British Journal of Radiology 75 (2002),S6-S15 © 2002 The British Institute of Radiology
Silvanus Thompson Memorial Lecture |
Molecular imaging with PET – the future challenges
T Jones
Wolfson Molecular Imaging Centre, Christie Hospital NHS Trust, Wilmslow Road, Manchester M20 4BX, UK
 |
Introduction
|
|---|
Silvanus Thompson was very enthusiastic about the technical applications of electricity, magnetism and optics and devoted most of his life to these areas. When Röntgen's discovery was first announced in the UK, Thompson repeated Röntgen's experiment the following day. He gave the first public demonstration of ‘Röntgen rays’ to the Clinical Society of London and later became the first President of the Roentgen Society. He was renowned for his clarity of presentation and could explain very difficult concepts in simple language. He was renowned for his "word pictures". Hence it is most daunting to be asked to deliver a lecture in his name when a presentation on future challenges has to rely on word pictures. It is also daunting to realise who were the previous presenters of this lecture. The first was Lord Rutherford, followed by other greats that include J J Thompson, Bragg Senior and Junior, Hevesey, Sievert, Gray and of course, Sir Godfrey Hounsfield, together with many other eminent, mainly British, scientists.
|
Molecular imaging with PET
|
|---|
I am commenting on the challenges to this area from my experience of working for over 30 years at the Hammersmith Hospital with short-lived positron emitting isotopes. In doing so, I call on my experience of developing methodology, sometimes in anticipation of, and sometimes in response to, clinical research questions. This has involved forging working relationships with a wide range of clinical specialties, sharing with them the risks of experimentation in clinical science and building multidisciplinary teams to support the programs.
Prior to discussing future challenges, I will first summarize what have we learnt to date about the use of PET, and from there identify future applications. The future challenges are addressed from the scientific, technical and practical standpoints. I am also going to present the means for encompassing these future challenges, which rests on creating dedicated molecular imaging centres based on the use of PET.
|
What have we learned about PET?
|
|---|
The advantages of PET lie in its specificity, its sensitivity and its ability to quantify tissue concentrations of tracer. Specificity comes from being able to label a given molecule with radioisotopes like carbon-11 without perturbing its biological function. Sensitivity arises from the use of coincidence, electronic collimation counting, and quantification from the ability to correct for signal loss due to tissue attenuation of emitted photons [1].
The use of PET technology in clinical research should not be seen to be isolatory in nature but as having a role to help translate basic knowledge derived in the laboratory or at postmortem to the living patient. Examples of this are measurements of tissue physiology, pathophysiology, pharmacokinetics and pharmacodynamics. The spin off from such research has been the use of PET for clinical management, an example being the use of 18-FDG for tumour localization.
Physiology
Figure 1
shows focal activation in response to subjects recalling telephone numbers, using PET and H215O measures of regional brain blood flow [2]. It demonstrates the existence of inner ear and inner speech areas in the brain, the location of which had been inferred earlier from clinical assessments of people who had suffered focal brain damage.
View larger version (129K):
[in this window]
[in a new window]
|
Figure 1. Focal activation of regional cerebral blood flow using H2O15, showing the inner speech and inner ear of a normal subject, silently keeping language stimuli, such as telephone numbers, in working memory [2]. Reproduced with the kind permission of Dr E Paulescu and Professor C Frith.
|
|
A more complex neurophysiology study using PET is its use to localize neurochemical changes which occur in the brain during a task–reward paradigm. Figure 2 shows the results of using [11C]-raclopride, a PET ligand for measuring dopamine receptor subtype 2 in the brain, in subjects playing a video game for which they were being rewarded. The figure shows that as subjects became more successful in the game there is an associated release of endogenous neurotransmitters that compete with the PET ligand for D2 receptor binding site [3]. Although this biological action was known from earlier animal brain dialysis studies, molecular imaging with PET has allowed such studies to be translated to humans, thereby confirming the existence and extent of this phenomenon in man.
