Ketogenic diet for super-refractory status epilepticus (SRSE) with NORSE and FIRES: Single tertiary center experience and literature data

Background and purposeKetogenic diet (KD) is an emerging treatment option for super-refractory status epilepticus (SRSE). We evaluated the effectiveness of KD in patients presenting SRSE including NORSE (and its subcategory FIRES).MethodsA retrospective review of the medical records was performed at the Necker Enfants Malades Hospital. All children with SRSE in whom KD was started during the last 10 years were included. A systematic search was carried out for all study designs, including at least one patient of any age with SRSE in whom KD was started. The primary outcome was the responder rate and Kaplan–Meier survival curves were generated for the time-to-KD response. As secondary outcomes, Cox proportional hazard models were created to assess the impact of NORSE-related factors on KD efficacy.ResultsSixteen children received KD for treatment of SRSE, and three had NORSE presentation (one infectious etiology, two FIRES). In medical literature, 1,613 records were initially identified, and 75 were selected for review. We selected 276 patients receiving KD during SRSE. The most common etiology of SRSE was acute symptomatic (21.3%), among these patients, 67.7% presented with NORSE of immune and infectious etiologies. Other etiologies were remote symptomatic (6.8%), progressive symptomatic (6.1%), and SE in defined electroclinical syndromes (14.8%), including two patients with genetic etiology and NORSE presentation. The etiology was unknown in 50.7% of the patients presenting with cryptogenic NORSE, of which 102 presented with FIRES. Overall, most patients with NORSE benefit from KD (p < 0.004), but they needed a longer time to achieve RSE resolution after starting KD compared with other non-NORSE SRSE (p = 0.001). The response to KD in the NORSE group with identified etiology compared to the cryptogenic NORSE was significantly higher (p = 0.01), and the time to achieve SE resolution after starting KD was shorter (p = 0.04).ConclusionsThe search for underlying etiology should help to a better-targeted therapy. KD can have good efficacy in NORSE; however, the time to achieve SE resolution seems to be longer in cryptogenic cases. These findings highlight the therapeutic role of KD in NORSE, even though this favorable response needs to be better confirmed in prospective controlled studies.