CHICAGO — Traditionally, brain tumors are identified and diagnosed under the microscope. Sometimes a small sample or subtle differences can make the identification of the exact type of cancer tricky. But now, for the first time, doctors are profiling tumors by using their DNA fingerprints.
When he was just five weeks old, Steve Stewart became easily irritable and started vomiting.
“Three weeks after him being home, actually, my wife delivering him, his head just started to swell,” said the baby’s father, Steve Stewart Sr.
The cause of that swelling turned out to be a congenital brain tumor. It measured 9.7 centimeters in length.
“If we look at the MRI images, it literally took up half to more than half of his skull,” said Dr. Angela Waanders, head of pediatric neuro-oncology at Lurie Children's Hospital.
Waanders is the executive board co-chair for the Children's Brain Tumor Network - an international research consortium.
Waanders says congenital brain tumors, like Steve’s, can be challenging for surgeons. The tumors bleed and can damage an infant’s early brain development.
“So, even doing a biopsy puts a patient at risk,” she said.
For pathologists, that meant very little tissue could be harvested from the tumor to identify it.
“In this case, it was an endoscopic biopsy. So, we're talking about millimeters of tissue,” said Dr. Nitin Wadhwani, director of neuropathology at Lurie Children’s Hospital.
Knowing exactly what kind of tumor they’re dealing with he says is important for setting up a treatment plan.
“In this scenario we knew a limited panel, immuno-stains would give us some guidance, but we needed something else,” said Wadhwani.
That something else was a process called DNA methylation profiling.
“Each individual spot in the DNA is either methylated or it's not. So, it's binary. It's either a one or a zero,” said Dr. Craig Horbinski, director of neuropathology at Northwestern University Feinberg School of Medicine.
“All of a sudden, a pattern emerges that reveals the identity of that tissue. It's very much like a fingerprint,” said Horbinski.
Horbinski’s team is among the first in the nation to use the technology for diagnostic testing of tumors. The inexpensive technique, he says, could adjust a brain tumor diagnosis 25% to 33% of the time, even with the smallest sample.
“All we need is about one nanogram of DNA, which is one-billionth of a gram of DNA. We can get that easily from a tissue the size of the head of a pin. So, it's extraordinarily small,” said Horbinski.
It also helps understand the origin of rare tumors and even predicts how effective specific therapies may be.
Using artificial intelligence, they’ve already discovered new types of tumor entities that they didn’t even know existed.
“That's the beauty of the artificial intelligence aspect of this is the more cases we have, the better the classification system becomes,” said Horbinksi.
For five-month-old Steve Stewart, the DNA fingerprint of his specific type of brain carcinoma helped doctors target their treatment.
“We got the results back rapidly. And that is a game-changer in terms of clinical diagnostics for difficult to diagnose brain tumors,” said Waanders.
It’s a game-changer that could give the even smallest patients a better chance of survival.