Failed Hemispherotomy: Insights from Our Early Experience in 40 Patients.

OBJECTIVE

To study the factors responsible for failure of hemispherotomy and outcomes of revision surgery. The effect of the surgeon's learning curve on failures was also analyzed.

METHODS

Forty consecutive patients, who underwent functional hemispherotomy through a 4-year period, from the inception of the single-surgeon epilepsy surgery program, were analyzed.

RESULTS

A total of 47 functional hemispherotomies were performed in the study period in 40 patients (7 revision surgeries on 6 patients). Mean age of the cohort was 9.45 ± 14.84 years and it included 7 infants (<2 years). Of the 9 patients (23.5%) who failed the first procedure, 6 qualified for revision surgery, all of whom belonged to the cohort of the first 15 patients treated during the first 2 years of the program. Hemimegalencephaly was the most common disease (n = 4). Ipsilateral temporal stem (n = 3), frontobasal connections (n = 2), splenium of corpus callosum (n = 2), and posterior insula (n = 2) were residual undisconnected substrates identified for revision on imaging. The substrates for failure were obvious in 5/6 patients and resulted from incomplete disconnection, implying surgical inadequacy. At the mean follow-up of 30 ± 13.17 months (range, 13-55 months), 35 of 40 patients (87.5%) remained seizure free (Engel class Ia), including 4/6 patients who underwent redo surgery. Revision did not benefit the remaining 2 patients (Engel class III). There was no mortality.

CONCLUSIONS

Surgical revision is more common in hemimegalencephaly and in the early days of a surgical program. Affirmative neuroimaging improves the outcomes of subsequent revision surgery.