Molecular imaging in oncology

Transcription

1 17 (Supplement 10): x287 x292, 2006 doi: /annonc/mdl275 Molecular imaging in oncology D. C. Sullivan National Cancer Institute, National Institutes of Health, Bethesda, MD, USA introduction In vivo medical imaging is performed by administering energy to the body and measuring, with spatial localisation, the energy that is transmitted through, emitted from or reflected back from various organs and tissues. The difference between the administered and the recorded energy provides information about the properties of the matter with which the energy interacted. The energy most commonly used is some form of electromagnetic energy, such as X-rays or light, but occasionally other forms are used, such as the mechanical energy used for ultrasound scans (Figure 1). The information extracted from the energy detected in clinical imaging technologies has generally been used to infer something about the underlying anatomy or structure. This has been, and continues to be, enormously important in oncology. For example, dramatic improvements have occurred in the past 25 years such that modern computed tomography (CT) and magnetic resonance imaging (MRI) scanners can now depict anatomic detail at sub-millimetre resolution. However as oncology moves into the molecular era, the property of matter that oncologists will increasingly want to know about is the biochemical makeup of normal and abnormal tissues. molecular imaging techniques There are a variety of imaging methods that can display information about a patient s biochemistry, and no single modality is superior to all others [1]. Collectively, these methods are referred to as molecular imaging. Molecular imaging agents and methods have been developed for a variety of systems using different forms of energy. These include nuclear medicine methods, such as positron emission tomography (PET) and single photon emission computed tomography (SPECT), MRI, ultrasound methods, CT scans and optical technologies. Although the term optical implies the use of visible light, it is often more broadly applied to include near-infrared (NIR) methods as well; the term photonics is sometimes used to describe both visible and non-visible radiation. These technologies have different advantages and drawbacks. For example, CT and MRI are able to portray anatomical detail exceptionally well, whereas nuclear medicine and optical methods have very high sensitivity for detecting specific molecules but cannot portray anatomical detail with high spatial resolution. Increasingly, methods with complementary strengths are combined in clinical practice, such as the CT PET systems that are now commercially available (Figure 2). image-enhancing agents In vivo molecular imaging can be thought of as a form of in vivo assay. Some in vivo imaging methodologies, such as magnetic resonance spectroscopy and optical spectroscopy, allow us to make direct inferences about underlying biochemistry simply from administering energy and analysing the recorded energy. However, the extent of biochemical information that can be obtained from energy alone is currently limited. Therefore, we commonly give patients diagnostic drugs, referred to as contrast agents or molecular probes, which interact in a specific way with the patient s underlying biochemistry and thereby alter the recoded energy in a way that tells us more about the patient s biochemistry than we could learn from administered energy alone. There is a wide variety of molecular mechanisms that can be used for developing imaging agents [2]. Some of these involve binding to cell-membrane structures, whereas others exploit transport mechanisms into the cell, and subsequent enzymatic or other biochemical reactions within the cytoplasm. Others may be localised to intracellular structures such as mitochondria or within the nucleus itself. No single molecular mechanism in the cell precludes all others for clinical utility (Figure 3). in vivo versus in vitro techniques In vivo imaging assays cannot currently provide the degree of genomic, proteomic and other phenotypic information that can be obtained from various in vitro assays on biopsied tissue or body fluids. However, in vivo imaging has at least three potentially important advantages that complement information from in vitro tests. First, imaging provides spatially localised information over large volumes of tissue or the entire body, whereas in vitro tests are usually performed on a very small volume of tissue. The term regional proteomics is sometimes used, indicating that imaging may reflect the heterogeneity of cancer better than in vitro techniques. Second, in vivo imaging can give dynamic information by being obtained serially or continuously for periods of time. In vitro assays provide information from a single point in time. Third, in vivo imaging depicts information from a tumour in its usual milieu or microenvironment. In vitro assays, on the other hand, will reflect the changes in gene expression patterns ª 2006 European Society for Medical Oncology

