Evolution Not Revolution: epilepsy in the last ten years

The new millennium ushered in the age of communication and ACNR has spearheaded high quality, accessible neurology. As a keen new consultant, I was delighted to be asked to write a series of articles about various aspects of epilepsy. Now ten years on, the editors in their desperation have charged me with looking back and asking what’s new, acorns, green shoots or grand oaks? There follows an unapologetically personal and perhaps hugely biased mosaic of those areas which have stuck in my mind.

I was grabbed by the identification of fast ripples (100-500Hz) described in animals and then in human intracranial EEG studies, described as “field oscillations composed of hypersynchronous action potentials” [1] possibly mediated via gap junctions, rather than synapses. Ripples may precede the onset of a seizure by a considerable period of time and are blocked by inhibitors of glutamate. They open a new vista in consideration of epilepsy mechanisms and treatment, which has yet to be realised and this reflects the core theme of my article; that, to date, our treatments deal with seizure expression, and not root causes. Of course, the most powerful tool we currently have to look into causes is genetics. Monogenic epilepsies (Table 1) have been identified [2], starting with autosomal dominant frontal lobe epilepsy (ADNFLE). Most of these monogenic epilepsies are disorders of ion channels or neurotransmitters, which is reassuring when considering the mechanisms of action of all the anti-epileptic drugs we possess. We have known since the work of William Lennox over half a century ago that there is a strong genetic component to the idiopathic generalised epilepsies but the search for their genetic basis is proving frustratingly slow. Some genes have been identified underlying monogenic pedigrees which are phenotypically similar to the sporadic diseases. These may inform the mechanisms underlying commoner disorders and we shall know more as the technology improves, and larger numbers of patients and genes can be tested quickly; I hope that, if I am asked to write another article in another ten years, the answer will be very different. Even with monogenic disorders, mapping clinical syndrome to gene defect has not proved altogether straightforward, with significant heterogeneity and a “many-to-many” map. So what influences the variable expression of genetic abnormalities? In some disease it may be the type of genetic abnormality; deletion or point mutation or the locus within the gene or else pre- sumably other variable genes playing a part.

Even now that we know more about ion channels they are becoming predictably more complex. In neonates GABA is an excitatory neurotransmitter [3] – who would have guessed that? This and other maturational changes, including the progressive myelination of the young brain, may go some way to explaining why childhood epilepsy is such a different disease to adult epilepsy with all those strange syndromes that no self- respecting adult neurologist can ever remember. In ADNFLE, seizures may result in a gain in inhibition in a homozygous mouse model. So a traditional view of drugs as uppers and downers of cortical activity, whilst appealing, is clearly simplistic and what may be important is the role of the neurotransmitter in neuronal synchronisation in the light of modified anatomy and physiology [4]. Identifying the gene is only one step, moving on to determining pathophysiology in different patients with different expressions and designing effective and safe drugs, is quite another. The next phase will be slower but therapeutically more rewarding. Genes may be tricky but are perhaps easier to investigate than environment and the clearest example of their interaction is the doubling of the risk of epilepsy in those with a major head injury who also have a family history of epilepsy.

There has been a burgeoning of anti-epileptic medications in the last ten years. But my guess is your clinic is as full as ever it was, even if NHS coffers are considerably emptier. So what do the new drugs offer? A few patients have become seizure-free who would never have been seizure-free before. Some patients have fewer or milder seizures. We can now achieve similar results with fewer adverse effects, less drug interactions and simpler pharmacokinetics. Why AED all-too-often do not work has also been the subject of work in the area of pharmacogenomics, but studies remain at a very preliminary stage. On a more positive note, the SANAD study has given us good quality data showing that lamotrigine, of the drugs tested, had the highest retention rate in focal epilepsy4 and valproate was far and away better in generalised epilepsy than lamotrigine or topiramate [5]. The results are interesting, both in their clinical relevance and also in telling us that focal and generalised epilepsies really are two different diseases and diagnosing them properly is very important. At the same time, the UK register of epilepsy and pregnancy, which collects prospective data on about half of pregnant women with epilepsy in the UK, and whose results are about to be expanded by other major registers, has given us the first robust and clinically useful data on AED and major malformations [6]. Only our data on lamotrigine, carbamazepine and valproate are statistically significant to date and whilst reassuring in the case of the first two, valproate emerges as the bad boy on the block. This combined with preliminary data suggesting that the verbal IQ of children exposed to valproate in utero may be reduced [7], is a major cause of concern for a drug which may be the only option for some women with, for example, severe juvenile myoclonic epilepsy.

