For someone with a dangerous tumor, surgical removal of the cancer may bring immediate relief but leave the person with an uncertain future. Sometimes, remnant cancer cells resurge, and there's usually no easy way to tell when this is happening. A new technique that uncovers genetic glitches specific to each person's tumor could let doctors monitor for cancer recurrence simply by testing a blood sample.
Cancer cells often have large-scale rearrangements of chromosomes that don't occur in normal cells and are easy to detect with current genetic methods. For example, patients with chronic myeloid leukemia, a blood cancer, nearly always have swapped pieces of chromosomes 9 and 22 in their cancer cells. Doctors can test for this glitch in blood cells to determine if a treatment is working. That hasn't been possible with solid tumors because no DNA rearrangement has been common enough to generate a widely useful test for cancer patients. Instead, people often learn that their cancer has come back or spread only when they feel symptoms or the tumors show up with an imaging test such as a CT scan.
Using next-generation DNA-sequencing machines, however, researchers at Johns Hopkins University in Baltimore, Maryland, have found a way to use chromosomal arrangements to detect DNA from an individual's solid tumor. The team took biopsies of breast or colon cancer tumors from six cancer patients and sequenced the entire genome in each tumor's cells. In each case, the researchers identified rejiggered chromosomes specific to the tumor but not seen in a person's normal DNA. All six tumors had at least four unique chromosomal rearrangements, and some growths had as many as 15, cancer biologist Victor Velculescu and his team report 24 February in Science Translational Medicine and at the annual meeting of the American Association for the Advancement of Science in San Diego, California (AAAS publishes ScienceNOW).
Next, the researchers showed that they could use a DNA-amplifying technique called polymerase chain reaction to detect the minute amounts of this distinctive tumor DNA in a sea of normal DNA. The team also showed that the technique could pick out people with cancer: The blood of two people who had colon tumors that had not yet been removed tested positive for their specific cancer biomarker, whereas blood from healthy people tested negative. The researchers then used the personalized probes to track one of the colon cancer patients' responses to various treatments. The amount of cancer-specific DNA in the person's blood dropped in the hours after surgery, rose again during the next few weeks, and then dropped again after chemotherapy and surgery for a secondary tumor in the liver. Oncologists could use these biomarkers to find out if a treatment is working or if surgery has missed some cancer cells, says Velculescu.
Many clinical researchers are already using a different blood test to monitor patients' response to therapy—counting circulating tumor cells. But these cells are extremely rare and can't be detected in some patients. The chromosomal biomarkers should be much more sensitive and are "extraordinarily specific. The chance of getting a false positive is essentially zero," says Velculescu.
Medical oncologist Klaus Pantel of the University Medical Center Hamburg-Eppendorf in Germany, a leader in using circulating tumor cells to monitor cancer, says his group has thought about the same sequencing strategy. "My colleague will cry. It is a great approach." But as Pantel and the authors themselves note, the costs--$5000 per patient just to find the unique sequences—are too high for routine use.
Geneticist Stephen Chanock of the National Cancer Institute in Bethesda, Maryland, also cautions that before personalized genome-based cancer detection moves to the clinic, the Johns Hopkins team needs to sequence many more tumors to make sure that they always have characteristic chromosomal rearrangements and that these glitches don't disappear as the tumor mutates.