2 Figure 1. Medical imaging involves administering energy to the body and measuring, with spatial localisation, the energy that is transmitted, emitted or reflected from the subject. The difference between the administered and the received energy provides information about properties of the matter with which the energy interacted. Increasingly the property of matter that oncologists need to know is the biochemical makeup of the tumour. Figure 3. Different imaging agents localise in or on cells by a variety of cellular mechanisms, including selective binding to receptors, trapping by enzymatic processes and incorporation into cellular structures by endogenous metabolism. Figure 2. Combined PET/CT scan from a patient with mandibular cancer. The white area (arrow) represents uptake of fluorodeoxyglucose (FDG) in the left mandibular tumour. The area of FDG uptake is superimposed on a computed tomography (CT) scan of the neck and mandible area acquired simultaneously with the PET scan. that occur very quickly after tissue is removed by biopsy. Information from in vivo and in vitro studies is therefore complementary, and both are essential in modern oncology research and clinical care. Investigators in drug development need in vivo assays to tell them whether a given patient has the appropriate molecular phenotype to benefit from a targeted therapy, to indicate whether the drug has hit its molecular target, to determine whether the drug has been given in the optimal biologic dose, and to ascertain whether the tumour is responding. Clinicians increasingly will have a series of targeted therapies to choose from for any given tumour, and will need in vivo assays to get an early determination as to whether their patient is responding to the chosen therapy. Early predictive assays will be important so that clinicians can change therapy quickly, thereby obviating unnecessary toxicity and expense, and increasing the chances of matching the patient to an effective therapy. In addition to knowing whether a given biochemical event is occurring or not, researchers and clinicians need objective, quantitative information about the biochemical events and need to monitor them quantitatively over time, before and after interventions. That level of quantification is generally not yet available in clinical imaging methods, but is an area of active research and development. imaging of gene function As the fundamental basis of cancer is at the gene level, molecular imaging methods that report directly on gene function would be x288 Sullivan Volume 17 Supplement 10 September 2006

3 particularly useful [3]. Reporter genes could be used in one of two ways. One method is to insert a gene that codes for an enzyme. A labelled substrate for that enzyme could later be administered and the trapped probe would identify those cells that express the reporter gene. The second method is to insert a gene that codes for a cell-surface receptor and later administer a labelled ligand specific to that receptor. Unfortunately, in most diagnostic and therapeutic situations it is not practical to insert reporter genes into a patient s native cells. However, genes could be inserted into genetically engineered cells or viruses that were administered to a patient, such as in stem-cell or vaccine therapies. The concept of nanoparticles as generic platforms for imaging agents is also under intense investigation. A single particle, such as a dendrimer, liposome or other construct, acts as a general platform to which a variety of signaling moieties is attached, along with one or more targeting molecules that can be substituted as required (Figure 4). Such imaging agents have already been developed for nuclear medicine, ultrasound, magnetic resonance and optical applications [4]. photonic techniques Photonic methodologies have also attracted much attention in recent years as they have the advantages of not involving ionising radiation and of being cheaper than traditional clinical imaging techniques. Photonic technologies that do not require the administration of any exogenous contrast materials, and which measure photon reflection, transmission, refraction or fluorescence from endogenous fluorophores, have already been developed. Other photonic methods use similar physical properties, but require the administration of agents that either fluoresce or bioluminesce. Organic dyes that fluoresce have been Figure 4. Diagram of a stylised liposome showing the concept of a multifunctional nanoparticle. Other nanoparticle platforms, such as dendrimers or polymers, would function in a similar way. The large ( nm diameter) nanoparticle serves as a general platform to which a variety of signaling moieties (such as gadolinium for magnetic resonance imaging or a radioisotope for nuclear medicine imaging) can be attached, along with one or more molecularly-specific targeting molecules, which could be small molecules, antibodies, etc. Therapeutic drugs can also be encapsulated into or attached to the nanoparticle. used clinically for some time, but their value in the molecular imaging of humans is limited because of tissue absorption. There is also considerable interest in using quantum dots or other nano-constructed particles that fluoresce much more intensely than organic dyes [5]. Such particles can be engineered to fluoresce at a variety of wavelengths, while still being stimulated by light of a single wavelength. A major drawback of bioluminescence is that it requires the combination of an enzyme (such