View larger version (88K):
[in this window]
[in a new window]
|
Figure 2. Focal neurotransmitter activation showing 11C-raclopride displacement during a video game based reward task, demonstrating the local release of dopamine [3]. Reproduced with the kind permission of Professor P Grasby and Macmillan Publishers Ltd.
|
|
Pathophysiology
An example of the use of PET for a translation study of pathophysiology is to investigate the role of microglia activation as a defence system of the brain in human disease. Microglia reside in the brain in a quiescent state and become activated when neurological damage occurs. Then they migrate to the damaged region, changing their shape and becoming macrophagic. They tend to isolate the damaged area of the brain by interposing themselves within the associated synapses. This is either to help it heal, or to prevent its malfunctioning from affecting the activity of the normal tissue. In order to be able to investigate this generic process in human disease, or as a sensitive marker of disease, a PET ligand [11C-PK]-11195 has been developed as a measure of the peripheral benzodiazepine binding site that is expressed on the mitochondria of activated microglia. Figure 3 shows how PET studies using this probe have demonstrated the presence of activated microglia [4] in the brains of early dementia patients before there are anatomical changes on the MRI scan [5]. This is translating what was already known from autopsy studies of the brains of advanced demented patients to early stages of the disease. It also demonstrates the potential of this technique for monitoring the effects of anti-inflammatory therapy for dementia and cerebral vascular disease.
View larger version (76K):
[in this window]
[in a new window]
|
Figure 3. Visualization of microglia activation in the brain of an early dementia patient using 11C-PK-11195 [5]. Reproduced with the kind permission of Dr R Banati and Elsevier Science.
|
|
Pharmacokinetics
This application of PET focuses on its use to help drug development. PET can measure the occupancy of a therapeutic molecular target at different administered doses of a drug. An example of such a dose ranging study is that of the occupancies of the 5HT1A receptor, measured with the PET ligand [11C]WAY-100635, in the brains of subjects to whom increasing doses of the drug pindolol were administered as part of an antidepression regime [6] (Figure 4).
View larger version (58K):
[in this window]
[in a new window]
|
Figure 4. The use of the ligand 11C-WAY100635 to measure the amount of the drug pindolol needed to be administered to achieve blockade of the 5HT1A receptor [6]. (a) Pindolol occupancy of pre-synaptic and post-synaptic 5-HT1A receptors. (b) 11C-WAY100635 binding before and after 20 mg of pindolol. Reproduced with the kind permission of Professor P Grasby and Elsevier Science.
|
|
The more direct pharmacokinetic application of PET is that of labelling the drug itself and using PET to measure the dose that is achieved within the tissue of interest and its time course. An example of such a study is measurement of the uptake of the anti-cancer drug temozolomide in brain tumours, which was possible using PET by labelling it with carbon-11 [7] (Figure 5). By recording in parallel the blood's plasma concentrations, it was possible to measure, as a pharmacokinetic parameter, the drug's volume of distribution in the tumour relative to that in blood.
View larger version (94K):
[in this window]
[in a new window]
|
Figure 5. The uptake of 11C-temozolomide in cerebral glioma [7]. Reproduced with the kind permission of Professor P Price.
|
|
Pharmacodynamics
Pharmacodynamics is concerned with measuring the functional response to therapy. Again this use of PET is finding application in drug development by providing proof of an agent's biological activity by measuring early metabolic response.
An example is when it was used to document metabolic changes in the glucose metabolism of brain tumours, treated with temozolomide before there were structural changes in the X-Ray CT [8]. For this, the tracer 18FDG was used as a measure of glucose utilization, where it is currently finding widespread application in the context of tumour detection. This is the topic of other chapters in this issue. While this use of PET is attracting much interest, especially for lung tumours, it is sobering to consider that the first demonstration of the ability to study glucose metabolism in these cancers was in 1985 [9]. This lag of interest, or maybe underachieving of PET, should be viewed in the context of other biological investigative areas where the latest cutting edge technology used tends to be only a few years old.
