Elias Zerhouni, M.D., Chair
The In Vivo Molecular/Functional Imaging Subgroup (MIS) is one of seven task forces that were created by the NCI Imaging Sciences Working Group to address pressing issues in functional tumor imaging. Chaired by Elias Zerhouni, M.D., the MIS is charged to link the vast array of emerging physiologic markers uncovered by powerful, new techniques of molecular genetics to existing and incipient imaging modalities. Drawing upon its composite expertise in molecular and cell biology, chemistry, clinical oncology, and the imaging sciences, the MIS will develop recommendations for potential cellular targets that will provide suitable tumor markers and assess the feasibility of converting those markers into viable in vivo imaging agents. In an era when chemists can visualize a single molecule through scanning tunneling microscopy, it seems not only possible but likely that medical imaging will soon provide a window into specific cellular processes, particularly events germane to neoplastic transformation and progression.
Dr. Klausner used the activities of the Developmental Diagnostics Working Group to illustrate how a Working Group has functioned and how the National Cancer Institute (NCI) has utilized Working Group recommendations. He emphasized that an important element of the Working Groups/Subgroups is that they bring together diverse communities that often do not interact with each other. Therefore, as this group explores ways to enhance the contribution of imaging sciences to the understanding of cancer, Dr. Klausner encouraged them to consider the relationships between traditional imaging and the concepts of the molecular basis of cancer, including the detection of subtle genomic and molecular changes.
Ms. Susan Waldrop provided guidance about Working Group/Subgroup meetings and interactions. Participants were reminded that discussions of the issues and needs related to high priority "investment opportunities" may include unpublished data and sensitive information that needed to remain confidential. Working Group/Subgroup recommendations are used by the NCI to develop and prioritize Institute goals and objectives, and serve as a platform for the development of operational plans by the NCI Divisions. The NCI staff develops specific requirements for any new initiatives or programs, which are reviewed as appropriate by NCI advisory committees. The activities of the Working Groups/Subgroups are conducted under the auspices of the Advisory Committee to the Director, NCI.
Overview of the Meeting
The meeting consisted of three sessions, the first of which commenced with self-introduction of the task force members who not only discussed their areas of expertise and interest, but also elaborated upon their questions about and desires for in vivo molecular imaging. At times "wish lists" were posed by the clinicians to the imaging scientists. The format for introduction itself provided fertile ground for discussion and set an energetic tone for the meeting. The four formal presentations that followed during the first two sessions were catalysts for further discussion. After each presentation, the participants provided written comments that form the backbone of this report. During the third and last session specific recommendations were consolidated by the group.
Major Points from Group Discussion
Communication: The issue of a communication gap between biologists and the imaging community was a major theme throughout both days of discussion. It was noted by all participants that, for a variety of reasons, there was little interaction between clinicians, biologists, and chemists, and the imagers (i.e., experts in MRI, PET/SPECT or optical imaging) at their own institutions. Partial solutions to the interdisciplinary communication issue have been found at The Beckman Institute at the California Institute of Technology and at the Center for Molecular Imaging Research at the Massachusetts General Hospital (MGH), but, as indicated by the name of the latter, each of those institutions was conceived with an interdisciplinary goal in mind. The key is to eliminate barriers where they have tacitly, if not structurally, existed for years both at and between institutions.
Target selection: This must occur at three levels. First, a clinical need must be defined, e.g., therapeutic monitoring. Second, a system that forms an integral part of the neoplastic process must be studied, e.g., gene transfer and expression. Third, once a suitable system has been chosen, a specific facet of that system must be targeted, e.g., thymidine kinase, because many potential enzyme, receptor, nucleic acid, and even carbohydrate-based targets exist in any given system. Once a clinical problem has been defined, it is incumbent upon the biologist to cite the appropriate system and specific target. The biologist, chemist, and imaging specialist next collaborate to determine the feasibility of imaging that target, with the chemist ultimately synthesizing the imaging agent.
