| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
2002 RSNA Annual Oration in Radiation Oncology |
1 From the National Cancer Institute, Radiation Oncology Branch, Bldg 10, B3-B69, National Institutes of Health, Bethesda, MD 20892-1002. Received November 29, 2002; revision requested December 16; revision received February 11, 2003; accepted February 20. Address correspondence to the author (e-mail: ccoleman@mail.nih.gov).
ABSTRACT
Among the areas defined by the National Cancer Institute as "Extraordinary Opportunities for Research Investment" that are highly relevant to the technology-oriented disciplines within the broad field of radiology are cancer imaging, defining the signatures (ie, underlying molecular features) of cancer cells, and molecular targets of prevention and treatment. In molecular target credentialing, a specific molecular target is imaged, the molecular signature is defined, a treatment is given, and the effect of the intervention on the image findings and the signature is then evaluated. Such an approach is used to validate the proposed target as a legitimate one for cancer therapy or prevention and to provide the opportunity to ultimately individualize therapy on the basis of both the initial characteristics of the tumor and the tumor’s response to an intervention. Therapeutic radiation is focused biology (ie, radiation produces molecular events in the irradiated tissue). Radiation can (a) kill cancer cells by itself, (b) be combined with cytotoxic or cytostatic drugs, and (c) serve to initiate radiation-inducible molecular targets that are amenable to treatment with drugs and/or biologic therapies. Focused biology can be anatomically confined with various types of external beams and with brachytherapy, and it can be used systemically with targeted radioisotopes. These new paradigms link diagnostic imaging, radiation therapy, and nuclear medicine in unique ways by way of basic biology. It is timely to develop new collaborative research, training, and education agendas by building on one another’s expertise and adopting new fields of microtechnology, nanotechnology, and mathematical analysis and optimization.
© RSNA, 2003
Index terms: Neoplasms, therapeutic radiology • Radiations, protective and therapeutic agents and devices • Radiobiology, cell and tissue studies • Radiology and radiologists, research • Therapeutic radiology
The era of molecular medicine is providing new insights that have a major effect on the diagnosis, treatment, and prevention of disease. Exciting challenges and opportunities in the technologically oriented disciplines of radiology and radiation oncology are leading in new directions. In establishing priorities for research investment, the National Cancer Institute (NCI) Bypass Budget program has identified "Extraordinary Opportunities for Research Investment" (1). These areas of investment opportunity include cancer imaging, defining the signatures (ie, underlying molecular features) of cancer cells, and molecular targets of prevention and treatment.
The process of molecular target credentialing was conceptualized by Robert Wittes, MD, and Richard Klausner, MD: A specific molecular target is imaged, the molecular signature is defined, a treatment is given, and the effect of the intervention on the image findings and the signature is then evaluated. Such an approach is used to validate the proposed target for cancer therapy or prevention and to provide the opportunity to ultimately individualize the therapy on the basis of both the initial characteristics of the tumor and the tumor’s response to an intervention.
Data on the molecular processes induced by radiation are emerging. This knowledge provides unique opportunities to use radiation as the cytotoxic agent along with cytostatic treatments and to use radiation to activate a molecular pathway that then becomes the molecular target for drug treatment. The term focused biology refers to the process by which radiation can be delivered to optimize the molecular events that radiation itself creates when used with an external beam, brachytherapy, or systemically administered radioisotopes.
The Radiation Oncology Sciences Program (ROSP) of the NCI is developing the new themes of focused biology and molecular target credentialing for both intramural and extramural programs. Created in 1999, the ROSP model is designed so that NCI radiation oncology programs can be highly interactive within the NCI and with extramural investigators and industries to help serve radiation oncology and radiology as a national and international resource. In this article, I discuss these new conceptual paradigms, which serve as a model for facilitating the new "three Ds" outlined by NCI Director Andrew von Eschenbach, MD: discovery, development, and delivery.
