Deviant AERP components

In all eight patients with BECTS cases no MMN was evident as assessed by visual inspection o f the subtraction waveforms (deviant - standard). In contrast a consistent MMN was observed in all control subjects. The MMN evoked in the control subjects was observed as a negativity between the latency range o f 1 0 0 - 2 0 0 ms (mean 160.72 ms ± 23.70) with an amplitude of 2.42 pV ± 1.22). The group ‘standard’ and ‘deviant’ AERPs for both the patents with BECTS and control subjects can be seen in Figure 5. 7.

A

B

C

-|LlV

100 m s

F igure 5. 7. G roup averaged data from Fz show ing w aveform s to frequent (blue) and deviant (red) A E R Ps in (A) controls and (B) patients w ith B E C T S. D ifference w aveform s in controls (brow n thick line) and patients (brow n thin line) are show n in (C). V ertical bar is 5 jLiV for (A) and (B) and 2.5 p,V for (C). N ote increased peak to peak am plitudes o f patient A E R P com ponents com pared to controls (A vs B) and the absence o f any m ism atch negativity in the patients (C).

5.5 D iscussion.

The main findings o f this study were: 1. the latéralisation o f the P85-120 component o f the AERP contralateral to the side o f unilateral spikes in five patients with BECTS. 2. The increased amplitude o f this component in all patients with BECTS compared to controls. 3. A bilateral absence o f M M N in all patients with BECTS. These findings suggest that more than one process may be involved in producing the language difficulties uncovered by Staden et al., (1998) and support the hypothesis of that study that deficits in central auditory processing occur in patients with BECTS contributing to their poor language performance. The first proposed process involves unilateral disruption o f auditory processing in the hemisphere ipsilateral to the side of discharges observed during sleep, while the second involves a more global deficit in auditory processing.

In all the patients with unilateral spikes, there was a clear dominance o f the P85-120 contralateral to the spike focus observed during sleep. All but one o f the patients displayed centro-sylvian discharges during sleep, and with the exception o f another case did not reveal any discharges during the wake recording. The one child who did have discharges during her wake record showed extreme increases in discharge rate during sleep.

B y definition children with BECTS have a functional lesion in the Rolandic areas of the cortex. It is therefore open to discussion as to how a functional lesion within these areas affects central auditory processing. Two possibilities will be discussed, the first involving the location o f cortex responsible for processing auditory information and secondly the stability o f the source o f the functional lesion.

Although AI and ATI are situated below the Sylvian fissure in cortical areas 41 and 42 respectively, intracranial recordings reported in this thesis (section 3) and previous recordings in adults (Celesia & Puletti, 1969) have demonstrated the possibility o f auditory processing by cortex above the Sylvian fissure . It is therefore possible that the changes in the AERPs in patients with BECTS are due to specific disturbance o f these areas. Secondly the topography o f the spikes observed in patients with BECTS have been shown to migrate from one cortical area to another (Gibbs & Gibbs, 1954). Therefore during natural sleep when the probability o f spiking is highest there may be migration towards the Sylvian fissure where A1 is situated.

Although epileptic discharges have been demonstrated to affect AERPs, most studies have concentrated on short temporal effects. For instance short-term effects o f spikes on auditory potentials have been demonstrated in children with Landau-Kleffner Syndrome (LKS) an acquired epileptic aphasia (Seri et al., 1998). In that study the N1 component evoked by pure tones triggered by spikes was investigated. A reduction in N1 amplitude and prolongation o f latency over both hemispheres was observed compared to control subjects. However, these effects were more marked with left sided spikes. This bilateral reduction in amplitude suggests that both auditory cortices were affected although to a greater extent with left-sided discharges. These findings differ from the findings o f the current study that demonstrate unilateral deficits in auditory processing in patients with BECTS with unilateral discharges. Seri et al., (1998) do not report the average ISl they used which may be important. According to the published illustration (page 507, figure 1.) a ten to 12 second delay occurred between spikes. With very long ISls exceptionally large responses are evoked, receiving a major contribution from the non-specific N1

component (Hari, et, 1982). The contribution o f the non-specific N1 to the AERP with short ISIs is minimal after the first presentation o f the stimulus. The difference in topography present in the AERP components o f our patients with unilateral BECTS is probably due to the disrupted processing within the auditory cortex. In contrast, the decrease in amplitude o f the N1 in LKS patients (Seri et al., 1998) may reflect a disruption o f the generators o f the non-specific N1 component that is associated with arousal and attention (see 3.2.2). Although LKS and BECTS have been described as the two extremes o f a continuum they have two different functional lesions. The classic LKS patient has a unilateral functional lesion in the area o f the auditory cortex, with projection to the other auditory cortex while a classic patient with BECTS has a functional lesion over the Rolandic area. Therefore comparison o f LKS and BECTS patients groups in terms source o f o f epileptic phenmena and changes in AERPs is difficult.