When treating our refractory patients, it is as well to bear in mind studies suggesting that the main determinants of quality of life are related to depression and adverse effects of medication and not to seizure frequency [8,9]. Poring over a seizure diary to see if seizures have been reduced by 35% or 50% is not of relevance to many patients but it is the measure used in regulatory studies prior to drug marketing. Smarter studies would be helpful but this needs a change in regulatory requirements. Sometimes the most important manifestation of a disease is not the most obvious; Parkinson’s disease doctors have learned to think about non-motor manifestations and epilepsy doctors must consider their co-morbidities. Some new ones have gained prominence in the last ten years, in particular new memory disturbances such as accelerated forgetting [10];the tendency for memories to drop out over a few weeks, which is not detectable in the short time frame of standard psychometric testing and transient epileptic amnesia [11], which has some similarities to transient global amnesia but episodes are briefer, often first thing in the morning and more prone to recurrence.

The approach to treatment is maturing, parallel with a change in culture across medicine, clinicians and patients partners in seeking goals of key importance to the patient. With this has come improvements in sharing of information, often led by patient groups and charities; we have a representative in our clinic giving out information to patients at each epilepsy clinic. Epilepsy nurses are now a widespread and invaluable resource, acting as a bridge for the dissemination of knowledge into the community. However, some key pieces of our knowledge jigsaw are missing. The natural history of epilepsy is, in my view, still not clear and without this we make guesses, which are only partly educated, in helping patients to make choices. Studies are conflicting but my overall impression is that if you follow patients long enough, many will enter into a remission but many will also relapse, with or without the withdrawal of treatment; for many, epilepsy seems to be chronic relapsing and remitting disease [12]. The suggestion that failure to respond to two AED [13] makes it unlikely that the epilepsy will respond to any drug is widely quoted and is more pessimistic than some studies [14] , but gives a basis for a sensible dialogue with a patient about the pros and cons of the “stick or twist” decision of trying new drugs from amongst the ever-increasing choice available. Sudden unexplained death in epilepsy, long recognised but under-rated is now given the prominence it deserves, thanks largely to the efforts of the charity “Epilepsy Bereaved” and although I still do not know the right way of handling the issue, it is something at the forefront of my mind. Investigation of epilepsy mortality has led to a practical change with a regular search for and occasional discovery of seizure-induced cardiac arrhythmia, particularly in epilepsy patients who develop drop attacks [15].

| Sign up to the ACNR email newsletter here |

High quality structural imaging continues to provide new insights into epilepsy [16]. A more powerful magnet (3T) may find structural abnormalities in 20% of patients with normal standard imaging and voxel-based morphometry can identify subtle changes, not obvious on standard imaging. The neurophysiology purists argue that conventional imaging looks at structure and epilepsy is a disturbance of function, but new methods such as diffusion tensor imaging and tractography may identify changes in connectivity which may be of significance in a disorder which is a disturbance of networks. Tractography is also of value in evaluation for temporal lobectomy, determining the location of Meyer’s loop and so being able to predict those patients more at risk of visual field defects [17]. But the approach is multifaceted and even in patients whose structural MRI is normal [18], investigators have become better at identifying the region of epileptogenesis with a combination of functional imaging and neurophysiology, such that successful outcomes of surgery are increasingly reported in these patients (Figure 1), although it can be argued that the definition of a normal scan depends on the type of scan used. Diffuse brain disease is also not a contraindication to surgery if it can be shown that the epilepsy satisfies conventional criteria of focal onset and good surgical outcomes can be seen in patients with apparently diffuse disorders such as tuberose sclerosis or severe learning disability and even in patients with apparently diffuse electrical abnormalities [19] but focal congenital structural lesions. In malformations of cortical development, the combination of EEG, structural and BOLD MRI is helping to elucidate the relationship of the lesion to the electrical disturbance; it appears that the overlying cortex is the key player in many cases, which has implications for surgical treatment [20]. So it seems that neuroimaging and EEG both have their place; neither should be over-interpreted in pre-surgical evaluation.

The assessment of cognitive function in epilepsy is often in preparation for surgery and for half a century the Wada test has been the gold-standard but this has been questioned with the development of functional neuroimaging techniques, which in some centres have taken over altogether from Wada testing and in others supplement it [21]. Not without risk, distressing for the patient and difficult to interpret, even though the Wada test must be one of the most dramatic investigations in medicine, its passing is a blessing for patients.

Where drugs and resective surgery do not work, it is now commonplace for patients to be considered for palliative surgery in the form of vagus nerve stimulation. I have moved from being something of a sceptic in the technique to a modest supporter on the basis that although it does not make patients seizure-free, it probably does give a useful improvement in seizures in 25%-40% of patients and does so without the cognitive and psychiatric side effects of anti-epileptic drugs which patients find so distressing [22].

