Cortical hyperexcitability has been reported consistently in transcranial magnetic stimulation (TMS) studies of focal epilepsy (Cantello et al., 2000; Werhahn et al., 2000; Hamer et al., 2005; Badawy et al., 2007). We recently found evidence to suggest that these changes are progressive in nature and extend to involve the unaffected hemisphere in patients with chronic refractory focal epilepsy over time (Badawy et al., 2013). Temporal lobe epilepsy (TLE) is the most common focal epilepsy and perhaps the most homogeneous, and therefore provides the best available model for further elucidation of the nature of electrophysiologic changes associated with epilepsy in vivo. TLE is also one of the most common forms of focal epilepsies associated with intractable seizures (Wiebe, 2000). In many of these patients, seizures are refractory from the onset (Kwan & Brodie, 2000), perhaps providing evidence for an underlying role for pharmacogenetic interactions (Loscher et al., 2009). In others, drug resistance may develop during the course of epilepsy after an initial positive response to antiepileptic drugs (AEDs). Furthermore, seizures may persist after the surgical removal of the epileptogenic focus despite what appears to be precise localization on both scalp and even invasive electroencephalography (EEG; Tellez-Zenteno et al., 2005, 2010). On many occasions, surgery is unsuccessful owing to failure of removal of additional regions of seizure origination that were not appreciated at the initial evaluation; however, this is not always the case. This suggests that there are multifaceted disturbances associated with epilepsy, which in some cases persist even if the origin or “generator” is removed.
In the current study we used TMS to assess cohorts with TLE with different levels of seizure control (medical and postoperative). We hypothesized that the pattern of disturbance in cortical excitability will depend on the clinical course of the disease and speed/degree of responsiveness to treatment. We also anticipated to find a difference in the cortical excitability profile depending on whether the management was medical (intact generator, but modulated by medication) or surgical (the generator is gone).
- Top of page
In the current study we show specific cortical excitability changes in different cohorts of patients with TLE at various stages of the disease. The primary difference is a significant interhemispheric difference in patients with TLE at early onset prior to starting AEDs, with a net increase in cortical excitability in the hemisphere ipsilateral to the seizure focus, whereas the unaffected hemisphere remains normal. According to our results, the course of the illness changes this pattern into one of two archetypes, both ultimately resulting in the loss of interhemispheric differences. Refractory seizures result in increased excitability in both hemispheres even though the focus is unilateral and seizure freedom results in reduction of the hyperexcitability that was present prior to treatment whether medical or surgical.
The refractory groups were taking a significantly higher number of AEDs compared to the seizure-free groups. Despite that, cortical excitability was still higher in all refractory groups. It is known that AEDs have a prominent effect on TMS measures (Ziemann, 2004). This effect is most marked on MT and would thus explain the increase in MT compared to controls that was observed in the current study as well as previous studies on patients with chronic focal epilepsy (Cantello et al., 2000; Hamer et al., 2005). However, this effect would be expected to increase as a function of the number of AEDs used (Cantello et al., 2000), but we did not find this to be the case. Values for MT were similar in both the refractory and seizure-free groups, even though the number of AEDs used by the refractory groups was much higher. The results suggest that patients with refractory seizures are resistant to the effect of AEDs. Conversely, although cortical excitability was normal or near normal in both hemispheres in patients who were seizure-free from onset, it remained mildly elevated even in the unaffected hemisphere at the 250 msec ISI in patients where seizure freedom was more difficult to achieve. These changes occurred irrespective of the specific AED(s) used. This suggests that despite what is known about particular mechanisms of each drug; whether it works on specific channels or receptors, a common mechanism possibly exists at the level of interneuronal interactions and synapses. This is likely due to the complexity of the alterations that are associated with the epileptic process. These disturbances cause widespread functional changes that extend beyond the epileptic focus to involve the unaffected hemisphere. This interneuronal interaction is probably affected by many physiologic and acquired factors, all of which determine the course and prognosis of epilepsy.
With successful surgery, cortical excitability decreased in both hemispheres in patients who became seizure-free. Postoperative changes in the contralateral hemisphere were also reported previously, in the form of increased intracortical facilitation (the late part of the short ISI curve) in one study (Kamida et al., 2007), and reduced intracortical facilitation together with shortening of the cortical silent period (denoting increased cortical excitability) in another (Lappchen et al., 2008). Another study also found increased interhemispheric inhibition on stimulation of the unaffected hemisphere postoperatively (Lappchen et al., 2011). Such contradictory findings may be due to sample size; however, they may also be due to the fact that each of these TMS parameters reflects activity in different intracortical circuits. Intracortical facilitation is most likely mediated by excitatory interneuronal circuits, possibly glutamate mediated (Ziemann, 2003); the early part of the short ISI curve is thought to be mediated by γ-aminobutyric acid (GABA)A (Boroojerdi, 2002), and long ISIs reflect GABAB mechanisms (Florian et al., 2008). The silent period may represent GABAB circuits (Ziemann, 2003). It appears that different circuits are modulated in different ways following surgery. This perhaps depends on the operative procedure and the level of cortical hyperexcitability in that hemisphere prior to surgery, as a net increase in excitability was found mainly in the study with patients without cortical hyperexcitability changes in the contralateral hemisphere prior to surgery (Kamida et al., 2007).
Our findings show that the characteristic pattern of cortical hyperexcitability associated with TLE is influenced by the natural course of the disorder. The results indicate that this abnormality can be reversed with successful epilepsy treatment (medical or surgical), which is interesting as there is a structural difference between medication- and surgical-related seizure freedom. With medication the presumed “generator” remains, but its activity is “damped down,” whereas with surgery the generator is removed. Although it is not clear whether this effect is due to a change within the brain's predisposition to generate seizures, or simply the loss of the secondary consequences of seizures, it seems the cortical hyperexcitable network is further reinforced by recurrent seizures. Successful treatment alters this relationship; when seizures stop regardless of whether this is the effect of medication or surgery functional reorganization within the previously hyperexcitable circuits takes place. Alternatively, synaptic reorganization following disruption of the epileptic network may elevate the threshold required for seizure generation and thus can also be the reason that seizures stop. Therefore, although we cannot ascertain that when seizure freedom is achieved the “abnormal” circuits are actually gone or simply inactivated via improved control (medical or surgical), our results suggest that TMS can provide a noninvasive insight into the dynamic nature of the condition at various stages.
The next step is to investigate whether the findings represent a potential objective surrogate measure to seizure control of the underlying abnormality within individuals. This would require large scale possibly multicenter studies to overcome the known high interindividual variability in TMS measures. Such studies would further benefit if the patients were followed with repeated TMS throughout the course of the disease. This is because it is known that TMS measures do not show significant intersession or interinvestigator variability (Mills & Nithi, 1997; Boroojerdi et al., 2000; Maeda et al., 2002; Badawy et al., 2012); therefore, any differences observed in the repeat TMS study(s) will be attributable to underlying electrophysiologic changes caused by the disorder.