Non-invasive imaging to reduce and refine the use of animals and monitor their welfare during the course of experimentations in oncology

Despite the efforts of the international community, including ours, faithfully and reliably mimicking human cancers in experimental animal models remains a fundamental requirement to improve our understanding of the disease and develop / assess more effective therapeutic strategies. It is generally admitted that orthotopic models, when cancer cells are located in the organ they originate from, are more informative and reliable than ectopic models (when the cancer cells are forced to grow in place they would normally not use, like under the skin). However, it is very challenging to analyse the tumour progression in orthotopic models since they cannot be seen with naked eyes. As a consequence, a batch of animals needs to be sacrificed at each time point for analysing the organs after dissection (bone marrow and lungs in the case of this project). In vivo imaging technologies offer exciting possibilities to conduct noninvasive and longitudinal studies over time to understand the dynamics of biological processes in live animals over long period of time. The basic idea is to use imaging approaches to quantify the amount of tumours and their distribution within the body of the animal over time, in a way that mimics diagnosis for human patients. In that case there is no need to sacrifice the subject to obtain these measurements and it is then possible to assess the evolution of cancer cells over time. Therefore in vivo imaging allows to reduce the number of animals needed to answer a biological question and / or the efficacy of a new treatment. Furthermore, different imaging techniques can be combined on the same animal in order to answer different questions at the same time instead of doing different experiments.

Xenotransplantation of human acute myeloid leukaemia (AML) initiating cells (AML-ICs) from samples coming directly from patients is a very powerful model to understand the development of human leukaemia and assess response to treatments. However to collect results over time, it is necessary either to sacrifice some animals, or to perform bone marrow biopsies, a well tolerated but sill invasive procedure. We can use bioluminescence (BLI) to track genetically engineered cell lines in the bone marrow, but this cannot be used for patients' samples. To our knowledge, no noninvasive approach has been reported so far to monitor primary human AML over time. We propose here to tests different strategies to achieve this goal, which would allow to decrease the number of animals needed to complete the studies in that field. Furthermore, being able to objectively quantify the "tumour load" would allow one to assess at the same time the health status of the animal, similarly to what is done for patients. Consequently, researchers would have the capability to refine their experimental conditions and stop the experiment as soon as they have obtained the answers they need.

We also use genetically engineered mouse models where a specific gene has been mutated and will induce the spontaneous apparition of a lung cancer, closely mimicking a specific type of human cancer. We can use x-ray computed tomography (CT) in mice to detect lung tumours, like it is done in humans, but we are trying to develop more sensitive techniques and also to collect more information that just the size of the nodules.

Therefore we need to test different types of probes (fluorescently labelled or coupled with radioactive tracers), which could be used to detect by modern in vivo imaging technologies (Near-infrared fluorescence imaging and Cerenkov luminescence imaging) the presence and the progression of the leukaemia / lung tumours. We also want to measure physiologic parameters (heart rate, breathing rate, oxygenation) to assess objectively the welfare of the animals during the course of the study.

Altogether, these different information would allow us to reduce the need for animals and refine the way they are used during the experiments

We have identified 7 probes for NEAR-INFRARED IMAGING (NIRI) (we are producing 1 of them)
- 2-Deoxy-Glucose (2 sources)
- MMPs sensitive probe
- Cathepsins sensitive probe
- Integrin alphov beta3 probe
- Blood pool agent
- Antibody directed against hCD33 (marker of AML)

Very recent publications have reported that probes labelled with the radionuclide [18F], routinely exploited for positron emission tomography (PET) imaging, can be detected thanks to a technique called CERENKOV LUMINESCENCE IMAGING (CLI). We will evaluate the following PET probes:
- 2-deoxy-2-[18F]fluoro-D-glucose
- [18F]-3-fluorodeoxythymidine
- [18] Fluorine

We will be using a MouseOx system to measure various PHYSIOLOGIC PARAMETERS like peripheral blood oxygen saturation, heart frequency or breathing frequency with a noninvasive sensor clip.

In the earliest stage of the project, we will use 1 or 2 luciferase-expressing leukaemic cell lines so that BLI can be used as a positive control and healthy mice will be used as negative controls. For each probe, 2-3 leukaemia bearing mice and 2 negative controls will receive the probe of interest and be imaged at different time points after administration of the probe. BLI will be co acquired with the NIRI / CLI signal. The operation will be repeated weekly until the apparition of clinical signs.

The general status of the animals will be checked weekly using the MouseOx system. This information, together with BLI, NIRI and CLI quantification, will reinforce our daily observation for clinical signs.

To be selected, a probe will have to show in all leukaemia or tumour bearing mice at least a two-fold increase of the signal as compared to the negative controls. Successful candidates will then be tested on primary human AML and sporadic mouse lung tumours. Eventually we will optimise the procedures in order to allow simultaneous analysis of the successful probes on the same animal to maximise the reduction of animal use.