Most Brain Metastases Have Actionable Mutations Not Found in Primary Tumor

More than half of brain metastases harbored clinically actionable genetic alterations that were distinct from those associated with the primary tumor.

Priscilla Brastianos, MD

More than half of brain metastases harbored clinically actionable genetic alterations that were distinct from those associated with the primary tumor, showed a review of more than 100 cases presented at the 2015 European Cancer Congress.

Half of the alterations were associated with sensitivity to cyclin-dependent kinase (CDK) inhibition. In 43% of cases, alterations predicted sensitivity to inhibition of the PI3K/AKT/mTOR pathway. Additionally, the study found that distant, non-brain metastasis did not harbor similar alterations and therefore could not be used as surrogates for predicting alterations in brain metastases, according to lead investigator Priscilla Brastianos, MD.

“Every brain metastasis displayed branched evolution,” said Brastianos, director of the CNS metastasis research program at the Massachusetts General Hospital Cancer Center in Boston. “Brain metastases harbored distinct clinically actionable genetic alterations, compared to their primary tumors. All brain metastasis regions harbored the same actionable alterations.”

The findings raise questions about the usefulness of so-called liquid biopsies for performing genetic analyses. Blood-based testing might detect genetic alterations associated with the primary tumor rather than brain metastases, she acknowledged. However, additional, larger studies are needed to determine the implications for liquid biopsies.

As many as 25% of cancer patients develop brain metastases, most commonly in association with lung cancer, breast cancer, and melanoma. Patients who develop brain metastases usually have a survival measured in months. A key to improving the outcome for those patients is to understand better the genetic and molecular makeup of the brain lesions.

“Currently, we have a limited understanding of how brain metastases genetically evolve from their primary tumor,” Brastianos said.

To inform the issue, investigators performed massively parallel sequencing analysis of 104 brain metastases from patients with corresponding specimens from the primary tumors. The specimens included 15 patients who had additional extracranial metastatic sites. The study was the largest of its kind conducted to date, according to Brastianos.

Investigators developed novel foundational tools to perform an integrated analysis of somatic mutations and copy number alterations. The analysis allowed Brastianos and colleagues to estimate the clonal architecture of the primary tumors and metastases and to reconstruct the phylogenetic tree for subclones in each patient.

The resulting data were consistent with branched evolution: a common ancestor shared by the primary tumor and brain metastasis but divergent evolution between the primary tumor and brain metastasis.

Investigators also wanted to know whether brain metastases harbored clinically actionable mutations not found in the associated primary tumors. To answer this question, they performed annotated analyses of the phylogenetic branches of trees for each brain metastasis and matched primary tumor.

“We did in fact find that brain metastases are genetically distinct from their matched primary tumor biopsies,” said Brastianos.

Overall, 53% of patients had clinically actionable alterations in the brain metastases branches but that were absent in the primary tumor. Subsequent analyses showed that brain metastases also had more genetically distinct alterations than lymph node and other extracranial metastases; meaning distal metastases are not reliable for predicting alterations in brain metastases.

“If one were to exclusively sample the primary tumor or regional lymph nodes for selection of the appropriate targeted therapies, one may miss potentially actionable mutations in the brain,” said Brastianos. “This presents a significant challenge to the application of precision medicine to brain metastases.”

Given the early nature of the results, Brastianos emphasized that the study results do not provide ample reason to change clinical practice with respect to analyzing brain metastases for actionable genetic alterations. More studies are needed to define the optimal approach to using the information for the greatest patient benefit.

Brastianos P, Carter S, Santagata S, et al. Genomic characterisation of brain metastases and paired primary tumours reveals branched evolution and potential therapeutic targets. Presented at: 2015 European Cancer Congress; September 25-29; Vienna, Austria. Abstract 2905.


View more from the 2015 European Cancer Congress