Risk factors for hypothalamic obesity after treatment for craniopharyngioma: A 10 years evaluation of all patients treated in the VUMC and AMC

Hypothalamic obesity is an major concern in children treated for craniopharyngioma. Not only the craniopharyngioma itself can cause hypothalamic obesity, but it can also be caused by the neurosurgical intervention or the subsequent radiation therapy. The primary goal of this study is to define risk factors for the development of hypothalamic obesity. Better understanding of risk factors for hypothalamic obesity may lead to an improvement in treatment decisions, prognosis and quality of life in survivors of craniopharyngioma. We retrospectively reviewed all charts of patients treated for craniopharyngioma between January 2002 and May 2012. Eligible cases were identified through our pathology database (n=42). For pediatric patients, obesity was defined as a body mass index (BMI) > + two standard deviations from the mean. In adult patients, obesity was defined as a BMI greater than 30 kg/m2. We reviewed all MRI studies to classify the hypothalamic involvement according to the Paris classification and all surgery reports to collect data about treatment strategies. Thirty-six patients (22 children) could be included with a mean follow-up of 53 months. Most common presenting symptoms were headaches (69.4%) and visual impairment (72.2%). Endocrine deficits were significantly more seen in pediatric patients (p=0.010). No significant difference was found in radiological features (p>0.05). Gross total resection was performed in 21 patients (58.3%) and partial resection or biopsy in 15 patients (41.7%). Immediate radiotherapy was performed in 7 patients with residual tumor. The 5- and 10-year event-free survival after complete resection was 90% and 75%, compared with 65% and 39% for partial resection (p=0.067). At last follow-up, obesity was present in 20 patients (55.6%), and more frequent in pediatric than in adult patients (68.2% versus 35.7%). Multivariate logistic regression showed age at diagnosis (p=0.028) and the degree of resection (p=0.046) to be significantly related to the development of obesity. Our findings shows hypothalamic obesity to occur more frequently in younger patients. Furthermore obesity seems to be more related to the degree of surgical resection than the size of the tumor and hypothalamic involvement at presentation. Considering the high rate of hypothalamic obesity after gross total resection and its negative influence on quality of life, we must recommend to consider less radical neurosurgery especially in pediatric patients, because obesity occurs more frequently after childhood craniopharyngioma. However, as long as there are no results of large prospective multicentre trials, there is no sufficient evidence available for a treatment algorithm. Further prospective research should be performed to develop a risk-based treatment algorithm.