Reduction & refinement in animal models of neuropathic pain: using systematic review & meta-analysis
Lead Research Organisation:
University of Edinburgh
Department Name: Centre for Clinical Brain Sciences
Abstract
The management of chronic pain is a major clinical problem that can severely affect the quality of life of sufferers. Neuropathic pain (NP), a common type of chronic pain, occurs after damage to the sensory nervous system and has a wide range of causes including diabetes, trauma, surgery, cancer and autoimmune disease. Current treatments often do not provide adequate pain relief for patients, or have intolerable side effects. Animals can be used to improve our understanding of a disease and to test the effectiveness of novel treatments. There are many different ways of modelling NP and these animals are subjected to many different tests, some of which cause more pain or distress than others. The main aim of this study is to refine the tests used in order to reduce the suffering of animals and to reduce the number of animals used.
Systematically identifying all available studies will provide an overview of the range of methods used in the modelling of NP. Aggregating these data together will give us direct evidence of the utility of high severity tests, whether multiple tests are necessary, whether lengthy experiments are of added value or whether those of shorter duration areas useful and will guide us on the number of animals required for future studies. This project will, for the first time, provide evidence from completed experiments in NP to reduce and refine the use of animals in this field of research.
Previously, there have been significant problems in translating findings from animal studies to humans in a clinical setting. We have shown that when you systematically scrutinise studies there are often flaws in their design, conduct and reporting. These shortcomings in the quality of experiments tend to result in studies exaggerating the magnitude of their findings. Using the dataset we will assess the quality of studies and the impact of this quality on their findings. In the past similar shortcomings in human clinical trials led to substantial reforms in their conduct and regulation. Unfortunately, the same standards for the conduct of laboratory animal studies do not exist; however, the systematic scrutiny of animal studies modelling neurological diseases has initiated the beginnings of a similar process. The impetus behind this is in response to the empirical evidence describing the impact of reduced quality and insufficient reporting in animal models of stroke. Using similar techniques and curating empirical evidence has the potential to reform and improve the utility of animal models of NP.
Within the scientific community there is a tendency towards publishing positive results. Studies reporting no or negative effects are often not published. This bias results in summaries of a research field, such as the proposed, overstating true effects. We plan to quantify the presence and magnitude of publication bias in the NP literature. If laboratory studies are to be used to improve human health their findings need to be both reliable and available irrespective of the results.
In summary, this study will provide an overview of all animal studies of NP. These data will allow us to assess the validity and utility of high severity tests in comparison to those inflicting less pain and distress to animals. In addition we will be able to quantify the value of varying durations of experiments and potentially improve the welfare of animals used in this field of research. Describing the quality and range of studies will allow to provide evidence based guidelines to improve the design, conduct and reporting of NP animal studies. Quantifying the magnitude of publication bias will highlight the severity of the problem and hopefully drive reform to improve the publicising of experimental data. We also plan to make the database publicly available for colleagues in the field; this will reduce unnecessary replication of experiments.
Systematically identifying all available studies will provide an overview of the range of methods used in the modelling of NP. Aggregating these data together will give us direct evidence of the utility of high severity tests, whether multiple tests are necessary, whether lengthy experiments are of added value or whether those of shorter duration areas useful and will guide us on the number of animals required for future studies. This project will, for the first time, provide evidence from completed experiments in NP to reduce and refine the use of animals in this field of research.
Previously, there have been significant problems in translating findings from animal studies to humans in a clinical setting. We have shown that when you systematically scrutinise studies there are often flaws in their design, conduct and reporting. These shortcomings in the quality of experiments tend to result in studies exaggerating the magnitude of their findings. Using the dataset we will assess the quality of studies and the impact of this quality on their findings. In the past similar shortcomings in human clinical trials led to substantial reforms in their conduct and regulation. Unfortunately, the same standards for the conduct of laboratory animal studies do not exist; however, the systematic scrutiny of animal studies modelling neurological diseases has initiated the beginnings of a similar process. The impetus behind this is in response to the empirical evidence describing the impact of reduced quality and insufficient reporting in animal models of stroke. Using similar techniques and curating empirical evidence has the potential to reform and improve the utility of animal models of NP.