|
Goals for future applications
|
|---|
Having briefly outlined what has been learnt about PET with respect to applications and before discussing hurdles, it is worth itemizing the expected future uses of PET. In summary these are:
Clinical research to study the pathophysiology of focal disease
Proof of concept within drug development to study, in focal disease:
- The expression of therapeutic molecular targets
- The occupancy of therapeutic molecular targets
- Drug concentrations and time courses
- Mechanisms of therapeutic action
- Functional response to therapy
Clinical applications:
- Spin off from clinical research/drug development programs
- Evidence based procedures
- Tailored therapy
Convergence with the post-genomic era:
- Functional genomics
- Phamacogenomics
- Cancer genomics
|
Challenge 1: Improving sensitivity and specificity
|
|---|
It is clear that the future developments and applications of PET will depend on advances made within the research domain. In turn, scientific advances rely heavily on improving the sensitivity and specificity of the methodology used for observation. This follows since they are fundamental drivers of the research frontiers in that they determine what advanced questions can be answered.
In order to advance the methodological base of PET there are opportunities for improvements through:
- The discovery of tracers and ligands
- Increased effective sensitivity of PET scanners and by making better use of the signals recorded by those tomographs
- Making better use of the recorded kinetic data
The discovery of tracers and ligands
Despite the flexibility for natural radiolabelling afforded for example by using carbon-11, no more than a dozen or so PET probes are available that provide interpretable in-vivo signals. This is despite the fact that there are many more tracer molecules and ligands that work well for in-vitro assays. The range of probes for in-vivo molecular imaging is poor owing to low penetration of endothelial barriers, extensive peripheral metabolism or non-specific binding. Most of these restrictions are circumvented in in-vitro work, including the ability to separate off the non-specific binding component.
Hence in order to be able to realise the full potential of the major investments made in the equipment and facilities required for PET, it will be necessary to discover and develop a much wider range of probes for in-vivo molecular imaging. The starting material for this discovery exists in that there are of the order of 107 molecular compounds that are already resident within the pharmaceutical industry. The challenge is to begin to develop a further understanding of what are the basic molecular criteria that make for a low level of non-specific binding. One way of approaching this is to develop a parallel to that used in the discovery of drugs by setting up high throughput in-vitro screens. These need to effect appropriate assays of the characteristics sought for molecular imaging, over and above signal specificity. The generation of large volumes of assay data could be used, through a bioinformatics approach, to develop an understanding of the molecular configurations that determine the characteristics needed for a good in-vivo imaging probe [10].
Increasing the effective sensitivity of PET scanners
The high sensitivity afforded by coincidence counting needs to be fully developed if maximum spatial and kinetic resolution is to be realised along with the detection of lower levels of binding site densities. The introduction of 3-D data collection has brought with it significant sensitivity gains for brain imaging [11]. Also, the solid angle for detection has increased appreciably for body imaging [12]. However, these developments have come up against the limitations of the detector/electronic configurations to deal adequately with the high counting and random coincidence rates that are present when imaging the body. To improve this, faster detector material is needed along with the maximum of in-parallel electronics to remove "bottlenecks" in data handling. The introduction of the crystal scintillator lutetium oxyorthosilicate (LSO) [13], which has faster light characteristics than the conventional BGO, is going some way to reduce these sources of effective dead time. It also has greater light output, which helps to resolve smaller detector elements within crystal arrays. This has been exploited in the latest tomograph dedicated to brain imaging, which has a 2.5 mm spatial resolution [14]. Further use of this detector material is the introduction of a 50 cm axial length, LSO detector based tomograph that will provide high effective sensitivity for screening the body (personal communication, CTI Positron Systems). The question still remains as to what is the optimal detector configuration for imaging a predefined site in the body.
While there is no doubt that 3-D coincidence data recording in the brain has provided major statistical advantage compared with the earlier studies with septa, there are those who are concerned that the gain for body imaging in this mode is marginal. However, by being careful not to overload 3-D cameras with activity, there are gains to be had. The future use of LSO will provide us with extra confidence to undertake 3-D in the body since it is subject to less dead time and registration of random coincidences. It has been projected that the currently most sensitive bismuth germinate (BGO) based tomograph will improve its effective sensitivity by a factor of three at higher count rates if it is equipped with LSO detectors and the electronics are optimized to minimize dead losses due to pile up [12].
