Official reprint from UpToDate®
www.uptodate.com ©2017 UpToDate, Inc. and/or its affiliates. All Rights Reserved.

Medline ® Abstract for Reference 10

of 'Clinical features, diagnosis, and management of von Hippel-Lindau disease'

Long-term natural history of hemangioblastomas in patients with von Hippel-Lindau disease: implications for treatment.
Ammerman JM, Lonser RR, Dambrosia J, Butman JA, Oldfield EH
J Neurosurg. 2006 Aug;105(2):248-55.
OBJECT: In the course of their lives most patients with von Hippel-Lindau (VHL) disease require treatment for several symptom-producing hemangioblastomas of the cerebellum, brainstem, or spinal cord. However, many tumors never produce symptoms and do not require treatment. Detection at an early stage of lesions that will later produce symptoms and ultimately require treatment would allow for earlier excision of hemangioblastomas of the spinal cord, brainstem, or cerebellum, and may identify cerebellar hemangioblastomas that can be treated with radiosurgery at a stage before treatment is contraindicated because of tumor size or the presence of an associated cyst.
METHODS: To identify features predictive of symptom development that might allow for earlier treatment of smaller, presymptomatic hemangioblastomas in patients with VHL disease, the authors reviewed and analyzed the serial clinical and imaging findings in all patients with VHL disease who were followed up at the National Institutes of Health for more than 10 years. Features predictive of symptom formation were determined by recursive partition and regression analyses. Nineteen patients (10 men and nine women; mean age 32.6 +/- 11.6 years) harboring a total of 143 hemangioblastomas were identified (mean follow-up duration 12.4 +/- 1.4 years). Hemangioblastomas were located in the cerebellum (68 hemangioblastomas, 48% of patients), brainstem (17 hemangioblastomas, 12% of patients), and spinal cord (58 hemangioblastomas, 40% of patients). Despite measurable growth in almost all hemangioblastomas (138 lesions, 97% of patients), only 58 (41% of patients) became symptomatic. Hemangioblastomas grew in a stuttering pattern. (mean growth period 13 +/- 15 months, mean quiescent period 25 +/- 19 months). Twenty-six (45%) of the hemangioblastomas that eventually produced symptoms were not among the tumors that were apparent on the initial MR imaging study. Depending on location, the hemangioblastoma size and/or tumor and cyst growth rates predicted symptom development and the need for treatment (p<0.05). Cerebellar hemangioblastomas growing faster than 112 mm3/ month or larger than 69 mm3 with associated tumor and cyst growth rates greater than 14 mm3/month became symptomatic (100% sensitivity, 72% specificity). Brainstem hemangioblastomas larger than 245 mm3 with growth rates greater than 0.1 mm3/month became symptomatic (75% sensitivity, 89% specificity). Spinal hemangioblastomas larger than 22 mm3 became symptomatic (79% sensitivity, 94% specificity).
CONCLUSIONS: Because hemangioblastomas exhibit a stuttering growth pattern, frequently remain asymptomatic, and do not require treatment for long intervals, unqualified radiographic progression is not an indication for treatment. Basing the decision to intervene in individual tumors solely on radiographic progression would have resulted in approximately four additional procedures per patient during the 10-year study period. Threshold values are presented for tumor size and/or tumor and cyst growth rates that can be used to predict symptom formationand future need for treatment.
Surgical Neurology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Maryland 20892-1414, USA.