Starting antiepileptic medication after a first unprovoked seizure

While 10% of the population will have a single seizure in their lifetime, the challenge is determining for which patients the risk and repercussions of recurrent seizure are great enough to warrant starting antiepileptic therapy. Although immediate treatment after a first unprovoked seizure may decrease seizure risk over the next few years, there is no evidence that early treatment improves longer-term prognosis.

Furthermore, such treatment comes at the expense of side effects and once antiepileptic medication is initiated there is reluctance to withdraw it. Therefore, current practice uses a treatment threshold of about 60% risk of recurrence over 10 years, based on the recurrence risk following two seizures defined by Hauser et al.. This threshold can be reached after one seizure with the presence of certain features: epileptiform EEG, abnormality on brain imaging, prior brain insult, and/or nocturnal onset, viewed in recent AAN guidelines by Krumholz et al.2

However, weighing an individual patient’s health status and lifestyle is also important. To that end, Bao et al. constructed a Markov model to compare immediate vs. delayed antiepileptic treatment following a first unprovoked seizure for three patients of differing seizure recurrence risks and quality of life ramifications3. Interestingly, this model demonstrated a slight benefit in treating a first unprovoked seizure patient with good health and no classic risk factors (immediate treatment 19.40 quality adjusted life years vs. delayed treatment 18.95 QALYs). The authors suggest a threshold recurrence rate of 38% over 10 years may be more appropriate for antiepileptic drug administration- a rate not dissimilar from the estimated recurrence rate of 21-45% over 2 years for all-comers with a first unprovoked seizure2.

Though intriguing, the utility of a slight gain in QALYs is difficult to appraise. This is particularly true as these QALYs estimates have a large range of plausible values. It would also be helpful to consider the monetary costs associated with immediate vs. delayed treatment to better explore the value of shifting practice towards earlier therapy. Antiepileptic medications can be exceedingly costly, up to several hundred dollars per month, and duration of treatment is commonly a lifetime. Furthermore, the appropriate threshold for AED treatment is unlikely the same for every patient; therefore, shared decision making is likely better than a one size fits all approach.

This study does highlight the great need for more sophisticated risk stratification for an individual epilepsy patient. Our estimates of recurrence and treatment risks should be informed by specific details: a named brain lesion and/or epileptiform pattern, the efficacy and side effect profile associated with a particular medication, and perhaps more clinical details such as seizure semiology. With growth of informatics capabilities, hopefully we can soon provide more tailored advice and personalized treatment to patients presenting with a first seizure.

  1. W. Allen Hauser, Stephen S. Rich, Ju R.-J. Lee. Risk of Recurrent Seizures after Two Unprovoked Seizures. NEJM 1998;338:429-434
  2. Allan Krumholz, Samuel Wiebe, Gary S. Gronseth. Evidence-based guideline: Management of an unprovoked first seizure in adults Report of the Guideline Development Subcommittee of the American Academy of Neurology and the American Epilepsy Society. Neurology 2015;84:1705-1713
  3. Erik L. Bao, Ling-Ya Chao, Peiyun Ni. Antiepileptic drug treatment after an unprovoked first seizure A decision analysis. Neurology Epub 2018 Sept 12


Chloe Hill

Chloe Hill

Chloe E. Hill, MD, MPH, Assistant Professor, University of Michigan, Michigan Medicine, Brighton Center for Specialty Care

More Posts