Journal of Clinical EEG & Neuroscience, January, 2009
Abstracts of presentations at the 5th Annual Joint Meeting of the EEG and Clinical Neuroscience Society (ECNS) and the International Society for NeuroImaging in Psychiatry (ISNIP), in Frankfurt, Germany, September 10-13, 2008.
Abstracts to be continued in Vol.40, No.2
1. Applications for Combined EEG and fMRI in Basic and Clinical Research
Relation Between Electrical Brain Activity and Slow Changes of Cerebral Blood Flow Estimated by Simultaneous EEG/fMRI Recordings
Recently fMRI BOLD signal fluctuations have been observed that characterize different cerebral circuits by synchronous low-frequency oscillations - so called resting state networks (RSNs). However, it is still debatable whether they are caused by neuronal activity or some to brain function more unrelated phenomena (e.g., breathing, heart beats, etc.). Evidence for a relationship between RSNs and neuronal activity may be demonstrated by simultaneous EEG/fMRI measurements. In recent studies BOLD correlates of different EEG rhythms, especially the alpha rhythm, have been investigated. Although the timescale of the BOLD fluctuations and the EEG rhythms are quite distinct, the similarity of the networks described by either method is evident.
Currently we have performed simultaneous EEG/fMRI recordings in healthy subjects during relaxed wakefulness. For individual datasets we could demonstrate correlations between RSNs and EEG rhythms suggesting that these BOLD oscillations are related to neuronal activity.
Understanding the relationship between BOLD signal fluctuations and EEG rhythms may also play a role in revealing psychopathological mechanisms on a biological level, since several psychiatric disorders show changes in BOLD signals and/or EEG characteristics.
With the support of the Swiss National Foundation (Grant no. 320000-108321/1)
Neural Correlates (ERP/fMRI) of Executive Control Processes in Psychiatric Patients
Executive functions comprise various cognitive abilities including the inhibition of prepotent responses and voluntary decisions. Inhibition-associated brain activations were demonstrated especially in frontal brain regions whereas voluntary selections showed activations in the rostral cingulate zone as well as parietal regions.
Aims of simultaneous the EEG-fMRI studies were to discriminate neural correlates of various executive functions, the investigation of behavioral control processes in patients with alcohol dependence syndrome as well as attention deficit/hyperactivity disorder (ADHD), and to determine the influence of comorbid psychiatric symptoms on these processes.
Cued go/nogo tasks were used comprising several conditions: go (behavioral response required); nogo (inhibition of the planned response); voluntary selection (free decision, whether to respond or not).
Distinct networks of brain regions have been demonstrated to be associated to various executive processes: inhibition-associated responses were found especially in the pre-SMA and in lateral frontal brain regions whereas voluntary selection was associated with activations in the medial part of BA 8. Patients with ADHD showed significant decreased frontal responses. Furthermore, alcohol dependent patients with high anxiety ratings showed increased inhibition-associated responses compared to those patients with low anxiety scores. These results emphasize the importance of monitoring comorbid symptoms and traits in studies with psychiatric patients.
2. Magnetic Resonance Spectroscopy in Psychiatry
Dysfunctional Glutamatergic Neurotransmission in Schizophrenia
Glutamate is the principal excitatory neurotransmitter in the human brain modulating neuronal activity, cognitive functioning and structural integrity. Glutamate was suggested to be associated with personality traits and plays a key role in the pathophysiology of schizophrenia. Research on glutamate function in humans can principally be conducted by challenges with glutamate active agents, the association of intermediate phenotypes with genetic variations in the glutamate system, and by in vivo measuring of regional brain concentrations of glutamate using magnetic resonance spectroscopy (MRS).
The talk presents data on the effects of genetic NMDA receptor variations on auditory stimulus processing in healthy subjects and schizophrenic patients. Further evidence for the key role of glutamate in cerebral activity is provided by the association of directly measured brain glutamate concentration and brain oscillations, event related potentials, cognitive performance and personality traits. The impact of the results is discussed with regard to the pathophysiology of schizophrenia.
3. Neuroimaging in Parkinson’s Disease
Imaging Biomarker Development for PET in Neuroinflammation and Neurodegeneration
In recent years the development of novel diagnostic imaging biomarkers for neuroinflammation and neurodegeneration has been in the forefront of PET imaging biomarker research and development. The social burden of various neurodegenerative diseases, including MCI (mild cognitive impairment) and AD (Alzheimers Disease), puts a special emphasis on the problem. Molecular imaging with PET has the clinical potential to perform high through-put diagnostic screening as well as refined differential diagnosis of AD patients in early as well as advanced stages.
The most commonly used PET marker of beta amyloids is PIB (Pittsburgh Compound-B), labeling deposits of amyloid in the diseased human brain. Though the mechanisms of PIB-amyloid interaction is not fully understood, the ligand has proved to be an effective marker of amyloid load. Recently, follow-up markers have been tested, with promising results.
The PBR or, as it is recently called, TSPO system is probably the “most popular” target in the search for a novel PET marker of neuroinflammation. It is not a classical membrane receptor system but is a part of the multifunctional mitochondrial membrane protein complexes. The “classical” ligands of the PBR system belong to isoquinolines. Among them, PK11195 (3-isoquinolinecarboxamide) is the “gold standard” of PET ligands for the PBR system. However, its brain disposition is low. Novel PBR ligands, including Ro5-4864, DAA1106 and its analogues, PBR28, DPA-713 and vinpocetine, have in recent years been tested with varying success. Of them the most promising ligand is DAA1106 with favourable brain disposition and relatively high affinity.
Of the “classical neuroreceptor systems”, predominantly the monoamine systems have been in the focus of recent research. Both the 5HT1A receptor system and the serotonin transporter system (5HTT or SERT), as well as the norepinephrine transporter system (NET) have been investigated in relation to their alterations in neurodegeneration.
More recently, the use of dedicated PET tracers for labelling neurotransmitter metabolic pathways has come to the fore. Recent research has indicated that some metabolic pathways responsible for the metabolism of monoamine neurotransmitters may be early affected in the disease. The MAO-A and MAO-B systems show early alterations and can therefore serve as surrogate imaging end-points. Consequently, these systems can be targeted specifically with PET radioligands, including, e.g., 11C-deprenyl or 18F-deprenyl analogues. A search for more specific enzyme tracers usable in molecular neuroimaging of neuroinflammation is underway.
In conclusion, the search for dedicated PET biomarkers of neuroinflammation and neurodegeneration has a major emphasis in now-a-days’ PET tracer development. Both neurodegeneration and neuroinflammation can be targeted by various ways, including the classical neuroreceptor systems, the PBR or TSPO system, metabolic enzymes, neurofibrillary tangles or amyloid plaques. Other targets may come into the focus of biomarker research, including tau proteins, apoptosis-related targets or other metabolic compounds specific for the pathophysiology of the disease.
Cortical Functional Anatomy of Voluntary Saccades in Parkinson’s Disease
In Parkinson’s Disease (PD) several aspects of saccades are affected. The saccade generating brainstem neurons are spared, however the signals they receive may be flawed. In particular voluntary saccades suffer, but the functional anatomy of the impairment of saccade related cortical control is unknown.
We measured blood-oxygenation-level-dependent (BOLD) activation with functional Magnetic Resonance Imaging (fMRI) while healthy participants and patients with PD performed voluntary saccades between peripheral visual targets or fixated centrally. We compared saccade related BOLD-activity vs. fixation in patients with PD and in healthy controls and correlated perisaccadic BOLD-activity in PD patients with saccade kinetics (multistep saccades). Saccade related BOLD-activation was found in both, PD and healthy participants in the superior parietal cortex (PEF) and the occipital cortex. Our results suggest remarkable hypoactivity of the frontal and supplementary eye fields (FEF and SEF) in PD patients. On the other hand, PD patients showed a statistically more reliable BOLD modulation than healthy participants in the posterior cingulate gyrus, the parahippocampal gyrus, inferior parietal lobule, precuneus and in the middle temporal gyrus.
Given abnormal frontal and normal PEF responses, our results suggest that in PD a frontal cortical circuitry, known to be associated with saccade planning, selection, and predicting a metric error of the saccade, is deficient. Current research aims to further analyze cortical control of voluntary and reflexive eye movements in PD.