PARADIGM 1: RADIATION ONCOLOGY AS FOCUSED BIOLOGY
Among the generalizations about the effect of radiation therapy on cancer treatment are that the primary target for cell killing is DNA and the distribution of the radiation dose is not constrained by pharmacologic barriers. Looking at track structure diagrams of ionizations produced by ionizing irradiation, one can see that although radiation is not limited by physical boundaries as drugs are, it does produce discrete areas of ionization (2) that can appear to be dramatic when high linear energy transfer–charged particles are used (3). The areas of dense ionization created by x rays are called multiply damaged sites (4).
The finding that cell killing correlates well with nonrepaired double-strand breaks indicates that DNA is a prime target (5,6). The results of many studies, however, have demonstrated that there are numerous changes induced by irradiation in a wide array of cellular targets—for example, signal transduction pathway changes, membrane changes, mitochondrial changes, programmed cell death, and cell cycle changes (7). Some of the changes may be responses to DNA damage (8–10); however, non-DNA targets, such as the cell membrane or mitochondria, may be critical components of cell death (11) or may even be the primary target (12–14). Furthermore, there is a phenomenon called the bystander effect in which the nonirradiated neighbors of an irradiated cell undergo biochemical changes (15–17) by way of molecules that may be transmitted through gap junctions in membranes or the extracellular microenvironment.
Recent studies have revealed that very low doses of radiation—as low as 2–10 cGy (this is 1%–5% of the standard single radiation fraction of 200 cGy)—can cause a sustained induction of stress response molecules (18) or a persistent increase in the production of reactive oxygen species (19). The potential consequences of these changes are numerous in that (a) they may be important in determining the fates of cells and tissue in response to ionizing radiation and thus in facilitating postradiation modifications that affect cell death, (b) they may become a unique therapeutic target for use with molecular-targeted therapies, and (c) they may aid the development of substantially new schedules for radiation therapy administration that are designed to optimize the inducible target.
Additional new biologic concepts for the molecular radiation oncology era include the potential for radiation to enhance the immune response for anticancer vaccination (20) and the potential to modulate the normal tissue response either during (21,22) or following treatment (23). Posttreatment modifications may be possible because late radiation injury is the result of a chronic persistent inflammatory process or cytokine-mediated process (24–27).
In paradigm 1, radiation as focused biology for cancer treatment, radiation therapy can cause direct cell killing, be combined with other cytotoxic and cytostatic agents, be used as an adjunct to immunotherapy, and induce biologic changes that in turn become the target for molecular-targeted therapy. Even more important for clinical care and drug development is the finding that radiation may be a crucial tool for use with many of the new molecular-targeted treatments, many of which do not produce cell death but rather merely alter biochemical pathways or cellular homeostasis. In this paradigm, the concept of radiation dose requires redefinition. While the gray value is necessary for determining the amount of energy delivered for treatment planning and quality assurance, in the future radiation dose may be determined in terms of the biologic changes produced.
The necessity to target the delivered radiation more precisely is occurring at a time when the ability to do so is at hand. The newer techniques for anatomically directed radiation include three-dimensional conformal radiation, intensity-modulated radiation therapy, various stereotactic techniques, tomotherapy, proton therapy, and brachytherapy (ie, temporary high-dose-rate and permanent low-dose-rate radioactive sources called seeds). Additionally, it is possible to use systemic radiation therapy either with a radioisotope alone or attached to a molecule, as in radioimmunotherapy. With the recent U.S. Food and Drug Administration approval of the use of yttrium 90 ibritumomab tiuxetan (Zevalin; IDEC Pharmaceuticals, San Diego, Calif) (28), the clinical application of this general approach is likely to increase. The carrier molecules may be molecules or constructs other than antibodies, such as antibody fragments, receptor ligands, and liposomes. Thus, the repertoires available for delivering radiation to sites within the body are being greatly enhanced by emerging technologic advances. What is needed now is the ability to define the target (29) and understand the biologic effect of the radiation.
