N of 1 trials in epilepsy: A systematic review and lessons paving the way forward

What is the main takeaway?

This systematic review examines the application of N of 1 trials in epilepsy management. N of 1 trials are prospective, single-patient crossover studies with repeated cycles of active and control interventions. The review identified five studies involving patients with focal seizures, absences, or cognitive impairment with electrographic discharges. Interventions included established or investigational antiseizure medications, off-label drugs, neurostimulation, or lifestyle modifications. The studies demonstrated strengths such as personalised treatment dosages and symptom-specific patient-reported outcomes, while limitations included minimal reporting of baseline characteristics and seizure burden.

What is the real-world impact?

Implementing N of 1 trials in epilepsy offers a pathway to personalised medicine, particularly for patients with rare epilepsies where large-scale randomised controlled trials are challenging. For the pharmaceutical industry, this approach can streamline the evaluation of anticonvulsants, accelerating the development of targeted therapies. In digital health, these trials can be conducted with platforms that monitor patient-specific data, facilitating real-time adjustments to treatment plans. Medical device companies can employ N of 1 trials to evaluate the efficacy of neurostimulation devices on a case-by-case basis, refining device settings for optimal outcomes. Clinical research organisations and universities can adopt this methodology to conduct rigorous, patient-centred research, advancing the understanding of epilepsy treatments and informing clinical guidelines.

What do we think about it?

This is a thoughtful article with in-depth analysis, important recommendations and broad applicability. The authors recommend that future N of 1 trials of anticonvulsants should take into account baseline seizure frequency, should apply statistical models suited to capture seizure frequency changes reliably and make predefined interim assessments. Baseline seizure frequency is clearly an important variable to compare to seizure frequency on the intervention. Statistical models that capture seizure frequency changes reliably, notably Bayesian models, are also of interest to N of 1 trials in other contexts because there are other conditions where frequency is an important outcome, such as migraine, urinary incontinence, and gastro-esophageal reflux. These statistical models could also be applied more widely to other neurological conditions and more broadly to other chronic stable conditions suitable for N  of  1 trials. 

Who are the authors & where can I read the paper?

The study was conducted by Victoria M Defelippe, Eva H Brilstra, Willem M Otte, Helen J Cross, Finbar O'Callaghan, Valentina De Giorgis, Annapurna Poduri, Holger Lerche, Sanjay Sisodiya, Kees P J Braun, Floor E Jansen, and Emilio Perucca. The paper can be accessed here.