4. Special Session on Transcranial Magnetic Stimulation: Basics
Near-Infrared Spectroscopy for the Guidance of Inhibitory rTMS Treatment of Auditory Verbal Hallucinations in Schizophrenic Patients
Auditory verbal hallucinations (AVHs) are among the most frequent and disabeling symptoms of schizophrenic diseases. In approximately one quarter of patients, AVHs have to be considered as therapy-refractory with regard to pharmacological treatment options. This group of patients may benefit from a treatment protocol with repetitive Transcranial Magnetic Stimulation (rTMS) aiming on an inhibition of AVH-associated increased activity of auditory brain areas in the temporal cortex. However, optimal protocols for the guidance and control of such innovative treatment regimens are still lacking.
We propose the application of a non-invasive optical imaging technique (functional Near-Infrared Spectroscopy; fNIRS) for the measurement of the AVH-related activity of the auditory cortex, for the guidance of the rTMS-treatment and for the control of a treatment success on the brain metabolic level.
In the reported patient, NIRS measurement indicated AVH-related activity in the left auditory cortex which strongly decreased after a period of three weeks with daily inhibitory rTMS treatment, in parallel with drastically diminished AVHs.
This is the first report of a NIRS-guided and –controlled inhibitory rTMS treatment of therapy-refractory AVHs in a schizophrenic patient. Given the excellent clinical applicability of the applied methods, the combination of fNIRS and rTMS might have the potential to establish new treatment options in psychiatry aiming on the modulation of pathological regional brain activity patterns.
Effects of Repetitive TMS on the Cortical Expression of Immediate Early Gene and Calcium Binding Proteins in the Rat Brain
Repetitive application of transcranial magnetic stimulation (rTMS) has been demonstrated to change cortical excitability for some time. Changes in synaptic transmission in terms of long-term potentiation and depression are possible mechanisms fundamental to these changes in cortical excitability since cortical depression is usually observed with low-frequency stimulation (~1Hz) but facilitation following application of high-frequency protocols (5Hz and higher). In addition, cortical activity is substantially controlled by many types of inhibitory interneurons, which differ in input/output characteristics, firing pattern and the expression of co-transmitters and calcium-binding proteins (CaBP). Just the latter seem to determine the temporal characteristics of firing and transmitter release (GABA).
To get more insights in the cellular mechanisms of the rTMS effects, we are currently studying changes in cortical protein expression induced by different rTMS protocols applied to the rat brain. In addition to the CaBP Parvalbumin (PV), Calbindin (CB) and Calretinin (CR), we analysed the expression of the immediate early gene (IEG) products c-Fos and zif268 (ERG-1). Both signal recent neuronal activity, but c-Fos is more related to a general increase in electrical and metabolic activity, while zif268 is involved in synaptic potentiation (LTP).
Application of the intermittent-type theta-burst protocol (iTBS), but not a 1 Hz rTMS, strongly reduced the PV expression in all cortical areas analyzed, while 1 Hz rTMS reduced the CB expression. The CR expression was not affected. Changes in the expression of the IEGs showed both, a dependence on the rTMS protocol applied and differences with regard to the type of cortical area. In general, 1 and 10 Hz rTMS increased the c-Fos expression in almost all cortical areas, while iTBS increased it only in limbic structures. Similarly, but less strong, also sham-stimulation increased c-Fos expression only in limbic structures but not in sensory and motor cortices. Enhanced zif268 expression could be found in all cortical areas only following iTBS and in sensory and motor cortices also after 10 Hz rTMS.
Our results indicate that rTMS might activate different populations of cortical neurons depending on the stimulation frequency used. iTBS specifically affected the fast-spiking PV-expressing neurons involved in theta and gamma oscillations. Interestingly, a reduced PV expression can also be found in schizophrenics and in a rat model of schizophrenia and preliminary data of our lab indicate a prolonged decrease in PV. On the other hand, increased zif268 expression following iTBS is indicative of the induction of synaptic plasticity.
Simultaneous TMS and fMRI in Spatial Cognition Research
Functional Magnetic Resonance Imaging (fMRI) is capable of visualizing in vivo associations between different cognitive functions and patterns of neural activity in the human brain, and has therefore developed into a popular and versatile tool in human neuroscience.
A parallel and complimentary development, Transcranial Magnetic Stimulation (TMS) allows controlled manipulation of brain activity, with a quantifiable impact on behavior or cognition and is now a well established tool, particularly in cognitive studies, for inducing transient changes in brain activity non-invasively in conscious human volunteers. Over the past couple of years, this ability of actively interfering with neural processing during behavioral performance has been increasingly used for the investigation of causal brain-behavior relations in higher cognitive functions.
Due to their complimentary contribution, the simultaneous combination of TMS with fMRI, however, promises to be of especially great value for our understanding of the human brain, as it provides the opportunity to stimulate brain circuits while simultaneously monitoring changes in brain activity and behavior. Such a multi-modal imaging approach could identify brain networks of functional relevance, and might allow for causal brain-behavior inferences across the entire brain.
Here, I will present a short introduction into the methodological and technical challenges such a simultaneous combination creates, followed by a sequence of research studies combining TMS and fMRI for investigating the neurobiology of spatial cognition.
Deep Transcranial Magnetic Stimulation: Comparison Between Effects of
The H-coils are a new development in transcranial magnetic stimulation (TMS), designed to stimulate deeper neuronal pathways. Following our previous study evaluating safety and cognitive effects of the H-coil in healthy volunteers, we studied the antidepressive response induced by repeated, high-frequency (20Hz) TMS treatment using the H-coils over the prefrontal cortex (PFC). We compared the effects of three versions of the H-coil designed to stimulate deep PFC regions unilaterally (H1, H1L) or bilaterally (H2) in medication free, drug-resistant depressive patients. Stimulation with the novel H-coils was well tolerated, with no major side effects. When using 120% of motor threshold to affect deep PFC regions, response rates (defined as at least 50% reduction in HDRS scores) were 47%, 60% and 30%, for the H1, H1L and H2, respectively. On the other hand, when a more superficial stimulation was applied (using the H1L coil at 110% of motor threshold), response rate was 0%. The patients’ subjective report using BDI showed a similar pattern of responses. Computerized cognitive tests (CANTAB) also indicated significant improvements induced by deep TMS over the PFC. This study is the first evidence for the efficacy and safety of deep TMS in major depression and indicate that high frequency stimulation over the left PFC is more effective then bilateral stimulation.
Modulation of Central Monoaminergic Neurotransmission During rTMS
Repetitive transcranial magnetic stimulation (rTMS) has been investigated as a novel biophysical treatment approach for patients with major depression. RTMS has been shown to locally affect the stimulated area and to transsynaptically modulate neuronal activity in adjacent brain regions. Its therapeutic efficacy in patients with depression has been proven in controlled studies, however, the exact mechanism of rTMS induced antidepressive effects remain unclear. Depression is a clinically and pathophysiologically heterogeneous disorder, but there is increasing evidence of neurotransmitter imbalances and central monoaminergic dysfunctions. Regarding the neurobiological effects of rTMS, it is postulated that changes of neuronal activity within the stimulated cortical region and subsequent modulations of cortical-subcortical processes contribute to the antidepressive properties. Both animal and human in vivo studies have shown that rTMS impacts on central monoaminergic pathways. PET and STECT studies using dopaminergic radioligands showed that central dopaminergic activity can be modulated by using frontocortical magnetic stimulation.
In studies with depressed patients and healthy subjects using neurophysiological and functional neuroimaging techniques, including molecular imaging, we further explored direct and indirect rTMS effects. Transcranial magnetic stimulation leads to changes in latencies and amplitudes of evoked potentials and there is increasing evidence of longterm dopaminergic effects. Transsynaptical cortical stimulation might further induce a marked dopaminergic stimulation of subcortical brain regions similar to pharmacologic dopaminergic stimuli which could contribute to the therapeutic effects. The relevance of these neurobiological mechanisms regarding pathophysiology and treatment approaches of psychiatric disorders will be discussed.