as luciferase) and its substrate (luciferin), and most methods insert the enzyme s gene into target cells using genetic engineering [6]. As a result, this method is unlikely to be clinically useful, but it has become an enormously valuable technique for basic research. Another potentially important characteristic of optical probes is the development of so-called activatable probes, which exploit the phenomenon of fluorescence resonance energy transfer (FRET). Two fluorescent molecules can be held in a steric configuration such that they will not fluoresce when stimulated by the appropriate wavelength of light. If something disturbs that configuration, they will fluoresce. In biological situations this is exploited by attaching the fluorophores to a substrate such as a polymer by peptide linkers, which are themselves the substrate for a particular enzyme, such as a protease. When the protease is present, for example in cancer cells that over-express it, the peptide linkers will be cleaved and the fluorophores will move away from the substrate and fluoresce, thereby signaling that the protease of interest is present and active [7]. Similar types of activatable agents that use magnetic resonance contrast materials have been reported [8]. Activatable agents that use radioisotopes are not feasible, because the radioactive decay phenomenon cannot be controlled to respond to a particular molecular event as in photonic or magnetic resonance situations. The main drawback of photonic imaging techniques is that visible light and NIR are highly absorbed by water and tissue. Therefore, the future role of photonic methods in patients, in which light may have to traverse many centimetres of tissue, is not yet clear. Nevertheless, photonic techniques could be valuable for the assessment of mucosal surfaces where the majority of human cancers originate. Engineering simulations have also suggested that it may be feasible to use photonic methods in deep tissue, although such applications are still far from clinical use [7]. biomarkers Remarkable advances in the understanding of neoplastic progression at the cellular and molecular levels have spurred the discovery of molecularly-targeted drugs. Although several new oncology drugs have reached the market over the past few years, >80% of drugs for all indications entering clinical development do not get marketing approval, with many failing late in development, often in phase III trials, because of unexpected safety issues or difficulty determining efficacy, including confounded outcomes. These factors contribute to the high costs of oncology drug development and clearly show the need for faster, more cost-effective strategies for evaluating oncology drugs and better definition of patients who will benefit from treatment. Volume 17 Supplement 10 September 2006 doi: /annonc/mdl275 x289

4 In oncology, the gold standard clinical trial endpoint is overall survival, which may require long-term studies and may be confounded by deaths from causes other than the patient s cancer. Over the years, the oncology community and the US Food and Drug Administration (FDA) in evaluating oncologic therapies have come to rely on other endpoints that, from a scientific perspective, are regarded as correlates of clinical benefit [1]. These endpoints are objective response (OR), timeto-progression (TTP), disease-free survival (DFS) and progression-free survival (PFS). All are determined by biomarkers measuring the cancer s extent. Anatomic imaging using one- or two-dimensional measurements to characterise cancers has been used traditionally to make these measurements in all aspects of cancer patient management from diagnosis and staging to monitoring response to therapy and disease progression. However, the measurements made using standard anatomic imaging techniques are often inadequate for characterising the cancer, especially for monitoring the effects of drugs that do not cause tumour shrinkage or for cancers that progress slowly or metastasise diffusely. Newer imaging modalities, including volumetric and functional imaging, show high promise as the basis for characterising better biomarkers of cancer. These issues highlight the need for faster, more efficient, and more cost-effective development of cancer therapeutics and for better definition of patients likely to benefit from treatment. As addressed by the recent FDA Critical Path Initiative ( collaborative interactions among such areas as bioinformatics, genomics, materials science and imaging technologies are needed to design and implement better drug development tools. Important among these tools are functional molecular imaging methods that enable visualisation of phenotypic expression of key targets in the cancer disease processes. Unlike anatomic imaging, functional imaging methods display biochemical and physiologic abnormalities underlying the cancer rather than the structural consequences of these abnormalities. Imaging-based biomarkers have many potential uses in all phases of the drug development process, from target discovery and validation to pivotal clinical trials for drug registration [9]. First, as disease biomarkers, imaging end points can be employed to define, stratify and enrich study groups. One such approach is to apply imaging-based