Prostate cancer is the second most common cancer in US men. Thanks to widespread screening with prostate specific antigen (PSA) tests, almost 90 percent of these cancers are caught early, before they have spread. But the behavior of individual prostate tumors can vary regardless of how early they are detected — some can be aggressive, while others may be cured with initial therapy or may remain inactive with no therapy at all.
There are currently several different systems for predicting the risk that a particular cancer will spread, based on PSA levels and the appearance of tumors under a microscope. But studies have shown that they don’t always accurately predict the aggressiveness of individual tumors.
Now a study from The Cancer Genome Atlas (TCGA) seeks to improve ways to diagnose prostate cancers.
“We decided to take a closer look at the diversity of molecular features in the genome across a large cohort of primary prostate cancer samples,” says Memorial Sloan Kettering investigator Nikolaus Schultz, who was a co-principal investigator of the TCGA study on prostate cancer published earlier this month in the journal Cell. “This diversity could be used to distinguish between aggressive and indolent [harmless] tumors. Our findings also provide new directions for targeted therapies.”
TCGA is a multicenter initiative that collects, processes, and characterizes the largest set of tumor samples to date using state-of-the-art genomic and molecular techniques. It is supported and managed by the National Cancer Institute and the National Human Genome Research Institute, both part of the National Institutes of Health. The study built on earlier work from MSK and other institutions looking at genomic changes in a number of different cancers, including endometrial cancer and ovarian cancer, among others.
Dr. Schultz and his fellow researchers — including MSK Computational Biology Program Chair Chris Sander; Barry Taylor, an Associate Director of MSK’s Center for Molecular Oncology; and Massimo Loda, a professor of pathology at Harvard Medical School — performed a comprehensive genetic analysis of 333 prostate cancer tumor samples from patients treated at a number of different centers.
They studied the tumors for different types of mutations, including gene fusions — in which two genes join together to make a new gene — and mutations that affect the number of times a certain gene is copied, as well as changes in the DNA sequence of individual genes. They also looked at epigenetic changes (changes that are not part of the DNA code), including DNA methylation, in which small chemical groups are attached to DNA molecules, affecting which genes are transcribed.
Seven Subtypes Identified
Based on their analysis, they were able to identify seven different subtypes of prostate cancer. Four of them were characterized by gene fusions and the other three were defined by mutations in the DNA sequence of certain genes.
“This study shows that there is an incredible molecular diversity within prostate cancer,” Dr. Schultz says. “This diversity helps to explain the wide range of outcomes seen in men with prostate cancer and could have implications for personalized medicine, in which therapies are targeted toward specific mutations. Further studies are under way to determine which subtypes may be more aggressive than others.”
Of those cancers characterized by changes in the DNA sequence, two of the subtypes — caused by mutations in the genes SPOP and FOXA1 — were very similar to each other. The third subtype, caused by mutations in the gene IDH1, was similar to certain types of leukemia and brain cancers that have IDH1 mutations. Brain tumors with IDH1 have been well characterized by investigators at MSK, and currently ongoing clinical trials are targeting not only brain tumors but also other types of tumors with IDH1 mutations.Back to top
Moving into Basket Studies
Basket trials group patients who are being tested with new targeted therapies by the individual mutations in their tumors, not by the location of the cancer in their bodies. In the future, patients with prostate cancer could be included in basket trials testing drugs that block IDH1.
“Our study also showed as many as 20 percent of prostate tumors have defects in how they repair their DNA,” Dr. Schultz said. “There are already treatments for these types of defects in clinical trials, and some are now available at MSK.”
The investigators say that the TCGA prostate data, which are open and available to the public, will continue to be mined for many years. “Our data show how heterogeneous prostate cancers are,” Dr. Schultz concludes. “It’s important to look not only at genetic changes, but also epigenetic changes and changes in gene expression.”Back to top