Gap Junctions and the Very Fast Oscillations (>70 Hz) That Precede Seizures
Electrographic seizure discharges are frequently preceded by runs (hundreds of ms to a few seconds) of very fast oscillations. This phenomenon has been observed in human patients with subdural grid and depth electrode recordings; in anesthetized cats in vivo, by the late Mircea Steriade and colleagues; and in in vitro models, using hippocampal and neocortical slices. Experimental and network modeling work indicate that very fast oscillations are generated by electrically coupled ensembles of principal neurons, with the site of electrical coupling in the axons. This conclusion is supported by simultaneous recordings of soma and axon in vitro, by the pharmacological block of very fast oscillations with halothane and other gap junction-blocking compounds (in vitro and in vivo), and by dye coupling captured with confocal imaging between axons. More recently, gap junctions have been imaged ultrastructurally, between mossy fiber axons, including freeze-fracture immunolabeling that demonstrates connexin36. It remains to be determined if the very fast oscillations themselves causally initiate seizures, or rather are a flag for the presence of altered cortical structure and cellular environment which independently initiate seizures.
5. Neural Synchrony and its Modulation With Training and Transcranial Magnetic Stimulation
Reticulo-Thalamo-Cortical Circuits and Auditory Processing in Schizophrenia – Results of EEG Analyses and Mathematical Modelling
Deficits in auditory sensory gating are well known in schizophrenia. Most research in this field is based on the analysis of mid-latency auditory evoked potentials P50 and N100. Time-frequency analyses enable a more precise investigation of stimulus-associated changes in brain oscillations. The relationship between neuronal synchronization and dysfunctional information processing in schizophrenia and prodromal states has been investigated in 39 subjects at risk to develop a psychosis (18 without transition to psychosis within a 2 year follow up, 21 with transition), 46 neuroleptic-naïve first episode patients, 20 chronic schizophrenia patients, and 46 healthy controls. Time-frequency analyses have been performed based on complex Morlet wavelets separately for 5 frequency bands (delta, theta, alpha, beta, gamma). Patients at risk with later transition to psychosis exhibited a more pronounced phase rest (gamma and beta) compared to healthy controls. In the at risk patients with transition and in first episode patients the reduction of the phase reset evoked by the second click was lacking. Reduced phase rest in lower frequency bands was present in all patients with predominance in chronic schizophrenia patients.
To contribute to a better understanding of the underlying neuronal circuits mathematical modelling is a helpful tool. Extending the two-phase oscillator model of Tass (2004), we constructed a mathematical model of four connected phase-oscillators, representing components of the thalamo-cortical circuit. Simulations showed that phase alignments occurred in the same frequency bands and during similar time intervals as in real EEG data. By modifying the strength of the coupling between thalamus and cortical areas differences between schizophrenia patients and controls could be simulated. The simulations from the mathematical model can give new insights into the understanding of real EEG/MEG data and the underlying time-adjusted neuronal processes.
Effects of Working Memory Training on Oscillatory Activity in Patients With Mild Cognitive Impairment and Age-Matched Elderly Controls
The stages of cognitive impairment that may lie between normal ageing and early dementia are of great interest if we want to develop strategies for preventing dementia. Recently, the construct of mild cognitive impairment (MCI) has been proposed to designate an early state of cognitive impairment that deviates from the age norm. We conducted a cognitive training using a delayed discrimination task over a period of 4 weeks with 20 MCI patients and 20 healthy elderly controls in order to investigate how far mechanisms of neural plasticity are preserved in these groups. Our particular interest was whether and to what degree cognitive training might lead to a change in task-related evoked and induced oscillations. We expected a relationship between performance improvements and changes in oscillatory activity.
Before and following the training we recorded EEG. We varied WM load by presenting one or three visual objects that were shown sequentially for 1 second each. Following a delay of 8 seconds subjects had to compare the memorized objects to a probe, which matched in 50 percent of the trials. Encoding, maintenance and the retrieval periods were analysed for oscillatory low (theta and alpha) and high-frequency beta and gamma oscillations using a sliding window FFT.
Training reduced the number of errors in both groups, but had no effect on reaction time. Overall, there was no difference between groups. However, we found a relationship between performance increase and changes in oscillatory activity. Furthermore, whereas training resulted in an increase in early visual activity during encoding and retrieval (evoked oscillatory activity), it resulted in a decrease in oscillatory activity during maintenance.
The results suggest WM training has beneficial effects in both groups. Furthermore, it indicates that patients with MCI show at least partially preserved mechanisms of training-related neural plasticity.
How Prestimulus Oscillations Act as an Attentional Filter in Visual Perception Tasks
Several recent studies have shown that brain activity, already present prior stimulus presentation, is capable of predicting perceptual performance. Alpha oscillatory activity (around 10 Hz) is the most dominant phenomenon in the ongoing EEG, and shows its highest amplitude in the prestimulus interval. This talk addresses the question whether brain oscillatory activity, in the alpha frequency range, prior to the presentation of a stimulus is capable of predicting the behavioural outcome in visual perception tasks. This question was explored in a set of three experiments. In the first experiment, it will be shown that inter-individual differences in alpha oscillatory activity are correlated with inter-individual differences in perception performance. The results of the second experiment show that alpha oscillations can be used to predict perception performance within subjects, on a single trial level. In a third experiment, the question was addressed whether alpha oscillatory activity can be modulated by mood induction, showing that mood modulates both, prestimulus alpha oscillations and perception performance. Together, the results of these experiments show that prestimulus alpha oscillations are closely linked to visual perception performance. Moreover, it could be demonstrated that these effects can be experimentally modulated by mood induction. The conclusion of this talk is that alpha oscillations represent an inhibitory filter mechanism for the brain, which dissociates top-down from bottom-up driven stimulus processing.
Behavioral Relevance of Oscillatory Activity During Auditory Memory Processing
Cortical oscillatory activity in the gamma-band range (>30 Hz) has been established as a correlate of cognitive processes including perception, attention and memory. Only few studies, however, have provided evidence for an association between gamma-band activity (GBA) and behavioral performance. Here we summarize findings from recent studies investigating oscillatory correlates of auditory short-term memory processing. During the memorization phase of an auditory spatial delayed match-to-sample task, stimulus-specific GBA components could be identified by comparing oscillatory activity between two sample sounds S1 presented at different lateralization angles. These components peaked during the delay period about 300 ms prior to the onset of the test sounds. Positive correlations between the amplitude of these components and correct response rate were found during the final 100 ms of the delay phase only. This suggested that the ability to maintain stimulus-specific activations until the end of this phase contributed to task performance. When investigating responses to test sounds S2, additional GBA components were observed that distinguished nonidentical from identical S1-S2 pairs. Here better task performers showed higher amplitudes than poorer performers. Apparently the decision about whether test stimuli matched the stored representation of previously presented sample sounds relied partly on the oscillatory activation of networks representing differences between both stimuli. The present findings demonstrate the behavioral relevance of GBA. Their significance will be related to suggestions of synchronization deficits in neuropsychiatric disorders including schizophrenia.
Modulation of EEG Coherence by Bifocal Transcranial Magnetic Stimulation
Interregional coupling of brain regions can be assessed by EEG coherence reflecting the spatial-temporal correlation between two oscillatory signals. It has been suggested that this coherence is a signature of functional integration of multimodal neuronal networks. Repetitive transcranial magnetic stimulation (rTMS) is a well established technique for non-invasive cortical stimulation. Its modulating effects outlast the train of stimulation and affect behavior. We have tested the hypothesis that cortico-cortical coherence between distant brain areas can be selectively enhanced by synchronous bifocal rTMS. Cortico-cortical coherence was assessed in healthy subjects before and after befocal high frequency (10 Hz) rTMS to the left primary motor cortex and the visual cortex simultaneously. Coherence and spectral power were measured between these areas on the stimulated and the homologue contralateral side. Synchronous bifocal rTMS induced an increase of interregional coupling in the alpha and lower beta band on the stimulated side without effects on spectral power. These data indicate that synchronous bifocal rTMS might be a new strategy for selective modulation of interregional coupling. Furthermore, they raise the hypothesis that interventional enhancement of long-range coherence may effectively modulate interregional integration with behavioral consequences.
