P300 amplitude may be used as an outcome predictor for severely impaired patients suffering from middle cerebral artery strokes or hemorrhages. Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction Stroke is the leading cause of disability and is the third leading cause of death worldwide (Murray and Lopez, 1997). According to ⇑ Corresponding author at: Klinikum Bremen-Ost, Department of Neurology, Züricher Str. 40, 28325 Bremen, Germany. E-mail address: [email protected] (H. Hildebrandt).
the World Health Organisation, a total of 15 million people suffer a stroke each year. Strokes can principally be classified into two main categories, ischemic and hemorrhagic (Donnan et al., 2008). 85% of all strokes are of ischemic origin (Qureshi et al., 2001) and most commonly occur within the territory of the middle cerebral artery (MCA). The MCA supplies the lateral portion of the cerebral cortex (about 60–70% of the hemisphere). Proximal occlusions of the MCA (M1 and M2 occlusions) usually lead to large lesions with
http://dx.doi.org/10.1016/j.clinph.2014.10.151 1388-2457/Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
M.R. Ehlers et al. / Clinical Neurophysiology 126 (2015) 1532–1538
severe functional impairments. In Germany, patients with large infarctions or hemorrhages are treated on specialised units for early rehabilitation if they show complications like swallowing disorders, disorientation and agitation, severe impairments of understanding among others and/or if the Barthel Index is below 30. Despite these enormous efforts many of those affected have a poor outcome (Barthel Index < 60) and depend on long-term institutional care (Ward et al., 2005). Aside from the high incidence of ischemic or hemorrhagic strokes, there is a lack of reliable and easy applicable tools to predict the patients’ outcome on an early rehabilitation unit (Balaban et al., 2011). In this context the most important question is, if and when a patient will be able to enter the next rehabilitation phase. A special version of the German Barthel Index (EBI), an assessment parameter to objectively evaluate a person’s ability to perform activities of daily living expanded by early rehabilitation specific factors (Schönle, 1995) is used to answer this question and thus to determine outcome. Different factors have been discussed as outcome predictors. For instance, increased age (Fiorelli et al., 1995), female gender (Di Carlo et al., 2003), paralysis (Kwakkel et al., 1996), post-stroke depression (Mödden and Hildebrandt, 2008) and cognitive impairment in early post-stroke phases (Hajek et al., 1997) have been associated with a poorer outcome (Kwakkel et al., 1996). But these parameters do not clearly predict a patient’s outcome and are in part not clearly distinguishable (Appelros et al., 2010). For example, women often show a more negative outcome but they are typically older when their first stroke occurs (Reid et al., 2008) and the damage is more severe (Santalucia et al., 2013). As a result, the small effects of gender (Denti et al., 2013) and extent of impairment may be in fact solely explained by age. Taken together, there is a strong need to establish a measure, which is easy to apply even in patients who have strong impairments in motor, visual or cognitive domains. In this scenario, the P300 as event-related potential (ERP) could serve as a possible candidate fulfilling these criteria. The P300 has been shown to predict the outcome of patients in low responsive state (Daltrozzo et al., 2007) or coma (Morlet and Fischer, 2014). It can provide information on conscious and unconscious cognitive functions (Risetti et al., 2013). The P300 was first described by Sutton et al. (1965) and is classically elicited by so-called oddball paradigms. These are characterised by the presentation of frequent standard stimuli randomly interspersed with infrequent targets (Squires et al., 1975). Its amplitude and latency are modulated by target probability (Duncan-Johnson and Donchin, 1977) and task difficulty (Picton, 1992). Furthermore, age (Polich, 1991) and gender (Conroy and Polich, 2007) have been shown to alter the component’s characteristics. Although the question which exact processes this component might reflect is not yet answered, there is a general consensus that it is related to context updating (Donchin, 1981), attentional resource allocation (Isreal et al., 1980) and working memory processes (Dolu et al., 2005). Changes associated with aging further suggest that the P300 reflects a person’s general cognitive ability (Linden, 2005). Various brain regions have been associated with the origin of the P300. Lesion studies revealed that damage to the temporoparietal junction severely reduces the P300 amplitude (Knight et al., 1989; Verleger et al., 1994). Notably, areas of the temporoparietal junction are supplied by the MCA. Simultaneous fMRI/EEG studies further demonstrated that a broad network of brain regions is involved in target detection (Friedman et al., 2001; Linden, 2005). Several studies found brain activation in areas strongly overlapping with attention networks (Bledowski et al., 2004a; Mantini et al., 2009) or working memory systems (Horn et al., 2003) such as the inferior and middle frontal gyrus, the insula
1533
(Mulert et al., 2004), the precentral gyrus, the supplementary motor area (Rusiniak et al., 2013) and the inferior parietal lobule (Horn et al., 2003). With the exception of the supplementary motor cortex, these areas are also supplied by the MCA. The aims of this study were (1) to investigate whether the P300 amplitude predicts the outcome of patients with a lesion in the territory of the MCA; (2) to identify which brain regions are essential for the elicitation of a normal P300 and (3) to determine if these areas are associated with a positive or a negative outcome.
2. Materials and methods 2.1. Patients Forty-seven patients from the early rehabilitation unit of the Klinikum Bremen-Ost, Germany with a mean age of 66.7 ± 10.4 (mean ± std) were included into the analysis. The study was approved by the ethics committee of the University of Oldenburg. Thirty-eight of these patients had had an ischemic infarction due to an occlusion of the MCA, eight patients had had a primary intracerebral hemorrhage and one patient had suffered from a hemorrhagic infarction (Table 1). The location of the brain regions affected in the ischemic and hemorrhagic stroke patients was comparable. The primary outcome variable was either discharge to the next stage of rehabilitation (phase C of the German system) (positive outcome) or transfer to a nursing home (negative outcome).
2.2. Clinical assessment All patients on the early rehabilitation unit were assessed using the German version of the Early Reha Barthel Index (EBI) (Schönle, 1995), which measures the ability to perform activities of daily living and also considers severe impairments in the early rehabilitation phase. In addition, the Functional Independence Measure (FIM) (Schulz, 2002; Dodds et al., 1993), which informs about a patient’s functional impairment, was used. All indices were obtained at admission and during rehabilitation. According to the German rehabilitation system, a patient is transferred to the next stage of rehabilitation as soon as an EBI of 30 is reached. To allow comparisons between studies and to meet criteria of international stoke care, the National Institutes of Health Stroke Scale (NIHSS) and the modified Rankin Scale (mRS) were further obtained to assess the patients’ disabilities. Moreover, the Neuropsychiatric Inventory (NPI) was used to screen the patients for post-stroke depression.
2.3. Auditory event-related potentials All participants performed the same classical auditory oddball paradigm. A frequent (80%) standard tone with a frequency of 1 kHz was randomly interspersed with an infrequent (20%) target tone presented with a frequency of 2 kHz. Both tones were bilaterally presented with a SPL of 80 db over headphones. The simple auditory oddball paradigm was repeated two times. In each run a total of 200 tones were presented for 100 ms with an interstimulus interval of 900 ms. The first run aimed at a reliable detection of the infrequent presented target tone, i.e. it was a passive presentation paradigm. In the second part, the patients were asked to mentally count them. Passive and active runs were performed separately, because most of the patients had a severe aphasia or severe attention deficits and were unable to count the targets.
1534
M.R. Ehlers et al. / Clinical Neurophysiology 126 (2015) 1532–1538
Table 1 Patient groups with positive and negative outcome after a lesion in the territory of the middle cerebral artery (MCA).
Sex Aetiology
Systemic thrombolysis Affected hemisphere Clearly visible P300
Age (years) EBI at admission NIHSS mRS Frontal N100 amplitude (lV) Frontal N100 latency (ms) Frontal P300 amplitude (lV) Frontal P300 latency (ms) Lesion load (mm3)
Positive outcome
Negative outcome
13 female, 17 male 24 ischemic infarctions: large artery atherosclerosis (12), cardioembolism (7), stroke of other determined etiology (2), stroke of undetermined etiology (3) 6 intracerebral hemorrhages 8 patients 16 left, 14 right 17 with, 13 without
9 female, 8 (7)a male 13 ischemic infarctions: large artery atherosclerosis (6), cardioembolism (4), stroke of undetermined etiology (3)
Mean ± std
Mean ± std
64.4 ± 11.1b 65.3 ± 39.7 11.4 ± 3.7 3.6 ± 1.0 4.24 ± 2.87 104.3 ± 15.0 4.13 ± 3.97b 439.87 ± 89.68
70.8 ± 7.7b 84.1 ± 32.5 14 ± 2.0 4.1 ± 0.4 3.76 ± 2.87 95.2 ± 24.2 2.27 ± 1.65b 461.06 ± 89.22
32.6 ± 18.7
3 intracerebral hemorrhages 7 patients 10 left, 7 right 10 with, 7 without
44.5 ± 22.6
Abbreviations: Standard deviation (std); Early Reha Barthel Index (EBI); modified Rankin Scale (mRS); National Institute of Health Stroke Scale (NIHSS). a Lesion load of one patient not available. b p < 0.05.