Feasibility issues: This area can be dismantled roughly into sensitivity, pharmacologic, and synthesis issues. Sensitivity can be addressed by first determining the system to be imaged. Calculations based on known receptor density (from, e.g., postmortem data) and tracer affinity (from in vitro studies) enable a good estimate of feasibility for imaging. Creative attempts at signal amplification for MR imaging, e.g., amplification of transferrin receptors to increase susceptibility, were discussed. Pharmacologic issues included agent delivery both macroscopically (adequate blood flow to the tumor) or microscopically (intracellular delivery). Delivery is a fundamental issue that is critical to the development of macromolecular imaging agents. It was noted that entire centers have been established largely to address delivery, e.g., the Steele Laboratory for Tumor Biology at MGH. Other pharmacologic issues discussed included nonspecific binding, masking, and metabolism of imaging agents. Feasibility of synthesis is an issue that has perhaps more significant constraints in the context of radiosynthesis for PET (short-lived isotopes) than in development of MR contrast agents.
Appropriate animal models: Some believe that the clinical underutilization of PET, particularly in its receptor measurement capacity, may be due to initial tracer development on faulty animal models. A human xenograft transplanted to a nude mouse provides an artificial tumor metabolic and structural milieu. A superior model would be a transgenic mouse that could be engineered to produce specific tumors spontaneously. The transgenic mouse was suggested as a model in which genetic imaging targets could be efficiently evaluated. Furthermore, MR and PET scanners are being adapted to image small animals, the former often at higher field than conventional scanners, enabling acquisition of higher resolution functional (metabolic) images. The promising new technique of in vivo optical imaging may currently only be performed using small animal models due to shallow beam penetration, although prototype tomographic units are under construction.
New versus old technology: This issue arose during discussion of short-term goals, i.e., what can be done now? "Old" technology generally refers to tracer methods (PET) or anatomic rather than physiologic imaging, whereas new technology may be given to include in vivo optical or electron spin resonance (ESR) imaging, for example. Tracer methods have not been used as widely to quantify oncologic processes, e.g., tumor burden or receptor occupancy for therapeutic dosing, as they have been to study neurologic or psychiatric problems. This has been, in part, due to the lack of highly selective therapeutics in oncology. Now, with angiogenesis inhibitors, antisense oligodeoxynucleotides, and ras farnesyltransferase inhibitors in clinical trials, it may be possible to take advantage of the capabilities of these approaches.
Recommendation I: Multidisciplinary Centers for Molecular/Functional Imaging
Recommendation #1: Multidisciplinary Centers for Molecular/functional Imaging Research
Develop and maintain centers dedicated to molecular/functional imaging research in accordance with the NCI. Staff such centers with personnel from a variety of fields who have a common goal to develop technology to image aspects of human cancer, noninvasively. Assure that the personnel in these centers are in sufficiently close physical proximity, engendering development of strong collaborative ties. Support nascent molecular/functional imaging centers with services and facilities already available in industry or at the NIH.
The recent confluence of advances in molecular and cell biology with high throughput chemical synthesis and new techniques for imaging living systems mandates formalization of an infrastructure that can capitalize on these important, new scientific achievements. A scientific gulf remains between basic scientists who discover new cancer genes and intracellular pathways, any of which could serve as a diagnostic or therapeutic target, and the imaging scientists who could transform those discoveries into greater understanding of neoplasia in humans, noninvasively. By fostering greater interaction between molecular and cell biologists with imaging researchers by mere proximity and providing an environment for training future imaging scientists, a dedicated molecular/functional imaging center (MFIC) would streamline the now somewhat lumbering process of bringing a new molecular target into the cancer imaging armamentarium. Such a center would legitimize serious investigation at the interface between what appear to be widely disparate fields. To our knowledge, the only such center organized under one roof is the Center for Molecular Imaging Research at the Massachussetts General Hospital.