6. Neural Synchrony and Connectivity in Schizophrenia
Disconnectivity, Misconnectivity or Hyperconnectivity? – Diffusion Tensor Imaging Evidence for White Matter Abnormalities in Schizophrenia
Schizophrenia is considered a disconnectivity disorder (Andreansen et al., 1998; Friston et al., 1998). However, increasing evidence suggests that, rather than being exclusively a disconnectivity syndrome, a combination of increased and decreased connectivity may be related to schizophrenia. This hypothesis is supported by the results of EEG (Lee et al., 2006) as well as fMRI experiments (Foucher et al., 2005; Rotarska-Jagiela et al., in preparation) as well as Diffusion Tensor Imaging (DTI) examinations of white matter integrity (Hubl et al., 2004; Rotarska-Jagiela et al., 2008; Rotarska-Jagiela et al., submitted). Increased anatomical connectivity has been moreover shown to be related to increased severity of auditory hallucinations (Rotarska-Jagiela et al., submitted) and increased positive and negative symptoms as measured by PANSS (Rotarska-Jagiela et al., 2008).
Andreasen NC, et al. Schizophr Bull 1998; 24: 203-218; Foucher JR, et al. Neuroimage 2005; 26: 374-388; Friston KJ, et al. Schizophr Res 1998; 30: 115-125; Hubl D, et al. Arch Gen Psychiatry 2004; 61: 658-668; Lee SH, et al. Schizophr Res 2006; 83: 111-119; Rotarska-Jagiela A, et al. Neuroimage 2008; 39: 1522-1532.
High-Frequency Gamma-Band Oscillations During Perceptual Organisation in Chronic and First-Episode Schizophrenia Patients
Recent evidence suggests that patients with schizophrenia are characterized by reduced synchronous, oscillatory activity in the beta- and gamma-band range that may index a core dysfunction in the coordination of distributed neural activity. However, it is currently unclear to what extent high-frequency oscillations (>60Hz) contribute to impaired neural synchronization as research has so far focussed on gamma-band oscillations between 30-60 Hz. Secondly, it is not known whether deficits in high-frequency oscillations are already present at the onset of the disorder and to what extent reductions may be related to the confounding influence of medication.
To address these issues, we employed magnetoencephalography (MEG), a method particularly suited for the examination of low-amplitude, high-frequency oscillations, during perceptual organisation in a sample of chronic patients with schizophrenia (N=10), a sample of first-episode, never-medicated patients (N=7), and in a group of healthy controls (N=20). Perceptual organisation was examined with Mooney Faces. MEG signals were analysed for spectral changes in oscillatory activity in the frequency range of 30-130 Hz. Compared to healthy controls, both groups of schizophrenia patients showed a highly significant decrease in gamma-band activity across a wide frequency range (30-130 Hz) over parieto-occipital sensors. Chronic patients were characterized by a pronounced deficit in gamma-band activity and perceptual organisation relative to first-episode patients. These results suggest that schizophrenia is associated with a widespread reduction in high-frequency oscillations that indicate local network abnormalities. These dysfunctions are independent of medication status and already present at onset, suggesting a possible progressive deficit during the course of the disorder.
7. Multicenter Studies With fMRI: Rationale and First Results
Neural Correlates of Classical Conditioning
Aversive conditioning is a widely employed method to explore the fear circuit in healthy subjects and patients with anxiety disorders. We investigated the neurobiological responses in conditioning and extinction of neutral stimuli in patients with panic disorder.
In our study 90 patients suffering from panic disorder are measured in 4 different fMRI-sites with a classic aversive conditioning paradigm before and after a CBT-intervention. We hypothesize that patients will show a hyperreaction in the amygdala and associated areas belonging to the so-called “fear circuit” during conditioning and extinction phase.
Rationale and Methodology of Multicenter fMRI Studies
One of the major disadvantages of fMRI studies is the use of small samples in clinical studies. One approach to overcome this limitation could be multicenter studies, where data of different fMRI-centers is pooled together and form bigger samples. We present a design of a multicenter fMRI study conducted in nine German fMRI centers and we show why quality control and reliability measures are essential for multicenter studies. Multicenter studies may offer a huge advantage for the examination of cerebral processes in psychiatric disorders due to the increase in patient numbers.
Interoception in Panic Disorder
Cognitive behavioural models of panic disorder stress the importance of an increased attention towards bodily symptoms in the onset and continuation of this anxiety disorder. We investigate these effects by directing the attention to internal physical perceptions in a group at risk for panic disorder.
Twelve healthy female students with high levels of anxiety sensitivity and twelve female students with normal levels of anxiety sensitivity were investigated by 3 Tesla MRI. Subjects had to focus their attention either on their heartbeats (internal condition) or on neutral tones (external condition). Task performance was monitored by reporting the number of heart beats or tones after each block. State of arousal and emotional valence were also assessed. Data analysis was performed using SPM5.
A network of cortical and subcortical regions that overlaps with known fear circuits showed higher activation in anxiety sensitive students than control subjects. A regression analysis indicated a positive association of cingulate cortex activation with anxiety sensitivity. Our results suggest that attention to internal body functions modulates the activity of fear relevant brain regions in anxiety sensitive persons, a high-risk group for anxiety disorders. The current paradigm is used to investigate the effects of increased attention towards bodily symptoms in patients with panic disorder.
Neural Correlates of Cognitive Behavioural Therapy in Schizophrenia
Schizophrenia has been thought of in the past as a disorder not suitable for psychotherapeutic intervention. However, in recent years clinical studies could demonstrate that these patients do benefit from a cognitive behavioural therapy (CBT). The neurophysiological processes underlying this potential amelioration of symptoms have hardly been investigated. To measure the neural correlates of therapeutic interventions it is useful to resort to functional magnetic resonance imaging (fMRI) paradigms assessing cognitive functions typically impaired in schizophrenia. In patients with delusions such cognitive markers include jumping to conclusions as well as an attributional bias in social/emotional situations.
We hypothesise, that brain activation differences between patients and healthy subjects at the first scan will be diminished in the course of therapy – in CBT more than in ST. Furthermore, the following questions are being addressed: 1) Is a possible psychotherapeutic success underlined by an alteration of the neural correlates associated with the dysfunctions investigated? 2) Which neural circuits can be altered by CBT? 3) Can future therapeutic success be predicted on the basis of specific brain activation patterns already present before treatment? In the German Multi-Centre BMBF-Project “POSITIVE” we will be able to delineate changes in neural activation patterns with regard to psychotherapeutic interventions. On clinical terms, we must rethink the common routine care of patients with psychotic disorders, and implement the benefits psychotic patients gain from modern psychotherapy.
This work was supported by the Federal Ministry of Education and Research and the German Aerospace Centre (01GV0622).
The Westphal Paradigm – A New Paradigm to Study the Neuronal Correlates of Agoraphobia
First described 1871 by C. Westphal, agoraphobia is a highly prevalent and disabling anxiety disorder. Disorder specific cues in fMRI studies allow to further characterize the neuronal mechanisms of the disorder but also of possible new treatment approaches. We therefore developed the first fMRI paradigm with agoraphobia specific stimuli.
More than 1000 pictures of potential agoraphobic situations were generated by the authors. Neutral pictures were collected from the International Affective Pictures System (IAPS). Two sets, each containing 48 pictures were selected on the basis of best distinguishing between 26 patients with agoraphobia and 21 healthy control subjects. Neutral pictures were taken from the IAPS and pictures are randomly presented in a cued and an uncued manner.
Arousal, valence and panic-inducing ratings of the agoraphobic pictures were significantly increased in patients with agoraphobia, when compared to healthy control subjects. First results of patients studied with a 3-T GE scanner suggest that anticipation of agoraphobic pictures is associated with an activation of the fear circuit including the amygdala, inferior frontal gyrus, hippocampus, anterior cingulate and insula.
Supported by the BMBF (01GV0612)
8. Cerebral Changes in Mild Cognitive Impairment
Perfusion and Diffusion Weighted MRI in Mild Cognitive Impairment
With the use of fast imaging sequences and the availability of contrast agents, the assessment and monitoring of cerebral processes using magnetic resonance imaging has become possible. Different MRI methods can be used for the assessment of tissue perfusion, vascularization and microcirculation. Diffusion MRI allows the noninvasive functional assessment of white matter viability and structure. It offers an in vivo view into the microarchitecture of the brain which enables very early detection of indicators that are suggestive of pathology.
