Epilepsy Surgery and Other Options When Medication Fails

Epilepsy surgery and the other options when medication fails

There is a treatment in epilepsy that can stop seizures permanently, is suitable for roughly a third of the people who have drug-resistant epilepsy, achieves seizure freedom in an estimated 62 per cent of those who receive it, and which some Australians wait up to 17 years to be offered.

That treatment is surgery, and the gap between what it can do and how often it is used is the largest single missed opportunity in Australian epilepsy care.

Australian research into why found the answer was not clinical. It was that GPs frequently did not recognise drug resistance, that referral pathways to tertiary centres were restricted, that patients did not know the option existed, and — strikingly — that people were frightened of surgery because nobody had given them accurate information about its risks, benefits and effects on their lives. Fear of an unexplained brain operation is entirely rational. It is also why people decline an assessment that would not have committed them to anything.

This guide covers what the assessment involves, the operations and what they can and cannot achieve, and the non-surgical options — vagus nerve stimulation, deep brain stimulation, the ketogenic diet and cannabidiol. It is written to be the accurate information the Australian research found missing. It is not a recommendation; the decision belongs to you and a tertiary epilepsy team.

The threshold for all of it: two adequately trialled, appropriately chosen medications without seizure freedom. See drug-resistant epilepsy.

The assessment, which is not the operation

The thing most people do not grasp: a surgical assessment is an investigation, not a commitment. Roughly two-thirds of people assessed turn out not to be candidates — and that is a useful outcome, because it converts an open question into a settled one and redirects effort to the treatments that might help.

What the workup involves, over months:

Video EEG monitoring. An admission of days with continuous EEG and video, usually with medication reduced to provoke events, to establish where the seizures actually start — and first, whether they are epileptic at all. Approximately one in five people referred for drug-resistant epilepsy turn out to have functional or dissociative seizures. See functional and dissociative seizures.

An epilepsy-protocol MRI. Thinner slices, specific sequences, oriented to the structures that matter. This finds lesions that general-protocol scans miss, and a visible lesion substantially improves surgical prospects.

Neuropsychological assessment. Detailed testing of memory, language and other functions, which does two things: it helps localise the epilepsy, and it predicts what an operation might cost cognitively.

PET and SPECT, which look at metabolism and blood flow to help localise.

MEG, in some centres.

Functional MRI and sometimes a Wada test, to establish which hemisphere handles language and memory before operating near those areas.

And in some cases, intracranial monitoring — electrodes placed on or in the brain surgically, for days of recording, when scalp EEG cannot localise precisely enough. This is itself an operation and it is how borderline cases become operable.

The multidisciplinary meeting, where neurologists, neurosurgeons, neuropsychologists, neuroradiologists and nurses review everything together and decide what, if anything, to offer.

What you should be told at the end: whether surgery is possible, which operation, the realistic chance of seizure freedom for your particular case, the specific risks given where the surgery would be, what the recovery involves, and what happens to your medication afterwards. If those numbers are not offered, ask for them. Generic statistics are a starting point; your own numbers are what the decision rests on.

The operations

Resective surgery — removing the seizure focus. The main option and the one with the best outcomes.

Anterior temporal lobectomy is the commonest and best-studied epilepsy operation, for temporal lobe epilepsy, which is the commonest drug-resistant focal epilepsy in adults. Where the seizures arise from one temporal lobe with hippocampal sclerosis on MRI, outcomes are genuinely good — this is the scenario behind the strongest seizure-freedom figures. Randomised evidence has shown surgery substantially outperforming continued medication in this group.

Lesionectomy, removing a specific abnormality — a cortical malformation, a cavernoma, a low-grade tumour, old scarring.

Extratemporal resection, from the frontal, parietal or occipital lobe, generally with somewhat lower seizure-freedom rates than temporal surgery.

Disconnection procedures — interrupting spread rather than removing the focus. These are palliative: the aim is fewer or less severe seizures rather than cure.

Corpus callosotomy, dividing the connection between the hemispheres, used mainly to stop atonic drop attacks in severe epilepsies. For somebody sustaining facial and dental injuries from drop attacks several times a week, this can be transformative even without seizure freedom.

Hemispherotomy or hemispherectomy, disconnecting one hemisphere, used in children with severe epilepsy arising from one damaged hemisphere. The outcomes in well-selected children are far better than the description suggests, because the brain was already not using that hemisphere productively.

Laser interstitial thermal therapy and radiofrequency ablation, minimally invasive techniques available in some centres, destroying a small target through a probe rather than open surgery.