2.4. EEG recordings EEG was recorded using a Nicolet Biomedical system. Two silver/silver chloride electrodes were placed on the scalp at Fz and Cz according to the 10–20 system and were referenced to the left mastoid. The electrodes were attached to the skin by the aid of adhesive collars. Impedance was kept below 10 kO. The EEG was averaged in a time window starting 100 ms prestimulus and lasting until 900 ms poststimulus. A 1 Hz high pass filter, a 30 Hz low pass filter and 50 Hz Notch pass filter were applied. The Nicolet Biomedical system has an automatised artefact control implemented, adding new trials until 200 artefact free trials are recorded. The N100 and P300 latencies and amplitudes were identified manually in the time window of 80–150 ms and 280–700 ms, respectively. For the primary focus of our study, the outcome prediction, the position of the N100 and P300 were detected by two independent observers. After agreement on the position, the latencies and amplitudes are calculated automatically by the Nicolet software. Amplitudes are determined compared to baseline with baseline defined as the average amplitude in the 100 ms prestimulus interval. Latency was defined as time between stimulus onset and peak amplitude. To compare the lesions of patients with and without P300 in MRIcron, their ERP waves were again inspected and judged by two independent observers. The component was determined as present if there was a clearly visible P300 in the range of 280– 700 ms at Fz and Cz. In order to control the somewhat subjective result of these inspections with the objective measure of amplitude, the mean of both groups (with and without P300) was calculated. Subsequently, all patients who had an amplitude higher than the mean of the patient group without the P300 were transferred to the group with P300. As a result, the 75 percentile of patients who were determined as not showing a clear P300 had smaller amplitude than the mean of the patients displaying a P300.
the axial cCT scans of the patients (slice gap of 6 mm), their lesions were transferred manually to the corresponding template slices and the lesion volume (in mm3) for each patient was calculated. For subtracting lesion localisation, left- and right-hemispheric lesion patterns were analyzed separately.
2.6. Statistical evaluation Univariate analysis of variance (ANOVA) with outcome (reaching the next stage of rehabilitation vs. nursing home) and affected hemisphere as independent variables and ‘‘P300 amplitude frontal’’ or ‘‘P300 latency frontal’’ as dependent variable was performed. Age and gender served as covariates, because of their known impact on the P300. Secondary exploratory correlational analyses were performed to further investigate the relationship of the frontal P300 with the scores (EBI, FIM, mRS) and with age, lesion load, etc. This was done because our primary outcome variable was binary (positive and negative outcome in terms of reaching the next step in neurological rehabilitation or not). Such a binary coding has the advantage of having a direct clinical meaning, but it ignores smaller improvements, which become more evident when using the scaled rehabilitation indices as variable. An ANOVA with the same factors but the dependent variables ‘‘N100 frontal’’ was performed additionally. Similarly, a post hoc exploratory correlational analysis was done to detect possible relationships with age, lesion load and different indices for ADL and functional impairment. Prior to the statistical analysis we tested the data for normal distribution. Since the depression domain of the NPI and the mRS deviated from normal distribution, we used non-parametric testing for these two variables.
3. Results 2.5. Lesion localization Size and location of the lesions were transferred to the standard brain of MRIcron (Rordon, 2012) according to cCT and MRT images. After rotating the MRIcron template (http://www.mccauslandcenter.sc.edu/mricro/mricron/index.html) into the same direction as
Patients with positive outcome were significantly younger than patients with negative outcome (Table 1), but did not differ in gender, localisation of stroke/hemorrhages (left or right hemisphere), and the functional status as measured with the NIHSS, EBI and FIM. There was also no clear pattern of different etiologies for
1535
M.R. Ehlers et al. / Clinical Neurophysiology 126 (2015) 1532–1538
strokes or hemorrhages comparing the patients with positive or negative outcome (Table 1). The ANOVA with the single dependent variable ‘‘P300 amplitude frontal’’ yielded an effect for outcome (F(1) = 4.216, p = .046), but no other significant effects. Patients being transferred to the next stage of rehabilitation showed higher P300 amplitudes then those transferred to a nursing home (Fig. 1). The same ANOVA with the latency of the P300 yielded no significant results. Secondary exploratory correlation analyses showed a correlation of the frontal P300 amplitude with EBI difference between admission and discharge (r = .31, p = .034). Correlational analysis with the frontal P300 latency showed a significant positive correlation with the FIM score at admission (r = .306, p = .037) and significant negative correlation with duration of stay at the rehabilitation unit (r = .336, p = .022). All patients showed N100 components indicating intact cortical processing of the tones. An ANOVA with frontal N100 amplitude or with frontal N100 latency as dependent variable did not yield any significant effects. Secondary exploratory analysis for the N100 amplitude revealed no significant correlations. But the N100 latency correlated with EBI difference (r = .366, p = .011), with the EBI at discharge (r = .421, p = .003), with FIM at admission (r = .298, p = .042) and at discharge (r = .416, p = .004).
3.1. Lesion overlap The lesion overlaps were created with 26 left (16 with and 10 without P300) and 20 right (14 with and 6 without P300) hemispheric lesions. After subtraction, a high lesion overlap was found in the superior temporal gyrus (85%) (Fig. 2). In the vast majority of patients (85%) without a visible P300 and involvement of the left hemisphere, the lesions encompassed Brodmann areas (BA) 44 and 45. The subtraction plot further revealed that up to 85% of patients showing no clear P300 had a lesion in the pre- and postcentral gyrus. In approximately 80% of these patients the middle frontal gyrus (BA 6) as well as the insula were damaged. In addition, we found that up to 80% of these patients had a lesion in the angular gyrus. In order to ascertain that neither lesion location nor etiology of lesions (i.e. ischemic versus hemorrhagic) influenced the results, we additionally created left hemispheric lesion overlaps only of patients suffering from ischemic strokes. The results are comparable although overlap in the superior temporal gyrus was less pronounced – probably due to reduced sample size. The overlap images of the right hemispheric lesions did not show an overlap, which reached or exceeded 80% match within the group. A chi-square test revealed that the lack of the P300
amplitude was not (v2 = 42.95, p = .516).
related
to
the
affected
hemisphere
4. Discussion In the present study we analyzed the late ERP component P300, which has been related to decision-making, context-updating and attention (Polich, 2007), in a group of patients treated on an early rehabilitation unit because of MCA strokes or haemorrhages. Our results demonstrate that the P300 amplitude at frontal electrode sites predicts if a patient will have a good outcome (defined as entering the next stage of rehabilitation) or a poor outcome (defined as transfer to a nursing home). We controlled for age and gender since both covariates have been shown to influence P300 amplitude (Polich, 1996; Hoffman and Polich, 1999). Our main finding leads to the assumption that patients are impaired in detecting differences in tone pitch although they are able to perceive them (as measured by the N100). In other words, the passive perception of auditory stimuli is intact, whereas the active target-detection is impaired in some patients. It further indicates that patients being discharged to a nursing home are impaired in tasks that require attention. Attention has been found to be diminished in 24% up to 51% of stroke patients leaving the hospital (Hyndman et al., 2008), but because we focused on patients with large lesions and severe impairment (EBI < 30) these figures are clearly too small for our patient group. As a necessary process for higher cognitive functions such as memory (Chun and Turk-Browne, 2007) or language (Myachykov and Posner, 2005), attentional deficits strongly affect daily life. McDowd et al. (2003) showed that a poorer attentional performance is associated with a stronger impairment in daily living following stroke. In addition, it has been shown that attention deficits may at least predict motor recovery (Robertson et al., 1997). Moreover, the amplitude of the P300 correlated with the difference of the EBI between admission and discharge. We can conclude that the amplitude of the P300 predicts the outcome of the patients during early rehabilitation unit in terms of reaching the next step of rehabilitation. On the other hand, P300 latency and duration of rehabilitation correlated negatively, which seems to contradict our other findings. One explanation is that short latency – indicating better cognitive performance (Polich, 2007) – led to increased effort in rehabilitation, being delayed for example by swallowing disorders or other factors. This would mean that the time given to these patients to reach the next step of rehabilitation is longer, leading to the observed negative correlation. Against the assumption that the P300 amplitude is simply reduced due to the extent of the damage, the amplitude did not correlate with lesion load. In line with this finding, motor recovery and functional outcome after a stroke depends more on lesion
Fig. 1. Mean (±standard deviation) of the P300 amplitude at electrode sites Fz (left panel) and Cz (right panel) for patients who were either transferred to the next phase of rehabilitation (positive outcome) or to a nursing home (negative outcome).
