• January 26, 2010
    OncoMap guides search for patterns in tumors

    meyerson-garraway.jpgMatthew Meyerson (left) and Levi Garraway found a new purpose for an existing technology that extracts information from tumors. 

    Conceiving a new use for existing technology can have as big an impact as inventing a novel technology itself.

    Witness the OncoMap program in Dana-Farber's Center for Cancer Genome Discovery (CCGD), which helps scientists tap information in tumor samples, some of them decades old, to advance understanding of adult and childhood cancers.

    For several years, scientists had used mass spectrometers and high-speed DNA dispensers to find "point mutations" – misspellings in a single letter of genetic code – in people with an inherited risk for various kinds of cancer. One advantage of the technique is that it can read DNA in tissue preserved in formalin and paraffin, a commonly used storage method that renders DNA invisible to most scanning devices.

    In 2005, Dana-Farber's Levi Garraway, MD, PhD, and Matthew Meyerson, MD, PhD, envisioned another, especially relevant use for the technology: scanning actual tumor samples for point mutations that can be targeted by existing therapies.

    At a stroke, the investigators converted a research technology whose practical benefits may have been years away to one with more immediate impact.

    "We realized, in principle, that we could adapt mass spectrometric screening of cell genes to extract highly relevant information from tumors that might guide therapy choices and yet cost only pennies per genetic variant," says Garraway, who directs the OncoMap program.

    To determine whether the approach could deliver on its promise, Garraway, Meyerson, and their colleagues used the technology to probe cancer DNA for 238 known mutations in 17 cancer-causing genes. "We observed mutations in tumor types that we would never have expected at the start," Garraway says.

    Word quickly spread that OncoMap could help cancer researchers identify different patterns of gene mutations within particular types of cancer, or enable them to discover which of nearly 1,000 known mutations are in a set of tumor tissue samples. Such research would go hand-in-hand with personalized medicine, which matches therapy to the unique set of genetic abnormalities in patients' tumors.

    Among the first to realize the potential of the technology were scientists at Dana-Farber. The CCGD's director, William Hahn, MD, PhD, and associate director, Laura MacConaill, PhD, have been closely involved with much of this research. Two teams, in particular, have used OncoMap to scientific advantage so far:

    Childhood brain tumors

    Mark Kieran, MD, PhD, of Pediatric Oncology and Charles Stiles, PhD, co-chair of Cancer Biology, have brought a unique collection of minds and machinery to bear on the problem of pediatric low-grade astrocytoma (PLGA), a relentless form of childhood brain cancer without effective treatment. Along with colleagues at the Dana-Farber/Harvard Cancer Center and Broad Institute of MIT and Harvard, they amassed hundreds of PLGA samples encased in paraffin and surveyed their DNA with OncoMap. In more than half the samples, they found mutations in the gene BRAF. Researchers are now testing drugs that target this mutation in animal cells.

    Langerhans cell histiocytosis

    For Chief Scientific Officer Barrett Rollins, MD, PhD, and colleagues, OncoMap has shown the way to a potential treatment advance for Langerhans cell histiocytosis, a rare disease in which too many Langerhans cells (a type of white blood cell) grow in certain tissues and organs, often the skin and bones. The rarity of the disease and scarcity of banked tissue has made it difficult to study, but Rollins and his associates found 60 samples, most stored in the tissue archives at partnering Boston Children's Hospital and Brigham and Women's Hospital. When they analyzed the specimens with OncoMap, they found nearly 60 percent had the same genetic mutation - an abnormality for which a drug is already available. The research also helped settle a longstanding question about Langerhans: Is it truly a form of cancer? The discovery of a cancer-causing gene in so many samples indicates that it is.

     –Rob Levy
    Robert_Levy@dfci.harvard.edu 

  • Email
  • Print
  • Share
  • Text
Highlight Glossary Terms