i. Develop a blueprint for the organization and physical plant of a MFIC. Initially centers "without walls" will be needed. Each center will be overseen by a scientific Steering Committee that will define the specific projects undertaken by the center. A center Director will be appointed by the Steering Committee. One example of such a center includes 5 interconnected core laboratories with the chemistry core more or less central to the operation. Core laboratories include: chemistry/radiochemistry, molecular biology, biochemistry, computing/database and animal/imaging. The flow of information between the cores can roughly be defined as proceeding according to the diagram below:
Discovery, characterization and refinement of new genetic targets will occur in the molecular biology core with that information provided to the chemistry core where potential new imaging agents directed toward those targets will be designed. The computing/database core will support the chemistry core through molecular modeling and structure searches but will also obviously interface with the other cores such as in data analysis or image display. The biochemistry core will be responsible for aspects of agent development that encompass cell biology and pharmacology as well. Receptor and enzyme purification and in vitro assays will occur in this core, in addition to development of kinetic models for studying complex issues involving agent delivery and metabolism. Animal work from rodents to primates will occur in the animal/imaging core, where the most promising potential agents will be tested. Although a freestanding facility is the ultimate goal, a confederation of existing laboratories at certain institutions must suffice initially.
ii. Scientific personnel in the MFIC must be professional scientists from a variety of fields including: chemistry, radiopharmaceutical chemistry, cell and molecular biology, pharmacology, computer science, radiology and biomedical engineering. Specialists in MR physics, immunology or neuroscience, for example, may also be involved. Through joint seminars periodically presented to the center by different cores and through subgroup (core) meetings, the members of the MFIC will collaborate to produce new cancer imaging agents and technologies. MFIC personnel must therefore be eager to collaborate outside of their own disciplines. In the future, a portion of MFIC personnel may be trained specifically to work at the interface of several disciplines engaged in imaging research.
iii. Ties to industry and existing centralized facilities, e.g., NCI chemistry laboratories, must be established. Industry and academics are interdependent for achieving their respective goals. Projects initiated by industry may require academic input, e.g., expertise in a specific field, access to large equipment (cyclotron) or patients. Projects initiated by academics may require compounds discovered by the high throughput techniques available only at large and at some specialty drug companies. Although the primary mission of the MFIC is intended to be an independent and creative endeavor, namely cancer imaging agent discovery, access to vast industrial chemical storehouses already in existence would not only prevent reinventing the wheel, but would also facilitate rapid introduction of imaging agents to the clinical arena. Either by purchase of access to industrial databases or by allowing industrial researchers to train in MFIC laboratories (or elsewhere in the corresponding university system) those ties may be strengthened. Representatives from industry will be invited to MFIC seminars and meetings in an effort to convince them of the utility of imaging to drug development. Likewise, a mechanism must be enacted whereby the NCI chemistry laboratories could open their doors to MFIC suggestions for synthetic targets to pursue, where that synthetic expertise needs to be supplemented at, for example, a nascent MFIC.
iv. To mature into an effective unit, the MFIC must progress through several stages that consist of the following critical elements:
1. The need for a planning phase that will require considerable lead time before the center becomes operative. The Dean of a candidate university will appoint an appropriate department chairman, e.g., radiology, oncology or genetics, who will be responsible for organizing an interdisciplinary meeting. After the initial organizational meeting, a retreat and/or series of further meetings during which clinically important targets and feasibility issues will be discussed would occur. The attendees of such a "consensus conference" may be two or three junior and two or three senior investigators from each interested department. They will eventually comprise the Scientific Steering Committee discussed above. Conferences will occur over a one year intellectual courtship period. At the end of that year, a written report to the NCI outlining a plan of action with specific project proposals will be submitted. The report must contain sound scientific ideas and demonstrate the ability for the consensus panel to collaborate effectively.
2. The need for pilot projects. These projects will be geared toward determining project feasibility, proof of principle and acquisition of preliminary data. A clear organizational structure to the MFIC must be in place at this time.
3. The need for industrial collaboration. Industrial input will be sought at the pilot plant stage. To entice industry further, the NCI may consider reimbursing the university for indirect costs that will consequently not need to be passed on to industry. One year will be allowed to accumulate pilot data and establish industrial contacts.