methods to identify appropriate patient populations in which to test targeted agents. An example would be the use of [ 18 F]estradiol (FES) positron emission tomography (PET) scans to identify patients for aromatase inhibitor trials [10]. Although it is routine to assess estrogen receptor status on the diagnostic biopsy of breast tumours, the result of this lab test will apply only to that piece of tissue tested and only to the time point in the course of the tumour at which the biopsy was obtained. Since tumours are heterogeneous, and change over time, the estrogen status of that biopsy may not be an accurate predictor of whether that patient s total tumour burden will respond to hormonal therapy or aromatase inhibitors. Figure 5 shows images where the FES-PET scan demonstrates that the majority of bone metastases in this patient express the estrogen receptor, and the post-therapy FDG-PET scan confirms that the patient had a good therapeutic response to hormonal therapy. Second, some clinical imaging methods have potential to facilitate early clinical pharmacokinetic/pharmacodynamic assessments, particularly in patients where traditionally there are no direct measures of pharmacokinetics/pharmacodynamics throughout the tissues of the body and at the target. These approaches could be used in early studies comparing lead candidates designed to interact with the same target. One example is the use of dynamic-contrast-enhanced magnetic resonance imaging (DCE MRI) as a measurement of the exposure-dependent effects of drugs targeting the tumour vasculature (e.g. anti-angiogenesis) occurring prior to tumour shrinkage (Figure 6) [11, 12]. A third area where imaging-based biomarkers have promise for speeding drug evaluation is by replacing or supplementing time- and labour-intensive dissection and histological analyses in both preclinical and clinical testing. These noninvasive approaches may enable longitudinal preclinical studies with greater relevance to future clinical study designs. Examples include several optical technologies, sometimes referred to collectively as optical biopsy. Another example is the use of magnetic resonance spectroscopy (MRS) to monitor total choline levels as a marker in functional imaging for adjuvant therapy for breast or prostate cancer, providing Figure 5. Pre-therapy and 6-weeks post-therapy posterior view PET scans from a woman with metastatic breast cancer treated with hormonal therapy. The pre-therapy fluoro-estradiol (FES) scan demonstrates that almost all the metastatic lesions identified by the pre-therapy FDG-PET scan (black areas on image labelled FDG) express estrogen receptors (black areas on image labelled FES). The post-therapy FDG scan indicates that the metastatic lesions have responded well to the hormonal therapy (i.e. most of the black areas have significantly diminished their uptake of FDG). x290 Sullivan Volume 17 Supplement 10 September 2006

5 will not be accomplished by individual scientists or research institutions. To this end, the US National Cancer Institute (NCI), FDA, academic researchers and industry have entered into collaborations to identify biomarkers that will provide clearer pictures of a patient s cancer and its response to therapy in a timely fashion. The NCI and FDA are developing a strategic plan for evaluation of biomarkers to address these needs for collaborative research to identify the best biomarkers for oncology, to standardise data collection and analysis, and to provide a pathway for establishing the use of new biomarker tools in oncology drug development and patient care. This plan includes public private partnerships with the pharmaceutical and imaging device companies. Furthermore, building on the recommendations of earlier workshops on these topics, NCI convened two workshops in the past year, one for MRS and one for FDG-PET, to develop consensus on standards for performing the imaging studies [15, 16]. As these workshop recommendations are transformed into recommendations for standardised image acquisition protocols in clinical trials, these uniform protocols will be available on the web site called Uniform Protocols for Imaging in Clinical Trials (upict.acr.org). Figure 6. (A) Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) scan from a woman with breast cancer (white arrows). (B) Time activity curve from one of the breast tumours. Y-axis is MRI signal intensity; x-axis is time (min). The amplitude and shape of the curve reflect blood flow and permeability in the tumour. Repeat DCE-MRI studies days or weeks after anti-angiogenesis therapy can be used to monitor the response of the tumour vascularity to the chemotherapy. information about therapeutic effects within days after treatment [13]. Finally, as biomarkers of tumour response, imaging endpoints can also serve as early surrogates of therapy success [14]. For example, clinical trials in breast cancer and other settings [e.g. non-small cell lung cancer (NSCLC) and oesophageal cancer] have demonstrated that 2-[ 18 F]-fluoro-2-deoxyglucose positron emission tomography (FDG-PET), a functional imaging modality, can provide an early indication of therapeutic response that is well-correlated with clinical outcome. FDG-PET thus has the potential to improve patient management, particularly