Particularly in early stages of dementing diseases, perfusion MRI and diffusion tensor imaging (DTI) are important since slight alterations are difficult to detect with structural methods. In Alzheimer´s disease (AD) and mild cognitive impairment (MCI), perfusion MRI demonstrated an alteration of cortical microcirculation in temporal and temporoparietal regions when compared to healthy controls. DTI also has great potential in this respect. Initial results suggest DTI can also be informative about the likely course of subsequent pathology among subjects with MCI. In a previous DTI study, we predicted which patients with MCI will convert to AD using a robust quantification of fractional anisotropy of the corpus callosum. The integration of knowledge concerning neuropsychological and neurobiological alterations associated with healthy ageing allows hypotheses for the differentiation of pathological ageing processes to be formulated. The use of perfusion and DTI in MCI is not widely studied, however, we will present initial clinical experience with these methods in healthy subjects and patients with neurodegenerative diseases, including MCI.
Functional MRI Studies on Learning and Memory in Mild Cognitive Impairment
Deficits of memory and learning are among the earliest symptoms in patients with mild cognitive impairment (MCI) but their neural correlates are only partly understood. From a clinical perspective, potential differences in learning capacities between cognitively impaired and healthy elderly people may be hypothesized but have not yet been investigated systematically. In a longitudinal study of working memory, Hempel et al. (2004) demonstrated an economisation with a decrease of cerebral activation following training in young adults. A similar effect was reported by Kodama et al. (2001) who investigated cortical activation during a motor task. That training effects may also apply to the elderly is supported by Erickson et al. (2007) who described an increase of hemispheric asymmetry with a reduction of age related differences in prefrontal cortices following practice of a dual-task. However, cross-sectional studies of declarative memory function, including autobiographical memory, were inconclusive in so far some studies showed increased or decreased cortical activation while others failed to demonstrate differences. We therefore investigated training effects in MCI under an explicit memory fMRI paradigm before and after one-week practice. After training, controls demonstrated a decreased neural activation whereas the MCI patients increased. We conclude training-induced mechanisms such as economisation and compensation may differ when incipient neurodegenerative processes are involved.
Hempel, et al. Am J Psychiatry 2004; 161: 745-747; Kodama S, et al. Psychol Med 2001; 31: 1079-1088; Erickson, et al. Neurobiol Aging 2007; 28: 272-283.
Quantitative MRI in the Diagnosis of MCI:
Quantitative analysis of MRI scans can be used to predict conversion to AD in at-risk subjects, such as patients with mild cognitive impairment (MCI). Manual volumetric techniques have been established in the past to measure atrophy of hippocampus and adjacent brain structures in vivo. More recently automated approaches to detect atrophy throughout the entire brain have been developed. These measures reach high accuracy in separating AD patients from non-demented controls, but the accuracy for the prediction of AD in at risk samples has still not finally been evaluated.
We will discuss how these techniques may usefully be applied for the prediction of the course of MCI. We will particularly focus on the potential of these techniques to assign an individual risk of conversion to a single subject. Multivariate analysis approaches, based on dimensionality reduction or feature selection, will be useful techniques in the future to solve this task. Another important issue that will be discussed is the effect of multicenter acquisition on the accuracy and effect sizes of MRI based volumetry. Presently, several national and international multicenter studies are underway to address this question.
In summary, MRI based quantification of brain structures is rapidly evolving to become an at risk marker of AD for clinical trials in the near future. The application of these techniques in clinical routine will depend on the availability of disease-modifying treatments. If these become available, we will need automated MRI analysis for preclinical diagnosis of AD.
Brain Morphological Changes in MCI, Clinical Symptoms and Neurobiological Marker
The term mild cognitive impairment (MCI) refers to minor cognitive deficits in elderly individuals which exceed those observed in the normal physiological aging process, but not yet fulfill the clinical criteria of dementia. Recent longitudinal studies demonstrated that MCI is associated with an increased risk of developing the full clinical picture of Alzheimer’s disease (AD).
While atrophic changes in the medial temporal lobe (MTL) of both MCI and AD patients (e.g., Pantel, et al., 2003) are a well established finding, their potential relation to characteristic neuropsychological or biological variables (e.g., tau-proteins, ApoE-genotype) has so far only scarcely been addressed. Associations between the aforementioned MCI/AD related variables and cerebral changes were investigated by using optimized voxel-based morphometry (VBM), an unbiased technique which allows examining morphological alterations throughout the entire brain. Our VBM results indicate both neuropsychological deficits and biological markers to be related to neurodegenerative changes typically found in MCI and in early AD.
Despite an involvement of the MTL, neuropathological research consistently revealed a deposition of neurofibrillary tangles and neuritic plaques in the olfactory bulb and tract (OBT) of AD patients. Based on these findings we performed region-of-interest (ROI) analysis of the OBT and found significantly smaller volumes in MCI and AD subjects (Thomann, et al., in press). Since the respective findings were associated with cognitive impairment, they may contribute to early recognition and diagnosis of the disease.
In another ROI study cerebellar lobes were found to be significantly smaller in AD patients when compared to healthy controls. In the AD group, atrophy of the posterior cerebellar regions was associated with poorer cognitive performance (Thomann, et al., in press). That the respective results did not apply to MCI subjects indicates cerebellar changes to be a feature of the more advanced stages of the disease.
Pantel J, et al. Am J Psychiatry 2003; 160:379-382; Thomann, PA, et al. Neurobiol Aging, in press; Thomann PA, et al. J Psychiatr Res, in press.
9. Electrophysiology in Psychosis
EEG Sources in Sinistral Schizophrenics Compared to Dextral Schizophrenics
Sixteen male schizophrenic, ambilateral and sinistral (or 24.6% of the 65 schizophrenics) were compared to 49 dextral male schizophrenics and 29 ambilateral and sinistral male controls. The LORETA sources were derived from a multi-channel EEG system of 48 electrodes (5 of these for artefact monitoring). The source analysis was carried out after factor analysis of the cross spectral matrix (30 factors accounting for over 99% of the variance) on the two factors with maximum variance on one group and minimum variance on the other and vice-versa. Compared to controls the sinistral schizophrenics have increased sources lateralized to the left hemisphere (frontal or temporal) in all frequency bands in both Eyes Open and Eyes Closed conditions. When the sinistral schizophrenics however are compared to dextrals then the dextral schizophrenics have increased sources in the right hemisphere (frontal or temporal) in all bands and conditions. When the dextral schizophrenics (n = 41) are compared to dextral controls (n = 61) the sources are increased in the right hemisphere either in the right frontal or right temporo-parietal region in all bands and conditions. The implications of these somewhat unexpected findings will be discussed.
Neuronal Generator Patterns of Olfactory Event-Related Potentials (OERP) in Schizophrenia
Deficits in odor threshold sensitivity, discrimination and identification are common in schizophrenia, presumably originating from brain structures also linked to their cognitive and emotional disturbances. However, the neurophysiological processes underlying olfactory dysfunction in schizophrenia have only been studied by Turetsky et al, who found reduced N1 and P2 amplitudes.¹
Nose-referenced 30-channel ERPs were recorded from 23 schizophrenic and 14 healthy adults (15/7 male) during an odor detection task. Hydrogen sulfide (H2S) stimuli (200 ms duration) at concentrations of 50% and 100% were presented to the left or right nostril by a constant-flow olfactometer (variable ISI 15-25 s). Time of odor stimulation was not cued. Subjects indicated whether they perceived a low or high odor intensity. To identify and measure neuronal generator patterns underlying ERPs, unrestricted Varimax-PCA was performed on their reference-free current source densities (spherical splines).
Patients’ behavioral performance was on par with that for healthy controls for high (25% vs. 26% misses) and low (43% vs. 44%) odor concentrations. Patients showed similar olfactory ERP and CSD waveforms when compared to controls, but their N1 sink (300 ms, bilateral frontotemporal maximum) and P2 source (615 ms, mid-parietal maximum) amplitudes were smaller. However, both groups had greater N1 sinks and P2 sources to high than low odor intensities.
