Abstracts

RATIONALE: Surgical treatment of hypothalamic hamartomas (HH) causing gelastic epilepsy is being increasingly advocated, although optimal surgical approach and long-term outcome remain uncertain.
METHODS: 17 children aged 4-17 yrs underwent stereotactically-assisted, transcallosal, interforniceal, trans-3^rdventricular resection of HH at our center since 1997. Seizure onset was at 0-34 mths (14[lt]6). All had gelastic seizures with or without complex partial features; 12 developed symptomatic generalized epilepsy (SGE) with tonic seizures and prominent spike-wave EEG patterns. Additional problems included precocious puberty (9), intellectual impairment (13), aggressive (11) and obsessive/autistic (5) behavior. Seizures were multiple daily in 16 and refractory to anticonvulsants (17), ketogenic diet (4), vagal nerve stimulator (2), partial subfrontal HH resection (2), thermocoagulation (1) and frontal corticectomy (1). All HH were sessile (size: 7x9x2mm to 26x30x39mm) with an intraventricular component. 7 were small and predominantly ([gt]90%) intraventricular. Hypothalamic attachment was unilateral (6), predominantly unilateral (6) or bilateral (5). Seizure, neurobehavioral and endocrine status are reported 1-45 (mean 14) mths after surgery.
RESULTS: 10 patients had 95-100% resection of the HH and 7 had 25-90% resection with complete or partial disconnection. 10 patients are seizure-free on reduced or no treatment (follow up 1-35 mths), 5 had [gt]90% seizure reduction, 1 had 75% reduction and 1 had [lt]50% reduction. Only 2 children had persistent major disabling seizures. Postoperative findings included a small thalamic infarct in 2 (mild residual hemiparesis in 1), appetite stimulation (9, persistent in 3), short-term memory complaints (6, persistent in 3), low thyroid or growth hormone (2), and transient somnolence (7), hypernatremia (Na⁺[gt]150)(8), temperature instability (3) and 3^rd nerve paresis (1). Precocious puberty remained in 7. 15 had marked improvements in behavior and quality of life. Seizure freedom was not associated with age at onset, pubertal status, HH size or attachment, age at surgery or [gt]95% resection. Normal intellect (p=0.09) and absence of SGE (p=0.04) were associated with seizure freedom.
CONCLUSIONS: Complete or near complete transcallosal resection/disconnection of HH can be achieved relatively safely. HH characteristics determine technical aspects of the surgery, but do not predict outcome. Although normal intellect and absence of SGE predicted seizure freedom, improvements in seizures, quality of life and behavior occurred in most children.
Support: Dr Freeman is supported by a Trainee Research Scholarship from the Murdoch Childrens Research Institute and by a Faculty Research Scholarship from the University of Melbourne, Faculty of Medicine, Dentistry and Health Sciences (Department of Paediatrics).