Tumor tracking gets personal
Bringing genomics to cancer care
By Eva Emerson
SAN DIEGO – A new way to identify cancer’s genetic scrambling may allow doctors to better monitor how individual patients respond to treatment and detect a recurrence of a tumor.
Wholesale juggling of chunks of DNA is common in cancer cells, but cataloging those changes hasn’t been easy. Now, a small study of colorectal and breast tumor cells shows that these genetic rearrangements can be reliably identified. Unlike with some other types of cancer, these genetic changes are unique to each individual’s tumor. While that individuality may stymie efforts to design therapies to target these rearrangements, their presence alone can be useful as a marker of cancer’s waxing and waning, scientists said during a news briefing February 18 at a meeting of the American Association for the Advancement of Science.
The team’s new technique to rapidly identify such rearrangements from individual tumors offers doctors a potentially powerful way to track tumor activity in patients. The research is scheduled to appear in the Feb. 24 issue of Science Translational Medicine
In a commentary in the same issue, Ludmila Prokunina-Olsson and Stephen Chanock of the National Cancer Institute in Bethesda, Md., call the new study a proof-of-principle “that lays an important foundation,” but caution that some issues still need to be worked out before the next-generation method used in the study becomes a “this-generation” medical tool.
Although the cost per patient – the team spent about $5,000 characterizing the specific genetic changes in each of the six people studied – is still high, most believe it will drop fairly quickly, said study author Victor Velculescu of the Kimmel Cancer Center at Johns Hopkins University in Baltimore. He and his team have applied for patents on the technology and hope that their work will lead to a commercially available, genomics-based blood test in the not far-off future.
Velculescu said that he is optimistic that the personalized tumor markers could be clinically available in two to three years.
In the study, scientists from Johns Hopkins and Life Technologies Corp. of Foster City, Calif., compared the genomic sequence from a tumor biopsy with that of normal tissue taken from the same person. Looking at four patients with colorectal cancer and two with breast cancer, scientists identified four to 17 unique genetic rearrangements present in each tumor. The rearrangements were dramatic in nature, Velculescu said, more like “switching the order of chapters in a book than a typo in a word in the text.” The team was able to find the rearrangements only by using powerful massively parallel sequencing methods capable of reading hundreds of millions of DNA letters as well as serious computer power to analyze the results.
For two of the colorectal cancer patients, the team took blood samples and, using bits of the patients’ tumor DNA sequences, probed the blood for signs of the tumor’s jumbled DNA. Scientists were able to detect very low levels of the tumor DNA, down to one tumor DNA fragment in 400,000 normal DNA fragments, Velculescu said.
The scientists tracked blood levels of the tumor genetic marker in one patient before and after surgery had removed the tumor. Levels of the genetic marker dropped immediately after the surgery, only to rise again, indicating the cancer was still present. The patient was then treated with chemotherapy and a second surgery and tested again. Levels went down substantially, but scientists were still able to detect a small amount of tumor DNA, consistent with a small part of the tumor that had spread to the patient’s liver.
Johns Hopkins oncologist Luis Diaz Jr., a coauthor of the study, noted that the blood levels of the biomarker “track very nicely with the clinical history of the tumor.”
The team is now testing other types of solid tumors to see if the method would also work on them. Also planned is a much larger, longer term study of patients to explore how the tumor markers relate to disease progression and other clinical issues.