BRAF mutation and CDKN2A deletion define a clinically distinct subgroup of childhood secondary high-grade glioma

Abstract

Purpose: To uncover the genetic events leading to transformation of pediatric low-grade glioma (PLGG) to secondary high-grade glioma (sHGG).

Patients and methods: We retrospectively identified patients with sHGG from a population-based cohort of 886 patients with PLGG with long clinical follow-up. Exome sequencing and array CGH were performed on available samples followed by detailed genetic analysis of the entire sHGG cohort. Clinical and outcome data of genetically distinct subgroups were obtained.

Results: sHGG was observed in 2.9% of PLGGs (26 of 886 patients). Patients with sHGG had a high frequency of nonsilent somatic mutations compared with patients with primary pediatric high-grade glioma (HGG; median, 25 mutations per exome; P = .0042). Alterations in chromatin-modifying genes and telomere-maintenance pathways were commonly observed, whereas no sHGG harbored the BRAF-KIAA1549 fusion. The most recurrent alterations were BRAF V600E and CDKN2A deletion in 39% and 57% of sHGGs, respectively. Importantly, all BRAF V600E and 80% of CDKN2A alterations could be traced back to their PLGG counterparts. BRAF V600E distinguished sHGG from primary HGG (P = .0023), whereas BRAF and CDKN2A alterations were less commonly observed in PLGG that did not transform (P < .001 and P < .001 respectively). PLGGs with BRAF mutations had longer latency to transformation than wild-type PLGG (median, 6.65 years [range, 3.5 to 20.3 years] v 1.59 years [range, 0.32 to 15.9 years], respectively; P = .0389). Furthermore, 5-year overall survival was 75% ± 15% and 29% ± 12% for children with BRAF mutant and wild-type tumors, respectively (P = .024).

Conclusion: BRAF V600E mutations and CDKN2A deletions constitute a clinically distinct subtype of sHGG. The prolonged course to transformation for BRAF V600E PLGGs provides an opportunity for surgical interventions, surveillance, and targeted therapies to mitigate the outcome of sHGG.

Fig 1.

Clinical, genetic and molecular characteristics in (A) secondary high-grade glioma and (B) pediatric low-grade glioma (PLGG) undergoing transformation. Frequency of genetic or molecular alterations indicated as percentage at the far right of each row. AA, anaplastic astrocytoma; AGG, anaplastic ganglioglioma; APXA, anaplastic pleomorphic xanthoastrocytoma; bMMRD, biallelic mismatch repair deficiency syndrome; CPS, cancer predisposition syndrome; Dx, diagnosis; GBM, glioblastoma; GG, ganglioglioma; HG, high grade; LFS, Li-Fraumeni syndrome; LG, low grade; LGA, low-grade astrocytoma; MT, malignant transformation; N/A, not applicable; PA, pilocytic astrocytoma; PXA, pleomorphic xanthoastrocytoma; RT, radiotherapy. p53 dysfunction is defined by more than 50% p53 immunopositive tumor cells and/or TP53 mutation.

Fig 2.

Correlative studies identifying that (A) BRAF V600E is enriched in pediatric low-grade gliomas (PLGGs) that later transform versus PLGGs that do not transform (P < .001); (B) BRAF V600E distinguishes secondary from primary high-grade gliomas (HGGs) in children (P = .0023); and (C) CDKN2A deletions are enriched in PLGGs that later transform versus PLGGs that do not transform (P < .001). Statistical significance was evaluated using Fisher's exact test.

Fig 3.

(A) Age distribution for BRAF mutant and BRAF wild-type patients at initial diagnosis of pediatric low-grade glioma (PLGG). Eighty-nine percent of patients who were BRAF positive were diagnosed between 5 and 10 years of age. (B) BRAF mutant PLGGs have a prolonged latency to malignant transformation (MT) compared with BRAF wild-type PLGGs (P = .0389). Horizontal lines indicate medians. Statistical significance was evaluated using an unpaired t test. (C) Kaplan-Meier survival estimates showing improved overall survival after initial diagnosis for BRAF mutant versus wild-type PLGGs that later transform. (D) Primary high-grade gliomas (HGGs) have a worse overall survival after initial diagnosis than BRAF wild-type secondary HGG (P = .0163). Nontransformed PLGGs have a favorable overall survival compared with all other groups, at 98% ± 0.5%.