4. The need for actual center development. Competition for, e.g., five year center development awards may be initiated. At this stage a center Director will be appointed by the Steering Committee. A primary function of the center Director will be to assure adherence to the proposed budget. The structure of the MFIC will be patterned after a typical NIH center.
5. The need for training programs. Current graduate programs are too narrowly focused and may be inadequate to train the next generation of imaging scientists. Development of in-house imaging-oriented training programs could be formalized once interdisciplinary laboratories have been consolidated into true MFICs and have developed productive track records.
Further Comments and Discussion
Imaging science is at a stage at which human, anatomic in vivo imaging can occur at submillimeter resolution using MR. Most research in oncologic imaging has focused on the CT and MR appearance pathology in various isolated organ systems. Functional imaging modalities, including MR spectroscopy, optical imaging and the nuclear medicine techniques of SPECT and PET have been applied only sparingly to oncology. A marriage between the myriad of new cancer-related genes and proteins uncovered at an increasing pace by molecular and cell biologists and the imaging sciences must occur to capitalize on what could be a windfall for studying this human infirmity noninvasively, and in many cases, quantitatively. A first step in that direction would be the establishment of MFICs which would a) place self-selected basic (biological) scientists and imaging researchers in close proximity for easy cross-fertilization of ideas, b) provide an infrastructure in which to train the next generation of oncologic imaging researchers and therefore c) streamline cancer imaging research.
As always, the critical components to ensure the success of such a center include personnel and funds. Imaging researchers need to know which of the many potential targets in any given cellular or subcellular system related to cancer to pursue. The first step in that process (see diagram) is to encourage cancer biologists to collaborate with imaging scientists by NCI funding directed toward such collaborations.
Pioneering approaches to be emulated already exist. For example, the incipient Harvard Institute for Chemistry and Cell Biology has been conceived to address issues at the interface for which it was named with a view to developing small molecule probes of biological processes. A MFIC would take that process one step further by undertaking the daunting task of converting such probes into viable noninvasive imaging agents. One such imaging center, also at Harvard, is actively involved in that process. Ultimately, the marker for its success and for the success of any similar center will be the productive clinical application of its imaging agents in cancer diagnosis and therapy.
Recommendation II: Training
Training and dissemination In collaboration with the NCI, develop and implement an efficient mechanism for postdoctoral training in molecular imaging sciences. Develop initiatives for interdisciplinary cross-training of senior investigators, while providing means for bringing researchers and teams with diverse backgrounds together. Establish a scientific forum that fosters dissemination of research at large, so that novel probes and techniques are efficiently used for the development of novel anti-cancer strategies.
Rationale The emergence of in vivo molecular imaging strategies has largely become possible by advances in molecular and cell biology techniques, the availability of transgenic animal models, highly specific and smart imaging drugs that are activated by target interaction, new methods of combinatorial drug design and the emergence of novel imaging techniques. These advances have brought with them new challenges to imaging research clearly transcending the traditional boundaries of anatomic imaging and basic science research. Unfortunately there exist very few interdisciplinary research programs in the US that have the necessary resources to harness this opportunity and provide effective training programs. Efficient training programs and a forum for dissemination of molecular imaging research are thus highly desirable in an effort to implement and foster the development of novel non-invasive imaging strategies.
i. Develop and implement postdoctoral training programs in molecular imaging. Currently there exist no supported training programs in molecular imaging research and consequently there is a lack of junior scientists familiar with both imaging and basic science training. Establishing formal training programs will ensure that the next generation of scientists is well versed in the field of the more rigorous imaging sciences as well as in the development of new probes and imaging techniques.
ii. Develop efficient initiatives for cross-training of senior investigators. Cross-training of established investigators with proven track records is highly desirable as such individuals are likely to comprehend complex biological issues and are uniquely suited to tackle important problems through peer reviewed research initiatives. Cross training will also ensure a high scientific standard and further facilitate the efficient training of postdoctoral fellows.
iii. Establish interdisciplinary molecular imaging conferences to bring researchers with diverse backgrounds together. Such national meetings are expected to be fertilizers for future research and at the same time present a forum to peer review research. Hand in hand with conferences are the establishment of mechanisms to efficiently disseminate research results, for example as web based discussion groups or special issues in scientific journals. Because of the interdisciplinary nature of molecular imaging research, traditional avenues of research dissemination often encountered barriers because of the interdisciplinary nature of the field.
vi. Establish an interdisciplinary NCI grant review committee that is uniquely suited to objectively evaluate training and faculty grants.