What the risks actually are. They depend entirely on location, which is why generic figures are of limited use. The general categories: infection, bleeding, the risks of any anaesthetic, and — the ones specific to epilepsy surgery — memory decline after temporal surgery, particularly where the operated side was carrying memory function, visual field loss after some temporal and occipital operations, which can affect driving eligibility independently of seizures, and language difficulty after dominant-hemisphere surgery. The neuropsychological assessment exists to predict these, and this is precisely where the candid conversation Australian researchers found missing needs to happen.

And after surgery: medication continues, usually for at least a couple of years, and is reduced cautiously if seizures stay away. Surgery is not an exit from epilepsy care.

Vagus nerve stimulation

A device implanted under the skin of the chest, with a lead to the vagus nerve in the neck, delivering regular stimulation.

What it does. Reduces seizure frequency. It does not usually produce seizure freedom, and anybody told it will is being oversold. The realistic expectation is a meaningful reduction in a proportion of people, with benefit increasing over the first couple of years.

Who it is for. People with drug-resistant epilepsy who are not candidates for resective surgery — frequently those with multifocal or generalised epilepsy.

The practical features: a magnet can be swiped over the device to deliver extra stimulation when a seizure starts or an aura is felt, which some people find genuinely useful, and newer devices can respond to the heart-rate rise that accompanies some seizures.

Side effects: voice change or hoarseness during stimulation, throat discomfort, cough, and sometimes breathing changes during sleep. Mostly dose-related and adjustable.

And it is a device, meaning battery replacement every several years and MRI considerations.

Deep brain stimulation and responsive neurostimulation

Electrodes implanted within the brain, most often targeting the anterior nucleus of the thalamus, connected to a generator.

What it does. Reduces seizure frequency in drug-resistant epilepsy, with the benefit growing over years in the long-term follow-up studies. Like vagus nerve stimulation, it is a reduction treatment rather than a cure.

Who it is for. People with drug-resistant focal epilepsy who are not resective surgery candidates.

Availability in Australia is limited to a small number of centres, and arrangements for funding and access change, so this is a question for a tertiary epilepsy centre rather than something to research independently.

Responsive neurostimulation, which detects abnormal activity and delivers stimulation in response, is established in some countries and its Australian availability differs — again, a question for the centre.

The ketogenic diet

A genuine medical treatment that sounds like a lifestyle choice, which is why it gets dismissed.

What it is. A strictly calculated high-fat, very low-carbohydrate diet that shifts the body into ketosis. The classical ketogenic diet, the modified Atkins diet and the medium-chain triglyceride diet are variants of differing strictness.

What it does. It has real efficacy in drug-resistant epilepsy, and particularly strong evidence in specific childhood epilepsies — Dravet syndrome, Lennox-Gastaut syndrome, infantile spasms, and it is the first-line treatment for glucose transporter type 1 deficiency and pyruvate dehydrogenase deficiency, where it is not an adjunct but the treatment.

Who it is for. Primarily children with drug-resistant epilepsy, and increasingly adults, where the modified Atkins diet is more practical than the classical version.

What it requires, and this is the part that gets underestimated: a specialist dietitian and medical supervision, not a recipe book. The Epilepsy Foundation’s guidance is explicit that these diets should only be used under medical supervision. It involves weighing food, monitoring ketones, supplementing vitamins and minerals, watching growth in children, and managing side effects including constipation, kidney stones, raised cholesterol and reflux. Done unsupervised it is both ineffective and unsafe.

And it is hard. Families describe it as a substantial undertaking, particularly for school-aged children. For the epilepsies where it works, it is worth it; the honest framing is that it is a treatment with a real adherence cost.

Cannabidiol, and the thing to be clear about

Cannabidiol (Epidyolex) is a pharmaceutical-grade CBD preparation approved for specific severe childhood epilepsies, including Dravet syndrome and Lennox-Gastaut syndrome, where randomised trials showed reductions in seizure frequency. It is a real medication with real evidence in those indications.

It is not the same thing as over-the-counter CBD oil, and the distinction matters more than almost anything else in this guide’s comments section. Consumer CBD products vary enormously in content, are not standardised, have not been trialled for epilepsy, and can interact with anti-seizure medications — cannabidiol raises clobazam levels, for instance, which is managed in a clinical setting and not managed at all when somebody buys a bottle online.

Availability and subsidy arrangements in Australia change, and anything written about them dates quickly. Ask your neurologist or epilepsy centre what currently applies to your situation.

