It feels that everyone is friendly in New Orleans, one of the most wonderful cities to visit and the location of this year’s American Epilepsy Society meeting, but I came away with a sense of paradox. As I sat in the curiously named Fritzel’s European Jazz Bar, sipping my beer and listening to a mixture of traditional jazz fused with honky-tonk and interspersed with the sophisticated banter of the band leader, on the two TV screens I could choose baseball or a commercial for a monoclonal antibody therapy.

Walking a couple of hundred metres down Bourbon Street took us past young women in shorts, fishnets and tank tops in the unseasonable 5o C, which drew from a local security guard, not surprise but the normalising comment, that he was cold, so how must they feel. The jazz and gentle humour were replaced by the slightly seedy and raucous, just a short distance away. Mixed in with all of this, the French quarter, spared by hurricane Katrina in 2005, has an amazing array of art and craft and very little tourist tat, with not a franchised fast food outlet in sight.

The passion for patient care in the AES was palpable and there is maturity of patient involvement in research and treatment, with numerous publications devoted to patient centred aspects and quality of life. This is a country in which there is orders of magnitude more charitable donation to health care than in the UK. But in the same city, one could hardly walk more than fifty metres in the French quarter without encountering beggars sleeping rough. The paradox of a society where the medical needs of the disadvantaged are met, not by right, but by philanthropy; the choice of the rich whether to help the poor.

This large meeting was an excellent balance between the highly specific, covering diverse basic science and clinical topics and the broadly clinically relevant, in a number of plenary sessions. The first day opened with the role of intracranial EEG in presurgical evaluation, highlighting the sampling bias inherent to the techniques and the importance of using them in selected cases to answer specific questions, such as: “Is epilepsy arising from right or left temporal lobes?” The merits of stereotactic EEG, with its ability to detect epilepsy from the depths were compared to the broader but more superficial sampling of subdural grids, which probably also have higher complication rates. A number of speakers described roughly 50% success rates in operating on non-lesional epilepsy and, although most of these cases will need intracranial EEG, it was pointed out that not all cases will, if there is highly focal surface EEG and concordant semiology, psychology and PET scanning. The session started with a case in whom there was such poor localisation prior to electrode implantation, that it was surprising to me that it was undertaken in the first place, underlining how for some centres, MRI is not quite so dominant in decision making.

In an excellent session entitled “Epilepsy on the consult service” Scott Mintzer described the impact of enzyme inducing anti-epileptic drugs (AED) on cardiovascular health, which can be deleterious. They cause a rise in LDL cholesterol of 0.1mmol/l and independent of their own direct effect, may interfere with the cholesterol lowering effect of statins; the only one they seem not to interfere with is rosuvastatin. Moreover, they alter the metabolism of too many cardiovascular active drugs to list here, but that should be borne in mind, particularly in the older population. The knowledge of AED in pregnancy continues to expand and Page Pennell drew heavily on the latest data from the European Register for her recommendations.1 This study states that daily doses of lamotrigine above 325mg per day have an increased risk of major malformations and that the rate with high dose lamotrigine and higher dose carbamazepine is comparable to low dose valproate ≤650mg. The safest drugs in pregnancy are low dose lamotrigine, levetiracetam and oxcarbazepine. The evidence, in combination with its more favourable pharmacokinetics is sufficient, that in my view, oxcarbazepine should displace carbamazepine in the UK from its current position in treatment of epilepsy and lamotrigine needs to be used somewhat more cautiously in women of childbearing age. The problem of drug levels in pregnancy was addressed and it is clear that as well as the known effects on lamotrigine, many other drugs are affected. A fall in AED levels to below two thirds of pre-pregnancy values is associated with a deterioration in seizure control. However, there is so much variability between women that the consensus is in favour of monitoring for most drugs where possible, including levetiracetam. In the UK, levels are not universally available.

Jeanne Young gave a dermatologist’s perspective on skin rashes with AED and helpfully delineated three groups. The commonest is a mild morbilliform reaction with no systemic features and is such a common presentation, even in those not on medication, that if clinically indicated, the drug might safely be re-started to see if the rash was indeed drug related. The more serious drug rash with eosinophilia and systemic, symptoms (DRESS) and the very worrying Stevens-Johnson syndrome or toxic epidermal necrolysis necessitate stopping drugs with varying degrees of urgency. But what is the risk of a rash with another drug under those circumstances? The main paper here is old2 but the key message is that the risk is high. Levetiracetam is relatively rash-free. And, if the patient is seizure-free, what is the risk that seizures will recur with switching to a new drug? Approximately 15% according to two studies.

