Multi-user, multi-centre MRI to reduce and refine the use of mice in cancer and trauma research
Lead Research Organisation:
Queen Mary University of London
Department Name: Barts Cancer Institute
Abstract
Non-invasive imaging techniques allow researchers to reduce the numbers of animals used in scientific research through following disease in the same animal over time and obtaining more information from each animal. This information can be anatomical (e.g. tumour size and location, brain damage) or functional (e.g. measuring what the cells within the tissue or organ are doing, such as whether or not they are growing). Our studies often require the use of genetically modified (transgenic) models where mice naturally develop cancers in a fashion similar to that of humans. The main challenge in such studies, which is shared in human disease, is to measure, accurately, the development of the cancer and the response of the tumour to new treatments and compare it with the current standard.
In one of our genetically engineered mouse models of pancreatic cancer, the animals develop tumours spontaneously after birth over a time period of 80-150 days. The rate at which these tumours develop is variable and because the tumours normally develop deep within the tissues of the body they are not externally visible and are only apparent when they become palpable or the animal becomes ill. If we could detect smaller tumours, every animal could be used at an earlier stage of tumour development, which would reduce suffering and waste of animals. With the use of Magnetic Resonance Imaging (MRI) we will be able to detect tumours earlier and size match with other animals in order to have a relevant comparative groups. These tumours are detectable using imaging protocols that show excellent contrast between tissues according to their fat and water content, ideal for looking at the soft tissues in the abdomen. This will enable us to determine treatment schedules and therapeutic responses - similar to the way patients are treated.
In addition, using a custom-made mouse holder, the animal can be transferred under anaesthetic from the MRI instrument to our other imaging cameras. In this way MRI can be combined with radiotracer imaging (positron emission tomography (PET) and single photon emission tomography (SPECT)). The high resolution anatomical MRI scans will allow us to identify whether the radioactive signal originates in tumour tissue in or surrounding tissues such as intestine and kidney. We will then be able to quantify the tumour radiotracer accumulation. By using radioactive probes that measure biological functions such as proliferation, we will be able to measure tumour function in response to different therapeutics at different time-points. Our current methods of anatomical imaging don't allow us to distinguish between tissues such as intestine and tumour but MRI will enable us to do this work.
MRI will also be invaluable in our research in mouse models of neuro-trauma and multiple organ failure following trauma. The Neuro-Trauma group has a particular interest in modulating acute neuroinflammation, minimising long term tissue damage and testing new therapeutic approaches. Having access to information gained from MRI (e.g. brain swelling, haemorrhage, blood brain permeability, white and grey matter damage), combined with neuro-behavioural testing would help us to reduce the numbers of animals studied and be more accurate with our endpoints. This would ensure that the experiments would end as soon as possible. MRI is ideal for looking at anatomical structural changes and inflammatory lesions, such as oedema, in organ dysfunction models and again would reduce the numbers of animals being entered into these studies.
We will also modify our current method of bioluminescence imaging (BLI) using MRI. BLI is a 2D technique and gives no information about tumour depth within the animal. Combining these images with 3D-MRI will allow us to correct for tumour depth and quantify the BLI signal more accurately. This will enable us to match animals with similar tumour burdens, reduce biological variation and decrease group numbers.
In one of our genetically engineered mouse models of pancreatic cancer, the animals develop tumours spontaneously after birth over a time period of 80-150 days. The rate at which these tumours develop is variable and because the tumours normally develop deep within the tissues of the body they are not externally visible and are only apparent when they become palpable or the animal becomes ill. If we could detect smaller tumours, every animal could be used at an earlier stage of tumour development, which would reduce suffering and waste of animals. With the use of Magnetic Resonance Imaging (MRI) we will be able to detect tumours earlier and size match with other animals in order to have a relevant comparative groups. These tumours are detectable using imaging protocols that show excellent contrast between tissues according to their fat and water content, ideal for looking at the soft tissues in the abdomen. This will enable us to determine treatment schedules and therapeutic responses - similar to the way patients are treated.
In addition, using a custom-made mouse holder, the animal can be transferred under anaesthetic from the MRI instrument to our other imaging cameras. In this way MRI can be combined with radiotracer imaging (positron emission tomography (PET) and single photon emission tomography (SPECT)). The high resolution anatomical MRI scans will allow us to identify whether the radioactive signal originates in tumour tissue in or surrounding tissues such as intestine and kidney. We will then be able to quantify the tumour radiotracer accumulation. By using radioactive probes that measure biological functions such as proliferation, we will be able to measure tumour function in response to different therapeutics at different time-points. Our current methods of anatomical imaging don't allow us to distinguish between tissues such as intestine and tumour but MRI will enable us to do this work.