There still remains the need to correct for the recording of the higher levels of scattered coincidences experienced in the body than the brain. Here, great care is needed, since such corrections can result in noise amplification. The challenge is to effect scatter corrections that are not only accurate but impose the minimum of statistical uncertainty. This should be possible by capitalizing on the fact that the spatial distribution of the scattered component is low in spatial frequency, and hence can be averaged across the field of view to effect a low noise correction. In addition, improvements in detector energy resolution are projected which will reduce the registration of scattered coincidences.
Over the last 10 years, there has been an increasing interest in developing and using PET cameras for small animal studies [15]. The stimulation here is to effect non-destructive studies of the time course of tracer and ligand distributions in small animals, as well as regional tissue function. This is particularly important if the experimental animal model, either of normal or diseased tissue, is poorly reproducible with respect to the underlying function or how a tracer or ligand is distributed. Clarity on this point is needed since the biological sciences and the pharmaceutical industry use animal sacrifice procedures to study tissue biology, pharmacokinetics and pharmacodynamics. However, as more emphasis is being placed on functional challenge studies in animal models of disease, e.g. therapeutic response, in-vivo monitoring offers the opportunity to extend the sensitivity and scope of such experiments. The other use of small animal PET studies is to develop and validate the tracer protocols and analytical procedures destined for use in humans. These include, in particular, analysis of the kinetic data to derive functional parameters, validation of which is possible through independent ex-vivo assays of the function or binding site being imaged. It is here that small animal imaging could have an important role in forging scientific links between the pharmaceutical industry and academic/clinical PET groups. This follows since for the former it provides a means to gain insight into the strengths and weaknesses of in-vivo imaging within an experimental medium well known and understood by them. This should lead to industry contributing to the formulation of human PET protocols and data analysis. It in turn will not only help to realise PET's potential for assisting drug development but also stimulate the introduction of new probes for molecular imaging, the chemical entities of which owe their origins to industry.
The challenge to developing the technology for imaging small laboratory animals and in particular the mouse is that of achieving the necessary spatial resolution, which in the first instance is comparable with that realised in man. Taking 5 mm as an achievable PET resolution in humans with an average weight of 70 kg, the scaled up resolution to achieve anatomical parody for a 30 g mouse is clearly beyond the realms of physical possibility. This follows from the finite positron range and statistical information needed to reconstruct at such high resolution. Nevertheless, 1 mm and less resolution has been achieved and in order to make the most of such spatial definition the challenge is to maximize sensitivity without compromising this resolution. Unlike the human situation, a small object like a mouse can be more easily surrounded with detector with the aim of maximizing detection solid angle. The challenge here is the cost of such arrays of detectors plus the fact that resolution is needed within the depth of the detector. This is required in order to avoid parallax errors and hence deterioration of spatial resolution. While the review of Marsden [25] addresses the various configurations of detectors used in small animal PET cameras, it is worth highlighting the multilayer detector approach developed using the high density avalanche chamber (HIDAC) technology. This uses a relatively low cost detector, based on lead converter/multi-wire read out technology. As a result, it provides a spatial resolution of 1 mm and a depth of interaction resolution of 3 mm [16]. While providing the physical readout to ensure uniformity of spatial resolution across the field of view, the sensitivity of this device is around 1.5%. This is comparable with other state of the art small animal PET cameras based on LSO scintillating crystal [17]. Hence, in order to achieve such resolution in practice, especially when reconstructing short duration time frames to follow kinetic time series, the challenge is to realise significant increases in sensitivity.
Image reconstruction
Having discussed the challenges to optimize the physical performance of PET cameras, the next challenge is to optimize the tomographic reconstructions of the recorded data. Advances have been made based on iterative techniques that incorporate prior information on the physical resolution of the PET camera and its variation across the field of view. An example of this is the work of Reader et al who reconstructed data from the HIDAC camera [18].