by signalling the need for early therapeutic changes in non-responders and partial responders, thereby obviating the side effects and costs of ineffective treatment. As an early surrogate for clinical benefit, the modality also has the potential to facilitate oncologic drug development by shortening phase II trials and detecting clinical benefit earlier in phase III investigations. Challenges to the development and implementation of imaging modalities in drug development include the lack of validation and standardisation of new as well as established imaging methods. The identification and evaluation of biomarkers require access to and systematic analysis of large amounts of data, new technologies and extensive research resources. Further, there is a requirement for convergence of research, regulatory and drug developer thinking an effort that conclusions While major technological advances in current imaging modalities and introduction of new techniques into the clinical arena will provide improved resolution, advances in functional imaging are likely to provide the greatest growth in the area of oncology. In many cases, it will be the fusion of functional data with the improved imaging capabilities that will provide the greatest overall benefit. Not only will this improve overall accuracy, initially by improved specificity, but will also allow interrogation of metabolic activity irrespective of anatomical changes. Functional knowledge will derive in large part from specific molecular probes that allow quantitative correlation between imaging and a specific molecular process. This is a time of explosive growth in biology, technology and medicine, during which research is yielding a clearer understanding of the genome and its phenotypic expression. Accompanying this understanding is the challenge to view the human body and its diseases as a complex adaptive system under genetic control. The implication for oncology, as it is for almost all of medicine, is to begin thinking at the molecular level. Efforts to detect and diagnose cancer, to target therapies and to monitor results should be directed at the molecular level. Biomedical imaging is the key technology for accomplishing these goals. Imaging practitioners should think of themselves primarily as biologists whose task is to image what is happening functionally and structurally at the molecular level. Although the challenges are significant, the opportunities for imaging to be central to modern health care delivery are correspondingly enormous. references 1. Thomasson DM, Gharib A, Li K. A primer on molecular biology for imagers. Acad Rad 2004; 11 (Suppl): Volume 17 Supplement 10 September 2006 doi: /annonc/mdl275 x291

6 2. Danthi SN, Pandit SD, Li K. A primer on molecular biology for imagers: VII. Molecular imaging probes. Acad Rad 2004; 11 (Suppl): Min JJ, Gambhir SS. Gene therapy progress and prospects: noninvasive imaging of gene therapy in living subjects. Gene Ther 2004; 11: Sullivan DC, Ferrari M. Nanotechnology and tumor imaging: seizing an opportunity. Review. Mol Imaging 2004; 3: Michalet X, Pinaud FF, Bentolila LA et al. Quantum Dots for Live Cells, in Vivo Imaging, and Diagnostics Review. Science 2005; 28: Contag C, Bachmann MH. Advances in in vivo bioluminescence imaging of gene expression. Annu Rev Biomed Eng 2002; 4: Ntziachristos V, Bremer C, Weissleder R. Fluorescence imaging with nearinfrared light: new technological advances that enable in vivo molecular imaging. Eur Radiol 2003; 13: Louie AY, Huber MM, Ahrens ET et al. In vivo visualization of gene expression using magnetic resonance imaging. Nat Biotechnol 2000; 18: Kelloff GJ, Sigman CC. New science-based endpoints to accelerate oncology drug development. Eur J Cancer 2005; 41: Mankoff DA, Peterson LM, Petra PH et al. Factors affecting the level and heterogeneity of uptake of (18F) fluoroestradiol (FES) in patients with estrogen receptor positive (ER+) breast cancer. J Nucl Med 2002; 43: 286P. 11. Leach MO, Brindle KM, Evelhoch JL et al. The assessment of antiangiogenic and antivascular therapies in early-stage clinical trials using magnetic resonance imaging: issues and recommendations. Br J Cancer 2005; 92: Morgan B, Thomas AL, Drevs J et al. Dynamic contrast-enhanced magnetic resonance imaging as a biomarker for the pharmacological response of PTK787/ZK222584, an inhibitor of the vascular endothelial growth factor receptor tyrosine kinases, in patients with advanced colorectal cancer and liver metastases. J Clin Oncol 2003; 21: Meisamy S, Bolan PJ, Baker EH et al. Predicting response to neoadjuvant chemotherapy of locally advanced breast cancer with in vivo 1H MRS: A pilot study at 4 Tesla. Radiology 2004; 233: Kelloff GJ, Hoffman JH, Johnson B et al. Progress and Promise of FDG-PET Imaging for Cancer Patient Management and Oncologic Drug Development. Clin Can Res 2005; 11: Evelhoch J, Garwood M, Vigneron D et al. Expanding the use of magnetic resonance in the assessment of tumor response to therapy. Workshop Report. Cancer Res 2005; 65: Shankar LK, Hoffman JM, Bacharach S et al. Consensus recommendations for the use of 18F-FDG PET as an indicator of therapeutic response in patients in National Cancer Institute trials. J Nucl Med 2006; 47: x292 Sullivan Volume 17 Supplement 10 September 2006