And the general principle for any complementary treatment: tell your neurologist what you are taking. Several supplements and herbal products interact with anti-seizure medications or lower the seizure threshold, and the risk is not in the product being taken but in it being taken secretly.

How to get assessed

Say the sentence. “Two anti-seizure medications have been properly trialled and my seizures are continuing. I would like a referral to a comprehensive epilepsy centre, including assessment for surgery.”

If your GP cannot refer directly, ask for referral to a neurologist specifically for onward referral to a tertiary centre, and say so in the request.

Bring the medication list — every drug, the highest dose, the duration, why it stopped.

Bring the seizure diary and a video.

Ask for the MRI to be repeated on an epilepsy protocol if the last one was general or years old.

Ask what your own numbers are, not the generic ones, once an assessment is done.

And ring a free specialist service before and after. Epilepsy Action Australia on 1300 37 45 37 offers telehealth with an epilepsy nurse; the National Epilepsy Support Service on 1300 761 487 is the Epilepsy Foundation’s line. Both will tell you where the tertiary centres are, and both will talk through what an assessment involves without any stake in the answer.

Questions worth asking before you decide

Australian research found that people declined surgical assessment largely because nobody had given them accurate information. These are the questions that produce it, and a tertiary team should be able to answer every one.

“What are my numbers, not the general ones?” The chance of seizure freedom varies enormously by operation, location and whether there is a visible lesion. Temporal lobe epilepsy with hippocampal sclerosis sits at one end; lesion-negative extratemporal epilepsy sits at the other. Ask for your figure, and ask what it is based on.

“What exactly would be removed or disconnected, and what does that part of the brain do?”

“What is the risk to my memory, my language and my vision?” These are the epilepsy-specific risks, they depend on location, and the neuropsychological assessment exists to predict them. Ask what the assessment showed and what it implies.

“Would I lose any visual field, and would that affect my licence?” A genuine and under-discussed consequence — some temporal and occipital operations cause a visual field defect that can affect driving eligibility independently of seizures.

“What happens to my medication afterwards?” The answer is usually that it continues for at least a couple of years. People who expect to stop immediately are disappointed.

“What does recovery actually involve?” Hospital stay, time off work, driving restrictions, and what the first three months feel like.

“What is the chance this makes things worse?”

“What happens if I do nothing?” The honest comparison is not surgery versus perfect health. It is surgery versus continuing drug-resistant epilepsy, with its own risks — injury, SUDEP, cognitive effects of ongoing seizures, and the life the seizures are already costing.

“If I am not a candidate, what else is there?” Vagus nerve stimulation, deep brain stimulation, dietary therapy, newer medications, and trials.

“Can I speak to somebody who has had this operation?” Both Australian epilepsy organisations can frequently arrange peer contact, and it answers things no clinician can.

And take somebody with you. These appointments contain more information than anybody absorbs alone, and a second person remembers the half you missed.

Clinical trials, and how to think about them

For people with drug-resistant epilepsy who are not surgical candidates, trials are a legitimate part of the landscape, and they are worth approaching with clear expectations.

What is being trialled. New anti-seizure medications, new stimulation devices and targets, gene therapies for specific genetic epilepsies, and in some cases new surgical and ablation techniques.

Where to find them. The Australian New Zealand Clinical Trials Registry and ClinicalTrials.gov list what is recruiting, but the realistic route is through a tertiary epilepsy centre — the centres run the trials, and being known to one is how people get offered a place. This is an additional argument for getting into a comprehensive centre even if surgery turns out not to be an option.

The honest framing. A trial is research, not treatment. You may receive a placebo or an inactive comparator. The drug may not work. There will be extra visits, extra monitoring and extra paperwork. Being in a trial is a contribution to knowledge that may also benefit you, and anybody presenting it as a last-chance treatment is misrepresenting it.

What is genuinely good about them, though, and it gets understated: trial participants are monitored unusually closely, by people who specialise in difficult epilepsy, with frequent review. Several people describe the surveillance itself as the most useful part, regardless of what the investigational treatment did.

Questions worth asking: what is the trial testing and what phase is it; what are the chances of receiving the active treatment; what happens if it works and the trial ends; what are the known risks; how many extra appointments and how far will I travel; are travel costs covered; can I withdraw at any time, and what happens to my usual treatment if I do.

And the thing to be careful of. Treatments offered outside a registered trial, for payment, for drug-resistant epilepsy, are worth extreme caution. Stem cell clinics and unregistered cannabis products marketed for epilepsy both exist and neither has the evidence they imply. Ask your neurologist before paying anybody anything.