Two Unique Glioma Subtypes Revealed

A comprehensive molecular analysis of 1,122 diffuse gliomas from The Cancer Genome Atlas has shed new light on an often deadly cancer that accounts for 80% of malignant brain tumors (Cell 2016;164:550–63). Notably, the emergence of two previously unknown, divergent glioma subtypes has been hailed by the study researchers as a clinically relevant discovery with implications for patient management.

Pathologists have long classified diffuse gliomas by their appearance under the microscope: Grade II and grade III tumors, including oligodendrogliomas, are relatively slow-growing, while grade IV tumors, such as glioblastomas, are the most aggressive. More recently, gliomas have also been designated IDH-mutant or IDH–wild-type, based on the status of the IDH1 gene, determinable by immunohistochemistry.

“Generally speaking, two glioblastomas—one IDH-mutant, the other IDH–wild-type—may have the same grade [IV] and look very similar under the microscope, but the former tends to have a better clinical outcome,” says co–senior author Antonio Iavarone, MD, a professor of pathology and neurology at Columbia University Medical Center in New York, NY.

Given that IDH-mutant gliomas are frequently marked by the CpG island methylator phenotype (G-CIMP), or widespread DNA methylation, the researchers decided that epigenetic analyses would be a key part of their study. They were surprised to discover a subtype of IDH-mutant tumors “with lower methylation levels than would be expected, based on our traditional understanding of this phenotype,” Iavarone says. It turns out that not all IDH-mutant, G-CIMP gliomas are equally methylated; they can be further divided into “G-CIMP-high” and “G-CIMP-low” subtypes. Far from having the relatively favorable outcome of most IDH-mutant gliomas, G-CIMP-low tumors “displayed the molecular and clinical features of a more aggressive glioma,” Iavarone notes, “and were associated with significantly worse survival.”

The researchers also observed that, after treatment, recurring IDH-mutant gliomas that were previously G-CIMP-high could emerge as G-CIMP-low tumors. “It suggests that demethylation is a marker for disease progression, although we haven't figured out if it's a cause or consequence,” says co–senior author Roel Verhaak, PhD, an associate professor of bioinformatics and computational biology at The University of Texas MD Anderson Cancer Center in Houston. “We only looked at 10 recurrent gliomas in this study. It would be interesting to see how often this curious loss of methylation occurs in a larger, longitudinal analysis.”

Iavarone, Verhaak, and their colleagues also discovered another subtype, this time in the IDH–wild-type group. Most IDH–wild-type gliomas are grade IV and highly aggressive, but the researchers found a small percentage that, instead, resembled pilocytic astrocytoma—a curable (through surgical resection) pediatric brain tumor with a far more favorable prognosis. They dubbed this new subtype “PA-like.”

“These two glioma subtypes we've uncovered represent clinically actionable information that will help identify [G-CIMP-low] IDH-mutant patients who need stronger drugs, and [PA-like] IDH–wild-type patients who can be spared unnecessarily intensive treatment,” Iavarone says. “It's not that pathology is no longer important in grading gliomas. Rather, by molecularly segregating these tumors as well, we can characterize and treat them more accurately.”

This study, which was co-led by Houtan Noushmehr, PhD, from the University of São Paulo in Brazil, “makes the case for DNA methylation profiles as a real diagnostic tool,” Verhaak adds. “That's what I'm hoping to achieve at MD Anderson within the next 2 years. I'll also gladly be a little optimistic and predict that within, say, 10 years, methylation profiles will be standard in clinics nationwide.” –Alissa Poh

For more news on cancer research, visit Cancer Discovery online at http://cancerdiscovery.aacrjournals.org/content/early/by/section.