Genome Studies Start to Unravel Prostate Cancer’s Complexity
Researchers at Memorial Sloan-Kettering Cancer Center (MSKCC) have completed the largest genomic analysis of prostate tumors to date. The results, based on clinical and genomic information collected from 218 patients, provide an overview of the common genetic changes in the disease and point to new directions for research, including a way to potentially differentiate aggressive tumors from those that are not life threatening. The researchers have made the data available to the community through a public Web site, and a summary of the results appeared online last week in Cancer Cell.
“We now have a much better picture of the common genetic alterations in prostate cancer,” said lead investigator Dr. Charles Sawyers. Although more samples need to be analyzed, he continued, the results could provide a roadmap for designing future clinical trials in this disease. “When it comes to developing and testing targeted cancer drugs, you need to be able to subclassify patients, and you can’t do this intelligently until you know what the alterations are.”
—Dr. Jonathan Simons
Genome studies have yielded insights into glioblastoma and cancers of the lung, colon, pancreas, and breast, but prostate cancer has been a challenge. Prostate tumors are relatively small, and expert pathologists are needed to obtain adequate samples of tumor tissue. With a large prostate cancer program and skilled pathologists, MSKCC was able to overcome these challenges.
“Dr. Sawyers and his colleagues have made an extraordinarily important contribution to the field of prostate cancer research,” said NCI Deputy Director Dr. Anna Barker. “Certain tumors are going to be difficult to collect and analyze, but this study—which used high-quality samples and multidimensional data—now gives the community new opportunities to understand the disease.”
The researchers sequenced 157 genes that were known to be associated with prostate and other cancers. Mutations in these genes were rare. But when the researchers included additional information, such as DNA copy-number changes (gains and losses of DNA), they identified several genetic pathways, including the PI3K pathway, that were altered in nearly all metastatic tumors and many primary tumors.
“Just as we’ve seen in brain tumors, when you combine all these different sources of information, there is a uniform consistency of pathways that are altered in the disease,” said co-author Dr. Peter Scardino, chairman of the Department of Surgery at MSKCC. “We found many more abnormalities in localized prostate cancer than we expected.”
The researchers also identified a gene called NCOA2 that appears to play an important role in about 11 percent of prostate tumors. The protein encoded by this gene may drive prostate cancer by amplifying signals from the androgen-receptor pathway; this pathway plays a critical role in early- and late-stage prostate cancer.
The analysis also revealed a striking association between changes in DNA copy number and the risk of recurrence after surgery, and this association could not be explained entirely by Gleason score. “This was one of the most exciting findings from the study,” said Dr. Scardino. “It offers the possibility of a biomarker that could be used to characterize the aggressiveness of prostate cancer, which is something we greatly need.”
Doctors currently do not have a way to distinguish between prostate cancers that require aggressive treatment and those that will cause no harm if left alone. Consequently, many men receive treatment unnecessarily. Genomic tests can provide prognostic information in breast cancer, for example, but none yet exists in this disease.
The new findings, if confirmed, represent a prototype for developing these kinds of prognostic tests for prostate cancer, said Dr. Jonathan Simons, CEO and president of the Prostate Cancer Foundation. “This would change how doctors talk to patients about the disease and the need for adjuvant therapy or not, which is why this finding is so exciting,” he added.
Dr. Scardino runs the Specialized Program of Research Excellence (SPORE) in prostate cancer at MSKCC, and his group has launched follow-up studies. The current work was done using frozen tumor specimens collected during prostatectomies. The researchers will now see whether copy-number changes are informative using paraffin-embedded tissues. If the answer is yes, they will test cells obtained from a needle biopsy.
The genome analysis also revealed that some patients whose tumors include a fusion of the genes TMPRSS2 and ERG are also missing part of chromosome 3. This fusion gene occurs in about half of all prostate cancers, and researchers have suspected that other genes also play a role in these cases.
“This deletion on chromosome 3 appears to be very strongly associated with the fusion,” said Dr. Sawyers. “The next steps are to see which genes in the region that is deleted are involved in the disease. We have a clear path forward.”
The TMPRSS2-ERG fusion was discovered in 2005 by University of Michigan researchers supported by NCI’s Early Detection Research Network. At the time, fusions were thought to be limited to cancers of the blood, but it is now known that these alterations are present in common cancers as well. About two dozen have been identified in prostate cancer.
New Class of Gene Fusions
Earlier this month, the Michigan group, led by Dr. Arul Chinnaiyan, reported a new class of prostate cancer gene fusions derived from the RAF pathway. One of these fusions involves the gene BRAF, which plays a role in melanoma. Drugs targeting BRAF are in clinical trials, and it appears, based on experiments in cells, that these drugs may be active in up to 2 percent of patients with prostate cancer, the researchers reported.
“The clinical promise of this discovery is that patients who have these RAF gene fusions may be candidates for drugs that target these changes,” said Dr. Chinnaiyan. “Some of the newer inhibitors, in particular, might be useful in treating this molecular subtype of prostate cancer as well as some other cancers.”
In the future, every man whose prostate cancer is biopsied is going to have his DNA read for gene fusions, just as women have their breast cancers tested for overexpression of the HER2 protein to determine whether they should receive trastuzumab (Herceptin), predicted Dr. Simons. “To cure every man of advanced prostate cancer, we’ll need at least 24 strategies,” he continued. “Now that we know what we’re facing, we can make research plans and do the work.”
Two prostate cancer genome sequences have been presented at scientific meetings and will likely be published later this year, Dr. Simons said. “It’s a complicated disease, but key properties of the disease are going from being in total darkness to full clarity. And that’s what is so amazing.”
For Dr. Barker, a founder of The Cancer Genome Atlas (TCGA) project with colleagues from the National Human Genome Research Institute, the MSKCC study is both exciting and gratifying. Dr. Sawyers and his colleagues essentially followed the TCGA approach by profiling high-quality tumor samples and integrating clinical and multiple kinds of genomic information into their analysis. As with TCGA, they made the results public so that investigators in the community can now mine the information for new insights.
The hope from the beginning of TCGA has been that the approach would be adopted by investigators performing cancer genomics studies in the community, said Dr. Barker. And now that this study has been completed, TCGA investigators will use the data when they launch an even larger study of prostate tumors in the future.
“This study gives us a much better starting point for prostate cancer than we’ve ever had before,” Dr. Barker said. “It’s a great day for cancer research, and an even better day for patients.”
—Edward R. Winstead