MRI will also be invaluable in our research in mouse models of neuro-trauma and multiple organ failure following trauma. The Neuro-Trauma group has a particular interest in modulating acute neuroinflammation, minimising long term tissue damage and testing new therapeutic approaches. Having access to information gained from MRI (e.g. brain swelling, haemorrhage, blood brain permeability, white and grey matter damage), combined with neuro-behavioural testing would help us to reduce the numbers of animals studied and be more accurate with our endpoints. This would ensure that the experiments would end as soon as possible. MRI is ideal for looking at anatomical structural changes and inflammatory lesions, such as oedema, in organ dysfunction models and again would reduce the numbers of animals being entered into these studies.
We will also modify our current method of bioluminescence imaging (BLI) using MRI. BLI is a 2D technique and gives no information about tumour depth within the animal. Combining these images with 3D-MRI will allow us to correct for tumour depth and quantify the BLI signal more accurately. This will enable us to match animals with similar tumour burdens, reduce biological variation and decrease group numbers.
Technical Summary
We aim to use a low field (1T) small animal MRI instrument to improve our current studies in spontaneous orthotopic mouse models of cancer as well as in models of neuro-trauma and trauma. In cancer we have multiple programmes, many in collaboration with the pharmaceutical industry, exploring new therapies for a treating cancer, most notably pancreas and breast. In fact, in addition to longstanding collaborations with Astra Zeneca/Medimmune, GSK and smaller biotech companies, AZ-Medimmune have identified BCI as the site where they will direct their pre-clinical testing. All of these studies ultimately require the use of genetically modified (transgenic) models where mice naturally develop cancers deep within the body of the animal, reflecting the process in humans. However tumour burden is very difficult to assess if the tumours are not palpable. The soft tissue contrast afforded by MRI will allow us to size match groups and carry out longitudinal imaging to assess response to treatment. By accurate staging of deep tissue disease with MRI, reduced numbers of animals would be required for studies and they could be entered into studies more accurately; we suggest up to 50% reduction in animal usage. This will allow us to use fewer animals and end studies sooner, thus reducing suffering. Co-registration of MRI with radiotracers using our PET and SPECT machines would improve the value of these studies. Our mouse models of neuro-trauma and trauma would greatly benefit from MRI monitoring of brain and spinal cord injury as well as organ dysfunction after trauma in longitudinal studies.
Our collaboration with Bruker/Aspect and InviCRO will allow us to set up a 3D-BLI method which uses a reconstruction algorithm to greatly improve quantitative data arising from 2D BLI as well as provide co-registered anatomical 3-D visualisation of tumour burden. This would allow tumour size-matching of mice, improving outcomes and reducing group size.
Our collaboration with Bruker/Aspect and InviCRO will allow us to set up a 3D-BLI method which uses a reconstruction algorithm to greatly improve quantitative data arising from 2D BLI as well as provide co-registered anatomical 3-D visualisation of tumour burden. This would allow tumour size-matching of mice, improving outcomes and reducing group size.
Planned Impact
The 3R's benefit of obtaining small animal MRI imaging at QMUL will be in reduction in numbers of animals used and refinements to reduce suffering.
Cancer: We currently use a minimum of 1000 transgenic animals annually that are entered into therapy studies or studies of metastatic disease and these numbers are increasing. This is because we are increasingly moving from using injectable models of cancer to these spontaneous orthotopic models that better mimic the human disease. However, their use is more challenging than injectable models due to the large timeframe over which tumours develop and the difficulty of monitoring them. However, we can reduce the numbers of these animals in two ways:
1) By using MRI to do longitudinal studies in order to assess the effect of treatments/interventions. This allows us to follow each animal over the course of the experiment
2) Entry into therapy studies based on MR imaging: sizing tumours and entering each animal into the study at an individual optimum time based on tumour burden (instead of after a pre-determined historically-based period of time, or after tumours become palpable).
Both approaches will translate into a reduction in animal numbers (fewer groups, less waste, less biological variability and therefore smaller groups) and we are hoping to cut the numbers of animals used by at least 50 %. If we could do this across all our studies, we could cut the current annual usage of transgenic animals to 500. This will also apply to our work with the pharmaceutical industry in these mouse models, so our future use of animals will be minimised.
MR imaging of these orthotopic mouse models of cancer will also have the 3R's benefit of refinement in that the animals will be used earlier, with lower tumour burden, the studies will end sooner at a less severe stage of disease, thus reducing suffering.
