Discussion
We identified an adolescent with Dravet syndrome who developed B-cell lineage ALL. In addition, there are at least two other cases of ALL in children with Dravet syndrome (personal communication, Veronica Hood, PhD). One with Dravet syndrome due to SCN1A variant was diagnosed with B cell ALL at age 7 years, having been on multiple anti-epileptics including valproate. The incidence of Dravet syndrome among live births is estimated at 1/15,000. According to the National Cancer institute, the incidence rate of ALL in those below the age 19 in US is 1 per 1,000.
Dravet syndrome commonly appears during the first year of life usually related to fever or vaccination. The first seizure typically is generalized clonic then the child develops multiple seizure types (convulsive seizures, myoclonic seizures, atypical absence seizures, focal seizures, or rarely, tonic seizures).1 Patients often develop psychomotor delay, behavioral disturbances, and cognitive impairment. Autistic features usually persist into adulthood.6
Eighty percent of patients with Dravet syndrome harbor a heterozygous SCNA1 gene mutation,1 resulting in its loss-of-function. SCN1A encodes the alpha subunit of a sodium channel NaV1.1. In disease-associated SCN1A variants, interneurons can no longer inhibit the electrical activity.7 Loss-of-function mutations of SCN1A  cause functional impairments in the GABAergic neurons and lead to a loss of appropriate inhibition in neuronal networks and a predisposition to seizures. Over 650 missense and nonsense SCN1A mutations have been described in these patients; some are de novo .8
Although rarely reported, non-central nervous system cancers could present in the context of seizure and/or autism disorders (AD). One case reported ALL in a nine-year-old girl with isodicentric chromosome 15 syndrome, which is associated with early central hypotonia, developmental delay, cognitive dysfunction, autism spectrum disorders, and seizures.9 A retrospective cohort study of 8000 Taiwanese children and adolescents showed that cancer occurred more frequently in pediatric patients with AD. The standardized incidence ratio’s estimate is 1.94 (95% CI 1.18-2.99). During the observation period, 20 patients with AD developed cancers. Five patients had leukemia, two of them (one with intellectual disability, the second one seizures) had ALL.10 Multiple copy number variants identified in autism patients are found with individuals with cancer, suggesting a link with chromosomal abnormalities and carcinogenesis.11 According to a cohort study done on patients with AD, there was an increased risk of developing cancer by midlife.12 Other studies have not suggested an increased association.13,14 The prevalence of autism in a North Carolina population found that out of 702 cancer patients, seven (1%, 95% CI of 0.4% - 2.04%) were labeled as AD in their medical records. The diagnosis of AD was made prior to cancer presentation in five of the seven of the patients; in the other two, obvious behavioral issues at cancer presentation led to an evaluation which suggested AD.13
While we cannot at this time conclude that patients with Dravet syndrome are at increased risk for developing acute leukemia, it is tempting to speculate that associated metabolic disturbances could produce intracellular stress.15 Seizures can induce endoplasmic reticulum stress, and mouse models suggest that this results in the expression and distribution of Mdm2 and down-regulation of Tp53.16 A complex host-environment interaction would include long-term use of anti-seizure drugs. For instance, hematologic side effects of valproate are well known17 and can be associated with transient neoplasia.18 Occurrence of cancer in children, a rare event, with non-hematologic monogenic disorders, a rare event, should be considered as a candidate predisposition syndrome. Hence, the number of pediatric and adult patients with cancer predisposition syndromes may be even greater by expanding the list of candidate genes not classically associated with hematologic malignancies.
Conflict of Interests : The authors declare no conflict of interests.
Acknowledgements : This work was supported in part by Rally Foundation for Childhood Cancer, V Foundation for Cancer Research, Alex’s Lemonade Stand Foundation (A-Award), and NIH/NCI K22 CA251649-01 to CH, VeloSano (HM), and Hyundai Hope Scholar Award to SJC.