Long-term functional outcomes and their predictors after hemispherectomy in 115 children

To examine the long-term functional outcomes and their predictors using a patient/family centered approach in a cohort of children who had hemispherectomy. Functional outcome measures studied were the following: ambulation ability, visual symptoms, spoken language, reading skills, and behavioral problems.

We reviewed 186 consecutive children who underwent hemispherectomy between 1997 and 2009 at our center. Preoperative clinical, electroencephalography (EEG), imaging, and surgical data were collected. One hundred twenty-five families completed a structured questionnaire to assess the functional status and seizure outcome. Prognostic predictors were examined using a multivariate regression analysis.

At a mean follow-up of 6.05 years after hemispherectomy, 70 patients (56%) were seizure-free and 45 (36%) had seizure recurrence; 10 patients (8%) were free of their preoperative seizures but had new-onset nonepileptic spells and were excluded from further analysis. Of 115, at follow-up (mean age at follow-up 12.7 years, range 2-28 years), 96 patients (83%) walked independently, 10 (8.7%) walked with assistance, and 9 (7.8%) were unable to walk. New visual symptoms that were not present preoperatively were reported only in 28 patients (24%). Eighty patients (70%) had satisfactory spoken language skills but only 44 (42%) of the 105 children older than 6 years had satisfactory reading skills. Significant behavioral problems were reported in 30 patients (27%). Only five (6.2%) of the 81 children aged between 6 and 18 years attended mainstream school without assistance; 48 (59%) were in mainstream school with assistance and the rest were in special school for disabled or home cared. Five (21%) of the 24 patients older than 18 years of age were gainfully employed. Multivariate logistic regression analysis identified the following factors as independently associated with poor functional outcome. (1) Seizure recurrence negatively affected all functional domains--ambulation ability, spoken language and reading skills, and behavior (p < 0.05). (2) Abnormalities in the unoperated hemisphere on magnetic resonance imaging (MRI) (p < 0.05) and preexisting quadriparesis (p < 0.01) correlated with poor motor outcome. (3) Multilobar MRI abnormalities in the contralateral hemisphere (odds ratio [OR] = 13.9, p = 0.001) and young age (indeterminate preoperative language status) at hemispherectomy (OR = 11.1, p = 0.01) also correlated with poor language outcome. (4) Younger age at epilepsy onset correlated with poor reading skills (p = 0.01) but not with spoken language skills.

This study highlights the long-term functional status of patients after hemispherectomy. The majority of patients were ambulant independently; however, impairments in reading and spoken language were frequent. Seizure recurrence after hemispherectomy and contralateral hemisphere abnormalities on MRI were the major predictors of poor outcome in ambulation, spoken language, and reading abilities. This study will assist in presurgical counseling using simple understandable functional outcome measures and may help in planning early interventions after hemispherectomy to improve functional outcome.