Key Takeaways:
- Researchers have developed a novel tumor-profiling tool they say captures a wider array of cancer-related genomic changes in a patients’ tumor and germline DNA, potentially enhancing drug treatment decisions.
- The Molecular Oncology Almanac combines an algorithm for interpreting tumor alterations in an individual cancer with a catalogue of molecular features that are associated with tumor sensitivity or resistance to drugs and disease prognosis.
- Together, the algorithm and the knowledge base will, compared with current profiling methods, help oncologists obtain a bigger picture of the molecular changes in a patient’s tumor, aiding the matching of targeted drugs to the patient’s cancer.
Researchers at Dana-Farber and the Broad Institute of MIT and Harvard have created a novel Molecular Oncology Almanac that combines an algorithm, or computational method, for interpreting genomic alterations in a patient’s tumor, along with a knowledge base of molecular changes reported to be associated with tumor behavior.
They say that the components of this tool will capture a more comprehensive array of a tumor’s abnormal features than is obtained with current methods to evaluate how an individual’s cancer will likely respond to certain targeted drugs or be resistant to them.
“All of this is wrapped up in the Molecular Oncology Almanac and made easy to use and report,” says Brendan Reardon, a researcher at Dana-Farber and the Broad Institute and first author of the publication in Nature Cancer describing the new tool.
The future of precision oncology
The scientists tested the use of the algorithm and knowledge base using records of cancer patient cohorts. They say that by casting this wider net, the method “increases clinical actionability over conventional approaches.” It has not yet been tested in real-world clinical applications to inform treatment decisions.
“We think this represents the future of where precision oncology is going,” says Eliezer Van Allen, MD, of Dana-Farber and the Broad Institute. He is the corresponding author of the publication.
“We’ve learned over the last few years that for all the excitement about precision oncology, we still have a long way to go,” Van Allen says. “The patient’s cancer genome is very complicated. And there’s a lot of interactive molecular machines that are contributing to advanced cancers and drug resistance. Putting that puzzle together and making that puzzle unique for each person is a very hard problem.”
[Learn more about precision medicine and immunotherapy.]
The goal of precision oncology — as distinguished from one-size-fits-all standard chemotherapy treatments — is to identify specific molecular changes in an individual patient’s tumor that can be attacked or blocked by highly targeted drugs. Some tumor genetic changes, such as mutations or excess copies of a segment of DNA, make the tumor more vulnerable to a targeted drug or, in some cases, confer resistance to such a drug. Other changes may predict the prognosis, or how the tumor will behave, a particular patient. In certain forms of cancer, identifying patient tumor profiles and matching them with “smart” drugs have improved outcomes.
Currently, say the researchers, the tumor changes used to guide treatment are generally “first-order” features — variations impacting a single gene, for example, that might make the tumor vulnerable or resistant to a particular targeted drug. But cancers have myriad other molecular events — which the scientists are calling “second-order” features, such as the number of cancer-related mutations in a tumor cell’s genome, or the relationship between different first-order events — for example, overlap between findings in the DNA and RNA.
“One actually needs to consider those two things together — different layers of information — and those second-order phenomena are what we’re trying to capture with the Molecular Oncology Almanac,” explains Van Allen.
The knowledge base component of the almanac is a database or catalogue of cancer-related molecular features that are cited in the medical literature. This includes U.S. Food and Drug Administration approvals, clinical guidelines, academic research reports, clinical and preclinical studies, and others inferred from mathematical modeling. It is “provided as a data source to identify molecular features that are clinically relevant when evaluating a molecular profile,” says Reardon.
As a simulation to test the almanac, the researchers used records of several groups of patients to note what results the almanac would have revealed in terms of clinically relevant molecular changes. In the case of four cohorts of patients with metastatic melanoma, metastatic prostate cancer, kidney cancer, and pediatric osteosarcoma, they compared those findings relative to a prior method. In another experiment, they evaluated their method on a group of patients who received targeted therapies through a prospective clinical trial and compared their findings to what was done in the actual trial. The researchers found that the almanac highlighted a median of two therapies per patient and, in about half of patients, recommended a therapy of the same therapeutic strategy as was administered in the trial.
While this new tool is not yet ready for guiding therapy in cancer patients, the researchers “are thinking very deeply how one actually tests a precision oncology tool like this across large sets of patients, different kinds of cancer, and where and how it might actually be useful in the real world,” says Van Allen.