In conclusion, OERP amplitude reductions to H2S stimuli in schizophrenia appear to reflect reduced activity in frontocentral, midline frontopolar, and parietal regions.
(1) Turetsky, et al. Biol Psychiatry 2003; 53(5): 403-411.
MMN Impaired - Then What? ERP Sequelae and Function in Schizophrenia
Mismatch negativity (MMN) results from the early detection of a change in an auditory pattern or stimulus sequence. Many patients with schizophrenia show attenuated MMN amplitudes. Activity from bilateral sources in the temporal and frontal lobes (auditory, cingulate and inferior frontal cortices) contributes to the MMN. Changes of source location in schizophrenia appear to reflect altered maturational changes in early onset patients and later, 14 years after illness onset.1 Temporal lobe sources reflect a sensory memory component while frontal sources may reflect the need to switch from automatic to controlled attentional processing. The ERP sequelae (Nd, negative difference; N2b and P3a) reflect the further assessment of the relevance, “targetness” and the need to orient to the stimulus, and the sources of these components reflect this change from bottom-up to top-down processing (superior frontal cortex and temporo-parietal junction2). In schizophrenia a lateralized imbalance in these contributions becomes evident3 that may reflect altered connectivity between the regions concerned (e.g., uncinate and longitudinal fasciculi). Poor MMN function is particularly associated with the poverty syndrome, and thought disorder, and is reflected in poor performance where attention and working memory is required. The immediate functional sequelae are seen in neuropsychological measures of cognitive flexibility and executive processes. In turn these abilities and poor MMN are associated with problems of social function and community living.4 But these social sequelae (and associated MMN impairments) can be ameliorated with training programs.5
(1) Oades, et al. BMC Psychiatry 2006; 6: 7; (2) Jemel, et al. Brain Topography 2003; 15: 249-262; (3) Oades, et al. Pharmacopsychiatry 1994; 27: 65-67; (4) Light & Braff. Arch Gen Psychiatry 2005; 62: 127-136; (5) Kawakubo, et al. Psychiatry Res 2007; 152: 261-265.
Mechanisms of Attention Control in Schizophrenia
Corbetta and co-workers (Corbetta and Shulman, 2002) have proposed a subtle interplay between neurologically distinct goal-driven and stimulus-driven attention networks. The stimulus-driven system can cause the goal-driven system to orient to a new stimulus if salience is high. However the stimulus-driven system can be inhibited, to some degree, by the goal-driven network. The operation of these mechanisms in patients with a diagnosis of schizophrenia was studied using a combination of ERP and EEG frequency measures during the performance of a simple number decision task interspersed with potentially distracting stimuli. The event-related potential and EEG frequency findings in this pilot study suggest the presence of both generalised abnormality of processing as well as specific abnormality of attention control in schizophrenia.
Corbetta, et al. Nat Rev Neurosci 2002; 3(3): 201-215.
Sleep and EEG-LORETA Studies in Young and Elderly Normal Subjects Under First- and Second-Generation Antipsychotics
Eighteen normal volunteers (9 young, 9 elderly) spent an adaptation, a baseline and 3 randomized treatment nights with weekly oral single doses of placebo, 50 mg Q (Seroquel®) and 80 mg P (Dominal forte®) in the sleep laboratory. Objective and subjective sleep and awakening quality was measured by polysomnography and psychometry. A 3-min V-EEG was obtained 10 h post drug and was analyzed by EEG-mapping and low-resolution brain electromagnetic tomography (LORETA).
Regarding sleep initiation and maintenance, in young subjects Q improved sleep latency, total sleep time and nocturnal awakenings, while P showed no changes as compared with placebo. In elderlies, both drugs improved all variables, with the placebo-corrected differences being more pronounced than in young subjects. Further differences in drug-induced changes between young and elderly volunteers were observed regarding sleep architecture, snoring, respiration, periodic leg movements and subjective sleep and awakening quality. After Q, morning EEG-LORETA showed changes typical of sedative antipsychotics, after P changes indicative of regulatory CNS effects after a short half-life drug (Q: 7 h, P: 2.5 h).
In conclusion, our EEG-LORETA findings are in agreement with the fMRI findings identifying the PFC and the limbic system as key regions in the pathogenesis of schizophrenia but also as target areas of Q.
10. Neurocognitive Effects of Second Generation Antipsychotics: ERP Findings
Impact of Antipsychotic Drugs on Topographical ERP Measures of Frontal Lobe Function: the NoGo-Anteriorization in the Treatment of Schizophrenias
First and second generation antipsychotics (FGAs, SGAs) are thought to exert their pharmacological effects via different neurotransmitter pathways in the brain, with a presumably more positive influence of SGAs on function and metabolism of the frontal cortex. However, studies directly investigating the effect of those drugs on neurophysiological measures of frontal lobe function are sparse. We therefore conducted a series of experiments investigating the impact of different antipsychotic drugs on an electrophysiological marker of prefrontal brain function, the so-called NoGo-Anteriorization (NGA), and neuropsychological test performance in different groups of schizophrenic patients. The NGA is a topographical EEG marker that quantifies the frontalization of the P300 field under conditions of response inhibition relative to response execution in Go-NoGo tasks. The NGA has been associated with an increased activation of prefrontal brain areas, particularly the anterior cingulate cortex (ACC), and has been found to be altered in chronic schizophrenic patients. Our results show 1) a consistently more positive influence of SGAs than FGAs on neuropsychological measures of frontal lobe function; 2) a significant positive correlation between the NGA and the daily dosage of SGAs in a group of patients suffering from acute psychotic disorders; 3) a more positive influence of SGAs than FGAs on functional activation of the ACC; and 4) a significantly different development of the NGA over a 6-week treatment period with SGAs vs. FGAs, with a trend towards decreased values after treatment with FGAs, and stabilized or increased values after treatment with SGAs or a combination of both types of drugs. In a final study, we examined the hypothesis that – because of their “frontal mechanism of action” – SGAs would yield particularly good treatment results in patients suffering from particularly pronounced frontal lobe deficits. To this end, the NGA was used to subdivide a sample of 76 schizophrenic patients into individuals with a relatively strong vs. weak frontal brain function at the beginning of treatment. The results confirm that baseline values of the NGA significantly predicted the patients’ clinical response to 6 weeks of treatment with either SGAs or FGAs: Low values of the NGA at baseline were associated with a particularly strong improvement under “atypical” medication (SGAs), whereas high initial values of the NGA predicted a particularly good response to conventional antipsychotics (FGAs). In summary, our findings indicate that the NGA is a topographic ERP parameter sensitive to the effects of a treatment with first vs. second generation antipsychotics and might be a valuable tool for developing individualized treatment strategies based on pathophysiological aspects of schizophrenic illnesses.
Event Related Activity and Symptom Improvement in Schizophrenia
A reduced amplitude of the auditory evoked P300 was interpreted as a trait marker of schizophrenia but reports about correlations between schizophrenic psychopathology and P300 amplitude indicate also a state character. To shed light upon these trait and state aspects a longitudinal study was performed to investigate the influence of symptom improvement and atypical neuroleptics on the amplitudes of the P300 and their subcomponents.
P300 was recorded in 17 schizophrenic patients before and after 4 weeks under either clozapine or olanzapine in a double-blind controlled design. For comparison, 17 ageand sex-matched healthy subjects were investigated. Parietal and frontal P300 subcomponents were investigated separately using dipole source analysis.
Schizophrenic patients had smaller parietal (temporo-basal dipole) but not frontal subcomponent amplitudes (temporo-superior dipole) than controls. For the whole sample subcomponent amplitudes did not change over 4 weeks despite clinical improvement but patients with a pronounced improvement of the PANSS positive score showed a slight enhancement of both subcomponents. This was not significant when the P300 amplitude was measured at a single electrode (Pz). No significant difference between clozapine and olanzapine concerning effects on P300 amplitudes were observed.
In conclusion, the results indicate that P300 subcomponents are modulated by changes of positive but not by changes of negative symptoms or different neuroleptics. This result was obvious for P300 subcomponents but not for Pz electrode measurement, which may be due to a higher reliability of the dipole source activity. The results can be integrated into a hypothetical model containing two athophysiological subgroups of schizophrenia.
