Malignant Transformation of Low-Grade Gliomas in Children: Lessons Learned From Rare Medical Events

Gliomas, the largest group of pediatric CNS cancers, are histologically classified as lowand high-grade tumors. Low-grade gliomas (LGGs), which are often mislabeled as benign tumors, can cause significant morbidity and sometimes even death, particularly in the case of midline tumors that are not amenable to gross surgical resection. High-grade gliomas (HGGs) are among the deadliest pediatric CNS cancers, despite the use of multimodality therapy. Several publications have recently unveiled the genome-wide molecular characteristics of pediatric gliomas, which include abnormalities in copy number, gene sequence and expression, and methylation. These data showed that most of the molecular characteristics of pediatric LGGs and HGGs differ from those of their adult counterparts and are dependent on tumor location. Long before the availability of whole-genome molecular studies, clinical characteristics and limited genetic information had already suggested major differences between adult and pediatric gliomas. For example, whereas 50% to more than 90% of adult LGGs undergo malignant transformation to secondary HGGs, malignant transformation is a rare phenomenon in children. In adults, isocitrate dehydrogenase 1 (IDH1) and IDH2 mutations are common in LGGs and secondary HGGs. The analysis of the cascade of molecular effects as a result of IDH mutations have generated substantially improved understanding of the biology of adult gliomas. However, in children, it has remained unknown whether malignant transformation of LGG is driven by the molecular mechanisms observed in adult tumors. In the article that accompanies this editorial, Mistry et al report the results of detailed genome-wide and/or targeted molecular analyses in the tumors of 26 children who experienced malignant transformation of LGGs. This was not a task for the faint of heart. First, the rarity of this event required a study period of almost 30 years, and some patients were observed for more than 10 years before they experienced malignant transformation. Second, the radiologic and histologic diagnosis of malignant transformation is far from simple. The appearance on magnetic resonance imaging of LGGs undergoing malignant transformation often does not reveal any characteristics that suggest a change in histology and aggressiveness. Therefore, an unequivocal diagnosis is dependent on physicians’ awareness of the possibility of this phenomenon and their pursuit of histologic confirmation. Surgery for confirmation is commonly fraught with obstacles depending on the location of the tumor, the associated technical difficulty of the procedure, and the willingness of patients/parents to undergo the operation. Finally, the histologic diagnosis and differentiation of lowversus high-grade tumors may be controversial, even among experts in the field. These obstacles, together with the rarity of malignant transformation of LGGs, have prevented the molecular analysis of this phenomenon in large cohorts of children. Mistry et al performed whole-exome sequencing of seven secondary HGGs and found mutations in several genes, the most common of which was BRAF V600E. Targeted sequencing of genes in the remaining high-grade tumors showed some findings that overlapped with those of primary pediatric HGGs (eg, the H3F3A K27M mutation in midline tumors) and confirmed the high prevalence of BRAF V600E in hemispheric secondary HGGs. CDKN2A deletion was also observed in more than half of the secondary HGGs. BRAF V600E and CDKN2A deletion were commonly identified in matched lowand secondary high-grade tumors and occurred concomitantly in 75% of transformed LGGs. When LGGs that underwent malignant transformation were compared with a control group of nontransformed tumors, the BRAF V600E mutation and CDKN2A deletion were significantly associated with malignant transformation of LGGs. By further comparing BRAF-mutated HGGs with those with wild-type BRAF and with a separate cohort of primary pediatric HGGs, the authors showed that BRAF-mutated HGGs may comprise a separate entity with distinct clinical characteristics and a comparatively improved prognosis. These new findings are likely to generate lively interest and additional research in the field of pediatric gliomas. Many centers in North America already test for BRAF V600E status by genotyping and/or immunohistochemistry. Approximately 10% to 15% of pediatric glioblastomas are known to carry the mutation. Is it possible that a subset of these BRAF-mutated tumors are secondary HGGs with an unidentified, preceding low-grade tumor? Likewise, BRAF V600E occurs in approximately 6%, 23%, 33%, and 70% of pediatric pilocytic astrocytomas, diffuse astrocytomas, gangliogliomas, and pleomorphic xanthoastrocytomas, respectively. The results reported by Mistry et al suggest that it would be interesting to prospectively evaluate the natural history of BRAF-mutated pediatric LGGs, including their likelihood of malignant transformation. As noted by the authors, the JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 33 NUMBER 9 MARCH 2