PARADIGM 2: MOLECULAR TARGET CREDENTIALING
Figure 1 is an illustration of the three "Extraordinary Opportunities for Research Investment" of the NCI Bypass Budget (1) that are grouped together under the concept of molecular target credentialing. Since the goal with these opportunities is successful cancer treatment and prevention, the concept of credentialing is focused on a specific molecular pathway that one wishes to use as a therapeutic target. To use a therapeutic agent optimally, one needs to understand the underlying biologic mechanisms of the agent. Also, to guide anatomically directed therapy, imaging is essential. A change in the image findings may be used for diagnosis and to evaluate the response to a local or systemic treatment and thereby perform real-time assessment and alter treatment, if necessary.
|
I believe that the revolution in imaging is nothing short of sensational and will have a major effect on radiation therapy. Radiation oncology treatment is based on knowledge of the anatomy of the tumor and of the normal tissue, with treatment plans designed to enable one to deliver the highest possible dose to the tumor and/or the critical subregions while safely treating the normal tissues by keeping below their tolerance limit, as defined according to both dose and volume. Improvements in computed tomographic (CT) technology and the ability to plan treatments directly on the basis of image findings and transfer the treatment data from the planning system to the treatment equipment have enabled the use of conformal fields that precisely follow the tumor outline. As clinicians and scientists, we require both precision and accuracy.
Accuracy in determining the extent of the tumor is improving dramatically with the advent of molecular imaging. Techniques include CT, magnetic resonance (MR) imaging, MR spectroscopy, positron emission tomography (PET), ultrasonography with new contrast agents, and developing modalities such as optical imaging, electron paramagnetic imaging, and Overhauser enhanced MR imaging. To use these technologies for radiation therapy treatment planning, one must have the ability to fuse images, usually onto CT images because CT offers excellent depiction of anatomic relationships. Although image fusion is conceptually simple, it requires an output code from the imaging units, which "talk" with one another, and it requires mathematical models to account for any distortion from the true anatomic relationships among structures—that is, it requires morphing or warping of the images. Such accuracy is not necessary in functional imaging if the result is only the finding that a certain biochemical or molecular event either is or is not taking place—for example, the PET scan is either positive or negative. However, the radiation oncologist’s need for accurate anatomic localization of both the tumor and the surrounding normal tissue represents an added dimension to what is required of the imaging modality.
Assuming that one is satisfied with the accuracy (discussed more later in the text), one should then ask, what about the precision of treatment delivery? This requires knowing where the target is on each day of treatment, and for some patients, immobilization and/or real-time imaging is required to glean such information. The use of three-dimensional conformal and intensity-modulated radiation therapy for prostate cancer has yielded many studies, the results of which indicate that there is both interfraction (30) and intrafraction (31) organ motion. The newer techniques involving online portal imaging with fiducial markers (32) or real-time CT (ie, tomotherapy [33]) may be helpful for precision, but they require additional time for patient treatment. The added time and expense must be factored into any equation for clinical application.
The target volumes for radiation therapy include the gross target volume, the clinical target volume (which includes microscopic disease), and the planning target volume (which includes added margins to account for beam characteristics and motion). Ling et al (34) have suggested the concept of a biologic targeted volume to indicate the potential use of knowledge about the subregions within the tumor that may have hypoxia, increased cell proliferation, altered pH, and other characteristics. I suggest adding the term molecular-targeted target volume to define the molecular processes that one wishes to image, target, treat, and assess the responses of in real time.
The ability to be accurate depends on the ability of the image to represent the true location of the cancer. The latter requires understanding the underlying molecular signature that produced the image. Therefore, the middle extraordinary opportunity illustrated in Figures 1–3, molecular signatures, is critical. Molecular imaging with different modalities yields different abnormal regions within patients, depending on the imaging modality. There are regions where the abnormalities overlap and others where only one imaging agent is abnormal (35,36). To understand why one image is abnormal in the same place that another is normal requires a microscopic and molecular-biologic analysis of what is producing the image findings.