1536
M.R. Ehlers et al. / Clinical Neurophysiology 126 (2015) 1532–1538
Fig. 2. Subtraction plots of lesion overlaps for patients with and without a clearly visible P300 at Fz and Cz for left (upper panel) and right (lower panel) hemispheric lesions. Colours ranging from black to red indicate 0% up to 100% lesion overlap. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
location than on lesion size (Chen et al., 2000). Our results demonstrate that although large parts of one hemisphere had been destroyed, the P300 may still be elicited. The majority of our patients were aphasic or suffered from severe attentional dysfunctions. Due to this fact, we cannot assure whether a run was performed under active or passive task conditions. However, the P300 amplitude does not seem to differ during active or passive auditory paradigms at frontal electrode sites (Bennington and Polich, 1999). For this reason, the run with a higher P300 amplitude was used for the analysis. This way, we were able to obtain the best possible task results for each participant. Even if the P300 was more pronounced due to the ability to actively track the tones, the higher amplitude would still point towards a positive outcome, as already implicated by the advanced task performance. From a clinical viewpoint, the most important question is whether the P300 can still be elicited although the patients show severe cognitive impairments. Studies with comatose patients (Fischer et al., 2008) showed that the P300 can be used as an outcome predictor even or especially for severely impaired patients. In contrast to the P300 amplitude, the N100 latency showed no direct relationship with positive or negative outcome, as defined in this study. Nevertheless, similar to the P300 amplitude, an increased N100 latency correlated with a larger EBI difference in addition to different ADL scores at discharge. Previous studies showed that N100 latency increases with enhanced fatigue and boredom (Callaway and Halliday, 1982) and is shortened as a result of diversion by reading (Roth et al., 1976). Generally, the N100 latency is prolonged due to increased attentional effort (Peeke et al., 1980). Similar to the P300 amplitude, the N100 latency is an indicator for attentional resource allocation (Ford et al., 1976). The correlation of increased N100 latencies with larger EBI differences and different ADL scores at discharge at least partially support the main conclusions drawn from P300 measurements. The lesion localization for the left hemisphere revealed a high overlap in the superior temporal gyrus. According to previous studies activation of parts of the superior temporal gyrus, presumably of the primary auditory cortex, has been related to auditory oddball paradigms, whereas the other neural generators seem to be modality-independent (Rogers et al., 1991; Tarkka et al., 1995; Opitz et al., 1999). The results of our overlay analyses further suggest that an intact precentral gyrus or supplementary motor cortex are necessary for the P300 generation (Rusiniak et al., 2013). The activation of the precentral gyrus in non-motor working memory tasks has led to the hypothesis that it is part of the working memory system
(Fiez et al., 1996) and may play a more general role in attention processes (Downar et al., 2000). In addition, lesions of the SMA have been related to attentional deficits such as hemineglect (Vallar, 1998). The involvement of the insula in patients with reduced P300 amplitude is also in line with previous studies, which have clearly shown its role in P300 generation during an oddball paradigm (Tarkka et al., 1995; Linden et al., 1999). Moreover, the insula is part of the core network, which is thought to be necessary for the maintenance of activity during continuous cognitive and behavioural tasks (Dosenbach et al., 2007). This network is likely to be engaged in target detection tasks (Mantini et al., 2009). In addition, it appears that the insula is necessary for overlearned, automatic responses (Fiez et al., 1996). It follows that a lesion in the region of the insular cortex reduces or diminishes P300 amplitude, because target detection is an automatic process, which happens to work even in patients in coma (Fischer et al., 2008). Previous studies demonstrated that the inferior parietal lobule is involved in P300 generation (Bledowski et al., 2004b). The lesion overlap in this study revealed the angular gyrus as part of this structure as critical for P300 production, but it has to be mentioned that the overlap was not as pronounced as in the frontal, insular and superior temporal area. In contrast to previous studies providing evidence for P300 generation in both hemispheres (Linden, 2005), the overlap and subtraction of right hemispheric lesions revealed no sufficient overlap to draw clear conclusions about right hemispheric neural generators of the P300. Nonetheless, there was no main effect for left/right hemispheric lesion location in the ANOVA for analyzing the P300 amplitude. A chi-square test further showed that there is no relationship between the affected hemisphere and the P300 amplitude. The lack of a clear relationship between right hemispheric lesions in this study might be due to the sample size. It could also be speculated that the contribution of brain regions within the right hemisphere is more globally related to attentional mechanisms and therefore less consistent. There are several limitations which have to be mentioned. First of all, we only used two electrodes to register the P300. Second, the overall number of patients included in our analyses was small, especially with respect to the subtraction plots for both hemispheres. Most patients had large lesions and this might limit the accuracy of the MRIcron subtraction procedure. Finally, they were cognitively severely impaired making more refined electrophysiological measures impossible. On the other hand this is also a strength of the study, because, as far as we know, there are only
M.R. Ehlers et al. / Clinical Neurophysiology 126 (2015) 1532–1538
a very few number of studies about the outcome of such a severely impaired patient group. In conclusion, this study shows that the P300 amplitude may serve as a predictor for positive or negative outcome in patients with ischemic or hemorrhagic lesions within the territory of the MCA during early rehabilitation. This might implicate that patients with a negative outcome are comparatively more impaired in attention and working memory related processes, an interpretation, which is supported by the lesions’ overlap of patients without P300. Acknowledgement We would like to thank Anja Tramp for performing the P300 recordings, Daniela Galashan for helpful comments on the manuscript and Markus Ebke for insisting on the relevance of ERPs for the assessment of early rehabilitation patients. Conflict of interest: None of the authors have potential conflicts of interest to be disclosed. References Appelros P, Stegmayr B, Terent A. A review on sex differences in stroke treatment and outcome. Acta Neurol Scand 2010;121:359–69. Balaban B, Tok F, Yavuz F, Yasßar E, Alaca R. Early rehabilitation outcome in patients with middle cerebral artery stroke. Neurosci Lett 2011;498:204–7. Bennington JY, Polich J. Comparison of P300 from passive and active tasks for auditory and visual stimuli. Int J Psychophysiol 1999;34(2):171–7. Bledowski C, Prvulovic D, Goebel R, Zanella FE, Linden DEJ. Attentional systems in target and distractor processing: a combined ERP and fMRI study. Neuroimage 2004a;22:530–40. Bledowski C, Prvulovic D, Hoechstetter K, Scherg M, Wibral M, Goebel R, et al. Localizing P300 generators in visual target and distractor processing: a combined event-related potential and functional magnetic resonance imaging study. J Neurosci 2004b;24:9353–60. Callaway E, Halliday R. The effect of attentional effort on visual evoked potential N1 latency. Psychiatry Res 1982;7:299–308. Chen CL, Tang FT, Chen HC, Chung CY, Wong MK. Brain lesion size and location: effects on motor recovery and functional outcome in stroke patients. Arch Phys Med Rehabil 2000;81:447–52. Chun MM, Turk-Browne NB. Interactions between attention and memory. Curr Opin Neurobiol 2007;17:177–84. Conroy MA, Polich J. Normative variation of P3a and P3b from a large sample (N = 120): gender, topography, and response time. J Psychophysiol 2007;21:22–32. Daltrozzo J, Wioland N, Mutschler V, Kotchoubey B. Predicting coma and other low responsive patients outcome using event-related brain potentials: a metaanalysis. Clin Neurophysiol 2007;118:606–14. Denti L, Artoni A, Scoditti U, Caminiti C, Giambanco F, Casella M, et al. Impact of gender–age interaction on the outcome of ischemic stroke in an Italian cohort of patients treated according to a standardized clinical pathway. Eur J Intern Med 2013;24:807–12. Di Carlo A, Baldereschi M, Gandolfo C, Candelise L, Ghetti A, Maggi S, et al. Stroke in an elderly population: incidence and impact on survival and daily function. The Italian longitudinal study on aging. Cerebrovasc Dis 2003;16:141–50. Dodds TA, Martin DP, Stolov WC, Deyo RA. A validation of the functional independence measurement and its performance among rehabilitation inpatients. Arch Phys Med Rehabil 1993;74:531–6. Dolu N, Basßar-Erog˘lu C, Özesmi Ç, Süer C. An assessment of working memory using P300 wave in healthy subjects. Int Congr Ser 2005;1278:7–10. Donchin E. Presidential address, 1980. Surprise!. . .Surprise? Psychophysiology 1981;18:493–513. Donnan GA, Fisher M, Macleod M, Davis SM. Stroke. Lancet 2008;371:1612–23. Dosenbach NU, Fair DA, Miezin FM, Cohen AL, Wenger KK, Dosenbach RA, et al. Petersen SE distinct brain networks for adaptive and stable task control in humans. Proc Natl Acad Sci U S A 2007;104:11073–8. Downar J, Crawley AP, Mikulis DJ, Davis KD. A multimodal cortical network for the detection of changes in the sensory environment. Nat Neurosci 2000;3:277–83. Duncan-Johnson CC, Donchin E. On quantifying surprise: the variation of eventrelated potentials with subjective probability. Psychophysiology 1977;14:456–67. Fiez JA, Raife EA, Balota DA, Schwarz JP, Raichle ME, Petersen SE. A positron emission tomography study of the short-term maintenance of verbal information. J Neurosci 1996;16:808–22. Fiorelli M, Alpérovitch A, Argentino C, Sacchetti ML, Toni D, Sette G, et al. Prediction of long-term outcome in the early hours following acute Ischemic stroke. Italian acute stroke study group. Arch Neurol 1995;52:250–5. Fischer C, Dailler F, Morlet D. Novelty P3 elicited by the subject’s own name in comatose patients. Clin Neurophysiol 2008;119:2224–30.
