Abstracts

Rationale: The left hemisphere is usually the dominant hemisphere for language in most patients. In patients with early extensive brain injury, language may shift to the other hemisphere and almost all children with early large brain lesions in the left hemisphere do not have any language regression related to hemispherectomy. However, prior to surgery, physicians and families often raise concerns about language functions in the affected hemisphere and the potential risk of regression leading to language lateralization studies. Anecdotally, we have not found this testing to affect our management, however, this has not been systematically studied.Methods: The following data was collected on each patient between the ages of 5 and 18 years old with hemispheric lesions considered for hemispherectomy as a treatment for intractable epilepsy within the last 10 years. Demographic data, age of seizure onset, age at time of brain injury, age at surgery, pre-op language skills, type and extent of lesion involvement- specifically whether there was involvement of the inferior frontal gyrus, superior temporal gyrus, supramarginalgyrus and/or heschlgyrus , if there were contralateral abnormalities on MR, if Wada and/or fMRI was done and post op language skills. The information was assessed to determine associations between patient characteristics and post op language.Results: 103 patients with left sided hemispherectomies were identified. 54 met our age criteria, 7 had language lateralization studies, 47 had none. In the 7 patients who had language lateralization studies there was no common characteristic that would suggest a need for testing. Out of the 47 patients who were not tested, two had documented worsening, one had subjective worsening, 9 had improvement and 35 had no major impact in speech or language. Two out of the three who worsened had Rasmussen’s encephalitis and later age of epilepsy onset. Out of 79 patients who had Wada, 13 were considered for hemispherectomy, 5 went on to have a hemipsherectomy and 3 were excluded because of the Wada results. 102 patients had a functional MRI, 2 were considered for hemispherectomy, one went on to surgery and one was excluded because of the findings. Of the 3 cases that were excluded 2 had Rasmussen’s encephalitis and late onset epilepsy. One had a large, early onset lesion and was considered an excellent candidate for hemispherectomy, however, the major difference between him and the other patients whose language shifted to the other side was that the inferior frontal gyrus was not involved on MRI.Conclusions: It is important to look at specific characteristics of each patient in order to decide the risk of worsening language after a hemispherectomy. In our study we found that worsening of language was unlikely in patients with early onset brain injuries whose lesions are large and specifically if the injury involved the inferior frontal gyrus.