Abstracts

Rationale: The association between systemic disease manifestations and epilepsy in Tuberous Sclerosis Complex (TSC) has not been thoroughly investigated. Such associations may assist in understanding the factors driving disease expression and have implications for disease prognosis and screening. Methods: Utilizing the TSC Natural History Database, a multicenter, retrospective database, we performed univariate and multivariable analyses to identify systemic or neurological disease manifestations associated with the presence of epilepsy. Results: Of 1816 TSC patients in the database, we included subjects older than 3 years and for whom epilepsy data was available (n=1615). Nearly 88% of individuals had a history of epilepsy. A history of focal seizures was reported in 81.8% of subjects, with the mean age of onset being 2.9 years (range 0 years to 51.2 years, SD 4.7 years). A history of infantile spasms was reported in 49.2% of subjects. On univariate analysis, patient characteristics associated with the presence of epilepsy included younger age (mean age 19.7 years compared to 25.8 years, p < 0.001) and male sex (p < 0.001). TSC mutation testing was available for 680 subjects. Those with a TSC2 mutation had a higher odds of epilepsy compared to those with no mutation identified (NMI) (p < 0.001), those with a TSC2 mutation had a higher odds of epilepsy compared to those with a TSC1 mutation (p=0.002), and those with a TSC1 mutation had a higher odds of epilepsy compared to those with NMI (p=0.02). Individual logistic regression models were built to examine each candidate systemic or neurological disease variable, controlling for the patient characteristics found to be significant on univariate analysis. Cardiac rhabdomyomas (p=0.002), retinal hamartomas (p=0.04), renal cysts (p=0.002), renal angiomyolipomas (p < 0.001), shagreen patches (p=0.04), and facial angiofibromas (p=0.03), as well as neurological variables including subependymal nodules (p < 0.001) and cortical tuber hamartomas (p < 0.001) were significantly associated with an increased likelihood of epilepsy. A global multivariable logistic regression model was then built including the variables found to be significant in the individual logistic regression models, again correcting for the patient characteristics found to be significant on univariate analysis. In this multivariable model, cardiac rhabdomyomas (p=0.04) and cortical tuber hamartomas (p=0.002) remained significantly associated with an increased likelihood of epilepsy. Renal angiomyolipomas approached statistical significance (p=0.05). Conclusions: We identified a novel association between systemic disease manifestations, including multiple non-neurological tumors such as cardiac rhabdomyomas, retinal hamartomas, renal angiomyolipomas, and facial angiofibromas, and the presence of epilepsy in TSC. Genetic mutation status did not fully account for these associations. Such associations may ultimately help to stratify risk and allow for tailored treatment strategies, especially as the approach to treatment in TSC moves towards early, potentially pre-symptomatic and disease-modifying interventions. Funding: National Institutes of Health (T32 Training Grant NS007205 to Washington University, R01 NS056872 to MW) and the Missouri State Tuberous Sclerosis Fund. The authors thank Jo Anne Nakagawa, the Tuberous Sclerosis Alliance, and all contributors to the TSC Natural History Database.