1537
Ford JM, Roth WT, Kopell BS. Attention effects on auditory evoked potentials to infrequent events. Biol Psychol 1976;4:65–77. Friedman D, Cycowicz YM, Gaeta H. The novelty P3: an event-related brain potential (ERP) sign of the brain’s evaluation of novelty. Neurosci Biobehav Rev 2001;25:355–73. Hajek VE, Gagnon S, Ruderman JE. Cognitive and functional assessments of stroke patients: an analysis of their relation. Arch Phys Med Rehabil 1997;78:1331–7. Hoffman LD, Polich J. P300, handedness, and corpus callosal size: gender, modality, and task. Int J Psychophysiol 1999;31:163–74. Horn H, Syed N, Lanfermann H, Maurer K, Dierks T. Cerebral networks linked to the event-related potential P300. Eur Arch Psychiatry Clin Neurosci 2003;253:154–9. Hyndman D, Pickering RM, Ashburn A. The influence of attention deficits on functional recovery post stroke during the first 12 months after discharge from hospital. J Neurol Neurosurg Psychiatry 2008;79:656–63. Isreal JB, Chesney GL, Wickens CD, Donchin E. P300 and tracking difficulty: evidence for multiple resources in dual-task performance. Psychophysiology 1980;17:259–73. Knight RT, Scabini D, Woods DL, Clayworth CC. Contributions of temporal–parietal junction to the human auditory P3. Brain Res 1989;502:109–16. Kwakkel G, Wagenaar RC, Kollen BJ, Lankhorst GJ. Predicting disability in stroke – a critical review of the literature. Age Ageing 1996;25:479–89. Linden DEJ. The P300: where in the brain is it produced and what does it tell us? Neuroscientist 2005;11:563–76. Linden DE, Prvulovic D, Formisano E, Vollinger M, Zanella FE, Goebel R, et al. The functional neuroanatomy of target detection: an fMRI study of visual and auditory oddball tasks. Cereb Cortex 1999;9:815–23. Mantini D, Corbetta M, Perrucci M, Romani G, Delgratta C. Large-scale brain networks account for sustained and transient activity during target detection. Neuroimage 2009;44:265–74. McDowd JM, Filion DL, Pohl PS, Richards LG, Stiers W. Attentional abilities and functional outcomes following stroke. J Gerontol B Psychol Sci Soc Sci 2003;58:P45–53. Mödden CH, Hildebrandt H. Poststroke depression (PSD): diagnose, verlauf und psychotherapeutische behandlungsmöglichkeit. Neurol Rehabil 2008;14:175. Morlet D, Fischer C. MMN and novelty P3 in coma and other altered states of consciousness: a review. Brain Topogr 2014;27:467–79. Mulert C, Jäger L, Schmitt R, Bussfeld P, Pogarell O, Möller H-J, et al. Integration of fMRI and simultaneous EEG: towards a comprehensive understanding of localisation and time-course of brain activity in target detection. Neuroimage 2004;22:83–94. Murray CJL, Lopez AD. Mortality by cause for eight regions of the world: global burden of disease study. Lancet 1997;349:1269–76. Myachykov A, Posner MI. Attention in language. In: Itti L, Rees G, Tsotsos JK, editors. Neurobiology of attention. Academic Press; 2005. p. 324–9. Opitz B, Mecklinger A, Von Cramon DY, Kruggel F. Combining electrophysiological and hemodynamic measures of the auditory oddball. Psychophysiology 1999;36:142–7. Peeke SC, Callaway E, Jones RT, Stone GC, Doyle J. Combined effects of alcohol and sleep deprivation in normal young adults. Psychopharmacology (Berl) 1980;67:279–87. Picton TW. The P300 wave of the human event-related potential. J Clin Neurophysiol 1992;9:456–79. Polich J. P300 in the evaluation of aging and dementia. Electroencephalogr Clin Neurophysiol Suppl 1991;42:304–23. Polich J. Meta-analysis of P300 normative aging studies. Psychophysiology 1996;33:334–53. Polich J. Updating P300: an integrative theory of P3a and P3b. Clin Neurophysiol 2007;118:2128–48. Qureshi AI, Tuhrim S, Broderick JP, Batjer HH, Hondo H, Hanley DF. Spontaneous intracerebral hemorrhage. N Engl J Med 2001;344:1450–60. Reid JM, Dai D, Gubitz GJ, Kapral MK, Christian C, Phillips SJ. Gender differences in stroke examined in a 10-year cohort of patients admitted to a Canadian teaching hospital. Stroke 2008;39:1090–5. Risetti M, Formisano R, Toppi J, Quitadamo LR, Bianchi L, Astolfi L, et al. On ERPs detection in disorders of consciousness rehabilitation. Front Hum Neurosci 2013;7:775. Robertson IH, Ridgeway V, Greenfield E, Parr A. Motor recovery after stroke depends on intact sustained attention: a 2-year follow-up study. Neuropsychology 1997;11:290–5. Rogers RL, Baumann SB, Papanicolaou AC, Bourbon TW, Alagarsamy S, Eisenberg HM. Localisation of the P3 sources using magnetoencephalography and magnetic resonance imaging. Electroencephalogr Clin Neurophysiol 1991;79:308–21. Rordon C. MRIcron. . Roth WT, Krainz PL, Ford JM, Tinklenberg JR, Rothbart RM, Kopell BS. Parameters of temporal recovery of the human auditory evoked potential. Electroencephalogr Clin Neurophysiol 1976;40:623–32. Rusiniak M, Lewandowska M, Wolak T, Pluta A, Milner R, Ganc M, et al. A modified oddball paradigm for investigation of neural correlates of attention: a simultaneous ERP–fMRI study. MAGMA 2013;26:511–26. Santalucia P, Pezzella FR, Sessa M, Monaco S, Torgano G, Anticoli S, et al. Sex differences in clinical presentation, severity and outcome of stroke: results from a hospital-based registry. Eur J Intern Med 2013;24:167–71.
1538
M.R. Ehlers et al. / Clinical Neurophysiology 126 (2015) 1532–1538
Schönle PW. The early rehabilitation Barthel Index – an early rehabilitationoriented extension of the Barthel Index. Rehabilitation (Stuttg) 1995;34:69–73. Schulz H. FIM Manual. Messung der Funktionalen Selbständigkeit (Functional Independence Measure). Meerbusch 2002: 48 S. Internet: http://www.fimpflegeplanung.de/kitteltaschenbuch.pdf, 2004. Squires NK, Squires KC, Hillyard SA. Two varieties of long-latency positive waves evoked by unpredictable auditory stimuli in man. Electroencephalogr Clin Neurophysiol 1975;38:387–401. Sutton S, Braren M, Zubin J, John ER. Evoked-potential correlates of stimulus uncertainty. Science 1965;150:1187–8.
Tarkka IM, Stokic DS, Basile LF, Papanicolaou AC. Electric source localisation of the auditory P300 agrees with magnetic source localisation. Electroencephalogr Clin Neurophysiol 1995;96:538–45. Vallar G. Spatial hemineglect in humans. Trends Cogn Sci 1998;2:87–97. Verleger R, Heide W, Butt C, Kompf D. Reduction of P3b in patients with temporo– parietal lesions. Brain Res Cogn Brain Res 1994;2:103–16. Ward A, Payne KA, Caro JJ, Heuschmann PU, Kolominsky-Rabas PL. Care needs and economic consequences after acute ischemic stroke: the Erlangen Stroke Project. Eur J Neurol 2005;12:264–7.
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
The P300 in middle cerebral artery strokes or hemorrhages: Outcome predictions and source localization Mana R. Ehlers a,b,c, Carmen López Herrero a, Andreas Kastrup a, Helmut Hildebrandt a,d,⇑ a
Klinikum Bremen-Ost, Department of Neurology, Züricher Str. 40, 28325 Bremen, Germany University of Bremen, Center for Cognitive Sciences, 28359 Bremen, Germany c University of British Columbia, Department of Psychology, Vancouver, BC V6T 1Z4, Canada d University of Oldenburg, Institute of Psychology, 26111 Oldenburg, Germany b
a r t i c l e
i n f o
Article history: Accepted 26 October 2014 Available online 1 November 2014 Keywords: Stroke Severe impairment Early rehabilitation Outcome prediction P300 Lesion localization
h i g h l i g h t s The P300 amplitude predicts outcome after lesions in the territory of the middle cerebral artery. Impairment of attention may impede successful rehabilitation. The left superior temporal lobe and the left premotor/prefrontal areas are essential for P300
generation.
a b s t r a c t Objectives: There are no reliable outcome predictors for severely impaired patients suffering from large infarctions or hemorrhages within the territory of the middle cerebral artery. This study investigated whether the amplitude of the event-related potential (ERP) component P300 predicts if a patient will be transferred to the next stage of rehabilitation (positive outcome) or to a nursing home (negative outcome). The second goal was to look for lesion locations determining the generation of the P300 amplitude. Methods: Forty-seven patients performed an auditory oddball task to elicit the P300 and were assessed with different scores for activities of daily living (ADL). Patients were divided in two groups according to their outcome. P300 amplitudes were compared between these groups controlling for age and gender. Post-hoc analyses were performed to analyse the relationship between P300 amplitude and neurological outcome scores. In addition, lesion overlaps were created to detect which lesion pattern affects P300 generation. Results: Patients with a positive outcome showed higher P300 amplitudes at frontal electrode sites than those with a negative outcome. P300 amplitude correlated with ADL score difference. Lesions in the superior temporal gyrus, middle and inferior frontal and prefrontal regions led to visibly diminished P300 amplitudes. Conclusions: The findings suggest that an impairment of attention (P300 amplitude reduction) negatively influences successful neurological rehabilitation. Left superior temporal lobe and the left premotor/prefrontal areas are essential brain areas for the generation of the P300. Significance: P300 amplitude may be used as an outcome predictor for severely impaired patients suffering from middle cerebral artery strokes or hemorrhages. Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction Stroke is the leading cause of disability and is the third leading cause of death worldwide (Murray and Lopez, 1997). According to ⇑ Corresponding author at: Klinikum Bremen-Ost, Department of Neurology, Züricher Str. 40, 28325 Bremen, Germany. E-mail address: [email protected] (H. Hildebrandt).
the World Health Organisation, a total of 15 million people suffer a stroke each year. Strokes can principally be classified into two main categories, ischemic and hemorrhagic (Donnan et al., 2008). 85% of all strokes are of ischemic origin (Qureshi et al., 2001) and most commonly occur within the territory of the middle cerebral artery (MCA). The MCA supplies the lateral portion of the cerebral cortex (about 60–70% of the hemisphere). Proximal occlusions of the MCA (M1 and M2 occlusions) usually lead to large lesions with
http://dx.doi.org/10.1016/j.clinph.2014.10.151 1388-2457/Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
M.R. Ehlers et al. / Clinical Neurophysiology 126 (2015) 1532–1538
severe functional impairments. In Germany, patients with large infarctions or hemorrhages are treated on specialised units for early rehabilitation if they show complications like swallowing disorders, disorientation and agitation, severe impairments of understanding among others and/or if the Barthel Index is below 30. Despite these enormous efforts many of those affected have a poor outcome (Barthel Index < 60) and depend on long-term institutional care (Ward et al., 2005). Aside from the high incidence of ischemic or hemorrhagic strokes, there is a lack of reliable and easy applicable tools to predict the patients’ outcome on an early rehabilitation unit (Balaban et al., 2011). In this context the most important question is, if and when a patient will be able to enter the next rehabilitation phase. A special version of the German Barthel Index (EBI), an assessment parameter to objectively evaluate a person’s ability to perform activities of daily living expanded by early rehabilitation specific factors (Schönle, 1995) is used to answer this question and thus to determine outcome. Different factors have been discussed as outcome predictors. For instance, increased age (Fiorelli et al., 1995), female gender (Di Carlo et al., 2003), paralysis (Kwakkel et al., 1996), post-stroke depression (Mödden and Hildebrandt, 2008) and cognitive impairment in early post-stroke phases (Hajek et al., 1997) have been associated with a poorer outcome (Kwakkel et al., 1996). But these parameters do not clearly predict a patient’s outcome and are in part not clearly distinguishable (Appelros et al., 2010). For example, women often show a more negative outcome but they are typically older when their first stroke occurs (Reid et al., 2008) and the damage is more severe (Santalucia et al., 2013). As a result, the small effects of gender (Denti et al., 2013) and extent of impairment may be in fact solely explained by age. Taken together, there is a strong need to establish a measure, which is easy to apply even in patients who have strong impairments in motor, visual or cognitive domains. In this scenario, the P300 as event-related potential (ERP) could serve as a possible candidate fulfilling these criteria. The P300 has been shown to predict the outcome of patients in low responsive state (Daltrozzo et al., 2007) or coma (Morlet and Fischer, 2014). It can provide information on conscious and unconscious cognitive functions (Risetti et al., 2013). The P300 was first described by Sutton et al. (1965) and is classically elicited by so-called oddball paradigms. These are characterised by the presentation of frequent standard stimuli randomly interspersed with infrequent targets (Squires et al., 1975). Its amplitude and latency are modulated by target probability (Duncan-Johnson and Donchin, 1977) and task difficulty (Picton, 1992). Furthermore, age (Polich, 1991) and gender (Conroy and Polich, 2007) have been shown to alter the component’s characteristics. Although the question which exact processes this component might reflect is not yet answered, there is a general consensus that it is related to context updating (Donchin, 1981), attentional resource allocation (Isreal et al., 1980) and working memory processes (Dolu et al., 2005). Changes associated with aging further suggest that the P300 reflects a person’s general cognitive ability (Linden, 2005). Various brain regions have been associated with the origin of the P300. Lesion studies revealed that damage to the temporoparietal junction severely reduces the P300 amplitude (Knight et al., 1989; Verleger et al., 1994). Notably, areas of the temporoparietal junction are supplied by the MCA. Simultaneous fMRI/EEG studies further demonstrated that a broad network of brain regions is involved in target detection (Friedman et al., 2001; Linden, 2005). Several studies found brain activation in areas strongly overlapping with attention networks (Bledowski et al., 2004a; Mantini et al., 2009) or working memory systems (Horn et al., 2003) such as the inferior and middle frontal gyrus, the insula
1533
(Mulert et al., 2004), the precentral gyrus, the supplementary motor area (Rusiniak et al., 2013) and the inferior parietal lobule (Horn et al., 2003). With the exception of the supplementary motor cortex, these areas are also supplied by the MCA. The aims of this study were (1) to investigate whether the P300 amplitude predicts the outcome of patients with a lesion in the territory of the MCA; (2) to identify which brain regions are essential for the elicitation of a normal P300 and (3) to determine if these areas are associated with a positive or a negative outcome.
2. Materials and methods 2.1. Patients Forty-seven patients from the early rehabilitation unit of the Klinikum Bremen-Ost, Germany with a mean age of 66.7 ± 10.4 (mean ± std) were included into the analysis. The study was approved by the ethics committee of the University of Oldenburg. Thirty-eight of these patients had had an ischemic infarction due to an occlusion of the MCA, eight patients had had a primary intracerebral hemorrhage and one patient had suffered from a hemorrhagic infarction (Table 1). The location of the brain regions affected in the ischemic and hemorrhagic stroke patients was comparable. The primary outcome variable was either discharge to the next stage of rehabilitation (phase C of the German system) (positive outcome) or transfer to a nursing home (negative outcome).
2.2. Clinical assessment All patients on the early rehabilitation unit were assessed using the German version of the Early Reha Barthel Index (EBI) (Schönle, 1995), which measures the ability to perform activities of daily living and also considers severe impairments in the early rehabilitation phase. In addition, the Functional Independence Measure (FIM) (Schulz, 2002; Dodds et al., 1993), which informs about a patient’s functional impairment, was used. All indices were obtained at admission and during rehabilitation. According to the German rehabilitation system, a patient is transferred to the next stage of rehabilitation as soon as an EBI of 30 is reached. To allow comparisons between studies and to meet criteria of international stoke care, the National Institutes of Health Stroke Scale (NIHSS) and the modified Rankin Scale (mRS) were further obtained to assess the patients’ disabilities. Moreover, the Neuropsychiatric Inventory (NPI) was used to screen the patients for post-stroke depression.
2.3. Auditory event-related potentials All participants performed the same classical auditory oddball paradigm. A frequent (80%) standard tone with a frequency of 1 kHz was randomly interspersed with an infrequent (20%) target tone presented with a frequency of 2 kHz. Both tones were bilaterally presented with a SPL of 80 db over headphones. The simple auditory oddball paradigm was repeated two times. In each run a total of 200 tones were presented for 100 ms with an interstimulus interval of 900 ms. The first run aimed at a reliable detection of the infrequent presented target tone, i.e. it was a passive presentation paradigm. In the second part, the patients were asked to mentally count them. Passive and active runs were performed separately, because most of the patients had a severe aphasia or severe attention deficits and were unable to count the targets.
1534
M.R. Ehlers et al. / Clinical Neurophysiology 126 (2015) 1532–1538
Table 1 Patient groups with positive and negative outcome after a lesion in the territory of the middle cerebral artery (MCA).
