Skip to main content
Premium Trial:

Request an Annual Quote

CHOP Researchers Find Gene Fusions are Harbingers of Aggressive Disease in Pediatric Thyroid Cancer


NEW YORK – Children with thyroid cancers harboring RET and NTRK fusion oncogenes may be less likely to achieve remission and more likely to experience metastases than patients with cancers driven by BRAF mutations, according to research out of the Children's Hospital of Philadelphia (CHOP).

The research, published recently in the Journal of Clinical Oncology, could have implications for treatment, encourage greater uptake of comprehensive genomic profiling for younger cancer patients, and impact the way the field approaches pediatric thyroid cancer on the whole.

"It emphasizes that thyroid cancer in kids is different than thyroid cancers in adults," said Aime Franco, a CHOP investigator and a lead author on the paper. "We can't always apply the rules from one population to another."

Adult papillary thyroid cancers are often grouped into molecular subtypes based on whether their RNA expression signatures are RAS-like or BRAF-like. On the other hand, gene fusions are far less common in adult tumors, making it less relevant from a prognostic standpoint to categorize cancers based on fusion status for these patients.

"On an oncogene basis, if you just say RAS-like or BRAF-like, which are the two adults' buckets, it hides away the fusions, which are less common in adults," explained Andrew Bauer, director of CHOP's Pediatric Thyroid Center and another lead researcher on the JCO study, adding that BRAF-driven cancers among adults tend to be more treatment refractory.

When Franco, Bauer, and colleagues analyzed genomic profiles from 131 pediatric thyroid cancer tumors, however, new prognostic categories emerged. They identified RET or NTRK fusions in 43 samples, while 26 harbored BRAF mutations and just nine harbored RAS mutations.

Looking at the genomic profiles alongside clinical information, the researchers found the fusion-driven pediatric cancers correlated with higher risk. A year after undergoing surgery, 42 percent of patients with RET/NTRK fusions, which were grouped together based on their known similarities, were in remission, whereas the same was true for significantly more patients (61.5 percent) with tumors driven by BRAF mutations. There were too few RAS-mutated tumors to draw definitive conclusions about outcomes.

In terms of the cancers' aggressive behavior, the clinical profiles associated with the analyzed tumors showed that far more RET/NTRK fusion-positive cancers (60 percent) spread throughout the neck and lymph nodes than did BRAF-mutated cancers (17 percent). The same was true with distant metastases, with 36 percent of tumors in the RET/NTRK fusion-driven subgroup spreading widely, while not one of the BRAF-mutated tumors did.

Of the 131 pediatric thyroid tumors that Franco, Bauer, and colleagues analyzed in their study, 65 had already been sequenced using Asuragen's miRInform Thyroid Test, which the researchers described as a first-generation assay. The miRInform test interrogates the presence of common mutations in BRAF, HRAS, KRAS, and NRAS as well as three fusion transcripts, but is limited in that it does not analyze DICER1 mutations, NTRK fusions, or any novel RET partners.

The other 60 tumors, meanwhile, were sequenced using the more comprehensive CHOP Solid Tumor Panel and CHOP Cancer Fusion Panel, which together assess 238 cancer-associated genes, and probe 673 exons within 110 such genes.

The different tests used in the study impacted the prognostic classifications of the tumors. For example, in the samples analyzed by miRInform, the percentage of tumors with RET and NTRK fusions — the higher risk tumors, according to the researcher's findings — was far lower than it was among the samples assessed by the CHOP panels. Half of the tumors analyzed by miRInform were classified as indeterminate, versus just over 30 percent with the more comprehensive CHOP assays.

The analysis illustrates how the specific test used for prognostic classification can have downstream treatment implications. "The [miRInform] panel that we used years ago was pretty good at the time, but two important thyroid drivers weren't on that panel [and] we didn't have NTRK fusions on that panel at all," Bauer said, pointing out that missed RET partners could be important for patients with aggressive disease that might benefit from a RET inhibiting agent.

While there are clear benefits with using the broader panels, Bauer acknowledged that more comprehensive tests have their own challenges, particularly when they turn up incidental findings with unclear significance.

According to Franco and Bauer, part of what makes this research so significant is the fact that several targeted drugs have entered the market for fusion-driven thyroid cancers in recent years.

Bayer's Vitrakvi (larotrectinib) and Genentech's Rozlytrek (entrectinib), for example, are both available for heavily treated, refractory adult and pediatric patients with NTRK fusion-positive solid tumors. Eli Lilly's Retevmo (selpercatinib) and Genentech's Gavreto (pralsetinib) are approved for patients over age 12 with treatment-refractory tumors with RET fusions.

"Now that there are [these drugs], pediatric endocrinologists and oncologists are … trying to figure out who benefits most and how to bring that into practice," Bauer said.

The finding that NTRK and RET fusions in pediatric thyroid cancer may be harbingers of more aggressive disease and risk of relapse after surgery may also encourage more oncologists to use comprehensive genomic profiling in younger patients, Franco added.

"This is pretty routine clinically at major centers … but when you get into community centers and smaller hospitals, it's not as routine," she said, noting that insurance coverage for these tests is inconsistent for the pediatric population since much of the data on the utility of genomic profiling to prescribe precision care comes from studies involving adults, rather than children.

To this end, CHOP researchers have been increasingly advocating for broader fusion testing in pediatric cancer patients.

Going forward, Franco, Bauer, and their colleagues hope to validate their findings in larger patient populations and in prospective trials. In addition to studies that stratify these patients for earlier intervention with targeted therapies, Bauer envisions studies in which patients can also be risk stratified to de-escalate surgery.

For example, if genomic profiling suggests a lower-risk cancer, that patient might not need a full neck dissection or full thyroidectomy. "If you know it's a RAS mutation that rarely goes to lymph nodes, [you could] potentially save half the patient's thyroid," Bauer said. "You could save half their thyroid, which has a lifelong impact on quality of life, and reduces the potential complications of surgery."

On the translational research side, Franco is also conducting studies that go beyond the genomic landscape to better define what drives pediatric thyroid cancers to respond better to therapy, or to behave aggressively.

"This is now looking beyond the genetics and at differences in the tumor microenvironment," she said. "Are there different cells that are recruited into these tumors? Do the gene expression patterns look different? We want to understand how and why the mutations that are present in pediatric and adult patients seem to induce different responses in pediatrics versus adults."