Analysis of kinetic PET data
The central theme within molecular imaging is the study of regional tissue molecular pathways and interactions. By and large, unless one is using the equivalent of a chemical microsphere for that pathway or interaction, e.g. 18F-FDG, then the procedure requires the recording and analysis of kinetic data to derive the rate constants which define the molecular process. Even the establishment of an FDG-like substance requires, in the first instance, a kinetic approach to validate the simplified uptake measurement. Hence in order to analyse data, the challenge is to formulate a biological, kinetic model of the process being traced. This invariably requires subcellular, cellular and animal distribution studies of the tracer or ligand in order to build the model. The next step is usually that of describing these processes in a compartment model that is composed of the volumes and the rates of exchange between the respective compartments. The intention is to be able to fit kinetically recorded PET data to that model in order to derive quantitative values for the rate constants and the volumes of distribution. Having overcome this first modelling hurdle, the next challenge is that of ensuring the identifiability of the model's components within the data and to be able to realise this identification in a systematic and reproducible manner. This challenge has to be coupled with that of the finite statistics that are present in the data, which becomes especially challenging when analysing data at the voxel element level. The latter is important since the overall aim is, where possible, to derive parametric images of the functional entity being traced while still retaining the inherent spatial resolution of the recorded data set. Layered upon this task is that of implementing the voxel based kinetic analysis which, when involving fitting routines, is often a formidable and computer intensive process. For an in depth treatise on tracer kinetic modelling in PET the reader is referred to an up to date, in depth review by Carson [19].
Having discussed the challenges to defining and analysing the kinetic data, it remains to consider that in some cases the model, despite being valid for normal tissue, may not be applicable to a diseased tissue of interest. This introduces the challenge of defining the data in more general terms such as the tracer's mean transit time or volume of distribution between the plasma and the tissue. Such surrogates allow detailed compartmental modelling to be avoided and yet still provide quantifiable measures of the tracer's or ligand's fate within a tissue of interest. Another approach has been to derive the spectrum of kinetic components that are actually resident within the recorded PET data [20]. From this, volumes of distribution can be measured and surrogates derived such as the one hour uptake image.
Having itemized the analytical challenges, it is salient to note that there are relatively few groups internationally who are addressing these in a comprehensive way. Despite the fact that there are some hundreds of PET cameras in use, most of which are capable of recording high quality kinetic data, very few are doing so. This is due not only to the lack of specific tracers and ligands which require kinetic analysis in the clinical setting but also to the lack of robust and user friendly kinetic analytical techniques. As a result, there are many owners of high quality movie PET cameras who are only able to use them for still photography. One way that has been considered of meeting this challenge is to establish "photo shops" whereby kinetic data, recorded in well defined protocols, are transferred to a central facility. There the data could be processed to derive quantitative functional data, equivalent to the developing and printing of exposed photographic film.
|
Challenge 2: Logistics
|
|---|
Radioisotopes such as carbon-11 offer considerable scope for providing a wide range of naturally labelled probes for the in-vivo molecular imaging of biochemicals and pharmaceuticals. However, its short 20.1 min half-life represents a considerable challenge for application. While not wishing to downgrade the scientific achievements needed to devise rapid chemical reactions, many advances have already been made in this area to provide a wide armamentarium of rapid syntheses. It does seem now that the main challenge to widespread application is that of the logistics of having to handle large levels of isotope and undertake remote radiolabelling, which is more involved than encountered in the average nuclear medicine department using single photon emitters. This chemistry is further logistically burdened by the increasing need to carry out these syntheses under good manufacturing conditions and within a short space of time, implementing the necessary quality control for radiochemical and chemical purity. It never ceases to amaze one that the radiochemists are able to produce, within 50 min or so following the cyclotron's production of carbon-11, injectable amounts of labelled compound. However, the reality is that the ability to undertake this for a wide range of compounds and therefore realise the inherent full potential of PET is confined to relatively few research based centres. As a result, there are very few radiochemists trained in the field. The challenge therefore is how to train more chemists and also provide commercial modules that overcome the logistically difficult procedures outlined. This is happening slowly but with an emphasis on using fluorine-18 as a label because of its longer half-life and on the goals to overcome these logistics with distribution centres.
