A clinicopathologic study of diencephalic pediatric low-grade gliomas with BRAF V600 mutation.

Abstract

Among brain tumors, the BRAF (V600E) mutation is frequently associated with pleomorphic xanthoastrocytomas (PXAs) and gangliogliomas (GGs). This oncogenic mutation is also detected in ~5% of other pediatric low-grade gliomas (LGGs) including pilocytic astrocytomas (PAs) and diffuse astrocytomas. In the current multi-institutional study of 56 non-PXA/non-GG diencephalic pediatric LGGs, the BRAF (V600) mutation rate is 36%. V600-mutant tumors demonstrate a predilection for infants and young children (<age 3) and have a higher tendency for multicentricity. On neuroimaging, BRAF (V600)-mutant tumors appear as nodular, yet infiltrative contrast-enhancing masses. Morphologic examination reveals a monophasic, predominantly compact and partially infiltrative architecture. Due to the lack of classic morphologic features associated with PAs, pilomyxoid astrocytomas (PMAs), or diffuse astrocytomas, 75% of the BRAF (V600)-mutant tumors could not be definitively classified on initial histopathologic evaluation. At a median follow-up of 55months, the 5-year progression-free survival (PFS) rate for BRAF (V600)-mutant diencephalic low-grade astrocytomas (LGAs) was 22±12%, shorter than BRAF (V600)-WT PAs (52±13%) but higher than PMAs (10±6%). Of note, long-term PFS was observed in several adolescent patients with BRAF (V600)-mutant tumors. In children aged 0-12years, 5-year PFS rate and median PFS in BRAF (V600)-mutant LGAs are 9±9% and 19months (95% CI 3-37months), respectively. The PFS is comparable to that in BRAF (V600)-WT PMAs (5-year PFS rate: 10±9%; median PFS: 15months, 95% CI 3-32months; p=0.96) and significantly shorter than BRAF (V600)-WT PAs (5-year PFS rate: 46±13%; median PFS: 51months, 95% CI 20-∞ months; p<0.05). In summary, diencephalic BRAF (V600)-mutant pediatric LGAs are associated with unique clinicopathologic features and have a more aggressive clinical course, especially in children under age 13. The low rate of CDKN2A deletion also suggests that these tumors are molecularly distinct from secondary pediatric high-grade gliomas.