Three approaches will be used to provide
the discriminators we need: This year, over 300,000 men will be diagnosed with prostate tumors, yet less than half of them will develop progressive disease. We need to know what distinguishes these two groups and be able to differentiate between them. The behavior of each cancer--how it will respond to therapy, how it will change over time, and whether it will be a threat to the patient--is determined by molecular changes that occurred during tumor development. The methods we currently use to detect and diagnose tumors often do not allow us to determine these properties. The problem is one of discrimination. We need to know enough about a particular tumor to make correct choices about therapy and accurate predictions of prognosis. At present our ability to assign these important discriminators is just too limited. Too often we cannot distinguish between seemingly identical tumors, even when the ultimate danger to a patient is different. For example, two apparently identical breast tumors may respond differently to the same treatment, or the course of tumor development may be vastly different. In such cases, unfortunately, patient outcome also differs radically. Our discriminators for all tumors must become more sophisticated. We must be able to determine the traits of the tumor that dictate choice of treatment and predict prognosis.
Develop diagnostic tests that will allow treatment choice to be based on the fundamental properties of a tumor cell.
The first identification of a human cancer gene was reported approximately 20 years ago, and progress in this field has been rapidly expanding since then. This explosive increase in our knowledge of the mechanisms that drive tumor development is one of the success stories of modern biology. We now understand the molecular basis for many of the changes that drive tumor development. This explosion in our knowledge needs to be applied to the diagnosis of cancers. Molecular diagnostics provides one of the most obvious, and what promises to be one of the first, links between the molecular characterization of cancer cells and patient care. In its simplest terms, this new phase of cancer diagnostics will provide a snapshot of the properties of a tumor. This snapshot will detail the key differences between a normal cell and a cancer cell and will provide a molecular scorecard of the properties of a tumor cell. These properties can in turn be used to develop and choose effective therapies and to plan patient care.
Developing a rapid method to determine this snapshot of the tumor cell will be an essential step for major advances in several areas. These include:
* Tumor Development. A snapshot of cancer cell properties taken at various stages of tumor development will chart the steps of cancer development. We need to learn the order in which tumor changes occur. Does one alteration dictate which alteration follows? When do the most dangerous changes take place? Establishing a picture of the natural history of a tumor will now be possible for all human tumors.
* Tumor Stratification. The availability of new and more complete diagnostic tests will allow us to classify tumors into groups based on their underlying fundamental properties. This will allow us to identify which tumors will respond similarly to different therapies and which tumors pose similar threats to the patient.
* Treatment. The molecular changes that occur during tumor development will ultimately determine whether or not a particular therapy will be needed or will succeed. Diagnostic tests that determine the key changes in tumor development will make it possible to choose between currently available therapies and eventually to design new and more effective therapies.
* Prognosis. Having both a complete picture of the alterations that have promoted tumor development and a clear indication of the tissue of origin will lead to the most accurate prognosis possible and hence will be a key element for advances in patient management.
To increase the value of diagnosis to patients, our effort needs to be focused on two major areas: developing diagnostic testing and creating the infrastructure to correlate test results with clinical outcomes. The first area will be in the development of the diagnostic tests themselves. The clear goal of the next generation of diagnostic tests will be to classify tumors into groups that behave and respond in similar manners. As better discriminators are found, our view of tumors will change. A muddy picture will begin to clear.
Three approaches will be used to provide
the discriminators we need:
* The first approach is to detect the actual mutations that drive tumor development, establish a profile of the proteins that are expressed in the tumor, and determine how key regulatory controls are changed in the tumor. For mutation-based studies, the snapshot of the tumor's characteristics will produce a list of cancer gene alterations.
* The second diagnostic approach will give us a picture of key genes that are expressed in the tumor. When two tumors are compared in this way, the patterns of gene expression should also indicate the tumor's origin.
* The third approach will be to characterize changes in key signaling pathways in the cell. Signaling pathways could be likened to electrical circuits, and they are the method that cells use to make decisions. Just as one would test a few key electrical circuits of a household appliance to locate the malfunction, signaling pathways in the cell will be monitored for defects.
The goal of all three of these approaches will be to establish groups of tumors that behave similarly, thus giving us recognizable targets for therapy.
Another major area of the research plan for developing diagnostics will be to provide the national infrastructure needed to correlate the results of diagnostic tests with clinical outcomes. We need to support molecular diagnostic laboratories where the alteration, expression levels, and pathways for diagnostic approaches can be established. We will need repositories of tissue and tumor samples isolated from patients who wish to participate in this research. The repositories will store clinical records and will serve as a resource for samples and for comparing prognosis with the various markers. Clinical research will be needed to correlate the tumor profiles with response to different therapies. All of this must be linked through an informatics network that allows information exchange among all the facets of this approach.
Through this initiative, we have the opportunity to convert our growing knowledge of tumor cell biology into practical advances in patient care. Over time, these advances will transform many aspects of clinical cancer medicine. They will allow a diagnosis based on biology, not just on the microscopic appearance of a tumor. By revealing the molecular changes that occurred during tumor development, these advances will allow much more accurate prediction of how a tumor will behave and what the patient's outcome will be. They will eventually permit the selection of therapy based on the tumor's individual biological properties, rather than on empirical guesses. Moreover, through advances in this area, new therapeutic targets will be uncovered, and it will be possible to stratify clinical trials based on tumor characteristics and the patient's cancer risk.
Without national investment, this fundamental transformation of cancer medicine will occur far more slowly. A portion of the work necessary to effect these developments would proceed in academia and industry, but less rapidly and with less intensity. Certain necessary steps may not occur at all; it is unlikely that the essential research infrastructure to support a cancer diagnostics effort of this type (tissue banks and the informatics links between banks, tumor registries, and the clinical trials program) can be organized and supported except by the NCI.
Without this investment, cancer treatment choices will continue to be based on unsatisfactory and incomplete methods of diagnosis. We will not be able to separate tumors into classes based on their molecular differences. Since these differences are the reasons that tumors behave and respond differently, our insights into choice of treatment and outcome will continue to be blurred. Progress will not be stopped, but an important opportunity for real advancement will be missed.
| NEXT SECTION... | PREVIOUS SECTION... | CONTENTS PAGE... | |