Neurochemical Regulation of Auditory Information Processing Studied With EEG/MEG: Application to Schizophrenia
The neurophysiological and neurochemical changes in the auditory cortex are shown to underlie cognitive impairments in schizophrenia patients. In the number of consecutive studies we explored the neurochemical regulation of auditory information processing in relation to schizophrenia. By means of ERPs/ERFs we aimed to determine how neural substrates of auditory information processing are modulated by the antipsychotic medication in schizophrenia spectrum patients and by the neuropharmacological challenges in healthy human subjects.
First, we used auditory ERPs to investigate the effects of olanzapine1, risperidone2 and quetiapine3 in a group of patients with schizophrenia spectrum disorders. After 2 and 4 weeks of treatment olanzapine had no significant effects on the mismatch negativity (MMN) and P300, which have been proposed to reflect pre-attentive and attention-dependent information processing, respectively. Further, after 2 weeks of treatment risperidone had no significant effect on P300; however, risperidone reduced P200 amplitude. The latter effect of risperidone on neural resources responsible for P200 generation could be partly explained through the action of dopamine. Moreover, after 2 weeks of treatment, quetiapine reduced P300 latency in patients with schizoaffective disorder. This could be interpreted as a favorable profile of quetiapine action on attention-dependent processes in these patients.
Afterwards we used simultaneous electroencephalography/magnetoencephalography to investigate the effects of memantine4 and methylphenidate5 in healthy subjects. We found that memantine modulates the MMN response without changing other ERP components. This could be interpreted as a possible influence of memantine through the NMDA receptors on the auditory change-detection mechanism without otherwise changing the processing of auditory stimuli. Further, we found that methylphenidate does not modulate the MMN response. This finding could indicate no association between catecholaminergic activities and electrophysiological measures of pre-attentive auditory discrimination processes reflected in the MMN. The finding could possibly explain why antipsychotic treatment did not affect the MMN measures in previously reported studies, as well in our olanzapine study. However, methylphenidate decreases the P200 amplitudes. This may indicate that dopaminergic and noradrenergic systems modulate auditory information processing reflected in the P200, and is in line with the results of our risperidone study.
Taken together, our studies indicate a complex pattern of neurochemical influences produced by the antipsychotic drugs in the neural substrate of auditory information processing in patients with schizophrenia spectrum disorders and by the pharmacological challenges in healthy subjects studied with ERPs and ERFs.
(1) Korostenskaja M, et al. Prog Neuropsychopharmacol Biol Psychiatry, 2005; (2) Korostenskaja M, et al. Acta Neurobiol Exp (Wars), 2006; (3)Korostenskaja M, et al. under review, 2008; (4) Korostenskaja M, et al. Brain Res Bull, 2007; (5) Korostenskaja M, et al. Psychopharmacology (Berl), 2008.
Effects of Antipsychotic Drugs on Early Attentional Processes and Orienting: an ERP Study
Animal studies demonstrated that the second-generation (SGA) antipsychotics clozapine and risperidone, but not the first-generation antipsychotic (FGA) haloperidol, improve PCP-induced attention deficit and have favourable cognitive effects in animal models of schizophrenia. Event-related potentials (ERPs) might represent a valuable tool to characterize antipsychotic effects on cognition in human subjects. The present study investigated the effect of risperidone and haloperidol on ERP components related to automatic and effortful attention processes, using a randomized, placebo-controlled, cross-over design.
ERPs were recorded from 30 unipolar leads (0.5-70 Hz bandpass, 256 Hz sampling rate), during a three-tone oddball task in which target, standard and rare-nontarget tones were randomly presented. Subjects had to press a button when hearing a target tone, while ignoring both standard and rare-nontarget stimuli. N1 was identified at Cz, while P3 for target (P3b) and rare-nontarget stimuli (P3a) were identified at Cz and Pz leads. Amplitude maps at peak latency were then compared across conditions. If a significant drug effect was obtained, changes in the cortical sources of the corresponding ERP components were analyzed using Low-Resolution Electromagnetic Tomography (LORETA).
The amplitude of N1 for attended stimuli and of P3a was significantly increased only by risperidone. No significant change was observed in overall topographic features and in LORETA cortical sources of the same components.
In conclusion, these findings suggest a favorable effect of risperidone on early selective attention processes and automatic attention allocation.
11. Evidence Based Electrophysiology of Neurobehavioral Disorders
Evidence Based Medicine Evaluation of Electrophysiological Studies of the Anxiety Disorders
We provide an evidenced based medicine review of the use of electrophysiology in the anxiety disorders. It highlights functional similarities across the disorders and identifies patterns that differentiate disorder classifications. Electrophysiological measures offer reliable and objective clinical indicators of brain dysfunction within individuals, providing biomarkers for the improvement of diagnostic specificity and for informing on treatment decisions and prognostic assessments. Common to most anxiety disorders is a basal instability in cortical arousal, as reflected in measures of quantitative electroencephalography (qEEG). Resting EEG measures tend to correlate with symptom sub-patterns and be exacerbated by condition-specific stimulation. Also common are condition-specific difficulties with sensory gating and attentional control. These are clearly evident from event-related potential (ERP) measures of information processing in OCD, PTSD, PD, GAD and the phobias. Other ERP measures clearly differentiate the disorders, but with considerable variation across studies in relation to inclusion and exclusion criteria, medication status and control group selection. Historically, study numbers have tended to preclude the derivation of reliable diagnostic biomarker patterns but this is now being overcome with the development of databases of brain and cognition function, and the potential for inclusion of electrophysiological measures in disorder classification as foreshadowed for DSM-V.
Evidence-Based Medicine and Electrophysiology in Schizophrenia
The present contribution focuses on qualitative and quantitative EEG and P300 abnormalities as diagnostic tests for schizophrenia. The PubMed clinical query was used; key-words were schizophrenia and EEG/P3/P300. Inclusion criteria were: a diagnosis of schizophrenia confirmed by DSM-III/ICD-9 criteria or later editions; the inclusion of both a schizophrenia and a healthy control group; qualitative or spectral EEG findings or amplitude measures for P3. The included studies were reviewed to verify results homogeneity and the presence of information needed for systematic review and meta-analysis. Studies meeting all requirements were classified according to the Oxford Centre for Evidence-based Medicine (OCEBM) criteria.
No standard/clinical EEG study was OCEBM level 3b or better. For spectral EEG most studies qualified as level 4, and 25% as level 3b or better. An increase of delta and theta activity in patients was reported by most studies, while negative or discrepant findings were reported in less than 5% of them. For P3 amplitude reduction, 63% of the studies qualified as level 3b or better. Only 6.8% reported negative findings and none reported opposite results. Moderate effect sizes were found for spectral EEG delta and theta increase, and a large effect size for P300 amplitude reduction.
Evidence-Based Evaluation of Diagnostic Accuracy of EEG in Dementia and Mild Cognitive Impairment
Consistent reports on the validity of the EEG method in dementia work-up and an evidence based consensus on appropriateness of this method in the initial evaluation of patients with suspected cognitive disorder and dementia is missing. Using an evidence-based technique we searched for articles on diagnostic accuracy of spontaneous EEG in dementia disorders published from 1980 until June 2008. Forty-six articles were retrieved that strictly satisfied inclusion criteria. Figures on sensitivity and specificity across the studies varied widely, positive likelihood ratio in studies reporting classification accuracies between patients with Alzheimer’s disease and normal aging ranged between 2.3 and 38.5, and diagnostic odds ratios consequently showed large variations between 7 and 219. Although reported indexes of accuracy are in general high they are obtained and optimised on different clinical populations with variable sample size and disease severity, have large confidence intervals and results, therefore, cannot be extrapolated to other clinical populations.
Joint effort of preferably multicentre studies using uniform standards should develop optimised methods, investigate added diagnostic value of EEG in clinically established dementia diagnosis and, more importantly, predictive utility of EEG in mild cognitive impairment and questionable dementia.
