Objective To evaluate if sirolimus in tuberous sclerosis complex (TSC) patients prior to epilepsy relieve the later burden of new-onset seizures. Methods A real world matched case-control trial was conducted. For early sirolimus group （EG）, TSC children (<1year old) prior to seizures but taking sirolimus for other symptoms were eligible. For the late sirolimus group （LG）, age and genotype were used as the first-place stratifying criteria for 1:4 matching, other symptoms of baseline condition were also assessed between the two group. None preventative drug should be introduced before seizures or before 2 years old in those without seizures in LG. Subjects were enrolled from 2015 to 2018 and followed up until 2020.06. The main outcome was the change of characteristic of seizures and the final condition of seizures. Results A total of 42 patients were included in the EG, and 168 were matched in the LG. Early sirolimus significantly reduced seizure-onset especially for infants < 6month old. The mean onset-age was significantly delayed by sirolimus (11.34±7.93month vs. 6.94±6.03month, P < 0.001). The subtypes of seizures benefit most was infantile spasm (5/42, 11.90% vs 73/168, 43.45%, P < 0.001) which might be eliminated or alleviated. Adding sirolimus prior seizures performed better than adding after seizures in reducing refractory seizures (10/42, 23.81% vs 70/147, 47.62%, P = 0.004). Seizures related to tuberous sclerosis complex might be none fever-related. Interpretations: Early sirolimus intervention in TSC can prevent some seizures, especially infantile spasm. Early sirolimus modified the disease, delayed seizure-onset and relieved the severity. Keywords: tuberous sclerosis complex, sirolimus, infantile spasm, partial seizure, febrile seizure Yan-Wen Shen Department of Pediatrics, Chinese PLA General Hospital China Yang-Yang Wang Meng-Na Zhang Yong Xu Qian Lu Wen He Hui-Min Chen Li-Ying Liu Ling-Yu Pang Qiu-Hong Wang Shuo Dun Yu-Feng Li Jing Gao Fang Han Li-Ping Zou