MicroRNAs as potential biomarkers of response to modified Atkins diet in treatment of adults with drug-resistant epilepsy: A proof-of-concept study.

Background: Accurate predictors of response to modified Atkins diet (MAD) are needed. MicroRNAs are potential biomarkers in epilepsy. This study aimed to explore the value of circulating miR-146a, miR-155, miR-22, miR-21 and miR-134 levels in predicting response to MAD.

Methods: Patients who completed 3 months of MAD were selected from a prospective cohort of adults with DRE followed in a specialized MAD outpatient clinic. Patients were classified as responders if any reduction in seizure frequency at follow-up, calculated through seizure-calendars). The >50 % seizure reduction cut-off was also explored. Qualitative benefits in seizures and cognition were analysed. Blood samples were collected prior to initiate MAD and microRNAs were quantified by qRT-PCR.

Results: Thirty-nine patients were included (56 %males, mean age=33.1±8.5yo, 62 %focal epilepsies, 59 %structural aetiology): 20(51 %) were responders [mean reduction in seizure frequency=54 %(17-100 %); 10 had ≥50 % reduction]; 25(64 %) reported qualitative benefit in seizures and 21(54 %) reported cognitive benefits. At pre-treatment baseline, a panel combining serum levels of all studied microRNAs predicted seizure reduction (AUC=0.839, p<0.0001), qualitative benefit in seizures (AUC=0.683, p=0.048) and in cognition (AUC=0.751, p<0.01) at 3months. miR-146a was the only significant microRNA when evaluated in isolation. There was no statistical correlation in the biomarkers when a ≥50 % seizure reduction was compared to <50 %.

Conclusions: A panel combining pre-treatment serum levels of miR-146a, miR-155, miR-134, miR-21 and miR-22 predicted any reduction in seizures with MAD in adults with DRE at 3months. This panel may be a promising biomarker and a useful tool in the selection of patients.