Sex Aetiology
Systemic thrombolysis Affected hemisphere Clearly visible P300
Age (years) EBI at admission NIHSS mRS Frontal N100 amplitude (lV) Frontal N100 latency (ms) Frontal P300 amplitude (lV) Frontal P300 latency (ms) Lesion load (mm3)
Positive outcome
Negative outcome
13 female, 17 male 24 ischemic infarctions: large artery atherosclerosis (12), cardioembolism (7), stroke of other determined etiology (2), stroke of undetermined etiology (3) 6 intracerebral hemorrhages 8 patients 16 left, 14 right 17 with, 13 without
9 female, 8 (7)a male 13 ischemic infarctions: large artery atherosclerosis (6), cardioembolism (4), stroke of undetermined etiology (3)
Mean ± std
Mean ± std
64.4 ± 11.1b 65.3 ± 39.7 11.4 ± 3.7 3.6 ± 1.0 4.24 ± 2.87 104.3 ± 15.0 4.13 ± 3.97b 439.87 ± 89.68
70.8 ± 7.7b 84.1 ± 32.5 14 ± 2.0 4.1 ± 0.4 3.76 ± 2.87 95.2 ± 24.2 2.27 ± 1.65b 461.06 ± 89.22
32.6 ± 18.7
3 intracerebral hemorrhages 7 patients 10 left, 7 right 10 with, 7 without
44.5 ± 22.6
Abbreviations: Standard deviation (std); Early Reha Barthel Index (EBI); modified Rankin Scale (mRS); National Institute of Health Stroke Scale (NIHSS). a Lesion load of one patient not available. b p < 0.05.
2.4. EEG recordings EEG was recorded using a Nicolet Biomedical system. Two silver/silver chloride electrodes were placed on the scalp at Fz and Cz according to the 10–20 system and were referenced to the left mastoid. The electrodes were attached to the skin by the aid of adhesive collars. Impedance was kept below 10 kO. The EEG was averaged in a time window starting 100 ms prestimulus and lasting until 900 ms poststimulus. A 1 Hz high pass filter, a 30 Hz low pass filter and 50 Hz Notch pass filter were applied. The Nicolet Biomedical system has an automatised artefact control implemented, adding new trials until 200 artefact free trials are recorded. The N100 and P300 latencies and amplitudes were identified manually in the time window of 80–150 ms and 280–700 ms, respectively. For the primary focus of our study, the outcome prediction, the position of the N100 and P300 were detected by two independent observers. After agreement on the position, the latencies and amplitudes are calculated automatically by the Nicolet software. Amplitudes are determined compared to baseline with baseline defined as the average amplitude in the 100 ms prestimulus interval. Latency was defined as time between stimulus onset and peak amplitude. To compare the lesions of patients with and without P300 in MRIcron, their ERP waves were again inspected and judged by two independent observers. The component was determined as present if there was a clearly visible P300 in the range of 280– 700 ms at Fz and Cz. In order to control the somewhat subjective result of these inspections with the objective measure of amplitude, the mean of both groups (with and without P300) was calculated. Subsequently, all patients who had an amplitude higher than the mean of the patient group without the P300 were transferred to the group with P300. As a result, the 75 percentile of patients who were determined as not showing a clear P300 had smaller amplitude than the mean of the patients displaying a P300.
the axial cCT scans of the patients (slice gap of 6 mm), their lesions were transferred manually to the corresponding template slices and the lesion volume (in mm3) for each patient was calculated. For subtracting lesion localisation, left- and right-hemispheric lesion patterns were analyzed separately.
2.6. Statistical evaluation Univariate analysis of variance (ANOVA) with outcome (reaching the next stage of rehabilitation vs. nursing home) and affected hemisphere as independent variables and ‘‘P300 amplitude frontal’’ or ‘‘P300 latency frontal’’ as dependent variable was performed. Age and gender served as covariates, because of their known impact on the P300. Secondary exploratory correlational analyses were performed to further investigate the relationship of the frontal P300 with the scores (EBI, FIM, mRS) and with age, lesion load, etc. This was done because our primary outcome variable was binary (positive and negative outcome in terms of reaching the next step in neurological rehabilitation or not). Such a binary coding has the advantage of having a direct clinical meaning, but it ignores smaller improvements, which become more evident when using the scaled rehabilitation indices as variable. An ANOVA with the same factors but the dependent variables ‘‘N100 frontal’’ was performed additionally. Similarly, a post hoc exploratory correlational analysis was done to detect possible relationships with age, lesion load and different indices for ADL and functional impairment. Prior to the statistical analysis we tested the data for normal distribution. Since the depression domain of the NPI and the mRS deviated from normal distribution, we used non-parametric testing for these two variables.
3. Results 2.5. Lesion localization Size and location of the lesions were transferred to the standard brain of MRIcron (Rordon, 2012) according to cCT and MRT images. After rotating the MRIcron template (http://www.mccauslandcenter.sc.edu/mricro/mricron/index.html) into the same direction as
Patients with positive outcome were significantly younger than patients with negative outcome (Table 1), but did not differ in gender, localisation of stroke/hemorrhages (left or right hemisphere), and the functional status as measured with the NIHSS, EBI and FIM. There was also no clear pattern of different etiologies for
1535
M.R. Ehlers et al. / Clinical Neurophysiology 126 (2015) 1532–1538
strokes or hemorrhages comparing the patients with positive or negative outcome (Table 1). The ANOVA with the single dependent variable ‘‘P300 amplitude frontal’’ yielded an effect for outcome (F(1) = 4.216, p = .046), but no other significant effects. Patients being transferred to the next stage of rehabilitation showed higher P300 amplitudes then those transferred to a nursing home (Fig. 1). The same ANOVA with the latency of the P300 yielded no significant results. Secondary exploratory correlation analyses showed a correlation of the frontal P300 amplitude with EBI difference between admission and discharge (r = .31, p = .034). Correlational analysis with the frontal P300 latency showed a significant positive correlation with the FIM score at admission (r = .306, p = .037) and significant negative correlation with duration of stay at the rehabilitation unit (r = .336, p = .022). All patients showed N100 components indicating intact cortical processing of the tones. An ANOVA with frontal N100 amplitude or with frontal N100 latency as dependent variable did not yield any significant effects. Secondary exploratory analysis for the N100 amplitude revealed no significant correlations. But the N100 latency correlated with EBI difference (r = .366, p = .011), with the EBI at discharge (r = .421, p = .003), with FIM at admission (r = .298, p = .042) and at discharge (r = .416, p = .004).
3.1. Lesion overlap The lesion overlaps were created with 26 left (16 with and 10 without P300) and 20 right (14 with and 6 without P300) hemispheric lesions. After subtraction, a high lesion overlap was found in the superior temporal gyrus (85%) (Fig. 2). In the vast majority of patients (85%) without a visible P300 and involvement of the left hemisphere, the lesions encompassed Brodmann areas (BA) 44 and 45. The subtraction plot further revealed that up to 85% of patients showing no clear P300 had a lesion in the pre- and postcentral gyrus. In approximately 80% of these patients the middle frontal gyrus (BA 6) as well as the insula were damaged. In addition, we found that up to 80% of these patients had a lesion in the angular gyrus. In order to ascertain that neither lesion location nor etiology of lesions (i.e. ischemic versus hemorrhagic) influenced the results, we additionally created left hemispheric lesion overlaps only of patients suffering from ischemic strokes. The results are comparable although overlap in the superior temporal gyrus was less pronounced – probably due to reduced sample size. The overlap images of the right hemispheric lesions did not show an overlap, which reached or exceeded 80% match within the group. A chi-square test revealed that the lack of the P300
amplitude was not (v2 = 42.95, p = .516).
related
to
the
affected
hemisphere
4. Discussion In the present study we analyzed the late ERP component P300, which has been related to decision-making, context-updating and attention (Polich, 2007), in a group of patients treated on an early rehabilitation unit because of MCA strokes or haemorrhages. Our results demonstrate that the P300 amplitude at frontal electrode sites predicts if a patient will have a good outcome (defined as entering the next stage of rehabilitation) or a poor outcome (defined as transfer to a nursing home). We controlled for age and gender since both covariates have been shown to influence P300 amplitude (Polich, 1996; Hoffman and Polich, 1999). Our main finding leads to the assumption that patients are impaired in detecting differences in tone pitch although they are able to perceive them (as measured by the N100). In other words, the passive perception of auditory stimuli is intact, whereas the active target-detection is impaired in some patients. It further indicates that patients being discharged to a nursing home are impaired in tasks that require attention. Attention has been found to be diminished in 24% up to 51% of stroke patients leaving the hospital (Hyndman et al., 2008), but because we focused on patients with large lesions and severe impairment (EBI < 30) these figures are clearly too small for our patient group. As a necessary process for higher cognitive functions such as memory (Chun and Turk-Browne, 2007) or language (Myachykov and Posner, 2005), attentional deficits strongly affect daily life. McDowd et al. (2003) showed that a poorer attentional performance is associated with a stronger impairment in daily living following stroke. In addition, it has been shown that attention deficits may at least predict motor recovery (Robertson et al., 1997). Moreover, the amplitude of the P300 correlated with the difference of the EBI between admission and discharge. We can conclude that the amplitude of the P300 predicts the outcome of the patients during early rehabilitation unit in terms of reaching the next step of rehabilitation. On the other hand, P300 latency and duration of rehabilitation correlated negatively, which seems to contradict our other findings. One explanation is that short latency – indicating better cognitive performance (Polich, 2007) – led to increased effort in rehabilitation, being delayed for example by swallowing disorders or other factors. This would mean that the time given to these patients to reach the next step of rehabilitation is longer, leading to the observed negative correlation. Against the assumption that the P300 amplitude is simply reduced due to the extent of the damage, the amplitude did not correlate with lesion load. In line with this finding, motor recovery and functional outcome after a stroke depends more on lesion
Fig. 1. Mean (±standard deviation) of the P300 amplitude at electrode sites Fz (left panel) and Cz (right panel) for patients who were either transferred to the next phase of rehabilitation (positive outcome) or to a nursing home (negative outcome).