Further comments and discussion
Molecular imaging, the non-invasive mapping of cellular and subcellular events in living organisms, represents a field in imaging sciences that will have significant impact in biomedical research and clinical practice of diagnostic imaging. Molecular imaging has largely become possible by significant advances in molecular and cell biology techniques, the availability of transgenic animal models, highly specific smart imaging drugs that are activated by target interaction, new methods of combinatorial drug design and the emergence of novel imaging techniques. These advances have brought with them new challenges to imaging research clearly transcending the traditional boundaries of anatomic imaging or basic sciences which usually do not employ in vivo imaging strategies. However, understanding in vivo carcinogenesis, developing highly effective imaging strategies to phenotype tumors, developing tools to assess tumoral growth kinetics, invasion and metastatic spread and to develop methods which can assess the efficacy of novel therapies are all highly desirable goals in the quest to eliminate cancer. Currently there exist only a few research programs have the necessary interdisciplinary resources to harness the unique opportunity that molecular imaging will provide and consequently no effective training programs exist in the US for highly qualified M.D. and Ph.D. scientists. Whereas the current generation of molecular biologists, cell biologists, geneticists and immunologists is quite familiar with molecular aspects of neoplastic diseases, there is a limited knowledge in imaging techniques, data acquisition, 3D reconstruction, probe design, or barriers to probe delivery in vivo. Likewise, the current generation of scientists trained in imaging sciences is often unfamiliar with the molecular basis of neoplastic disease. For these reasons, efficient training programs and conferences need to be established that bring together multiple disciplines and foster biological imaging research. These programs should focus on developing scientists who can exploit molecular techniques to answer basic science questions through in vivo. The training program should include, but not be limited to, instruction in imaging sciences, chemistry, molecular biology, cell biology, tumor biology, and computer science. It is expected that the successful implementation of training will foster the development, validation, and application of novel in vivo imaging techniques such as receptor mapping, imaging of gene delivery and gene expression, imaging of cell trafficking in vivo, 3D imaging in developmental biology. This should further enhance our understanding of disease mechanisms and go hand in hand with the development of molecular medicine.
Recommendation III: Imaging Agent Discovery - Chemical Design and Synthesis of Imaging Agents
A barrier to the further development of imaging modalities (e.g. MRI) as diagnostic tools in experimental biology and clinical settings is the inability to obtain new classes of contrast enhancement agents. While steady progress has been made in the development of new imaging techniques and hardware, and the identification of genetic targets, there has not been a concerted effort to design, synthesize, and test imaging agents to exploit these advances and discoveries.
The In Vivo Molecular Imaging Subgroup discussed two possible approaches to facilitate the discovery and evaluation of new imaging agents: establishing chemistry "resource centers" where precursors to tracers and synthetic chemistry support for other agents could be tapped; and obtaining access to industry's chemical databases to search for viable cellular targets and specific compounds that may not be suitable for therapeutic purposes, but may be ideal for imaging.
Specific areas in which new agents would have significant impact include:
whole organism in situ hybridization studies
whole organism investigation of antigens and immunocytochemistry for the location of tumors
identification and localization of toxin and drug binding sites
performance of rapid screens of the physiological response to drug therapy.