Within the scientific community there is a tendency towards publishing positive results. Studies reporting no or negative effects are often not published. This bias results in summaries of a research field, such as the proposed, overstating true effects. We plan to quantify the presence and magnitude of publication bias in the NP literature. If laboratory studies are to be used to improve human health their findings need to be both reliable and available irrespective of the results.
In summary, this study will provide an overview of all animal studies of NP. These data will allow us to assess the validity and utility of high severity tests in comparison to those inflicting less pain and distress to animals. In addition we will be able to quantify the value of varying durations of experiments and potentially improve the welfare of animals used in this field of research. Describing the quality and range of studies will allow to provide evidence based guidelines to improve the design, conduct and reporting of NP animal studies. Quantifying the magnitude of publication bias will highlight the severity of the problem and hopefully drive reform to improve the publicising of experimental data. We also plan to make the database publicly available for colleagues in the field; this will reduce unnecessary replication of experiments.
Technical Summary
Neuropathic pain (NP) is a major clinical problem and current therapies do not provide adequate analgesia for many patients. Subjecting animals to lesions modelling NP, or to tests designed to assess responses to painful stimuli, is inevitably associated with distress and discomfort.
We plan to systematically identify all animal experiments modelling NP. Using meta-regression we will provide evidence as to whether less noxious tests are as predictive as more severe alternatives to refine their use to minimise pain and suffering. We will also investigate the effect of experimental duration and post-operative analgesia regimens to reduce the period of suffering of the animals. We will provide precise estimates of the observed variance of different tests. This will identify which tests require fewer animals and allow robust sample size calculations to be performed and reported to reduce the number of animals sacrificed in experiments which are too small for the effect sought or those which use more animals than required.
Limitations in the validity of experiments reduce their reliability and may compromise their utility. Preclinical studies rarely report measures to reduce potential sources of bias. We will use meta-analysis to quantify the impact of quality on estimates of treatment effects in the NP literature.
We will quantify the impact of publication bias in the NP literature; this will highlight the importance of the issue and encourage the dissemination of all data derived from the use of animals. We will also make the database publicly available; this will enable colleagues to identify whether experiments have already been conducted and reduce the unnecessary replication of experiments.
This project will generate robust empirical data without the use of animals. It will contribute towards a reduction in the number of animals used and the refinement of experimental design, tests used and has the potential to reform and improve animal welfare in this field.
We plan to systematically identify all animal experiments modelling NP. Using meta-regression we will provide evidence as to whether less noxious tests are as predictive as more severe alternatives to refine their use to minimise pain and suffering. We will also investigate the effect of experimental duration and post-operative analgesia regimens to reduce the period of suffering of the animals. We will provide precise estimates of the observed variance of different tests. This will identify which tests require fewer animals and allow robust sample size calculations to be performed and reported to reduce the number of animals sacrificed in experiments which are too small for the effect sought or those which use more animals than required.
Limitations in the validity of experiments reduce their reliability and may compromise their utility. Preclinical studies rarely report measures to reduce potential sources of bias. We will use meta-analysis to quantify the impact of quality on estimates of treatment effects in the NP literature.
We will quantify the impact of publication bias in the NP literature; this will highlight the importance of the issue and encourage the dissemination of all data derived from the use of animals. We will also make the database publicly available; this will enable colleagues to identify whether experiments have already been conducted and reduce the unnecessary replication of experiments.
This project will generate robust empirical data without the use of animals. It will contribute towards a reduction in the number of animals used and the refinement of experimental design, tests used and has the potential to reform and improve animal welfare in this field.
