SANTE - Stimulation of the Anterior Nucleus of the Thalamus for Epilepsy

A generalized estimating equations (GEE) analysis was used to evaluate the treatment effect on seizure frequency. With one outlier subject removed, the GEE model for this alternative analysis included treatment effect, log of the baseline seizure count, log of age, visit (categorical), and the offset (the number of days the diary was recorded in each month).

Time frame: Through the end of the three-month blinded phase

Population: This analysis used the protocol-prespecified Primary analysis data set with one additional outlier subject removed from the active group. This subject had 210 stimulation initiated seizures within 48 hours of the device being turned on and was determined to be an outlier using both a statistical and clinical rationale.

A protocol-prespecified generalized estimating equations (GEE) analysis was used to evaluate the treatment effect on seizure frequency. The final GEE model for the primary objective evaluation included treatment effect, log of the baseline seizure count, log of age, visit (categorical), treatment-by-visit interaction (categorical), and the offset (the number of days the diary was recorded in each month).

Time frame: Through the end of the three-month blinded phase

Population: This analysis used the protocol-prespecified Primary analysis data set which required that subjects had at least 70 days of diary in the blinded phase. One control subject was not included in this analysis as they had 66 of the required 70 days.

The results are for the follow-up after device implantation through Year 2 and summarized are events that occurred in greater than 5% of subjects. Only events related to the device, therapy, or surgery are included. These abbreviations were used: * General dis...=General disorders and administration site conditions * Injury, poison...=Injury, poisoning and procedural complications * Ther.=Therapeutic. For this summary, adverse events are reported as 'MedDRA System Organ Class - adverse event'.

Time frame: Through Year 2 of the long-term follow-up phase

Population: This analysis used all subjects who were implanted and received stimulation. One subject was implanted but was not subsequently randomized, but did receive stimulation. This subject has been included for the purposes of this analysis for a total of 110 subjects, as opposed to the 109 stated in the participant flow.

Difference between active group and control group in percentage change in seizure-free days over the entire blinded phase as compared to the entire baseline phase. The number of seizure-free days was normalized to 84-day baseline and blinded phases for each subject.

Time frame: Through the end of the three-month blinded phase

Population: This analysis used the Primary analysis data set which required that subjects had at least 70 days of diary in the blinded phase. In addition, percentage change was not calculated for subjects who had no seizure-free days during the baseline phase.

The number presented is for Definite and Probable SUDEP. The rate is calculated per 1000 subject years of follow-up. The confidence interval is the 95% Poisson confidence interval. Per protocol, only definite and probable SUDEP classifications were included in the calculation. The results shown are for the entire study follow-up after device implantation.

Time frame: Inclusive of all study follow-up after device implantation (mean follow-up 3.7 years)

Population: This analysis used all subjects who were implanted and received stimulation. One subject was implanted but was not subsequently randomized, but did receive stimulation. This subject has been included for the purposes of this analysis for a total of 110 subjects included in this analysis.

Difference between active group and control group in percentage change in the maximum length of seizure-free intervals over the entire blinded phase as compared to the entire baseline phase.

Time frame: Through the end of the three-month blinded phase

Population: This analysis used the Primary analysis data set which required that subjects had at least 70 days of diary in the blinded phase. One control subject was not included in this analysis as they had 66 of the required 70 days.

A treatment failure was defined in the protocol as a subject who 1) required 3 or more doses of rescue medication within 48 hours, 3 times during the blinded phase; or 2) had 3 episodes of convulsive status epilepticus during the blinded phase.

Time frame: Through the end of the three-month blinded phase

Population: This analysis used the Primary analysis data set which required that subjects had at least 70 days of diary in the blinded phase. One control subject was not included in this analysis as they had 66 of the required 70 days.

A responder is defined as a subject with greater than or equal to 50% reduction in seizures as compared with baseline.

Time frame: Through the end of the three-month blinded phase

Population: This analysis used the Primary analysis data set which required that subjects had at least 70 days of diary in the blinded phase. One control subject was not included in this analysis as they had 66 of the required 70 days.

Seizures were recorded on daily seizure diaries. The subject recorded the number of seizures by seizure type on the seizure diary. The subject also noted at baseline, of those they had ever experienced, which seizure they considered to be most severe.

Time frame: Through the end of the three-month blinded phase

Population: This analysis used the Primary analysis data set. In addition, the protocol prespecified that if a subject did not experience the severe seizure in the baseline phase that they would not be included in the calculation of the blinded phase median seizure frequency percentage change from baseline.