Authors :
Presenting Author: Alina Ivaniuk, MD – Mayo Clinic in Florida
Emile Moura Coelho da Silva, MS, CGC – Department of Neurology, McGovern Medical School, The University of Texas Health Science Center at Houston, Houston, TX, USA
Manuela Ochoa-Urrea, MD – The University of Texas Health Science Center at Houston
Radhika Dhamija, M.B.B.S. – Mayo Clinic
Elaine Wirrell, MD – Mayo Clinic, Rochester MN, USA.
David Burkholder, MD – Mayo Clinic
Gregory Cascino, MD – Mayo Clinic, Rochester MN, USA.
William Tatum, DO – Mayo Clinic in Florida
Brin Freund, MD – Mayo Clinic in Florida
Elia Pestana Knight, MD – Cleveland Clinic
Dennis Lal, PhD – Department of Neurology, McGovern Medical School, The University of Texas Health Science Center at Houston, Houston, TX, USA
Anthony Fine, MD – Mayo Clinic
Rationale:
Haploinsufficiency of CACNA1A has been predominantly associated with episodic ataxia type 2 (EA). However, animal models and recent small case series suggest a broader phenotypic spectrum, including epilepsy. The phenotypes, longitudinal course, treatment responses, and penetrance of epilepsy in people with CACNA1A haploinsufficiency have not been systematically characterized.Methods:
We computationally screened electronic health records from Cleveland Clinic and Mayo Clinic for patients with epilepsy and PTVs in CACNA1A and verified presence of pathogenic variants and diagnosis of epilepsy manually. We identified 32 missense variants that change channel properties to loss-of-function (LoF), therefore, haploinsufficiency, from published electrophysiological studies. We additionally performed semi-automated literature review with LitVar API to query literature-linked LoF CACNA1A variants (n=99), filtered from 2709 total variants across 6567 publications. Cases with sufficient clinical data were included for phenotypic analysis. Variant annotation was performed using the Ensembl Variant Effect Predictor API.
Results:
We identified 46 individuals with epilepsy and 30 unique haploinsufficiency-associated CACNA1A variants (39 from 17 published studies, 7 from institutional cohorts). We observed variable rates of seizure phenotype expression in 4 families with 4 or more variant carriers, ranging from 0.5 to 0.8. Seizures were the presenting symptom in 21/40 (53%) cases with available onset information. Absence seizures (typical and atypical) were reported in 8/8 (100%) individuals with missense LoF variants and 21/39 (57%) individuals with protein-truncating variants (PTVs) (p=0.037), and GTCs for 3/8 (38%) missense LoF and 14/39 (38%) PTV cases (p=1). A median age of seizure onset was 3 years (IQR 1-4.5) versus 10 years (IQR 8-14) for EA (log-rank p = 0.00042). EEG features were reported for 34 individuals; 15/34 (47%) presented classical 3 Hz spike-and-wave ictal discharges. Interictally, 28/34 (82%) had generalized epileptiform discharges. Of 32 individuals with reported seizure outcomes, 16 (50%) were seizure-free (all PTV carriers, median age 5 years, range 2-41 years). Seizure-free individuals had earlier age of seizure onset (24 [IQR 10-72] vs. 6 [IQR 6-8], p=0.016), less commonly had absences (44% vs 81%, p=0.028), and had no or mild intellectual disability (p=0.009). ASMs with broad mechanism of action (valproate, topiramate, zonisamide) resulted in seizure decrease (20/29 use cases, 69%), whereas ethosuximide only led to seizure decrease in 2/7 (28.6%) cases. In one patient, VNS led to reduced absence seizure frequency, and DBS resulted in freedom from GTCs.
Conclusions:
Epilepsy in LoF CACNA1A variants presents with marked phenotypic variability and predominantly generalized phenotypes with typical and atypical absence seizures that are not responsive to ethosuximide. Earlier seizure onset and higher degree of intellectual disability may be associated with worse seizure control in this population. The presence of EA, though often delayed, along with developmental delay, can prompt considering genetic testing in individuals with history of generalized epilepsy.
Funding: None