In this chart review, 31/36 patients (86%) with LGS experienced a reduction in seizure frequency after the addition of CNB, including 22 patients (61%) with a ≥ 75% reduction in seizures and five (14%) who achieved seizure freedom. Three patients exhibited dramatic improvement initially followed by a recurrence of seizures at a lower level, often referred to in clinical practice as a temporary “honeymoon” from seizures when a new drug is first introduced [24]. These recurrences were further examined for duration of maximal impact and ultimate seizure reduction compared with baseline. At last follow-up, two of these patients remained at ≥ 90% reduction, and one had < 50% reduction.

Although there are limited reports examining CNB’s effectiveness in LGS, results from previous case series and retrospective studies indicate broad-spectrum efficacy of CNB for other types of seizures and epilepsy syndromes in adult and pediatric patients [11–15]. In a small case series (n = 4), adult patients with LGS experienced seizure frequency reductions ranging from 25% to 74% at 12 months, including two patients with ≥ 50% seizure reduction [11]. Other study cohorts that have included adult and pediatric patients with LGS have reported substantial reductions in seizure frequency and good tolerability, but did not examine specific LGS subgroup outcomes [12, 25]. Two ongoing prospective studies are currently evaluating CNB dosing in specific pediatric weight-based cohorts [16–18]. Our findings contribute additional data to support the potential efficacy of CNB in LGS.

Patients being treated for drug-refractory epilepsy may experience decreased quality of life and have a greater potential for toxicities due to the need for polytherapy and potentially high ASM doses [26, 27]. After initiating CNB, a substantial proportion of our patients were able to reduce their concomitant ASM burden. Cannabidiol, clobazam, and lacosamide were chosen to taper earlier due to potential medication interactions [28]. These interactions may manifest early in CNB titration, requiring close observation and dose adjustments of the concomitant ASMs [29]. CNB inhibits CYP2C19 and therefore increases plasma concentrations of CYP2C19 substrates such as clobazam and cannabidiol [28, 29]. Concomitant administration of CNB and the prescription formulation of cannabidiol may result in somnolence, especially when administered in conjunction with clobazam [28]. In particular, the proactive, early reduction of clobazam dosing may reduce CNS-related adverse events including somnolence [28]. In our study, a substantial proportion of patients reduced or discontinued cannabidiol or clobazam. CNB has a dual mechanism of action that combines a block of the persistent current of voltage-gated sodium channels and positive allosteric modulation of GABA_A receptors [30, 31]. Although clobazam and CNB both exhibit GABA agonistic activity, many patients needed 5–15 mg/day of clobazam to optimize seizure control. In these cases, the combination appeared to serve as a rationale for polypharmacy [32, 33]. The side effect profile of CNB in our patient population was consistent with that observed in clinical trials, with somnolence being the most common [7].

The retrospective single-center nature of the study presents inherent constraints that may limit its generalizability. These include the potential for incomplete or missing documentation. In our study, adverse event reporting and seizure frequency assessments were based on non-standardized caregiver reports, which may be prone to bias. In addition, our definition of LGS was broader than that outlined in the International League Against Epilepsy (ILAE) diagnostic criteria [34]. Although we required patients to have an EEG with a slow background and at least frequent interictal multifocal or generalized epileptiform discharges, we did not require the presence of generalized paroxysmal fast activity and/or slow spike/wave complex, as not all patients will display slow spike/wave at all points of illness, especially with increasing age [21–23]. All other diagnostic criteria were consistent with ILAE guidelines.

It is difficult for caregivers to monitor and quantify the polymorphic seizures associated with LGS (e.g., tonic seizures that occur during sleep and atypical absences). Although the efficacy of CNB in various seizure types is of great interest to clinicians, the clinical records did not further define changes in subtypes of seizures; therefore, only countable motor seizures were followed for change in frequency. Because this was a retrospective real-world study, titration patterns and reductions to concomitant medications were variable, and changes to concomitant treatments occurred. As treatment progressed, we based dose adjustments on individual patient responses and our increasing experience with managing concomitant ASMs during the addition of CNB to existing therapy. We measured changes in polypharmacy by counting the number of concomitant medication discontinuations and changes to clobazam and cannabidiol doses, which under-represents the impact of dose reductions on tolerability. Although the small sample size is a limitation of our study, it should be noted that LGS is a rare disorder, which by necessity decreases sample size.