Probing the Genomic Basis for Cancer

Agnès Viale and Faculty Member Marc Ladanyi collaborate with many members of Memorial Sloan-Kettering's research community on a variety of projects related to cancer genomics. Here they are pictured with a 454, a so-called next-generation DNA sequencer, which is changing the way investigators learn about tumors and their genetic defects.

Imagine a day when a patient could have a biopsy taken from a tumor, have the entire genome of the tumor characterized, and then be treated with drugs and other therapies that are tailored to the exact genetic changes present in the tumor cells. The scenario is not science fiction, but a goal that investigators at Memorial Sloan-Kettering Cancer Center and other institutions have recognized as one of the most promising avenues for treating cancer.

"The future of oncology is going to be in large part a drive toward more personalized medicine, which is more effective and has fewer side effects," said medical oncologist William Pao. "It would be similar to the way we treat infections today: We take cultures, identify the organism causing the infection, and figure out exactly which antibiotic to give. We can now envision something similar in cancer: obtain a tumor sample, identify the molecular lesions causing that tumor, and give specific targeted agents to kill it," said Dr. Pao, who is a member of the Human Oncology and Pathogenesis Program (HOPP).

"As we learn more about the genetic abnormalities in human cancers, we will find there are many different subtypes among tumors that, using conventional pathology techniques, might look quite similar," said molecular pathologist Marc Ladanyi. "However, more detailed analysis will reveal subsets of a certain cancer that can benefit from a particular drug. Even if the subgroup doesn't immediately connect to a known drug, it may help us determine which tumors are likely to behave more aggressively and which patients are therefore candidates for more intensive conventional therapy." Dr. Ladanyi is a HOPP member, and he leads the Center's contribution toward a nationwide effort to catalog comprehensively the genetic changes involved in human cancer.

For a few types of cancer, this concept is starting to become a reality. Several years ago, Dr. Pao and Memorial Sloan-Kettering President Harold Varmus, as members of an Memorial Sloan-Kettering team called the Lung Cancer Oncogenome Group (LCOG), helped determine molecular reasons why some lung adenocarcinomas (a subtype of non-small cell lung cancers) are either sensitive or resistant to the targeted therapies gefitinib (Iressa®) and erlotinib (Tarceva®). The team found that highly sensitive tumors harbor mutations in a gene called EGFR, whereas resistant tumors often contain mutations in a gene called KRAS. About 10 percent of patients with lung adenocarcinomas have EGFR mutations, and about one-quarter of lung adenocarcinomas have KRAS mutations. Based on these findings, testing for these mutations has become a part of routine care for lung cancer patients at Memorial Sloan-Kettering to help guide treatment decisions regarding the use of targeted therapies.

The knowledge gained from studying these lung cancer drugs -- along with the success of other targeted therapies -- illustrates why this area of research is becoming so important. Currently, for most types of cancer, there is still much to be learned about their molecular characteristics, including the genetic changes that cause them to form, grow, and spread (metastasize) and that make some tumors more aggressive and resistant to therapy than others. Identifying a genetic change that leads to the formation of a certain tumor, however, does not necessarily mean there will be compounds available today to counteract the defect. But identification of these changes could spur the development of such agents.

Faculty Member William Pao's research is focused on the genetic changes present in lung cancer cells.

One of the strengths that has put Memorial Sloan-Kettering at the forefront of the search for genomic changes in cancer is its extensive tumor bank, directed by molecular pathologist and HOPP member William L. Gerald. As early as the 1980s, Memorial Sloan-Kettering surgeons, led by then-Department of Surgery Chair Murray F. Brennan, began saving thousands of tumor specimens collected from patients with all different types of cancer. These specimens are stored anonymously to protect patient confidentiality but are linked to clinical information about the patients, such as whether they responded to certain treatments, where the cancer metastasized (if it did), and whether the patients were ultimately cured. The tumor bank already has contributed to significant findings, including work by cancer biologist Joan Massagué pinpointing genes that mediate the metastasis of breast cancer to the lungs and bones.

Another resource that gives Memorial Sloan-Kettering a distinct advantage in the field is its Genomics Core Laboratory and its DNA Sequencing Core Laboratory. The latter recently received a huge boost in sequencing capabilities through the purchase of an instrument called a 454 -- a "next generation" DNA sequencer. Both facilities are led by Agnès Viale. The genomics lab analyzes tumor cells using various microarrays (also called "chips"), which can monitor thousands of genes at the same time to look for changes in the number of copies of a gene or determine whether a gene is expressed, for example. Data from these microarrays might explain differences between tumor cells and normal cells or distinguish different subtypes of the same cancer.

