Urgent need to improve long-term outcomes for newly diagnosed epilepsy patients

Epilepsy affects approximately 250,000 people in Australia, and about 68 million people worldwide. In Australia it's the second most common neurological disease. But despite the wide availability of anti-epileptic drugs – AEDs – almost one-third of those newly diagnosed don't get immediate treatment.

Why? Anti-epileptic drugs, also called anticonvulsants, are usually the first choice of treatment for seizure management. However, more than 80 per cent of patients with epilepsy live in economically disadvantaged or poor countries, predominantly with no treatment.

This phenomenon, known as "treatment gap" in epilepsy, is generally attributed to socio-economic factors, including lack of access to and cost of trained clinicians and the AED therapy itself.

Superstitious and cultural beliefs, and the predominance of traditional treatments are also factors.

But the treatment gap hasn't been well studied in high-income countries such as Australia. Understanding treatment decisions in these patients and the reasons behind them is crucial for optimal care.

Our team performed a longitudinal cohort study (gathering data over a long period from the same group of people) in specialist hospital seizure clinics in Western Australia. We observed in this well-resourced, specialist clinic setting that more than 16 per cent of patients with newly diagnosed epilepsy weren't recommended treatment at the time of diagnosis, and 15 per cent declined therapy, resulting in more than 30 per cent of patients being untreated at diagnosis.

Most patients who were initially untreated were started on treatment later after further seizures.

Patients who were found to have epilepsy through neuroimaging suffered more seizures, were older and, perhaps counterintuitively, lived in lower socio-economic areas were more likely to start on an anti-epileptic drug.

Interestingly, epileptiform EEG findings (where electrodes were placed on the patient’s scalp to measure electrical activity) appeared to play a lesser role than neuroimaging abnormalities in decisions about when to start treatment. The reason for this is unclear, but might reflect the differences in “weighting” ascribed by neurologists to different investigations in the treatment decision process.

The decision to start or withhold treatment reflects the complex interplay between factors perceived to influence the predicted risk of seizure recurrence and personal factors. When patients have doubts about the need for treatment, it's important for neurologists to clarify any misunderstanding.

Does anti-epileptic drug treatment actually help?

Back in 2000, a now widely cited study by Kwan and Brodie first reported that anti-epileptic drug therapy could render about two-thirds of patients seizure-free.

Anti-epileptic drugs vary in their activity against different seizure types, and the discovery and development of newer drugs since the 1990s has greatly expanded the range of treatment options.

During the past two decades, more than two dozen new recognised anti-epileptic drugs have become available, and there's increased popularity in using newer anti-epileptic drugs as initial treatment for epileptic seizure.

Our study expanded and extended the 2000 study. We analysed 1795 adolescent and adult patients with newly diagnosed and treated epilepsy in Glasgow, Scotland, recruited since the 1980s.

The study demonstrated increased use of many new anti-epileptic drugs with differing mechanisms of action over the past two decades. The long-term prognosis of anti-epileptic drug treatment, however, hasn't improved, and remained at 64 per cent.

The probability of achieving seizure freedom diminishes with each unsuccessful anti-epileptic drug.

The study highlighted the urgent need for ways to improve long-term outcomes in patients with newly diagnosed epilepsy. Patients with high risk of drug-resistant epilepsy should be considered early for non-drug therapies.

The study highlighted the urgent need for ways to improve long-term outcomes in patients with newly diagnosed epilepsy.

While some modern anti-epileptic drugs have novel anti-seizure mechanisms, their increasing use didn't seem to have improved overall long-term seizure control. This may be attributed to deficiencies in the pre-clinical and clinical strategies of anti-epileptic drug development.

This finding also suggests that the modern anti-epileptic drugs still lack the capacity to rectify the underlying processes in the brain and reverse a gradual slide towards epilepsy.

Most importantly, the study also provides further evidence that current anti-epileptic drugs are seizure-suppressing and not disease-modifying. A new treatment approach is needed to improve prognosis of epilepsy.

More comprehensive predictive models are needed to allow early identification of patients who have a high risk of drug-resistant epilepsy and hence deciding on non-drug interventions.

‘Patient-centred’ epilepsy care

To formulate a more effective management plan for epilepsy, it's crucial to understand the long-term clinical courses and patterns of responses to anti-epileptic drugs.

The dynamic – or changeable – response patterns of anti-epileptic drug treatments in epilepsy is currently poorly understood. For a newly diagnosed patient, an anti-epileptic drug will be selected mainly based on reported seizure type or epilepsy syndrome, with switching or addition of another drug if therapy fails.

This current "disease-based" approach delays effective non-drug interventions. Our current project aims to create the first predictive scoring system in epilepsy management, which could potentially disrupt the current treatment pattern for epilepsy.

By applying machine learning techniques on population epilepsy databases and individual seizure clinic registries around the world, the study will provide this insight into dynamic response patterns in epilepsy and create the first scoring system that can predict the prognosis of anti-epileptic drug treatment based on pre-treatment medical histories.

Once a reliable system can be established, instead of being the last resort, non-drug interventions can become the first-line option for some patients with new-onset epilepsy. The shift from "disease-centred" to more personalised healthcare will greatly improve the prognosis and wellbeing of patients with epilepsy.