Personalizing Biomarkers for Cancer
Demonstrating a new use for DNA-sequencing technology, researchers at the Johns Hopkins Kimmel Cancer Center have developed individualized biological markers for patients with cancer. The biomarkers, which can be detected in blood, could potentially help doctors manage the disease in patients.
In a pilot study, the researchers used next-generation DNA sequencing tools to identify genetic changes called chromosome rearrangements. The changes, which occur when chromosome regions are swapped or merged inappropriately, were then used to create genetic fingerprints for individual tumors.
To test potential clinical applications, the researchers first identified rearrangements in the tumors of two patients with colorectal cancer. They then showed that blood tests could pick up these same changes in DNA shed into the circulatory system by tumors, allowing them to track levels of abnormal tumor DNA as a biomarker over time.
In one case, the patient’s biomarker levels dropped substantially after a tumor was removed, before rising again later. The levels declined once more after subsequent chemotherapy and a second surgery, though not to zero. This was likely because of a small lesion that remained in the patient’s liver, the researchers said.
“This study is a proof of principle, and it demonstrates that the approach worked well for these particular patients,” said lead investigator Dr. Victor Velculescu, who co-directs the cancer biology program at Johns Hopkins. The strategy represents one of the first applications of next-generation, whole-genome sequencing that could be clinically useful for patients, he said.
A report on the method, called Personalized Analysis of Rearranged Ends (PARE), will appear in the February 24 issue of Science Translational Medicine.
“This is an exquisitely thoughtful and useful application of DNA sequencing technology,” said Dr. Stephen Chanock, who directs the Laboratory of Translational Genomics in NCI’s Division of Cancer Epidemiology and Genetics and who wrote an editorial accompanying the study.
In addition to marking the tumor for forensic purposes, the strategy could reveal insights into the biology of tumors. “What this method allows us to do very nicely is to quickly identify sets of markers, and we can see how they change during the disease,” Dr. Chanock continued. “The presence or absence of these markers may reveal insights into the life cycle of a tumor.”
Because some rearrangements may disappear and others may emerge over time, it will be important to test sets of markers rather than individual rearrangements, Dr. Chanock noted. This would reduce the chances that a patient might be found incorrectly to have no cancer when in fact the disease had simply gone undetected.
The Johns Hopkins team did not set out to develop personalized biomarkers when the project began. They sequenced the genomes of six colorectal and breast tumors, hoping to find recurring chromosome rearrangements like those that are hallmarks of cancers of the blood. But they found none. Instead, the sequencing revealed that each tumor had these rearrangements and each one was different—on average, nine per tumor.
Right away the researchers realized that rearrangements were promising biomarkers. Not only were the changes specific to the tumors, but these were large, dramatic events in the genome. They would be easier to detect than point mutations, where a single letter of DNA is altered.
“This study is an exciting step forward in terms of how we might develop and use these types of biomarkers in patients,” said NCI Deputy Director Dr. Anna Barker. “While there is a lot more work to do, it opens up some new windows for investigation.” If the biomarkers are reproducible and highly sensitive, they could help address a range of clinical needs, such as improving the detection of residual disease following surgery, she added.
Dr. Velculescu and coauthor Dr. Luis Diaz, an assistant professor of oncology at the Kimmel Cancer Center, presented the findings last week at the annual meeting of the American Association for the Advancement of Science. They said the approach could be made broadly available to patients in several years if costs could be brought down and the findings were confirmed. The cost was about $5,000 per patient, primarily for DNA sequencing. But the cost of next-generation sequencing has declined sharply and is expected to keep falling, they noted.
The Johns Hopkins group has expanded their study to identify other common cancers that are amenable to this strategy. It will also be important to establish how early in the development of cancer the markers are detectable.
The researchers are hopeful that this approach might help distinguish patients who are at risk of relapse and need additional therapy from those who could be safely spared more treatment. Such questions will need to be addressed in prospective clinical trials.
“In terms of next steps, it will be important to show that the new strategy is not just novel but also useful in real clinical situations,” noted Dr. Sridhar Ramaswamy, an assistant professor of medicine at the Massachusetts General Hospital and Harvard Medical School, who directs a laboratory focused on translational cancer genomics.
“The million-dollar question when you have a patient who has undergone treatment for cancer is does the person have incipient disease,” said Dr. Ramaswamy. “Circulating tumor cells are one way to pick up microscopic cancers that are not readily apparent, and this is potentially a different way.”
What’s exciting about the strategy, he added, is that the researchers are using the powerful new genetic tools to go after individualized mutations in tumors. The tools are more commonly used to identify recurrent genetic changes shared among many patients.
Thousands of Genomes
The genomes of thousands of patients with cancer will be sequenced over the next few years. As the data from large-scale projects such as The Cancer Genome Atlas and the Pediatric Cancer Genome Project become available, researchers will increasingly focus on translating the results into knowledge and clinical tools, Dr. Barker predicted.
“The community is really starting to think about new ways to detect genomic changes in tumors and about how to use the changes as biomarkers,” she said. “All of these studies are building an invaluable cancer genomics knowledge base, and this will move us toward improving the care of patients with cancer.”
—Edward R. Winstead