|
|
, hexokinase, and glucose transporter 1. It is recognized that the relationship between image findings, underlying biochemical mechanisms, and molecular biologic mechanisms is very complex (39). Achieving an understanding of this relationship requires a hypothesis-driven investigation of specific pathways and the use of hypothesis-generating approaches involving the use of techniques such as DNA and protein microarrays. High-throughput molecular biology techniques such as microarray analysis are potential tools for analyzing a large number of molecular processes that facilitate understanding of the "engine" behind the malignant process (40) or the basic components of a biologic response—for example, oxidative stress (41). Given that tumors are heterogeneous, molecular analysis will require distinction of the tissue types seen at biopsy: invasive cancer, preinvasive cancer, and normal stroma. A technique called laser capture microdissection (42), which enables sampling of the individual cell populations within a biopsy specimen, can aid in this analysis. The interpretation of array data requires input from biomathematicians and statisticians (43). Given the intertumoral and intratumoral heterogeneity, interpretation of genomic and proteomic data is a daunting task that requires multidisciplinary research teams.
CREDENTIALING TUMOR HYPOXIA AS A DEMONSTRATION PROJECT
A number of groups are actively studying the imaging, biologic mechanisms, and treatment of hypoxia. Within the ROSP, we are endeavoring to conduct research to validate the paradigm of molecular target credentialing by using hypoxia as the demonstration project subject. Figure 2 shows aspects of the credentialing process that we are working toward by using collaborators within the NCI and the National Institutes of Health, as well as academic and industrial partners. (Many of these collaborators are listed in the Acknowledgments.) What is noteworthy in Figure 2 is the range of expertise required: clinicians with a range of specialties, basic scientists, chemists, imaging experts, engineers, mathematicians, technologists, entrepreneurial administrators and executives, and, most important, patients willing to participate in the clinical trials.
A key aim of ROSP is to create a locus within the NCI and the National Institutes of Health for bringing all of the necessary expertises together to acquire robust and extensive data sets on a limited number of patients. This collaboration would enable one to (a) compare imaging modalities and facilitate the development of new imaging and sensor technologies, (b) evaluate the effects of local and systemic radiation (eg, intensity-modulated radiation therapy, high dose rate, and isotopes), drug, and biologic treatments on image findings; and (c) perform tissue biopsy to study molecular signature changes. This concept is central to the ROSP goal of serving as a national and international resource that welcomes and facilitates collaboration with extramural investigators and industries. This nascent program is called the Radiation Oncology Molecular Assessment and Technology (ROMAT) Center.
There is, at the moment, no perfect molecule or pathway to credential. With the need to start somewhere, however, we selected oxygen and tumor hypoxia, given the role of tumor hypoxia in treatment outcome. Hypoxia can cause resistance to radiation and drug treatments. The hypoxic microenvironment can select for more aggressive tumor cells, and the presence of hypoxia can lead to the induction of factors that can promote tumor progression. In addition, if one assumes that almost all normal tissues have ambient oxygen concentrations higher than the hypoxic concentrations in tumors, therapy directed against hypoxia or the consequences of hypoxia can provide a relatively tumor-specific target (44). A recent editorial by our group (44) provides background information on hypoxia and its importance in oncology.
A number of techniques designed for measuring tissue and tumor oxygenation are emerging; these include blood oxygen level–dependent MR imaging and flow and oxygen–dependent MR imaging (45), PET (46), dynamic enhanced MR imaging (47), and, from the radiation biology branch of NCI laboratories, electron paramagnetic imaging and Overhauser enhanced MR imaging. The latter two techniques that are under development have the potential to enable noninvasive assessment of oxygen concentrations in vivo (48–51). Oxygen concentration can also be measured invasively (52,53) so that there is somewhat of a reference standard for comparison with the image findings.
