Targeted, radiolabelled near-infrared quantum dots for high sensitivity and resolution, dual modality imaging of human tumours in mice

The aim of the proposal is to use small fluorescent probes (nanoprobes), to improve the quality of imaging of cancer models that are used to test new drugs before they are moved forward into the clinic. Current models used for testing anti-cancer drugs are often not very representative of real patient tumours and so may not correctly identify drugs likely to work well in a patient. We have established a number of advanced models of cancer which are more useful for this purpose which use close-to-patient cells, incorporate additional supporting cells and are grown at site of origin of the tumour, rather than below the skin of the experimental animal, which provides them with a more relevant environment. However, imaging the growth of cancer cells in these models, and monitoring the way that the different cells within the tumours behave and interact with each other, which are important in determining drug response, is difficult with current methods. Standard cell-lines can be readily labelled genetically so that they are bioluminescent or fluorescent (produce light or fluoresce), but this is difficult to achieve with the close-to-patient cells, and the process of labelling them in this way can also result in alterations that mean that they no longer behave in the same way as patient tumours. In addition, use of luminescence involves additional procedures for the animals and the fluorescent probes currently used are easily blocked by the animal tissues. The probes we intend to develop will overcome these problems.
They will fluoresce at a higher wavelength than standard fluorescent probes which will allow the signal to be detected at deeper sites within the body and improve the sensitivity and quality of the images we can obtain. They will be designed so that they are readily taken up by specific populations of cells within the tumour which will allow us to follow the growth and drug response of e.g. cancer cells or supporting cells within the tumours. Lastly, they will have a radiolabel which will allow them to be used for PET as well as fluorescence imaging, which will provide a further improvement in the resolution of the images that can be obtained.
Thus, the successful development of these new nanoprobes will have significant 3Rs impact, by providing more refined pre-clinical cancer models, and reducing numbers of animals used in such models. Application of these new methods of imaging will allow more detailed information to be gained from each animal used, allow repeat imaging of the same animal over time, and allow the use of advanced models which provide more relevant information about how useful a drug is likely to be in the clinic. Such models are widely used by industry for drug development, as well as by academics for understanding the science underlying tumour development. Thus, the knowledge we will gain will have significant scientific impact; it will also have economic impact because it will streamline the process by which drugs are screened by industry, and reduce the price of anti-cancer drugs. This in turn will be good for patients by providing cheaper, more effective drugs for treatment of cancer. The probes are likely to also have further applications in non-cancer settings including in other diseases and in the development of methods for engineering normal tissues.

This proposal will use near infrared quantum dots (NIR-QDs), to improve sensitivity and resolution of imaging in advanced pre-clinical cancer models. Successful development of NIR-QDs will have 3Rs impact, enabling use of more refined, patient-relevant cancer models and reducing numbers of animals used. Academia and Pharma are focussed on developing new cancer drugs resulting in the use of >430,000 mice/year. We estimate that since more advanced patient-relevant cancer models will predict drug failure earlier in preclinical testing, numbers of animals used for continued evaluation of such failing drugs would be reduced by half (~170,000 animals/year).
Bioluminescent/fluorescent imaging despite advantages has drawbacks, especially for use in advanced cancer models. Bioluminescent imaging requires administration of substrate (an additional in vivo procedure); organic fluorophores photo-bleach and have limited sensitivity due to light absorption by tissues. For application in advanced models using close-to-patient tissues, standard labelling methods cause loss of cellular heterogeneity resulting in models unrepresentative of the original tumour.
The NIR-QDs we will develop do not photo-bleach, will be water-soluble and optically active in the NIR wavelength, providing improved tissue penetration and sensitivity, and will incorporate targeting ligands for uptake by specific populations of cells within complex xenografts, and a positron-emitting radionuclide for translation into the PET/CT imaging platform.
The knowledge gained will have scientific as well as 3Rs impact, leading to development of probes that can be more broadly applied to imaging complex tissues in diseases other than cancer, and in a non-disease context e.g. bioengineered tissues. The probes also have potential for economic impact, through improving the drug discovery cascade by refinement of patient-relevant pre-clinical cancer models that are better able to predict clinical outcomes.

Currently, ~80% of new cancer drugs fail in clinical trial, even though they passed preclinical (animal) testing. Thus, most animals (80% of 430,000/year, or 344,000 animals/year based on 2011 figures) used to test human cancer drugs do not contribute to the development of new therapeutics directly, because they did not inform on how the drugs would behave in humans. More refined preclinical models could reduce the number of failures by half (~170,000 animals/year). One of the major reasons for clinical failure is that drugs are tested preclinically either at subcutaneous sites which do not provide a relevant tumour microenvironment, or, even if tested orthotopically cancer cell lines are used. There are a limited number of such cell lines - e.g. most prostate studies use PC3 or LNCaP cells, and the models based on them do not reflect human cancers, as they have specific characteristics accumulated over years in the laboratory that set them apart from human cancers, and they are grown in a mouse microenvironment (mouse endothelial and mesenchymal cells). The reason these lines are used is that they can be monitored in the animal because they have been genetically labelled (e.g. with luciferase, or fluorescent proteins). To build a more accurate pre-clinical model of cancer, that reduces the number of failed animal testing protocols, we need to be able to image human tumours, not just human cancer cell lines, in orthotopic and metastatic cancer models in mice.
The successful development of NIR imaging with new nanoprobes will have significant 3Rs impact, largely by providing more refined pre-clinical cancer models, and reducing numbers of animals used in such models by enabling repetitive molecularly targeted in vivo imaging of patient-derived xenografts by standard low cost imaging modalities that are widely used worldwide. Success in this work will lead to development of probes which, as well as being applied in pre-clinical models of cancer can be more broadly applied to imaging complex tissues in disease settings other than cancer, and in a non-disease context e.g. for improving the resolution and imaging of populations of cells within complex bioengineered tissues. The knowledge generated is likely to be used by academic, industry and medical researchers: potential users will thus be approached following IP protection, if practical applications appear feasible.
A major area of interest in cancer biology currently is the role of the tumour microenvironment (tumour location and factors released by other cell-types, stromal cells, within a tumour) in determining the growth and progression of a tumour, as well as its response to therapy. Indeed, drugs targeting the stromal elements of tumours are being developed on the basis of this knowledge. However, the tumour microenvironment cannot easily be monitored with current technology due to limitations of sensitivity and resolution. The probes that we will develop during this proposal will address these issues, enable targeting of the probes to individual populations of cells within the tumour so that they can be monitored individually, and assess the resolution provided by such probes or by probes further modified to enable multi-modal imaging. The applicants are firmly committed to knowledge transfer without compromising industrial competitiveness. Although this work is developmental in nature, we have already received interest from a company developing hardware that will be required to further enhance the impact of our research outcomes. We will contact further potential downstream funders/partners as the work progresses when sufficient proof-of-concept data has been obtained.