New method offers snapshots of cancer genome

Scanning electron micrograph of a breast cancer cell
Photo courtesy of the National Cancer Institute

Efforts to map the genetic changes associated with cancer — known as the "cancer genome" — have intensified in recent years, driven by a potential to catalyze the development of new, molecularly targeted cancer drugs. Realizing this promise, though, will require techniques for analyzing tumors that are practical not only for the laboratory but for the clinic, where accuracy and cost-efficiency are paramount. As described in the February 11 advance online edition of Nature Genetics, a research team led by scientists from the Broad Institute, the Dana-Farber Cancer Institute and Harvard Medical School has developed a high-throughput method for analyzing genes in tumor cells that could help pierce the bench-to-bedside barrier.

Current methods for studying the cancer genome rely heavily on large-scale DNA sequencing. While this approach can help generate a full list of genes that are defective within a specific tumor, it is simply too expensive to be applied routinely to multiple different tumors. Levi Garraway, a Broad associate member and an assistant professor at Dana-Farber Cancer Institute and Harvard Medical School, together with Matthew Meyerson, a Broad associate member and an associate professor at Dana-Farber Cancer Institute and Harvard Medical School, set out to develop a more streamlined and cost-effective method. Rather than query the entire genome, the scientists and their colleagues devised a way to simultaneously measure a subset of the most pertinent and commonly mutated genes among different types of tumors.

The work is based on a genomic technique, now frequently used in other disease-based research, which uses mass spectrometry to identify single letter variations in the genetic code. Instead of using it to catalogue normal genetic variability, the researchers adapted the method to reveal defects in tumor DNA that change how certain cancer genes (called "oncogenes") function. Part of what makes the approach feasible is the distribution of oncogene mutations — they tend to cluster in relatively few, predictable locations, rather than being scattered haphazardly. Because of how they work to promote cancer, oncogenes are considered prime targets in the design of novel cancer drugs.

The scientists, including first authors Roman Thomas and Alissa Baker, engineered tests for more than 200 known mutations in 17 human oncogenes, effectively zeroing in on many of cancer’s usual suspects. These suspects include common mutations, those that are linked to particularly aggressive cancers, or ones that suggest sensitivity to specific anti-cancer drugs.

By using their method to analyze 1,000 human tumor samples, the scientists developed a panoramic view of oncogene mutations and their frequencies in a variety of common and rare cancers. They uncovered mutations in genes that had not been previously linked to certain types of cancers and in some cases, identified tumors with mutations in multiple oncogenes — an unusual phenomenon that would previously have been much more difficult to detect. Moreover, the approach proved as sensitive and, in some cases, even more sensitive, than comparable sequencing-based tests.

"It is now well-established that a knowledge of the key mutations in cancer DNA can in some cases speed the development of effective therapies. However, extracting the right kind of genetic information from each cancer patient poses a major challenge," said Garraway. "We hope that this study will help oncologists to address this issue, which is so crucial for the goal of individualized cancer treatment. The next step will be to apply this method systematically and in ‘real time’ to the types of tumor specimens that are collected in the clinic."

The scientists from the Broad who participated in this work include: Jordi Barretina, Rameen Beroukhim, Ralph DeBiasi, Amit Dutt, Whei Feng, Stacey Gabriel, Levi Garraway, Mary Goyette, Heidi Greulich, Charlie Hatton, Thomas LaFramboise, Jeffrey Lee, William Lin, Massimo Loda, Laura MacConnaill, Matthew Meyerson, Rick Nicoletti, Kinjal Shah, Roman Thomas, Meng Wang, Wendy Winckler, and Liuda Ziaugra.