Neuro-trauma and trauma: in these models the annual usage is approximately 300 animals and the main 3R's impact would be reduction in animal numbers through longitudinal non-invasive imaging studies, hopefully by ~ 50%. Since these models are moderately severe to severe, this reduction in animal use is important. Trauma models are intrinsically invasive so it is essential to promote minimally invasive ways to monitor the evolution of injury. Improved readouts using MRI will lead to more accurate data and allow us to finish studies earlier, a refinement which reduces the suffering of every animal in the study.
Our collaborative project to develop 3D-BLI/MRI could have significant impact on the use of injectable cancer models (most commonly IP models of ovarian cancer). Animals are currently monitored for tumour growth using 2D BLI which is very inaccurate. By improving measurement of tumour growth, the animals could be entered into studies in a much more efficient way, again with lower tumour burden. This will lead to lower variability and thus a reduction in group size, with experiments being terminated earlier, reducing suffering.
The impetus for this NC3R's application has come from QM research groups within cancer and trauma research who have clearly identified a need for MRI to improve the way our research is done, to reduce the number of animals used (for economic, scientific and 3R's motivations) and to reduce suffering in challenging models of injury. However, there are many groups within our Institutes who could benefit from using non-invasive MR imaging and we will promote its use through demonstration of improved outcomes. We will publicise the benefits of the technique and record the 3R's impacts.
Within the wider research community, both in industry and academia we will show what can be achieved using MRI through high impact publications and presentations. Our industrial partners are extremely open to using non-invasive imaging techniques in therapeutic studies, and MRI will enable us to make a 3R's impact in their drug development programmes.
Cancer: We currently use a minimum of 1000 transgenic animals annually that are entered into therapy studies or studies of metastatic disease and these numbers are increasing. This is because we are increasingly moving from using injectable models of cancer to these spontaneous orthotopic models that better mimic the human disease. However, their use is more challenging than injectable models due to the large timeframe over which tumours develop and the difficulty of monitoring them. However, we can reduce the numbers of these animals in two ways:
1) By using MRI to do longitudinal studies in order to assess the effect of treatments/interventions. This allows us to follow each animal over the course of the experiment
2) Entry into therapy studies based on MR imaging: sizing tumours and entering each animal into the study at an individual optimum time based on tumour burden (instead of after a pre-determined historically-based period of time, or after tumours become palpable).
Both approaches will translate into a reduction in animal numbers (fewer groups, less waste, less biological variability and therefore smaller groups) and we are hoping to cut the numbers of animals used by at least 50 %. If we could do this across all our studies, we could cut the current annual usage of transgenic animals to 500. This will also apply to our work with the pharmaceutical industry in these mouse models, so our future use of animals will be minimised.
MR imaging of these orthotopic mouse models of cancer will also have the 3R's benefit of refinement in that the animals will be used earlier, with lower tumour burden, the studies will end sooner at a less severe stage of disease, thus reducing suffering.
Neuro-trauma and trauma: in these models the annual usage is approximately 300 animals and the main 3R's impact would be reduction in animal numbers through longitudinal non-invasive imaging studies, hopefully by ~ 50%. Since these models are moderately severe to severe, this reduction in animal use is important. Trauma models are intrinsically invasive so it is essential to promote minimally invasive ways to monitor the evolution of injury. Improved readouts using MRI will lead to more accurate data and allow us to finish studies earlier, a refinement which reduces the suffering of every animal in the study.
Our collaborative project to develop 3D-BLI/MRI could have significant impact on the use of injectable cancer models (most commonly IP models of ovarian cancer). Animals are currently monitored for tumour growth using 2D BLI which is very inaccurate. By improving measurement of tumour growth, the animals could be entered into studies in a much more efficient way, again with lower tumour burden. This will lead to lower variability and thus a reduction in group size, with experiments being terminated earlier, reducing suffering.
The impetus for this NC3R's application has come from QM research groups within cancer and trauma research who have clearly identified a need for MRI to improve the way our research is done, to reduce the number of animals used (for economic, scientific and 3R's motivations) and to reduce suffering in challenging models of injury. However, there are many groups within our Institutes who could benefit from using non-invasive MR imaging and we will promote its use through demonstration of improved outcomes. We will publicise the benefits of the technique and record the 3R's impacts.
Within the wider research community, both in industry and academia we will show what can be achieved using MRI through high impact publications and presentations. Our industrial partners are extremely open to using non-invasive imaging techniques in therapeutic studies, and MRI will enable us to make a 3R's impact in their drug development programmes.
