Traditional methods of diagnosing certain brain cancers in children are deeply flawed, a new study shows. As a result, some children with these particular rare tumors have been getting the wrong diagnoses and, in some cases, the wrong treatment, the researchers say.
The errors were only revealed with the help of new tests that can look at tumor cells’ molecular profiles, said lead scientist Dr. Jim Olson of Fred Hutchinson Cancer Research Center. He urged his fellow pediatric brain cancer specialists to use such tests to diagnose their young patients to help make sure they receive optimal care. And families of young brain cancer patients, he said, should insist on the tests, which are called DNA methylation profiling.
“The biggest thing coming out of this is that we really need to do genomic analyses at the time of diagnosis for kids with brain tumors,” Olson said. “Tumors that can look exactly alike under a microscope can have very different biology and require entirely different forms of treatment.”
The particular brain cancers in question were known until recently as supratentorial primitive neuroectodermal tumors of the central nervous system, or CNS-PNETs. Traditionally, these cancers are diagnosed based on their locations in the brain and how they appear under a microscope.
But in the new study, published this week in the Journal of Clinical Oncology, the researchers showed that looks can be deceiving: A tumor that would be classified as a CNS-PNET based on its appearance could actually be any of one several other types of brain cancers based on its molecular profile. And the human experts missed it.
The new analysis used an emerging technology to inspect tumor samples from certain patients who had enrolled in a huge, ongoing trial led by Olson at more than 150 institutions worldwide. The trial compares two new intensive treatment combinations for pediatric patients with certain brain cancers.
The look back was prompted by new data in the field pointing toward unexpected diversity among pediatric brain cancers long thought to be the same. The newly published study is the first to assess diagnostic discrepancies for these particular cancers in the setting of a large multinational trial.
Thirty-one patients on the trial had been diagnosed in the traditional manner with CNS-PNET. But when the scientists went back and conducted DNA methylation profiling, they found 22 of the patients actually had cancers whose underlying biology was so different that they should never had been enrolled in the trial at all. (The scientists also found some errors in diagnosing another type of cancer included in the trial, pineoblastoma, which were not as common.)
“We were absolutely shocked to see how frequently the diagnosis made by the pathologists was in disagreement with the diagnosis made by the methylation studies,” Olson said, referring to the doctors who specialize in studying tissue samples in the lab.
The researchers also saw vast differences in outcomes open up between so-called CNS-PNET patients depending on what kind of tumor they actually had. For example, patients who in fact had a kind of brain cancer called a supratentorial embryonal tumor did much better than expected: More than three-quarters of them would live for at least five years after diagnosis, the scientists calculated. (Previous studies have reported that only about half of patients with supposed CNS-PNET survive that long.)
In contrast, the researchers discovered that 18 young patients on the trial actually had had a kind of brain tumor called a glioblastoma, which is terribly aggressive. Most patients with this diagnosis do not survive to five years. This was true of the patients on the trial as well, despite the intensive experimental treatment they received and the side effects they endured.
“We do not want to do that to any children,” Olson said. And, he said, this illustrates “the importance of finding things other than chemotherapy and radiation therapy to treat those kids.”
Olson’s hope is that this study’s findings will have an impact on the people he and his colleagues around the world care for.
“At very least the study helps us give proper prognosis to families,” Olson said. “Hopefully what it will do is get more kids treated properly.”
Other lead researchers on the study are Dr. Eugene Hwang of Children’s National Medical Center in Washington, D.C., and Dr. Marcel Kool of the German Cancer Research Center. The study also included 19 other collaborators from institutions throughout Germany and the U.S.
The National Institutes of Health, private philanthropic foundations and the German Cancer Consortium provided funding for the trial. The ongoing clinical trial is conducted by the Children’s Oncology Group, a publicly funded research network dedicated to finding better treatments for childhood cancers.
Susan Keown was a staff editor and writer at Fred Hutchinson Cancer Center from 2014-2022 who has written about health and research topics for a variety of research institutions. Find her on Twitter @sejkeown.