Why do we need Informatics?
Cancer is known to be a complex series of diseases, each of which has a range of clinical manifestations and molecular abnormalities. However, this enormous diversity is currently coupled with a homogenized approach to treatment. This makes it incredibly difficult for oncologist's to identify the most appropriate treatment for every individual with cancer. For example, we currently characterize cancers primarily by the site of origin, such as the breast or the colon, and do not take into account the molecular characteristics of the disease. Best practices in clinical management rely upon incomplete data and obsolete methods of data generation, management and analysis, while applying only a small fraction of available information to the management of cancer.
Recently, however, we have begun to employ more sophisticated means to subdivide cancers into types defined by their molecular properties. For example, we now know that breast cancer is in fact three distinct diseases – there are those that express the estrogen or progesterone receptors, those that express a mutated protein known as HER-2, and those known as “triple negative,” which express none of these molecular markers. These molecular sub-classifications allow us more effectively to deploy therapies that are designed to address specifically the molecular abnormality that is driving or maintaining a particular type of cancer. Yet, we have only begun to explore how these networks are working in a particular patient’s cancer. Using new tools to better understand cancer biology, we expect to discover that breast cancer is more than three diseases, and that treatment can be effectively customized to attack each of the critical pathways that lead to each type.
We know that this approach can be highly successful. Continuing with the breast cancer example, treating breast cancer patients with the monoclonal antibody Herceptin has a success rate of only 6 percent when applied uniformly to all women with advanced, metastatic disease. But when this treatment is selectively given to women whose cancers are known to be driven by their HER-2 gene, that group can expect a success rate of about 30%. In fact, combining Herceptin with standard chemotherapy increases cure rates for women whose high risk cancer has this genetic marker.
Imagine if we were able to apply even greater precision in understanding the mutations and dysregulated molecular pathways that drive human cancers. Knowing how these abnormalities are driving clinical behaviors, we can predict relapse and failure to respond to a treatment, and help patients make better choices. This knowledge will allow us to deploy existing drugs that target these abnormalities more effectively, and it will provide a vital tool for the development of new drugs that target newly defined cancer sub-types.