The other logistical challenge when using the shorter half-life radioisotopes is that of the patient imaging procedure itself. Not only does the patient need to be in place in the PET camera, often before the end of a radiochemical synthesis, but for many of the procedures involving kinetic analyses, blood samples, most of which need to be arterial, have to be withdrawn and analysed immediately. This challenge extends to having to measure the level of radiolabelled metabolites in the circulating plasma, which are present in low levels. One way to avoid the logistical complexities of blood withdrawal and analysis is to use internal reference tissues to provide the input functions necessary to quantify the molecular pathways or interactions of interest [21]. A very simplified means for deriving quantitative data is the use of the standard uptake unit (SUV), which is an expression of the tissue's tracer content as a ratio to the amount administered to the patient [22]. This has helped overcome the challenges of quantification to some degree. However, it does not provide for values of tracer flux through molecular pathways, volumes of distribution between plasma and tissue, and the binding characteristics of a molecular interaction in units of quantification understood by the biologists and pharmacologists. The challenge to derive such quantifiable values with the minimum of complexity needs to be borne in mind if PET is to realise its full role in translating between the laboratory and the clinic.
|
Challenge 3: Developments though a clinical science led programme
|
|---|
Having discussed some of the scientific and technical challenges to advancing the applications of PET, it will be clear that these encompass a wide range of disciplines [23]. It is also clear that to develop the field, either for translational studies or for clinical utility, such advances will initially occur primarily within the clinical research domain. This follows since as clinical science advances this, like any science, involves answering research questions which require higher specificity and sensitivity of the methodology used.
To effect a multidisciplinary approach, there needs to be a focus that relates to the contributions that each of the disciplines is capable of making. Such a focus is vital if the added value of the creativity often experienced at the interface of disciplines is to be realised. To address the challenges itemized will require long term, and in some cases, high risk programs of research and development. This is especially the case for discovering and developing new tracers and ligands for in-vivo molecular imaging. Because of the scale needed for such investment, careful consideration is required for the focus of activity. Based on past experience, it is concluded that a clinical science led program, in a major clinical specialty (e.g. oncology, neurology, psychiatry or cardiology) is necessary to provide the appropriate focus and long-term developmental theme needed to address the challenges facing the advancement of PET. This is needed in order to provide:
- The clinical research questions that are of generic importance to the clinical specialty, which could be answered by PET.
- The long term, enduring case and hence justification for financial support for the development of the advanced methodology needed to answer the big clinical research questions.
- Common consensus for research missions towards which the non-clinical expertise can focus within a team effort: pharmacology, physiology, radiochemistry, physics, mathematical modelling, computer science and technical support.
- The appropriate clinical/ethical investigative infrastructure to develop and validate, in patients, new methodologies for clinical research.
- Access to representative cohorts of patients for clinical research through referrals from clinical colleagues within the specialty.
- The clinical investigative environment needed by basic molecular and cellular biologists as well as the pharmaceutical industry to use in-vivo molecular imaging as a translational tool.
- The security to attract and nurture junior clinical researchers (MD or PhD) to commit themselves to an on the edge investigative area so that they do not lose contact with their mainstream clinical specialty.
- The link with national and international trials within the clinical specialty.
- The means to identify where clinical research procedures could translate into clinical management and care.
|
Challenge 4: Attracting and nurturing clinical research fellows
|
|---|
It is the clinical research fellows who are the drivers of the individual clinical research studies. In-vivo molecular imaging is complex, new and on the edge of clinical science. Hence we have to identify and attract some fairly self-generating clinical research fellows. This is especially the case since they are expected to make original contributions and champion the experiment from the beginning to the end and in doing so:- Refine experimental paradigms and protocols
- Administer cases to Ethics and ARSAC Committees
- Recruit, consent and prepare the patients for study
- Define and undertake quality control of the PET study
- Undertake data analysis
- Interpret, present and publish the results
- Provide on the ground links with the referring clinicians
Clearly the research fellows cannot undertake these roles in isolation and they have to receive clinical, scientific and technical support, but they do provide the common theme of a project, calling upon the respective resources of the centre. This role in a patient based research area is critical and much care is needed to nurture and mentor the clinical research fellows. It is ever important at present given the general concern of clinical science in general in the UK and incompatibilities of clinicians in training spending time in research [24]. Hence, mundane as it may appear research fellows do need:- Scientific supervision
- Attention to their clinical training needs
- Operational and methodological support
- Help, where appropriate, in launching their future clinical academic careers
The challenge of attracting and looking after clinical research fellows underlines one of the reasons for ensuring a clinical science led programme.