1536
M.R. Ehlers et al. / Clinical Neurophysiology 126 (2015) 1532–1538
Fig. 2. Subtraction plots of lesion overlaps for patients with and without a clearly visible P300 at Fz and Cz for left (upper panel) and right (lower panel) hemispheric lesions. Colours ranging from black to red indicate 0% up to 100% lesion overlap. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
location than on lesion size (Chen et al., 2000). Our results demonstrate that although large parts of one hemisphere had been destroyed, the P300 may still be elicited. The majority of our patients were aphasic or suffered from severe attentional dysfunctions. Due to this fact, we cannot assure whether a run was performed under active or passive task conditions. However, the P300 amplitude does not seem to differ during active or passive auditory paradigms at frontal electrode sites (Bennington and Polich, 1999). For this reason, the run with a higher P300 amplitude was used for the analysis. This way, we were able to obtain the best possible task results for each participant. Even if the P300 was more pronounced due to the ability to actively track the tones, the higher amplitude would still point towards a positive outcome, as already implicated by the advanced task performance. From a clinical viewpoint, the most important question is whether the P300 can still be elicited although the patients show severe cognitive impairments. Studies with comatose patients (Fischer et al., 2008) showed that the P300 can be used as an outcome predictor even or especially for severely impaired patients. In contrast to the P300 amplitude, the N100 latency showed no direct relationship with positive or negative outcome, as defined in this study. Nevertheless, similar to the P300 amplitude, an increased N100 latency correlated with a larger EBI difference in addition to different ADL scores at discharge. Previous studies showed that N100 latency increases with enhanced fatigue and boredom (Callaway and Halliday, 1982) and is shortened as a result of diversion by reading (Roth et al., 1976). Generally, the N100 latency is prolonged due to increased attentional effort (Peeke et al., 1980). Similar to the P300 amplitude, the N100 latency is an indicator for attentional resource allocation (Ford et al., 1976). The correlation of increased N100 latencies with larger EBI differences and different ADL scores at discharge at least partially support the main conclusions drawn from P300 measurements. The lesion localization for the left hemisphere revealed a high overlap in the superior temporal gyrus. According to previous studies activation of parts of the superior temporal gyrus, presumably of the primary auditory cortex, has been related to auditory oddball paradigms, whereas the other neural generators seem to be modality-independent (Rogers et al., 1991; Tarkka et al., 1995; Opitz et al., 1999). The results of our overlay analyses further suggest that an intact precentral gyrus or supplementary motor cortex are necessary for the P300 generation (Rusiniak et al., 2013). The activation of the precentral gyrus in non-motor working memory tasks has led to the hypothesis that it is part of the working memory system
(Fiez et al., 1996) and may play a more general role in attention processes (Downar et al., 2000). In addition, lesions of the SMA have been related to attentional deficits such as hemineglect (Vallar, 1998). The involvement of the insula in patients with reduced P300 amplitude is also in line with previous studies, which have clearly shown its role in P300 generation during an oddball paradigm (Tarkka et al., 1995; Linden et al., 1999). Moreover, the insula is part of the core network, which is thought to be necessary for the maintenance of activity during continuous cognitive and behavioural tasks (Dosenbach et al., 2007). This network is likely to be engaged in target detection tasks (Mantini et al., 2009). In addition, it appears that the insula is necessary for overlearned, automatic responses (Fiez et al., 1996). It follows that a lesion in the region of the insular cortex reduces or diminishes P300 amplitude, because target detection is an automatic process, which happens to work even in patients in coma (Fischer et al., 2008). Previous studies demonstrated that the inferior parietal lobule is involved in P300 generation (Bledowski et al., 2004b). The lesion overlap in this study revealed the angular gyrus as part of this structure as critical for P300 production, but it has to be mentioned that the overlap was not as pronounced as in the frontal, insular and superior temporal area. In contrast to previous studies providing evidence for P300 generation in both hemispheres (Linden, 2005), the overlap and subtraction of right hemispheric lesions revealed no sufficient overlap to draw clear conclusions about right hemispheric neural generators of the P300. Nonetheless, there was no main effect for left/right hemispheric lesion location in the ANOVA for analyzing the P300 amplitude. A chi-square test further showed that there is no relationship between the affected hemisphere and the P300 amplitude. The lack of a clear relationship between right hemispheric lesions in this study might be due to the sample size. It could also be speculated that the contribution of brain regions within the right hemisphere is more globally related to attentional mechanisms and therefore less consistent. There are several limitations which have to be mentioned. First of all, we only used two electrodes to register the P300. Second, the overall number of patients included in our analyses was small, especially with respect to the subtraction plots for both hemispheres. Most patients had large lesions and this might limit the accuracy of the MRIcron subtraction procedure. Finally, they were cognitively severely impaired making more refined electrophysiological measures impossible. On the other hand this is also a strength of the study, because, as far as we know, there are only
M.R. Ehlers et al. / Clinical Neurophysiology 126 (2015) 1532–1538
a very few number of studies about the outcome of such a severely impaired patient group. In conclusion, this study shows that the P300 amplitude may serve as a predictor for positive or negative outcome in patients with ischemic or hemorrhagic lesions within the territory of the MCA during early rehabilitation. This might implicate that patients with a negative outcome are comparatively more impaired in attention and working memory related processes, an interpretation, which is supported by the lesions’ overlap of patients without P300. Acknowledgement We would like to thank Anja Tramp for performing the P300 recordings, Daniela Galashan for helpful comments on the manuscript and Markus Ebke for insisting on the relevance of ERPs for the assessment of early rehabilitation patients. Conflict of interest: None of the authors have potential conflicts of interest to be disclosed. References Appelros P, Stegmayr B, Terent A. A review on sex differences in stroke treatment and outcome. Acta Neurol Scand 2010;121:359–69. Balaban B, Tok F, Yavuz F, Yasßar E, Alaca R. Early rehabilitation outcome in patients with middle cerebral artery stroke. Neurosci Lett 2011;498:204–7. Bennington JY, Polich J. Comparison of P300 from passive and active tasks for auditory and visual stimuli. Int J Psychophysiol 1999;34(2):171–7. Bledowski C, Prvulovic D, Goebel R, Zanella FE, Linden DEJ. Attentional systems in target and distractor processing: a combined ERP and fMRI study. Neuroimage 2004a;22:530–40. Bledowski C, Prvulovic D, Hoechstetter K, Scherg M, Wibral M, Goebel R, et al. Localizing P300 generators in visual target and distractor processing: a combined event-related potential and functional magnetic resonance imaging study. J Neurosci 2004b;24:9353–60. Callaway E, Halliday R. The effect of attentional effort on visual evoked potential N1 latency. Psychiatry Res 1982;7:299–308. Chen CL, Tang FT, Chen HC, Chung CY, Wong MK. Brain lesion size and location: effects on motor recovery and functional outcome in stroke patients. Arch Phys Med Rehabil 2000;81:447–52. Chun MM, Turk-Browne NB. Interactions between attention and memory. Curr Opin Neurobiol 2007;17:177–84. Conroy MA, Polich J. Normative variation of P3a and P3b from a large sample (N = 120): gender, topography, and response time. J Psychophysiol 2007;21:22–32. Daltrozzo J, Wioland N, Mutschler V, Kotchoubey B. Predicting coma and other low responsive patients outcome using event-related brain potentials: a metaanalysis. Clin Neurophysiol 2007;118:606–14. Denti L, Artoni A, Scoditti U, Caminiti C, Giambanco F, Casella M, et al. Impact of gender–age interaction on the outcome of ischemic stroke in an Italian cohort of patients treated according to a standardized clinical pathway. Eur J Intern Med 2013;24:807–12. Di Carlo A, Baldereschi M, Gandolfo C, Candelise L, Ghetti A, Maggi S, et al. Stroke in an elderly population: incidence and impact on survival and daily function. The Italian longitudinal study on aging. Cerebrovasc Dis 2003;16:141–50. Dodds TA, Martin DP, Stolov WC, Deyo RA. A validation of the functional independence measurement and its performance among rehabilitation inpatients. Arch Phys Med Rehabil 1993;74:531–6. Dolu N, Basßar-Erog˘lu C, Özesmi Ç, Süer C. An assessment of working memory using P300 wave in healthy subjects. Int Congr Ser 2005;1278:7–10. Donchin E. Presidential address, 1980. Surprise!. . .Surprise? Psychophysiology 1981;18:493–513. Donnan GA, Fisher M, Macleod M, Davis SM. Stroke. Lancet 2008;371:1612–23. Dosenbach NU, Fair DA, Miezin FM, Cohen AL, Wenger KK, Dosenbach RA, et al. Petersen SE distinct brain networks for adaptive and stable task control in humans. Proc Natl Acad Sci U S A 2007;104:11073–8. Downar J, Crawley AP, Mikulis DJ, Davis KD. A multimodal cortical network for the detection of changes in the sensory environment. Nat Neurosci 2000;3:277–83. Duncan-Johnson CC, Donchin E. On quantifying surprise: the variation of eventrelated potentials with subjective probability. Psychophysiology 1977;14:456–67. Fiez JA, Raife EA, Balota DA, Schwarz JP, Raichle ME, Petersen SE. A positron emission tomography study of the short-term maintenance of verbal information. J Neurosci 1996;16:808–22. Fiorelli M, Alpérovitch A, Argentino C, Sacchetti ML, Toni D, Sette G, et al. Prediction of long-term outcome in the early hours following acute Ischemic stroke. Italian acute stroke study group. Arch Neurol 1995;52:250–5. Fischer C, Dailler F, Morlet D. Novelty P3 elicited by the subject’s own name in comatose patients. Clin Neurophysiol 2008;119:2224–30.