Using next-generation sequencing, investigators are able to sequence each individual DNA molecule separately, rather than combining the sequences of all fragments, as older sequencing machines do. "What this means," explained Dr. Viale, "is that if only 1 percent of cells in a tumor have an additional mutation that confers resistance to a drug treatment, we can find it. That small number of cells is what likely would be responsible for disease recurrence, and up until now we had no way to detect that rare additional mutation. This technology is beautiful, absolutely cutting edge."

Recently the Center's tumor analysis capabilities were further enhanced by the creation of the Geoffrey Beene Translational Oncology Core Facility, under the leadership of Adriana Heguy. The facility, which is part of HOPP, extracts DNA from patient tumor samples, prepares the DNA for sequencing, and then outsources the material to high-volume sequencing facilities. It also performs mutation detection and data analysis using software designed by Memorial Sloan-Kettering's bioinformatics team and maintains a centralized database.

"The goal of our core is to identify novel mutations in different cancer types, as well as to characterize the samples for known mutations that can ultimately translate to new options for patient care," Dr. Heguy said. "Our laboratory has been in operation for less than six months, and we are already collaborating with a large number of researchers across Memorial Sloan-Kettering who are utilizing our technology in diverse tumor types."

The Center's research on genomic changes in lung adenocarcinoma is moving forward. A multidisciplinary team of clinicians and basic scientists, led by medical oncologist Mark G. Kris, has received a large grant from the National Institutes of Health to expand the work started by the LCOG. Members of the team are working on a variety of projects, including the development of newer targeted therapies; the search for additional genes that cause the formation of lung cancer, as well as genes that mediate the metastasis of lung cancer to the brain; and the study of genes related to disease persistence.

"Disease persistence means that even in patients for whom targeted therapies like gefitinib and erlotinib induce massive tumor shrinkage, you never get rid of all the cancer," explained Dr. Pao. "So we're trying to figure out why these cells persist."

Members of Memorial Sloan-Kettering's lung team also have participated in the Tumor Sequencing Project (TSP), a federally funded, multi-institutional collaboration that is seeking to map genomic changes in lung adenocarcinoma. In November 2007, the TSP team published a paper in Nature identifying more than 40 previously unknown genomic regions that are frequently altered in lung adenocarcinoma. The paper was based on a different approach than the one used by LCOG, a technique called single nucleotide polymorphism (SNP, pronounced "snip") profiling, which allows investigators to look for changes in the number of DNA copies that occur in tumors on a genome-wide level. [PubMed Abstract]

Faculty Member
Samuel Singer

Another type of cancer on which the Center is focusing its genomics efforts is soft tissue sarcoma, a broad category of rare but aggressive tumors that affect tissues such as fat, muscle, and nerves. Through a joint project with the Broad Institute of Harvard and MIT, now funded by a grant from the Starr Cancer Consortium, Memorial Sloan-Kettering scientists are characterizing the genetic changes in several subtypes of soft tissue sarcoma, which together make up about three-quarters of all cases. "Most types of sarcoma are not sensitive to traditional chemotherapy, and that is why we set out to look for molecular changes that we might be able to target with new therapies," said surgical oncologist Samuel Singer.

"We're analyzing patient samples to look for genes that are altered or amplified to see if they promote sarcoma formation," he elaborated. So far, the team has studied more than 200 sarcoma samples across seven subtypes for changes in the number of gene copies and changes in gene transcription.

Investigators have sequenced more than 225 genes in 48 tumor samples and have identified mutations that may serve as potential therapeutic targets. They plan to expand the mutational analysis to additional genes and additional sarcoma samples. "Now we need to determine which mutations are the most significant, in terms of those that can be targeted with drugs, as well as those that can teach us more about how tumors behave," he said.

In collaboration with several clinical departments at Memorial Sloan-Kettering, Dr. Gerald has begun a similar project with prostate cancer. The group is studying approximately 250 genes in 200 prostate tumor samples, and data generation is expected to be completed later this spring. Likewise, a project initiated by immunologist Alan N. Houghton seeks to characterize genomic changes that occur in melanoma cell lines.

"We have the expertise and resources at Memorial Sloan-Kettering to make an important contribution to the cancer genome efforts," Dr. Ladanyi said.