|
Challenge 5: The location and environment of a clinical research led PET based molecular imaging centre
|
|---|
Given that the point has been made for a clinical specialty focus for such a centre, it follows that it should be located adjacent to or incorporated into a hospital dedicated to that clinical area. Examples are Oncology, Psychiatry, Neurology or Cardiothoracic Centres. The case for this rests on the need for access, through site specialized clinicians, to representative cohorts of patients for research. Ideally such hospitals should also have a clinical academic culture to give added support to the investment made in molecular imaging. In addition, such specialized hospitals represent a pool of clinicians in training to be attracted as research fellows.
The environment of the centre itself needs special consideration as one is seeking to achieve interactive, multidisciplinary programs with innovation at the interfaces of the disciplines. Ideally, this is addressed through ensuring the appropriate amount of space to house together the clinical and non-clinical people to effect continuous "next bench" interaction and reinforcement of common missions. If one believes in the increasing importance of molecular imaging then room for expansion needs to be built into the design of a bespoke centre.
|
Challenge 6: Realising the funding
|
|---|
It is expensive to:- Establish and run a stand alone, all encompassing Molecular Imaging Centre based on PET.
- Undertake high risk, long term PET methodology development and its clinical research application.
Hence there is a formidable challenge to funding such an expensive research centre especially as the competition for public money is high. This is ever the case given that scientific research activity is growing at about twice the rate of that of the economy. Understandably, government-funding agencies perceive that there is more return on smaller research outlays than major multidiscipline centres of excellence with their high infrastructure costs. This is despite the fact, as has been explained, that there is little alternative to big science laboratory, multidisciplinary activity if cutting edge research in molecular imaging is to be undertaken. Without such centres, new methodology and their application will not be forthcoming for general use by smaller centres for research or clinical PET groups.
The solution to meeting the high costs for establishing a new centre of excellence in in-vivo molecular imaging can only be one of funding partnerships encompassing philanthropy, national charities, government and industry. However, to develop from a green field site requires additional investment to cover the fallow period before a centre can be self-funding through research grants etc.
It is encouraging that recently increasing investment is being made by industry in this area either within a business plan to develop new tracers and ligands as commercial products or by pharmaceutical companies to help drug development. Such investment is extremely welcome given the lack of probes for in-vivo molecular imaging. It now remains to be seen how industry will prioritise those application areas for which molecular probes could be of value. In addition, it remains to be seen how industry will develop partnerships to explore, refine and validate the use of new potential probes in an academic clinical research setting and/or a routine clinical environment.
|
References
|
|---|
- Jones T. The role of positron emission tomography within the spectrum of medical imaging. Eur J Nucl Med 1996;23:2207–11.[Medline]
- Paulesu E, Frith CD, Frackowiak RSJ. The neural correlates of the verbal component of working memory. Nature 1993;362:342–5.[Medline]
- Koepp MJ, Gunn RN, Lawrence AD, Cunningham VJ, Dagher A, Jones T, et al. Evidence for striatal dopamine release during a video game. Nature 1998;393:266–8.[Medline]
- Camsonne R, Crouzel C, Comar D, Maziere B, Prenant C, Sastre J, et al. Synthesis of N-(11C) methyl, N-(methyl-1-propyl), (chloro-2-phenyl)isoquinoline carboxamide-3 (PK11195): a new ligand for peripheral benzodiazepine receptors. J Labelled Comp Radiopharm 1984;21:985–91.
- Cagnin A, Brooks DJ, Kennedy AM, Gunn RN, Myers R, Turkheimer FE, et al. In-vivo measurement of activated microglia in dementia. Lancet 2001;358:461–7.[Medline]
- Rabiner EA, Gunn RN, Castro ME, Sargent PA, Cowen PJ, Koepp MJ, et al. Beta-blocker binding to human 5-HT1A receptors in-vivo and in-vivo: implications for antidepressant therapy. Neuropsychopharmacology 2000;23:285–93.[Medline]
- Brock CS, Matthews JC, Brown G, Osman S, Luthra SK, Brady F, et al. Temozolomide uptake in human astrocytomas demonstrated in vivo. Proc Am Soc Clin Oncol 1998;16:231a.