Planned Impact
The main objective of this study is to refine the testing of animal models of neuropathic pain (NP). Animal models of NP are used to investigate the pathophysiology underlying NP and to test novel therapies; there are diverse models of NP, and the outcome measures used to assess hypersensitivity and efficacy of therapies vary significantly. Subjecting animals to lesions modelling chronic pain, or to tests designed to assess the ability of drugs to modulate the response to painful stimuli, is inevitably associated with some distress and discomfort to the animals. The extent of this distress and discomfort varies with different experimental designs, and it is not clear that experiments of higher severity provide data which are substantially more useful than experiments using different approaches of lower severity. This study will establish a large data set from previous studies using animal models of NP and use systematic review and meta-analysis of to identify whether a test with a high burden of pain or distress might be replaced with one of lower impact. This will identify whether multiple tests are necessary and if less noxious tests are as predictive as more severe alternatives. Additionally, we aim to refine the duration of experiments using animal models of NP. Using meta-analysis we will determine whether the utility of a NP model is the same at an earlier time point compared to a later time point. Similarly, we aim to compare the various post-operative analgesia regimens in the models to determine if it is possible to use post-operative analgesics and only cease them for a window when measurements are being taken. This would refine the testing of animal models of NP to reduce the period of suffering of the animals and will provide a basis for experimental design of future studies.
Another key objective of this study is to reduce the number of animals used in studies of NP. If it is possible this will be achieved by identifying tests which, because of the variance observed, require fewer animals. We will also provide guidance on sample size calculations and therefore reduce the number of underpowered studies carried out. This should in turn reduce the number of animals sacrificed in experiments which are either too small reliably to detect the effect being sought or those which involve more animals than required.
We plan to assess for publication bias and the impact of this on outcome. In the in vivo stroke literature we identified that one in six studies remain unpublished and this leads to a 31% overstatement in efficacy9. Publication bias is a prevailing issue in the pre-clinical literature and we are of the opinion that all animal studies should be disseminated. Identifying the magnitude of this problem in the NP literature will highlight the importance of the issue and encourage the public availability of all data derived from the use of animals. Minimising the unnecessary replication of experiments will ultimately reduce the waste of animals and redundant studies.
Additionally, this study will attempt to identify which models show the best translation to the clinic. This data would enable researchers to choose the model most suitable for translating to the clinic and therefore reduce the number of animals used that do not provide translatable results.
This project will, for the first time, generate robust empirical data from completed experiments in NP to guide the reduction and refinement of the use of animals in this area. Adequate power and valid experiments would improve the utility of animals in pain modelling. By improving experimental design and statistical rigour this study will contribute towards a reduction in the number of animals used in pain modelling studies. This study will also contribute to the refinement of procedures, which would improve animal welfare in pain modelling. The area of animal pain modelling could greatly benefit from this study.
Another key objective of this study is to reduce the number of animals used in studies of NP. If it is possible this will be achieved by identifying tests which, because of the variance observed, require fewer animals. We will also provide guidance on sample size calculations and therefore reduce the number of underpowered studies carried out. This should in turn reduce the number of animals sacrificed in experiments which are either too small reliably to detect the effect being sought or those which involve more animals than required.
We plan to assess for publication bias and the impact of this on outcome. In the in vivo stroke literature we identified that one in six studies remain unpublished and this leads to a 31% overstatement in efficacy9. Publication bias is a prevailing issue in the pre-clinical literature and we are of the opinion that all animal studies should be disseminated. Identifying the magnitude of this problem in the NP literature will highlight the importance of the issue and encourage the public availability of all data derived from the use of animals. Minimising the unnecessary replication of experiments will ultimately reduce the waste of animals and redundant studies.
Additionally, this study will attempt to identify which models show the best translation to the clinic. This data would enable researchers to choose the model most suitable for translating to the clinic and therefore reduce the number of animals used that do not provide translatable results.
This project will, for the first time, generate robust empirical data from completed experiments in NP to guide the reduction and refinement of the use of animals in this area. Adequate power and valid experiments would improve the utility of animals in pain modelling. By improving experimental design and statistical rigour this study will contribute towards a reduction in the number of animals used in pain modelling studies. This study will also contribute to the refinement of procedures, which would improve animal welfare in pain modelling. The area of animal pain modelling could greatly benefit from this study.