The molecular signature of oxygen is complex. With regard to cancer and other diseases such as heart disease and stroke, there is interest in the genes regulated by hypoxia (54). Such gene products have different on-and-off patterns that may enable gene transcription and protein analysis for determination of an accurate molecular signature that defines the oxygen status of the tissue (54,55). More work is required to establish an optimal oxygen-hypoxia signature. Furthermore, although the hypoxia-inducible factors are hypoxia induced, some tumors have hypoxia-inducible factors that are abnormally "on" constitutively as part of their malignant transformation (44,56,57).
When considering the molecular therapeutic aspects of oxygen and hypoxia, there is a range of approaches (44,58). One can increase the oxygen delivery per se by using normobaric oxygen, hyperbaric oxygen, carbogen, or oxygen carrier molecules. One can provide oxygen replacements such as the oxygen-mimetic radiosensitizers or "attack" the hypoxic cells with hypoxic cytotoxic agents. The physiologic characteristics of a tumor can be modified with substances that alter intratumoral pressure or blood flow; these include antiangiogenetic agents, antivascular agents, or standard chemotherapeutic agents used with modifications (ie, metronomic therapy) (59,60).
The fact that so many molecular pathways may be subject to molecular target credentialing is obvious and daunting. Many new molecular therapeutic agents are being combined with radiation therapy for use as both sensitizers and protectors. A few examples of molecular pathways being studied for radiation sensitization include the cell cycle (61), growth factor receptors (62), signal transduction pathways (63), apoptosis (64,65), protein degradation (66), protein trafficking, stress response (67–69), and antiangiogenetic agents (70). Acute and chronic radiation injuries can be modified by understanding the molecular pathways of cellular injury (71) and tissue fibrosis (25). Imaging requires the use of technology and imaging agents, and there must be an appropriate sensitivity to small biologic changes. Microsensor and nanosensor technologies may yield implantable sensors with which to monitor treatment and disease status.
Targeted therapy requires designer molecules that are pathway specific or, when it involves the use of radiation, the ability to selectively target areas with the radiation. Theoretically, the ability to boost the radiation dose to hypoxic regions in tumors (ie, a biologic treatment volume) can lead to an improvement in local tumor control (72), making the intensity-modulated radiation therapy or brachytherapy boost approach to treating tumor subregions, as defined in terms of molecular imaging, a testable hypothesis. Encouraged by the success demonstrated with imatinib mesylate (Gleevac; Novartis, Basel, Switzerland), the Bcr-Abl tyrosine kinase inhibitor (73), pharmaceutical and biotechnology industries are working to develop molecular target therapies.
The new field of RNA silencing (74), by enabling individual genes to be inhibited, is allowing the study of cell function changes in the laboratory. This should facilitate understanding of the effect of targeting specific pathways. Owing to our understanding of the complex biologic characteristics of tumors and normal tissue that has been gained from the ability to simultaneously measure many genes, proteins, and biochemical states with array-type analyses, new methods of mathematical analysis are needed. Thus, there are many unique possibilities for using the concept of molecular-targeted treatment volume.
PARADIGM 3: THE NEW THREE Ds—DISCOVERY, DEVELOPMENT, AND DELIVERY
The attitude of a person with cancer is sandwiched between great enthusiasm for new science and the flood of potential breakthroughs reported in the media on the one hand and healthy skepticism based on recent overstated research reports and the rising costs of health care on the other. While waiting for breakthrough treatments to be developed, physicians and patients must do something today. To best serve patients, we must appropriately use our current tools, recognize the limitations of these tools, and conduct high-quality research that tests the best hypotheses. Support is needed in all aspects of a research portfolio, from novel ideas to clinical application.
NCI Director Andrew von Eschenbach, MD, has outlined a balanced portfolio concept that includes the three Ds: discovery, development, and delivery. Figure 3 illustrates how the fields of diagnostic imaging, nuclear medicine, and radiation therapy can be important contributors to science and health care within this new paradigm. In Figure 3, molecular target credentialing and the three paradigm components are compared. The underlying biologic mechanism—that is, molecular signatures—which is the fundamental underpinning, is akin to discovery. Imaging and therapeutics, although based on the underlying biologic mechanism, each comprise the combined components of development plus delivery.