- Brock CS, Young H, O'Reilly SM, Matthews J, Osman S, Evans H, et al. Early evaluation of tumour metabolic response using [18F]fluorodeoxyglucose and positron emission tomography: a pilot study following the phase II chemotherapy schedule for temozolomide in recurrent high-grade gliomas. Br J Cancer 2000;82:608–15.[Medline]
- Nolop KB, Rhodes CG, Brudin LH, Beaney RP, Krausz T, Jones T, et al. Glucose utilization in vivo by human pulmonary neoplasms. Cancer 1987;60:2682–9.[Medline]
- Abrunhosa AJ, Brady F, Luthra S, De Lima JJ, Jones T. Preliminary studies of computer aided ligand design for PET. In: Gjedde A, Hansen SB, Knudsen GM, Paulson OB, editors. Physiological imaging of the brain with PET. Academic Press 2001;51–6.
- Wienhard K. Applications of 3-D PET. In: Bendriem B, Townsend DW, editors. The theory and practice of 3D PET. Dordrecht: Kluwer Academic, 1998;137–67.
- Spinks TJ, Jones T, Bloomfield PM, Bailey DL, Miller MP, Hogg D, et al. Physical characteristics of the ECAT EXACT 3D positron tomograph. Phys Med Biol 2000;45:2601–18.[Medline]
- Melcher CL, Schweitzer JS. A promising new scintillator: cerium-doped lutetium oxyorthosilicate. Nucl Instrum Methods Phys Res A 1991;314:212–4.
- Wienhard K, Schmand M, Casey ME, Baker K, Bao J, Eriksson L, et al. The ECAT HRRT: performance and first clinical application of the new high resolution research tomograph. IEEE Trans Nucl Sci 2002;45:104–10.
- Myers R. The biological application of small animal PET imaging. Nucl Med Biol 2001;28:585–93.[Medline]
- Jeavons AP, Chandler RA, Dettmar CAR. A 3D HIDAC-PET camera with sub-millimetre resolution for imaging small animals. IEEE Trans Nucl Sci 1999;46:468–73.
- Tai YC, Chatziioannou A, Siegel SJY, Newport D, Goble RN, et al. Performance evaluation of the microPET P4: a PET system dedicated to animal imaging. Phys Med Biol 2001;46:1845–62.[Medline]
- Reader AJ, Ally S, Bakatselos F, Manavaki R, Walledge RJ, Jeavons AP, et al. Regularized one-pass list-mode EM algorithm for high resolution 3D PET image reconstruction into large arrays. In: Proceedings of the IEEE Nuclear Science Symposium and Medical Imaging Conference; 2001; San Diego, CA. IEEE, 2001.
- Carson RE. Tracer kinetic modelling in positron emission tomography. In: Valk P, Bailey D, Townsend D, Maisey M. Positron emission tomography. London: Springer 2002.
- Cunningham VJ, Jones T. Spectral-analysis of dynamic PET studies. J Cereb Blood Flow Metab 1993;13:15–23.[Medline]
- Lammertsma AA, Hume SP. Simplified reference tissue model for PET receptor studies. Neuroimage 1996;4(3 Pt 1):153–8.[Medline]
- Woodward HQ, Gigler RE, Freed B, Russ G. Expression of tissue isotope distribution. J Nucl Med 1975;16:958–9.
- Jones T. The imaging science of positron emission tomography. Eur J Nucl Med 1996;23:807–13.[Medline]
- Savill J. The tenure track clinician scientist: a new career pathway to promote recruitment into clinical academic medicine. London: Academy of Medical Science.
- Marsden PK, Strul D, Keevil SF, Williams SCR, Cash D. Simultaneous PET and NMR. Br J Radiol 2002;75:S53–S59.[Abstract/Free Full Text]
Top
Introduction
Molecular imaging with PET