In comparison, reductions in the amplitude o f the N1 recorded from sphenoidal electrodes have been reported ipsilateral to the side o f epileptic focus in adult patients prior to withdrawal from anticonvulsant medication (Tuunainen et al., 1995). No asymmetries in the N1 were seen in patients who did not have seizures prior to withdrawal o f medication. The deficits in N1 were not due to short-term effects of seizures, as AERP recordings were deferred for two hours after partial seizures and 12 hours after secondary generalised seizures. That study together with the results from the present study support the hypothesis that epileptiform activity can have long term effects on cortical function.

It is well established that transient cognitive impairment is observed in some epileptic patients with subclinical epileptiform discharges in the EEG ( Aarts et al., 1984; Binnie & Marston, 1992). It has been also shown that psychosocial function

improves with treatment o f interictal spikes (Marston et al., 1993). Transient cognitive impairments have been shown in BECTS patients during interictal spikes (Binnie & Marston, 1992). In the present study all o f the children were noted to have behavioural problems and/or cognitive dysfunction.

In the current study, spikes were not present in the on-going EEG during or immediately before the AERP recording session (with the exception o f one case). Patients had also been awake a minimum o f eight hours prior to auditory recordings and so it is unlikely that there were any short-term effects o f spikes observed during sleep on the AERPs. The normal topography o f the P85-I20 patients with bilateral discharges may be a result o f the small number o f discharges observed during sleep in these patients. Therefore the abnormality in the topography o f the P85-120 may be a result o f the frequency of the discharges observed during sleep.

Unlike the spike focus associated with structural cortical lesions, which remain localised to a relatively circumscribed area, the functional lesion focus may appear to move from one area to another and even shift between hemisphere. The phenomenon o f the migrating spike was first described by Gibbs & Gibbs (1954) and can occur over weeks or even days. Time and money constraints unfortunately precluded reinvestigation o f patients with BECTS in the present study, and it is therefore impossible to determine if the abnormality o f the P85-120 was temporary or resolved if the discharges were to migrate to the opposite hemisphere. Nonetheless, this abnormality was noted in all patients with BECTS, and so it seems likely that the relative abnormality is constant, though it may migrate according to the active spike focus. The migratory nature o f spikes in BECTS patients has lead to the concept that this epilepsy may derive from a gene-determined cellular defect.

Based on animal studies it is possible to speculate the probable cellular changes responsible for the epileptogenisis in BECTS. The ‘tottering m ouse’ (Noebels, 1948) with generalised epilepsy displays hyperinervation o f the forebrain mediated by locus ceruleus. The increased epileptogenicity in this mouse mutant is a consequence o f a noradrenergic decrease in hyperpolarization o f cortical neurones (Helekar, Noebels, 1991). Alterations in brain architecture have been demonstrated in patients with temporal lobe epilepsy including an indistinct boundary between the cortex and subcortical white matter and between laminae one and two o f the cortex and inappropriate columnar arrangement o f cortical neurones (Meencke, 1984; Meencke, & Janz, 1985; Hardiman et al., 1988). These brain alterations may result in faulty development o f synaptic connections resulting in a state o f cortical hyperexcitablilty. These faulty synaptic connections may be responsible for the abnormal processing of auditory information within the hyperexcited cortex, reflected by the presence o f an abnormal P85-120 component and increase in amplitude o f all the AERP components in the patients with BECTS. Although previous studies have reported an absence o f any structural lesion in patients with BECTS, recently a study investigating asymmetries o f brain structure has demonstrated distinct hippocampal asymmetries, with the affected smaller side being ipsilateral to the predominant epileptiform EEG manifestations, in six out o f 14 children (Lundberg et al., 1999). However, MRIs did not reveal any lesions in the BECTS patients in the current series.