There was an excellent session on epilepsy and memory. I learned how there is increasing evidence of the importance of sleep spindles, in particular fast ripple components, in linking thalamus and cortex in the formation of memory. At a molecular level BDNF expression may have a role in experimental models and DNA methylation is an important component with inhibitors adversely affecting memory. John Duncan gave an excellent review of the work using fMRI to delineate activation patterns in those with left or right mesial temporal sclerosis to predict the cognitive outcome of epilepsy surgery,3 but there are still centres using the Wada test, which also featured in a couple of posters.

Sudden unexplained death in epilepsy (SUDEP) has been a focus for over twenty years but research has been largely descriptive. Earlier studies suggested it was less common in children but a more methodologically robust study from Sweden refutes that view.4 Three main mechanisms have been postulated; arrhythmias, apnoea and suppression of brain electrical activity. In sessions devoted to this topic, hypotheses underlying these mechanisms were described. They include 5-HT3 and 5-HT4 receptors implicated in a mouse model of fatal apnoea, which are altered by existing medications  acting on these systems; hypoperfusion of the brain blocked by COX-2 inhibitors and genetic abnormalities with overlap with long Q-T syndrome. It is increasingly recognised that patients with these cardiac channelopathies may also suffer seizures, which had previously been attributed to syncope but provide a means for understanding interaction between seizures and cardiac events which may be fatal. Whilst heart rate and regularity are easy to measure in seizures, vascular reactivity and alterations in blood pressure are not and may be another relevant mechanism, again amenable to existing medications. Suppression of brain activity seems to be more likely in patients with seizures with a longer tonic phase, providing a potential means of predicting who may be more at risk.

This leads on to the question of seizure detection and prediction. A number of commercially available devices are postulated to identify when patients are experiencing a seizure. They are based on a range of technologies, including accelerometry, HR changes and changes in skin conductance. None is yet particularly reliable but one has been licensed by the FDA. In a fascinating session, Robert Fisher speculated where epilepsy would be in the next ten years. He predicted that it would be the decade of AI, with this being applied to seizure detection devices; to diagnostic algorithms and to treatment algorithms for patients with epilepsy. He predicted that clinicians would mostly be needed for the more global decisions, education and patient interaction. I am not sure if I am pleased or not that I shall not be practising to see it.

Cannabidiol (CBD) is the drug of the moment, although only licensed for Dravet’s and Lennox Gastaut syndromes. Tyler Gaston guided us through the differences between pharmaceutical CBD (pharmacologically consistent) and artisanal cannabis products (highly inconsistent). In terms of efficacy, CBD really is very little different from other drugs that have been trialled and there is less fanfare around fenfluramine in LGS, which is just as effective. CBD’s anti-epileptic activity appears to be at a range of sites unrelated to the cannabinoid receptor. Life would be so much easier if it were just called something else!

A session on febrile status epilepticus described how the FEBSTAT study5 has shown that prolonged febrile seizures are associated with hippocampal sclerosis, although the relationship appears complicated, for example more on the right than the left. This provides an opportunity to intervene to prevent epilepsy and work on mechanisms has identified the potential importance of interleukins and neuronal restrictive silencing factor, which are potentially mediated through alterations in micro RNA’s. A possible early marker of risk is increased T2* signal in the hippocampus, a sequence not standard in epilepsy protocols, but it might become so, if an intervention were possible.

Finally, no conference would be complete without some genetics. One of my practical enigmas is in how many of my patients with refractory epilepsy since childhood who are in my adult clinic, and who have normal imaging and no clear reason for their epilepsy, should I be requesting genetic testing? The answer is apparently all of them. I am not sure Matthew Hancock or particularly Philip Hammond would agree. Still, I prefer working in healthcare where access is relatively uniform even if that may not always match the very best in some less equal systems.


  1. Tomson T, Battino D, Bonizzoni E, et al. Comparative risk of major congenital malformations with eight different antiepileptic drugs: a prospective cohort study of the EURAP registry. Lancet Neurol 2018;17:530-8. doi: 10.1016/S1474-4422(18)30107-8
  2. Hirsch LJ, Arif H, Nahm EA, et al. Cross-sensitivity of skin rashes with antiepileptic drug use. Neurology 2008;71:1527-34. doi: 10.1212/01.wnl.0000334295.50403.4c
  3. Sidhu MK, Stretton J, Winston GP, et al. Memory network plasticity after temporal lobe resection: A longitudinal functional imaging study. Brain 2016;139:415-30. doi: 10.1093/brain/awv365
  4. Sveinsson O, Andersson T, Carlsson S, Tomson T. The incidence of SUDEP: A nationwide population-based cohort study. Neurology 2017;89:170–177. doi: 10.1212/WNL.0000000000004094
  5. Lewis DV, Shinnar S, Hesdorffer DC, et al. Hippocampal sclerosis after febrile status epilepticus: The FEBSTAT study. Ann Neurol 2014;75:178-85. doi: 10.1002/ana.24081