Levetiracetam Administration Is Correlated with Lower Mortality in Patients with Mitochondrial Encephalomyopathy, Lactic Acidosis, and Stroke-like Episodes: A Retrospective Study
Mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes (MELAS) is a progressive mitochondrial disorder characterized by recurrent stroke-like episodes (SLEs), seizures, and multi-organ dysfunction. Epileptic seizures, which occur in up to 90% of patients with MELAS, often necessitate long-term antiepileptic drug (AED) therapy. However, the efficacy and safety of AEDs in MELAS remain understudied, with some drugs like valproate (VPA) and carbamazepine (CBZ) showing potential mitochondrial toxicity. In contrast, levetiracetam (LEV), a newer AED, has demonstrated neuroprotective properties in non-mitochondrial neurological disorders. This retrospective cohort study investigated the association between LEV administration and clinical outcomes in patients with MELAS, focusing on disability and mortality.
Study Design and Patient Selection
The study analyzed 102 patients with MELAS and a history of seizures from a single-center cohort. Participants were followed for a median of 4 years (range: 1–8 years) and stratified into two groups: those receiving LEV (n=48) and those treated with non-LEV AEDs (n=54). MELAS diagnosis required confirmation through mitochondrial DNA mutations or muscle biopsy findings, such as ragged-red fibers. Patients were excluded if they had incomplete AED records or changed AED regimens during follow-up. Baseline demographics, clinical features (e.g., deafness, diabetes, cortical blindness), and AED usage patterns were recorded. Disability was assessed using the modified Rankin Scale (mRS), with favorable outcomes defined as mRS scores of 0–1.
AED Utilization Patterns
LEV was the most commonly prescribed AED (47.1% of patients), followed by CBZ (36.3%), benzodiazepines (17.6%), topiramate (12.7%), oxcarbazepine (10.8%), VPA (8.8%), and lamotrigine (8.8%). In the LEV group, 20 patients received combination therapy, most frequently with benzodiazepines (n=6) or CBZ (n=3). In the non-LEV group, CBZ monotherapy (n=23) and polytherapy involving CBZ + VPA (n=4) were common. No significant differences were observed in AED duration between groups (LEV: 4.1 ± 2.6 years vs. non-LEV: 5.5 ± 4.1 years; P=0.11).
Disability Outcomes
At the final follow-up, the LEV group exhibited significantly lower mean mRS scores (2.79 ± 1.47) compared to the non-LEV group (3.83 ± 1.93; P=0.006). However, the proportion of patients achieving favorable outcomes (mRS 0–1) did not differ significantly between groups (16.7% vs. 9.3%; P=0.37). Multivariate logistic regression confirmed LEV as an independent protective factor for reduced disability (OR=0.32; 95% CI: 0.15–0.68; P=0.003) after adjusting for age of onset, gender, disease duration, deafness, and diabetes.
Seizure Control and Dose Response
Seizure reduction within the final year of follow-up was superior in the LEV group: 56.3% achieved complete seizure freedom (level 2), compared to 33.3% in the non-LEV group (P=0.02). Intriguingly, LEV dosage inversely correlated with seizure control: patients with complete freedom used lower doses (0.024 ± 0.012 g/kg/day) compared to those with partial (0.040 ± 0.016 g/kg/day; P=0.02) or no response (0.031 ± 0.010 g/kg/day; P=0.03). AED duration did not influence seizure outcomes.
Mortality and Survival Analysis
The LEV group demonstrated markedly lower mortality (8.3%, 4/48) than the non-LEV group (37.0%, 20/54; P=0.001). Causes of death in the LEV group included SLEs (n=2), sudden unexpected death (n=1), and pseudo-intestinal obstruction (n=1). In the non-LEV group, status epilepticus (n=6) and pseudo-intestinal obstruction (n=4) were leading causes. Kaplan-Meier analysis revealed significantly better survival in the LEV group (log-rank P=0.04). Cox proportional-hazards modeling identified LEV as an independent survival predictor (HR=0.24; 95% CI: 0.08–0.74; P=0.013), even after adjusting for confounders like deafness and diabetes.
Mechanistic Insights and Limitations
The neuroprotective effects of LEV in MELAS may involve dual mechanisms: improved seizure control and direct mitochondrial modulation. LEV targets synaptic vesicle protein 2A (SV2A), which localizes to mitochondria and may stabilize mitochondrial membrane potentials. Preclinical studies suggest LEV enhances complex I activity and reduces oxidative stress, potentially counteracting mitochondrial dysfunction in MELAS. However, the study’s retrospective design limits causal inference. Selection bias could not be fully excluded, as LEV’s higher cost in China might have skewed its use toward economically advantaged families. Additionally, the absence of baseline mRS scores and detailed seizure phenotyping (e.g., distinction between epileptic and non-epileptic events) complicates outcome interpretation.
Clinical Implications and Future Directions
This study provides the first clinical evidence supporting LEV’s association with reduced mortality in MELAS. While disability outcomes were not normalized, the survival benefit underscores the importance of AED selection in mitochondrial disorders. Prospective trials should validate these findings, clarify optimal dosing, and explore LEV’s long-term effects on mitochondrial biogenesis and SLE recurrence. Concurrently, in vitro and animal studies are needed to dissect LEV’s molecular interactions with respiratory chain complexes and reactive oxygen species pathways in MELAS models.
doi.org/10.1097/CM9.0000000000000061
Was this helpful?
0 / 0