1537
Ford JM, Roth WT, Kopell BS. Attention effects on auditory evoked potentials to infrequent events. Biol Psychol 1976;4:65–77. Friedman D, Cycowicz YM, Gaeta H. The novelty P3: an event-related brain potential (ERP) sign of the brain’s evaluation of novelty. Neurosci Biobehav Rev 2001;25:355–73. Hajek VE, Gagnon S, Ruderman JE. Cognitive and functional assessments of stroke patients: an analysis of their relation. Arch Phys Med Rehabil 1997;78:1331–7. Hoffman LD, Polich J. P300, handedness, and corpus callosal size: gender, modality, and task. Int J Psychophysiol 1999;31:163–74. Horn H, Syed N, Lanfermann H, Maurer K, Dierks T. Cerebral networks linked to the event-related potential P300. Eur Arch Psychiatry Clin Neurosci 2003;253:154–9. Hyndman D, Pickering RM, Ashburn A. The influence of attention deficits on functional recovery post stroke during the first 12 months after discharge from hospital. J Neurol Neurosurg Psychiatry 2008;79:656–63. Isreal JB, Chesney GL, Wickens CD, Donchin E. P300 and tracking difficulty: evidence for multiple resources in dual-task performance. Psychophysiology 1980;17:259–73. Knight RT, Scabini D, Woods DL, Clayworth CC. Contributions of temporal–parietal junction to the human auditory P3. Brain Res 1989;502:109–16. Kwakkel G, Wagenaar RC, Kollen BJ, Lankhorst GJ. Predicting disability in stroke – a critical review of the literature. Age Ageing 1996;25:479–89. Linden DEJ. The P300: where in the brain is it produced and what does it tell us? Neuroscientist 2005;11:563–76. Linden DE, Prvulovic D, Formisano E, Vollinger M, Zanella FE, Goebel R, et al. The functional neuroanatomy of target detection: an fMRI study of visual and auditory oddball tasks. Cereb Cortex 1999;9:815–23. Mantini D, Corbetta M, Perrucci M, Romani G, Delgratta C. Large-scale brain networks account for sustained and transient activity during target detection. Neuroimage 2009;44:265–74. McDowd JM, Filion DL, Pohl PS, Richards LG, Stiers W. Attentional abilities and functional outcomes following stroke. J Gerontol B Psychol Sci Soc Sci 2003;58:P45–53. Mödden CH, Hildebrandt H. Poststroke depression (PSD): diagnose, verlauf und psychotherapeutische behandlungsmöglichkeit. Neurol Rehabil 2008;14:175. Morlet D, Fischer C. MMN and novelty P3 in coma and other altered states of consciousness: a review. Brain Topogr 2014;27:467–79. Mulert C, Jäger L, Schmitt R, Bussfeld P, Pogarell O, Möller H-J, et al. Integration of fMRI and simultaneous EEG: towards a comprehensive understanding of localisation and time-course of brain activity in target detection. Neuroimage 2004;22:83–94. Murray CJL, Lopez AD. Mortality by cause for eight regions of the world: global burden of disease study. Lancet 1997;349:1269–76. Myachykov A, Posner MI. Attention in language. In: Itti L, Rees G, Tsotsos JK, editors. Neurobiology of attention. Academic Press; 2005. p. 324–9. Opitz B, Mecklinger A, Von Cramon DY, Kruggel F. Combining electrophysiological and hemodynamic measures of the auditory oddball. Psychophysiology 1999;36:142–7. Peeke SC, Callaway E, Jones RT, Stone GC, Doyle J. Combined effects of alcohol and sleep deprivation in normal young adults. Psychopharmacology (Berl) 1980;67:279–87. Picton TW. The P300 wave of the human event-related potential. J Clin Neurophysiol 1992;9:456–79. Polich J. P300 in the evaluation of aging and dementia. Electroencephalogr Clin Neurophysiol Suppl 1991;42:304–23. Polich J. Meta-analysis of P300 normative aging studies. Psychophysiology 1996;33:334–53. Polich J. Updating P300: an integrative theory of P3a and P3b. Clin Neurophysiol 2007;118:2128–48. Qureshi AI, Tuhrim S, Broderick JP, Batjer HH, Hondo H, Hanley DF. Spontaneous intracerebral hemorrhage. N Engl J Med 2001;344:1450–60. Reid JM, Dai D, Gubitz GJ, Kapral MK, Christian C, Phillips SJ. Gender differences in stroke examined in a 10-year cohort of patients admitted to a Canadian teaching hospital. Stroke 2008;39:1090–5. Risetti M, Formisano R, Toppi J, Quitadamo LR, Bianchi L, Astolfi L, et al. On ERPs detection in disorders of consciousness rehabilitation. Front Hum Neurosci 2013;7:775. Robertson IH, Ridgeway V, Greenfield E, Parr A. Motor recovery after stroke depends on intact sustained attention: a 2-year follow-up study. Neuropsychology 1997;11:290–5. Rogers RL, Baumann SB, Papanicolaou AC, Bourbon TW, Alagarsamy S, Eisenberg HM. Localisation of the P3 sources using magnetoencephalography and magnetic resonance imaging. Electroencephalogr Clin Neurophysiol 1991;79:308–21. Rordon C. MRIcron. . Roth WT, Krainz PL, Ford JM, Tinklenberg JR, Rothbart RM, Kopell BS. Parameters of temporal recovery of the human auditory evoked potential. Electroencephalogr Clin Neurophysiol 1976;40:623–32. Rusiniak M, Lewandowska M, Wolak T, Pluta A, Milner R, Ganc M, et al. A modified oddball paradigm for investigation of neural correlates of attention: a simultaneous ERP–fMRI study. MAGMA 2013;26:511–26. Santalucia P, Pezzella FR, Sessa M, Monaco S, Torgano G, Anticoli S, et al. Sex differences in clinical presentation, severity and outcome of stroke: results from a hospital-based registry. Eur J Intern Med 2013;24:167–71.
1538
M.R. Ehlers et al. / Clinical Neurophysiology 126 (2015) 1532–1538
Schönle PW. The early rehabilitation Barthel Index – an early rehabilitationoriented extension of the Barthel Index. Rehabilitation (Stuttg) 1995;34:69–73. Schulz H. FIM Manual. Messung der Funktionalen Selbständigkeit (Functional Independence Measure). Meerbusch 2002: 48 S. Internet: http://www.fimpflegeplanung.de/kitteltaschenbuch.pdf, 2004. Squires NK, Squires KC, Hillyard SA. Two varieties of long-latency positive waves evoked by unpredictable auditory stimuli in man. Electroencephalogr Clin Neurophysiol 1975;38:387–401. Sutton S, Braren M, Zubin J, John ER. Evoked-potential correlates of stimulus uncertainty. Science 1965;150:1187–8.
Tarkka IM, Stokic DS, Basile LF, Papanicolaou AC. Electric source localisation of the auditory P300 agrees with magnetic source localisation. Electroencephalogr Clin Neurophysiol 1995;96:538–45. Vallar G. Spatial hemineglect in humans. Trends Cogn Sci 1998;2:87–97. Verleger R, Heide W, Butt C, Kompf D. Reduction of P3b in patients with temporo– parietal lesions. Brain Res Cogn Brain Res 1994;2:103–16. Ward A, Payne KA, Caro JJ, Heuschmann PU, Kolominsky-Rabas PL. Care needs and economic consequences after acute ischemic stroke: the Erlangen Stroke Project. Eur J Neurol 2005;12:264–7.