12. Affect and Affectregulation: Basic Research and Clinical Applications
Neural Effects Following Training Using the “Frankfurter Test und Trainings Fazialen Affekts (FEFA)” in Autism Spectrum Disorders
One of the most consistent findings in the neuroscience of autism spectrum disorders (ASD) is hypoactivation of the fusiform gyrus (FG) during facial affect processing. In addition, reduced activation of the amygdala (AMG) has been associated with emotion perception alterations in ASD. In this study we examined whether basic computer-aided facial affect recognition training using the “Frankfurter Test und Trainings fazialen Affekts (FEFA)” is associated with an increased activation of the FG and AMG in high functioning ASD.
Eight FEFA training sessions lasting one hour each were applied over a period of five to six weeks. Preliminary findings in 11 trained versus 10 untrained participants aged 15 to 30 years (mean age: ~ 20.5 y, mean IQ: ~ 105) show enhanced BOLD-fMRI signal changes in the FG and AMG during implicit facial affect processing pre-post FEFA training. Behavioral facial affect recognition measures yield high effect size improvements for identical and similar tasks, but only minor effects for more general emotion recognition and social cognition tests.
In conclusion, in contrast to a previous pilot study (Bölte et al., 2006), the present data indicate that circumscribed behavioral gains in basic facial affect recognition are indeed correlated with robust and expected neural activation changes in the FG and AMG. Nevertheless, on the behavioral level, of a lack of sufficient generalization of acquired affect processing skills was confirmed.
Neural Foundations of Affective and Cognitive Theory of Mind
A growing body of research explores the neural foundations of the ability to infer on other persons mental states, frequently referred to as Theory of Mind (TOM). The functional system that is rendered by most studies comprises the anterior medial prefrontal cortex, superior temporal sulcus and the temporal poles. However, while this issue is of high importance for the understanding of social interaction, the question how TOM judgements about cognitive and emotional states differ from a neurofunctional scope is still unsolved. Hence we designed an experiment to explore cognitive and affective Theory of Mind generation in healthy subjects with the intention to use this method in subsequent studies of changes in TOM generation induced by psychotherapy.
Our experiment comprised an epoch related fMRI design. Healthy participants were presented 32 different cartoon stories. In each subject 4 different types of instructions were randomly assigned to the stories: Participants had to make judgements about the feelings of the stories´ main character, about their own feelings and about the number of living beings perceived from their own view and from the perspective of the main character. Thus the design comprised a 2 by 2 design exploring the main effects of a.) change of cognitive perspective and b.) emotional versus numeric judgements. Data were acquired with a 3 Tesla MRI-Scanner.
Preliminary data analysis revealed a main effect of other versus self perspective of cognition in bilateral temporoparietal junctions. In contrast self versus other judgements activated central parts of the pons and bilateral insular cortices. Compared to numeric judgements, focussing on emotional content resulted in a widespread activation of bilateral temporoparietal, temporal, medial and lateral prefrontal cortices as well as different parts of the limbic system. On the contrary a main effect of numeric compared to emotional judgements was only observed in the right intraparietal sulcus.
Most interestingly judgements about emotional states activated most regions of the previously described TOM network, while the main effect of change from the self to the other person’s perspective was only observed at the temporoparietal junction. Functional activation of the brainstem and the insular cortex induced by changes to the self perspective is in line with theories assuming that enteroceptive areas are involved in the generation of self reference.
Neural Correlates and Genetic Modulation of Affectregulation
Many of our behavioural and psychological dispositions are determined by genetic factors to a considerable extent, ultimately mediated by influencing brain structure and function. For example, it has been shown in several independent studies that a polymorphism in the serotonin transporter gene (5-HTTLPR) may influence amygdala reactivity to aversive stimuli. On the other hand it has been shown that human subjects are able to downregulate their emotions by conscious emotion regulation strategies thereby downregulating activity in the amygdala. The aim of this study was to find out if conscious emotion regulation has an impact on genetically determined differences in amygdala reactivity or would leave differences in amygdale activation in different genetic groups unchanged.
Thirty-seven young, female healthy subjects (mean age 22.64 years) performed an emotion regulation task. Subjects were presented with neutral or negative pictures from the IAPS and had either to only watch the pictures or to regulate their emotion by detachment, a reappraisal strategy. Functional imaging data were acquired on a 3 Tesla Siemens Trio Scanner. For the assessment of 5-HTTLPR influences on neural activity, participants were divided a priori into two groups. Individuals carrying one (n = 7) or two (n = 14) short alleles were appointed to the S-group, while the L-group consisted of long allele homozygotes (n = 16).
During passive viewing, s-allele carriers showed significantly more activation of the amygdala in reaction to aversive stimuli (p<0.05, s.v.c.). During intentional emotion regulation there was no more difference in amygdala activation between the genetic groups.
In conclusion, intentional emotion regulation strategies is a means to compensate for genetically determined differences in amydala reactivity to aversive stimuli. This has implications for affect regulation in general and psychotherapy in particular.
13. Modulation of Neuroplasticity in the Human Brain
Task-Related Modulations of Neuroplastic Changes Induced in the Human Motor Cortex
Transcranial direct current stimulation (tDCS) appears to be a promising tool in neuroplasticity research with perspectives in clinical neurophysiology. Its effect is closely related to modulation of cortical excitability and activity which are key mechanisms for learning and memory processing. The aim of our study was to test if cognitive task and motor exercise practiced during the stimulation are able to modify tDCS-induced plasticity in the left primary motor cortex in healthy subjects. Motor evoked potentials were recorded before and after 10 minutes of anodal, cathodal and sham stimulation. In experiment 1, subjects were required passively sit during the stimulation; in experiment 2 subjects attention was directed toward a cognitive test; whilst in experiment 3 subjects were instructed to push a ball in their right hand. Both the cognitive task and motor exercise modified tDCS-induced plasticity: performing the cognitive task during stimulation the motor cortex excitability was lower after anodal stimulation and higher after cathodal stimulation, when compared to the passive condition. When performing the motor exercise, the motor cortex excitability was lower both after anodal and cathodal stimulation, when compared to the passive condition. Our results show that transcranial direct current stimulation induced plasticity is highly dependent on the state of the subject during stimulation.
Homeostatic Metaplasticity of Human Cerebral Cortex
Evidence from animal data and slice preparations suggests that homeostatic mechanisms exist in order to keep neuronal excitability of the cerebral cortex within a physiologically useful range. Combining two different transcranial stimulation techniques, i.e., transcranial direct current stimulation (tDCS) and repetitive transcranial magnetic stimulation (rTMS), we demonstrated that homeostatic-like mechanisms could also be found non-invasively in the human motor cortex (M1) of healthy subjects. The amount and direction of rTMS-induced excitability changes, indexed by the amplitudes of motor evoked potentials (MEP), depended on the type of preconditioning with tDCS. rTMS, that was ineffective after sham tDCS, induced lasting inhibition when given after facilitatory anodal tDCS and lasting facilitation when given after inhibitory cathodal tDCS, irrespective of rTMS-frequency. These mechanisms turned out to be different in primary visual cortex (V1). Here, phosphene thresholds (PT) could be modulated with tDCS and rTMS, like MEP in M1. However, PTs only showed a modest tendency towards inhibition when rTMS was given after facilitatory tDCS of V1, and no homeostatic-like effect could be observed after preconditioning with cathodal tDCS, indicating substantial differences between the two cortices regarding their control of excitability. Finally, we tested homeostatic-like mechanisms in M1 in migraine patients. We found that migraineurs were fully capable of developing inhibition with facilitatory counter-regulation after cathodal tDCS and consecutive rTMS. However, in contrast to the control group patients were not able to develop intrinsic inhibitory counter-regulation when rTMS was given on a raised excitability level after anodal tDCS, indicating deficient homeostasis of cortical excitability with impaired inhibitory mechanisms in migraine. This deficit may be used as a physiological parameter for therapeutic drug monitoring in future studies.
Pharmacological Modulation of Neuroplasticity
Understanding development of plasticity in the human brain is of crucial importance for developing rational therapeutic strategies against diseases of the central nervous system which involve altered brain activity. In man pharmacological interventions are well suited to explore transmitters and neuromodulators involved in human neuroplasticity induced by non-invasive brain stimulation. An overview of the current state of research on the interaction of transcranial stimulation and neuropharmacology of central nervous plasticity in the human brain will be given. For this an order along the main neurotransmitters involved is pursued.