The design and synthesis of imaging agents would be greatly enhanced by the development of "Multidisciplinary Centers for Molecular/Functional Imaging Research." It is envisioned that these centers, which would have multidisciplinary research teams, would develop new technologies and reagents for the imaging of biological structure and function. Two examples of this framework are at Harvard Medical School and The Beckman Institute at the California Institute of Technology. (see Recommendation #1 - Multidisciplinary Centers for Molecular/Functional Imaging Research)
i. The NIH/NCI should support basic research dedicated to the design, synthesis, and testing of new contrast enhancement media for in vivo imaging of cellular processes and tumor identification.
ii. The NIH/NCI should create a mechanism to fund the chemical synthesis of contrast agents for use by researchers in the clinical and basic research settings. This goal may be accomplished by a consortium of research laboratories or established NCI facilities. The specific goals of the chemistry core would be:
Synthesis - organic and inorganic coordination chemistry of new goal directed agents
Characterization - purification and physical measurements of the new agents, including biodistribution and toxicity data
Dissemination - make the agents and images obtained available to the scientific community in convenient formats, such as a catalogued library of complexes.
iii. The NIH/NCI should facilitate access by the research community to the chemical storehouses, databases, and the high-throughput capabilities of industry, which would accelerate imaging agent discovery.
iv. The NCI should establish a yearly meeting to address the need for communication to be established and encouraged between clinicians, biologists, and chemists.
Recommendation IV: Small Animal Imaging Facility
I. Recommendation: The NIH/NCI should support dedicated small animal imaging facilities focusing on the study of genetically engineered tumor models.
Neoplasms have an intrinsic spatially distributed nature.
That is, tumors develop in different sites, metastasize to other sites and are internally heterogeneous. To study tumors one must make spatially distributed measurements. Imaging is nothing more than a means of making and displaying spatially coherent measurements and is therefore a key resource for studying the development, growth and therapeutic response of neoplasms.
A major limitation to studying tumors with current imaging techniques is their limited applicability to contemporary tumor models. Key understanding of neoplastic behavior is being derived from molecular biological techniques which are often related to small animal models, particularly genetically engineered mice. Most biomedical imaging devices have been optimized for human studies and have inadequate spatial resolution for small animals and their tumors. However, most imaging techniques can be scaled down to yield very high resolution and signal sensitive in vivo images of mouse sized samples. Furthermore, there are some imaging techniques which could provide valuable information in small animal models, but are not applicable to human subjects. Therefore, in order to take advantage of the small animal tumor models being developed, it is recommended that dedicated small animal imaging laboratories be developed.
B. Specific Recommendations
1. Multimodality mouse imaging laboratories should be developed. While there are other transgenic animal models, including rabbits and pigs, they are less numerous and more expensive to develop, maintain and study. The momentum is for using mice in mammalian gene work.
a. Support several multidisciplinary mouse imaging 'Centers of Excellence.'
b. Develop novel, affordable small animal in vivo imaging devices providing anatomic and functional images, starting with MR, PET/SPECT, ultrasound and possibly x-ray CT, complemented with more novel techniques such as optical and EPR imagers.
c. Support appropriate personnel such as veterinarian pathologist, small animal anesthesiologist, veterinary and imaging technologists.
d. Develop image processing and analysis techniques for high-throughput, automated screening of whole animal/organ scans.
1. Integrate the in vivo imaging techniques with in vitro and ex vivo imaging techniques, including gene expression maps.
2. Develop animal specific radiopharmaceuticals and contrast agents.
C. Comments and Discussion
1. Is a survey of actual services needed by potential users required?
2. Should the small animal imaging facilities provide all, most or one critical imaging technique(s)?
3. Are 'dirty' and 'clean' animal facilities required for an imaging center?
4. Small animal veterinary skills are critical. For example, vascular access in small animals is a challenge, especially chronic vascular access.
5. Because commercial animal imaging devices are available in some instances, procurement of commercial devices when available would speed the development of a central facility.
6. The small animal imaging centers offer a unique opportunity for new imaging technique development.
7. The facilities may need support for a radiopharmacist and/or chemist since animal biochemistry and physiology isn't always the same as human's.
8. Is it appropriate to "pilot" one or a few of these concepts with available technology?