The overall increase in amplitude o f the P85-120 in the BECTS patients compared to controls is o f interest in that it may represent disinhibition or hypersynchronization of auditory cortex or a larger area o f cortical recruitment. Famarier (1988) found that the VEP and SEP were enlarged compared to control subjects without any changes in

morphology or latency o f the responses, although Manganotti et al., (1998a) concluded that apparent enlargement o f somatosensory evoked potentials seen in some children with BECTS could be explained on the basis that they shared the same cortical generator as the spontaneous rolandic spikes. These findings do not explain the increase in amplitude o f the VEPs in the Famarier at al., (1988) study and the AERPs in the present study.

Recent genetic studies have found evidence for the linkage o f BECTS to a region on chromosome 15q l4 (Neubauer et al., 1998) encoding for a subunit o f the neuronal nicotinic acetylcholine receptor (nAChR) ‘a-7-A ch R ’. The a-7-A chR subunit gene has been previously linked to abnormalities in AERPs and auditory processing assessed neuropsychologically (Freedman et al., 1997). In the Freedman et al., (1997) study schizophrenic patients displayed a decrease in inhibition o f the P50 auditory evoked response to the second o f paired stimuli.

The nAChR belongs to the family o f ligand-gated ion channels that are widely distributed in the brain. Specific nicotinic receptor ligands that interact with specific subtypes o f nAChRs can be used as pharmacological tools to investigate the distribution of various nAChR subtypes. ( + )-[^H]epibatidide isolated from frog skin has a very high affinity to human as, 0 4 and a? nAChR sub-units. In a study investigating the distribution o f cortical nAChRs using ( + )-[^H]epibatidide (Marutle et al., 1998) two binding sites in the temporal cortex were isolated. Samples from autopsied brains were taken from Brodmann area 21. The concentration o f the nAChR in the temporal cortex ‘area 2 1 ’ corresponds to an area that has been shown during intracranial recordings to evoke auditory discrimination potentials.

There is substantial evidence that neuronal nAChR plays a major role in cognitive processing, such as attention, performance and memory in humans and animal

models (Newhouse et al., 1994; Levin & Simon, 1998). Nicotine has been shown to increase electrocortical arousal, which is accompanied by a release o f acetylecholine at the cortex that results from the action of nicotine on cholinergic neurones in the mesencephalic reticular formation (Wesnes & Warburton, 1983). Recently the involvement o f nAchR in pre-attentive memory processes have demonstrated in a study that revealed the MMN increases in size after chewing a gum containing nicotine (personal communication T. Baldeweg). These results provide a direct link with nAchR function and MMN.

It is postulated that the overall increase in amplitude o f the AERPs as well as the absence o f the MMN reflects disturbances in bilateral nAchR function within the temporal lobes and probably is irrespective o f the functional lesion imposed by the spikes. Chromosome 15 has also been linked to spelling disabilities in children with dyslexia (Grigorenko et al., 1997; Schulte-Kome et al., 1998); in both these studies there was evidence o f a linkage between chromosome 15q21 and dyslexia.

One implication o f these results is that there may be a case for prescription of antiepileptic medication in children with BECTS. In accordance with present practice in paediatrics, treatment is not recommended after the first or second convulsive seizure (Panayiotopoulos, 1999). Treatment is based on the clinical history and not the electrographic findings. Therefore patients with BECTS not suffering from seizures with a possibility o f continuous spikes observed during sleep are generally not medicated.

This study has demonstrated that BECTS patients have abnormal auditory processing and supports previous behavioural studies that have demonstrated language deficits in patients with BECTS. The electrophysiological techniques employed to identify deficits in central auditory processing in one o f the most benign forms o f epileptic

syndromes and may provide a tool to investigate BECTS patients susceptible to auditory processing deficits. In order to determine the feasibility o f this electrophysiological technique as a diagnostic tool further studies involving patients with BECTS who do not demonstrate behavioural deficits in speech and language will need to be assessed. The study however has not been able to determine if the changes in the AERPs are modality specific and/or permanent. Longitudinal studies involving subjects with BECTS would establish if the deficits remain after the clinical and electrographic markers have been resolved.

Patients with BECTS have a clear functional lesion and auditory processing dysfunction. Other children have been described with similar auditory problems in the absence of clear abnormalities. Children who have no abnormalities in their EEG and normal peripheral hearing assessed clinically can present with subtle auditory deficits. They have trouble hearing in certain environments, and this may be due to abnormalities within their auditory pathways or higher auditory processing. These children will be examined in the next chapter.

6 Neurophysiological abnormalities in children with