Motivation: Many children have epilepsy, and many look remarkably well otherwise. They eat, play, fight, go to school, and do otherwise regular things. But, I have always wondered what happens to the children in the future? Do they live happily ever after?
Recently, a group in Finland published one of the longer prospective follow-ups on childhood onset epilepsy.
Paper: Long-Term Mortality in Childhood-Onset Epilepsy. Sillanpaa, M. and Shinnar, S. NEJM (2010) 363: 2522-9. http://www.nejm.org/doi/pdf/10.1056/NEJMoa0911610
Methods: The study included all children younger than 16 living in Turku, Finland with epilepsy (at least two unprovoked seizures). The study was restricted to children who were seen in Turku University Hospital between 1961-1964. At that time, epilepsy was apparently a reason to hospitalize children. Also excluded were children with febrile seizures, single episode of seizure, or seizure judged to be from acute causes (like hyponatremia or hypoglycemia). Since the start of the study, a follow-up examination was conducted every five years till 2003!
Results: The study followed a cohort of 245 patients. During the 40 year follow-up, only five subjects migrated out of Finland. Of the rest, information on death and cause of death were ascertained from the national registry, which covers all patients in the country. At the time of last follow-up, 45% were in 5-year remission without drugs, 11% were in remission with drugs, and 44% were not in 5-year remission.
Overall mortality: The overall mortality rate during the forty year follow-up was 24%. The overall mortality rate in comparison to mortality rate of the general population was 5.5 (CI: 4.6 to 6.6) for patients who suffered their first episode of epilepsy between 1961-1964. The median age of death was 23 years. No significant difference was found between genders. 85% of the total deaths occurred in patients who were not in 5 year remission. Among the 60 deaths in the overall group, 33 deaths (55%) were epilepsy related (including 18 patients suffering sudden, unexplained death, 13 from seizures, and 6 drownings). The rest of deaths were non-epilepsy related.
Subgroup analysis: When the overall cohort was subdivided into two subgroups consisting of (1) idiopathic or cryptogenic seizures and (2) remote symptomatic causes such as developmental delay, cognitive impairment, cerebral palsy, or neurologic insult like stroke, clear differences emerged. The overall mortality rate was 11.1 person per 1000 person-years in the group with remote symptomatic cause while the mortality rate was 3.2 per 1000 person-years in the group with idiopathic or cryptogenic seizures (p<0.001). In multivariate analysis, the most consistent predictor of mortality was absence of 5-year remission of seizure.
Discussion: The primary strength of the study was the long follow-up time. This study paints a clear picture that children who particularly have long-term uncontrolled seizure are at significantly increased risk of mortality as adults. While patients with remote symptomatic causes of seizure have higher overall mortality than idiopathic causes, both groups have higher overall mortality. There are many interesting results of the study. One that I found surprising is that a substantial number of patients die of sudden unexplained death. The authors comment on the side that in most of these patients, pulmonary edema was found during autopsy. I wonder why some patients who were apparently not having active seizures would die from pulmonary edema.
The most encouraging result clearly was that in patients in 5-year remission, the mortality rate was significantly less. While the association cannot be taken for causality, the next question is whether more aggressive management such as with surgery can stop the trend. I guess we would need another heroic 40 year study to answer that!
Recently, a group in Finland published one of the longer prospective follow-ups on childhood onset epilepsy.
Paper: Long-Term Mortality in Childhood-Onset Epilepsy. Sillanpaa, M. and Shinnar, S. NEJM (2010) 363: 2522-9. http://www.nejm.org/doi/pdf/10.1056/NEJMoa0911610
Methods: The study included all children younger than 16 living in Turku, Finland with epilepsy (at least two unprovoked seizures). The study was restricted to children who were seen in Turku University Hospital between 1961-1964. At that time, epilepsy was apparently a reason to hospitalize children. Also excluded were children with febrile seizures, single episode of seizure, or seizure judged to be from acute causes (like hyponatremia or hypoglycemia). Since the start of the study, a follow-up examination was conducted every five years till 2003!
Results: The study followed a cohort of 245 patients. During the 40 year follow-up, only five subjects migrated out of Finland. Of the rest, information on death and cause of death were ascertained from the national registry, which covers all patients in the country. At the time of last follow-up, 45% were in 5-year remission without drugs, 11% were in remission with drugs, and 44% were not in 5-year remission.
Overall mortality: The overall mortality rate during the forty year follow-up was 24%. The overall mortality rate in comparison to mortality rate of the general population was 5.5 (CI: 4.6 to 6.6) for patients who suffered their first episode of epilepsy between 1961-1964. The median age of death was 23 years. No significant difference was found between genders. 85% of the total deaths occurred in patients who were not in 5 year remission. Among the 60 deaths in the overall group, 33 deaths (55%) were epilepsy related (including 18 patients suffering sudden, unexplained death, 13 from seizures, and 6 drownings). The rest of deaths were non-epilepsy related.
Subgroup analysis: When the overall cohort was subdivided into two subgroups consisting of (1) idiopathic or cryptogenic seizures and (2) remote symptomatic causes such as developmental delay, cognitive impairment, cerebral palsy, or neurologic insult like stroke, clear differences emerged. The overall mortality rate was 11.1 person per 1000 person-years in the group with remote symptomatic cause while the mortality rate was 3.2 per 1000 person-years in the group with idiopathic or cryptogenic seizures (p<0.001). In multivariate analysis, the most consistent predictor of mortality was absence of 5-year remission of seizure.
Discussion: The primary strength of the study was the long follow-up time. This study paints a clear picture that children who particularly have long-term uncontrolled seizure are at significantly increased risk of mortality as adults. While patients with remote symptomatic causes of seizure have higher overall mortality than idiopathic causes, both groups have higher overall mortality. There are many interesting results of the study. One that I found surprising is that a substantial number of patients die of sudden unexplained death. The authors comment on the side that in most of these patients, pulmonary edema was found during autopsy. I wonder why some patients who were apparently not having active seizures would die from pulmonary edema.
The most encouraging result clearly was that in patients in 5-year remission, the mortality rate was significantly less. While the association cannot be taken for causality, the next question is whether more aggressive management such as with surgery can stop the trend. I guess we would need another heroic 40 year study to answer that!
No comments:
Post a Comment