George Mason University researcher Emmanuel Petricoin is developing a database detailing phosphorylation levels in a variety of protein-signaling pathways linked to common cancers.
The database could help physicians make better treatment decisions, aid pharmaceutical firms' drug-development efforts, and uncover new indications for existing therapeutics, Petricoin, who is co-director of GMU's Center for Applied Proteomics and Molecular Medicine, told ProteoMonitor.
The data is based on analyses done in Petricoin's lab of more than 500 tumors from a variety of malignancies, including cancers of the lung, breast, prostate, kidney, and brain. The researchers isolated the tumors via laser-capture microdissection and used reverse-phase protein microarrays to measure protein-phosphorylation levels within them.
This approach allowed them to map activation levels of key signaling pathways across a range of cancers, he said.
Petricoin likened the project to the Cancer Genome Atlas, which set out to sequence a range of different tumor genomes. "We have almost like a TCGA initiative where we're just doing massive pathway mapping of human cancers in our laboratory," he said. "We're microdissecting hundreds of [tumors of the] breast, colon, [and] prostate so that we can start answering the question of which pathways are activated."
The database covers activation levels of between 150 and 200 phosphoproteins over some 12 to 15 specific pathways associated with cancer, Petricoin said. He noted that results from the project suggest that cancers may be better classified by pathway activation than by the site of the tumor's origin.
"What it's revealing is that patients are grouping based on pathway activation, not by site of location," he said. "We'll have breast and colorectal and lung cancer patients grouping together because they have a common pathway arrangement. If I didn't tell you which was lung, colon, and breast cancer, you wouldn't be able to tell it based on the pathway groupings."
Such groupings could create new ways of selecting cancer treatments and identify potential new indications for existing drugs.
"We've been giving therapy based on the site of origin — Herceptin for breast cancer, Gleevec for leukemia," Petricoin said. "But what we're learning is that cancer is a pathway disease, not a site-of-origin disease. So you could have breast-cancer patients [whose tumor on the molecular level] looks like a renal cell cancer."
This means that drugs typically used on one type of cancer might be suited to a variety of other cancers. This is the notion behind a clinical trial Petricoin and Lance Liotta, co-director of CAPMM, started with Novartis last summer to examine whether the company's leukemia drug Gleevec might also be useful against colorectal cancer (PM 1/22/2010).
Colorectal cancer, the third-most common cancer in the US with roughly 100,000 new cases diagnosed annually, could represent a significant revenue channel for Novartis if Gleevec were shown to be an effective treatment for the disease.
In February, Petricoin and Liotta launched a trial that will analyze protein-signaling pathways to identify potential targeted treatments for breast cancer.
With funding from Virginia non-profit The Side-Out Foundation, the researchers have been looking at the activation levels of drug-targeted proteins in metastatic breast cancer patients in hopes of better tailoring their courses of therapy.
"We can measure the phosphorylation levels of all the direct drug targets, the proteins that the drugs hit," Petricoin said. "We're measuring whether or not those are active in the patient sample."
For the study, the partners perform laser-capture microdissection, lyse the cells, and print arrays. "Everything is developed as a quantitative, calibrated assay," said Petricoin. "We have high and low controls and calibrators, just like an FDA-approved clinical-chemistry assay, and then we have cut points that we've generated from population data so we know what's high and low."
He noted that his team can generate a physician's report showing which pathways are or are not activated within 72 hours after receiving a tumor sample. Physicians can then make treatment decisions based on the pathway-activation data.
The goal, Petricoin said, is to "divine the molecular circuit map of that tumor" so that a physician "can … make a molecularly rationalized decision."
The breast and colorectal-cancer studies are two-year trials that could potentially be extended to three years. Although patients in the trials are currently being treated based on pathway-activation information obtained during the study, it's still too early to determine whether the approach has led to improved outcomes, Petricoin said.
In the meantime, his lab is compiling the pathway information collected in these trials and its other independent mapping efforts.
"We have a very big and growing database of microdissected tumors from patient biopsies," said Petricoin. "Especially once we start knowing who responded and who didn't respond to therapy we can ascertain whether or not we can find a cutpoint in the data that can predict response or not. So that will be kind of an evolving, growing database."
Petricoin said that he is "gearing up" to publish a paper based on this pathway data, although he added that there are "proprietary aspects to the data that will be subjected to pre-existing commercialization arrangements" with companies such as Theranostics Health. Petricoin and Liotta launched the biotech firm in 2006 to commercialize their laser-microdissection and reverse-phase protein microarray technologies.
He suggested that Theranostics might use this proprietary information both to develop in-house assays and to build a database that it could charge drug makers to access.
[ pagebreak ]
Theranostics is not directly involved in either the breast or colorectal cancer trials, or in Petricoin's database work. However, Ron Hencin, the firm's vice president of business development, told ProteoMonitor that it has compiled similar information through collaborations with undisclosed pharma companies and its own diagnostic-development projects.
"We have a parallel, fairly large, effort similar to [Petricoin's database work] that's really around some of the supporting validation work that we're doing for our tests," he said. "If you're looking at activation status within a particular type of cancer, you need to understand what the population dynamics of that is."
"So you need to profile a significant number of samples from various individuals to understand the dynamic range of expression of the proteins in different kinds of cancers if you want to make an assessment of how activated that particular pathway may be in a cancer," he explained. "It's all part of our validation work as well."
Theranostics currently has no plans to package and sell these data, he said.
Theranostics has used its laser microdissection and reverse-phase protein microarray platforms to develop validated panels of assays for three undisclosed drug firms that are using them in clinical trials of various cancer treatments.
The company is also developing an in-house, CLIA-certified companion diagnostic for an undisclosed cancer it hopes to bring to market early next year.