The subtitle of this presentation, "Now That Therapy and Diagnosis Have Separated, It’s Time to Get Together Again!" reflects my opinion that there is a need for collaboration in research and training among the radiology disciplines. Radiation oncologists need to understand imaging and the underlying biologic mechanisms that produce the image findings; diagnostic radiologists and nuclear medicine physicians need to understand how images are used by radiation oncologists. All trainees need a solid background in basic biology and the new molecular biology techniques, and they should be sufficiently intrigued and excited by the emerging science to maintain an active interest in clinical and laboratory investigations. Those who enter clinical practice should have the background necessary to keep up with basic science concepts throughout their careers.
We in the ROSP are working to implement a program by which the NCI plays an important role in national and international radiation oncology research agendas by means of active participation in addition to its necessary roles in research administration. The intent of the ROSP is to build on the expertise of many fields by being an active participant and helping to develop and encourage new models of collaboration, such as the Radiation Oncology Molecular Assessment and Technology Center. An additional aim is to foster a collaboration among research and industrial partnerships that facilitates and accelerates the introduction of new science and technology into the clinic and that brings knowledge from clinical experiences back to the laboratory. Without knowledge, collaboration, information, and support from others, none of the individual components could succeed as well as they do. The technologically oriented radiologic fields—diagnostic imaging and radiation oncology—have the opportunity to be leaders in these new paradigms (molecular imaging, signatures, and therapy) and to aim our efforts in terms of helping with the newest extraordinary opportunity for 2004, cancer survivorship.
OXYGEN: A TRIBUTE TO WILLIAM E. POWERS, MD
The 2002 RSNA Annual Oration in Radiation Oncology was dedicated to William E. Powers, MD, one of the major figures in the formation and development of academic radiation oncology in the United States. Among his early contributions was a groundbreaking study in collaboration with Leonard Tolmach, PhD; their study results demonstrated the relative radioresistance of hypoxic tumor cells in vivo. Figure 4 shows the biphasic survival curve published in Nature 40 years ago (75). The change in slope is caused by the relative radioresistance of the hypoxic cells once the more sensitive oxic cells are killed. Dr Powers would appreciate the fact that the fundamental biologic mechanisms of hypoxia are being elucidated and that the combination of imaging and therapy to target the physiologic properties and molecular changes of hypoxia is an area of intense research. The oxygen effect is a time-honored and worthy paradigm for molecular target credentialing and for exploiting the focused biologic mechanisms of radiation therapy.
|
ACKNOWLEDGMENTS
The accomplishments of ROSP are due to the work of the members of the Radiation Oncology Branch and Radiation Biology Branch of the Center for Cancer Research and the Radiation Research Program of the Division of Cancer Treatment and Diagnosis. Collaborators on the projects described in this article include the following: From the Radiation Oncology Branch are Kevin Camphausen, Cynthia Menard, Robert Miller and the physics group, and Martin Brechbiel and the chemistry group; from the Radiation Biology Branch are Jim Mitchell, Murali Krishna and his team, Angelo Russo, and David Wink; from the Radiation Research Program are Jim Deye and Phil Tofilon; from the Biomedical Imaging Program of the NCI are Dan Sullivan, Ed Staab, and colleagues; from the National Institutes of Health Clinical Center Imaging Sciences Program are Peter Choyke and King Li; from the National Heart Lung and Blood Institute are Bob Balaban and Robert Lederman; from the National Institute of Neurological Disease and Stroke is Alan Koretsky; and from the National Institute of Biomedical Imaging and Bioengineering are Rod Pettigrew and colleagues. The collaboration with and enthusiasm from various radiation oncology and imaging companies were essential. The support of the ROSP program from the leaderships of the NCI, the Division of Cancer Treatment and Diagnosis, and the Center for Cancer Research and from the Clinical Center was essential and is greatly appreciated.
FOOTNOTES
Abbreviations: NCI = National Cancer Institute, ROSP = Radiation Oncology Sciences Program
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