So my summary of the last ten years is that we do what we always did and I think we do it a little better and more sensitively with a growing understanding of what is tediously called patient-focussed (other-wise known as good) practice, but we have not had any quantum leap in understanding, which would really allow us to make inroads into the enormous psychosocial morbidity of uncontrolled epilepsy, still the commonest serious neurological disease. We have many new strings to our pharmaceutical bow, but we need a new instrument. Where Alzheimer’s disease modifying drugs are around the corner and MS disease modifying drugs are with us already, epilepsy disease modifying drugs don’t even know where the bus-stop is. So my hopes for the next ten years: firstly, not another drug that controls seizures by blocking electrical activity, although a really good one would do no harm, but one that interferes with epileptogenesis that we can give patients after major head injuries, encephalitis and haemorrhages, which we know carry a high risk of later epilepsy. Arguably, we may have filled in some gaps in the picture left to us by Hughlings Jackson over a century ago but our conceptual framework has advanced only a little and my second hope is of increasing knowledge of the mass action of neurons, what makes them synchronize into pathological networks and how we can interfere with that process. Hopefully knowing the genes will start to inform what goes on inside cells as well as on their surfaces. We need to treat epilepsy and not just seizures. Finally, we need to keep doing epidemiological studies, because without knowing the natural history of the disease we are treating, we are navigators without a map.

References

1. Dzhala VI and Staley KJ. Mechanisms of fast ripples in the hippocampus. J Neurosci 2004:24;8896-906.

2. Helbig I, Scheffter IE, Mulley JC, Berkovic SF. Navigating the channels and beyond: unravelling the genetics of the epilepsies. Lancet Neurology 2008:7;231-45.

3. Mann EO, Mody I. The multifaceted role of inhibition in epilepsy: seizure-genesis through excessive GABAergic inhibition in automsomal dominant nocturnal frontal lobe epilepsy. Current Opinion in Neurology 2008:21;155-60.

4. Marson AG, Al-Kharusi AM, Alwaidh M et al. The SANAD study of effectiveness for carbamazepine, gabapentin, lamotrigine, oxcarbazepine or topiramate for treatment of partial epilepsy: an unblended randomised controlled trial. Lancet 2007:369;1000-15.

5. Marson AG, Al-Kharusi AM, Alwaidh M et al. The SANAD study of effectiveness of valporate, lamotrigine or topiramate for treatment for generalised and unclassified epilepsy: an unblended randomised controlled trial. Lancet 2007:369;1016-26.

6. Morrow J, Russell A, Guthrie E et al. Malformation risks of antiepileptic drugs in pregnancy: a prospective study from the UK Epilepsy and Pregnancy Register. 2006;77:193-98.

7. Meador KJ, Baker GS, Browning N et al. Cognitive function at 3 years of age after fetal exposure to antiepileptic drugs. NEJM 2009:360;1597-605.

8. Boylan LA, Flint LA, Labovitz DL, Jackson SC, Strarner RN and Devinsky O. Depression but not seizure frequency predicts quality of life in treatment-resistant epilepsy. Neurology 2004:62;258-61.

9. Gilliam F. Optimizing health outcomes in active epilepsy. Neurology 2002:58;S9-S20.

10. Blake RV, Wroe SJ, Breen EK, McCarthy RA. Accelerated forgetting in patients with epilepsy. Brain 2000:123;472-83.

11. Butler C and Zeman A. Syndromes of transient amnesia. ACNR 2006:6(4);13-14.

12. Schiller Y. Seizure relapse and development of drug resistance following long term seizure

remission. Archives of Neurology 2009:66;1233-39.

13. Kwan P, Brodie MJ. Early identification of refractory epilepsy. NEJM 2000:342;314-319.

14. Callaghan BC, Anand K, Hesdorffer D, Hauser WA and French JA. Likelihood of seizure remission in an adult population with refractory epilepsy. Ann Neurol 2007:62;382-89.

15. Britton JW, Gearing GR, Benaroch EE, Cascino GR. The ictal brdycardia syndrome: localization and lateralization. Epilepsia:47;737-44.

16. Duncan J. The current status of neuroimaging for epilepsy. Current Opinion in Neurology 2009:22;179-84.

17. Tayoka T, Sakamoto M, Nakagawa H. Diffusion tensor tractography of the Meyer Loop in cases of temporal lobe resection for temporal lobe epilepsy: correlation between postsurgical visual field defect and anterior limit of Meyer Loop on tractography. AJNR 2008 2008:29;1329-34.

18. Cohen-Gadol AA, Wilhelmi BG and Collignon F. Long term outcome of epilepsy surgery among 399 patients with non-lesional epilepsy including mesial temporal sclerosis. J Neurosurg 2006:104;513-24.

19. Tyvaert L, Hawoo C, Kobayashi E et al. Different structures involved during ictal and interictal epileptic activity in malformations of cortical development: an EEG-fMRI study. Brain 2008:131;2042-60.

20. Wyllie E, Lachwani DK, Gupta A. Successful surgery for epilepsy due to early brain lesions, despite generalized EEG abnormalities. Neurology 2007:69;389-97.

21. Baxendale S, Thompson P, Duncan JS. The role of the Wada test in the surgical treatment of temporal lobe epilepsy: an international survey. Epilepsia 2008:49;715-20.

22. Theodore WH and Fisher RS. Brain stimulation for epilepsy. The Lancet Neurology 2004:3;111-18.