Volatile Non-Invasive Biomarkers of Epileptic Seizures

Our funders

Approximately 20-30% of patients are misdiagnosed with epilepsy [Nat Rev Dis Primers. 2018, 4, 18024]. Inaccurate diagnosis can result in unnecessary treatments, with potential adverse effects as well as social issues. Currently, the gold standard method for diagnosing epilepsy is the ictal video electroencephalogram (EEG) [Expert Rev Neurother. 2015, 15, 425-444]. However, this investigation is not readily available in emergency settings. Another unmet requirement for the management of epilepsy is the unpredictability of seizure onset, posing daily life challenges for people living with epilepsy. It can represent a risk for their safety, and also considerably affects their personal activities with a significant psychosocial impact.

Altogether, accurate diagnosis and seizure predictability could drastically improve the quality of life of people with recurring seizures. The identification of seizure-specific biomarkers would enable the development of diagnostic and predictive tests, but a specific panel of biomarker compounds is yet to be developed.

Tests based on volatile organic compound (VOC) measurement are non-invasive and well-accepted by patients. For this reason, they are an ideal testing modality to be performed during unpredictable acute medical circumstances, such as seizures.

The potential role of volatile molecules in epilepsy is already known – it has been shown that dogs possess the ability to predict oncoming seizures [Epilepsy Behav. 2023, 150, 109563]. Furthermore, research has demonstrated that a dog’s perception of seizures does not depend on body movement, as they can differentiate between samples collected from subjects who have experienced a seizure compared to samples collected from those who have performed physical exercise [Sci Rep. 2019, 9, 4103]. This suggests the existence of a possible ‘odour’ linked to seizures which we aim to investigate through the analysis of VOCs.


To (i) define a VOC panel in breath and skin which distinguishes between epileptic and non-epileptic seizures; and (ii) identify VOCs in breath and skin whose concentrations change prior to seizure onset.


VIBES is a prospective cohort study. Patients who have been admitted to either the National Hospital for Neurology and Neurosurgery (NHNN) telemetry unit or Chalfont Centre for Epilepsy, aged 18 or above are eligible for recruitment. The recruitment target for participants in the seizure cohorts (both epileptic and non-epileptic) is 100. In addition, 50 age-matched healthy participants who never experienced a seizure in their lifetime will be recruited as controls.

For participants in the seizure cohorts, baseline samples will be collected when participants arrive at the hospital, with further breath and skin VOC samples being collected at pre-determined time points following any witnessed daytime seizure. Additional pre-seizure samples will be collected for the prediction study, based on the subject clinical history. For participants in the control cohort, only one baseline breath sample is collected.

Breath samples will be collected in thermal desorption (TD) tubes and sent to the Imperial College London VOC Laboratory to be analysed with two different gas-chromatography time-of-flight mass spectrometry instruments (polar and mid-polar analytical columns). Skin VOCs will be collected by swabbing the participant’s upper back using a medical gauze.

Participating centres:

For more information relating to the study, please get in touch with Ilaria Belluomo, via email: