Home » Non-human primates » Level 3 » Question 3

Non-human primates in research and safety testing

3. Are there alternatives to the use of primates in research and safety testing?

  • 3.1 Are there alternatives to the use of primates in safety testing of pharmaceuticals?
  • 3.2 Are there alternatives to the use of primates in research on infectious diseases?
  • 3.3 Are there alternatives to the use of primates in research on the human brain?
  • 3.4 Are there alternatives to the use of primates in research on organ transplants?

The SCHER opinion states:

Tests are first carried out on cells grown in the laboratory
Tests are first carried out on cells grown in the laboratory
Source: Jean Scheijen

SCHER recognises that there are promising developments that have replaced NHP use. A number of alternative methods (either in vitro or using other animal species) have been developed and implemented over the last decade (e.g. the TgPVR21 transgenic mouse model for neurovirulence and potency testing of poliomyelitis vaccines) (EDQM).

The position of the SCHER relating to the use of animal testing in the context of the assessment of hazards and health risk assessment of chemicals, the “Three Rs” concept of Replacement, Refinement and Reduction of animal use for experimental purposes has been stated before. The continuation of the high level of human health and environmental protection is identical to the position of the former CSTEE. This position is outlined in detail in opinions by the CSTEE in 2004 and by SCHER in 2005 (CSTEE, 2004; SCHER, 2005). This position can also be extended to areas of basic research where information generated will have a major influence on understanding basic physiological functions and mechanisms of pathophysiology, when a benefit to prevent or treat humans diseases can be expected in the longer term.

In the opinion of SCHER, animals should only be used in medical research when it is unavoidable and when appropriate and validated alternative methods are not available. Replacing animals in medicine research is a long and difficult process and application of in vitro or in silico methods are often not yet feasible due to highly complex systems and limited knowledge of basic biology and pathophysiology. In addition, experimental models not using animals are often developed in medical research as complementary methods as they may only address questions at sub-cellular or single cell level, or, at best, at the level of interactions between a very limited number of cell types. When whole body integrated systems need to be examined, animal models have to be used in order to better understand the interactions between different cells in an intact organ, and between different organs. The importance of combining all approaches at the cellular, organ and whole body level are vital to a full understanding of the scientific issues.

SCHER also recognises that when animals are used as models of human conditions or as surrogates for humans, there are limitations to the accuracy with which the animal model reflects the pathophysiology, pharmacology or toxicological susceptibility of humans. In the cases examined in this opinion, the use of NHP is considered essential because other species provide demonstrably unsatisfactory models in crucial respects.

It should not be forgotten that humans are also used in experiments whether healthy human subjects, patients participating in clinical studies, and tissues from bio-banks. Furthermore, it is important that there is a constant feedback and iteration between human and animal research, as well as in vitro studies, to improve our knowledge and to make animal and human experiments more meaningful.

Source & ©: SCHER,  The need for non-human primates in biomedical research, production and testing of products and devices (2009),
Section 3.2 The currently available possibilities to replace NHP use either with methods not entailing the use of animals or by resorting to other species of animals including genetically altered animals, p.21.

3.1 Are there alternatives to the use of primates in safety testing of pharmaceuticals?

The SCHER opinion states:

Safety testing of pharmaceuticals

In safety testing, regulatory requirements and scientific considerations may almost mandate the use of NHPs if NHPs represent the non-rodent species resembling humans most closely regarding pharmacodynamics and pharmacokinetics. It needs to be noted that testing of new pharmaceuticals in NHPs represents only a very small part of the total safety and efficacy testing. Results obtained in NHPs are introduced into the risk assessment process, which integrates all information from safety testing based on a weight of evidence approach. The total replacement of animals, including NHPs in testing for safety, is not possible based on present knowledge. Arguments against phasing out NHPs in safety testing of pharmaceuticals are therefore identical to those regarding using rodents for toxicity testing, i.e. incomplete knowledge of integrated body systems and pathophysiology, poor representation of pharmacokinetics by in vitro systems, and the absence of NOAEL or benchmark doses vital for human risk assessment (SCHER, 2005).

Regarding safety testing of the highly specific monoclonal antibodies and the other biotechnology derived products, NHPs are often the only relevant model for humans. In certain cases, genetically modified rodents, carrying the human pharmacological target, may replace NHPs. This requires, however, that downstream signalling is relevant for humans and that the alternatives are sufficiently well characterised. At present, genetically modified rodents as well as testing of the homologous protein in rodent species are usually considered as supportive data and not as replacements for the use of NHPs by regulators (Anonymous, 2008).

Micro-dosing is sometimes postulated to be able to replace some animal testing. Microdose studies in humans are considered to be clinical trials in accordance with the EU Clinical Trials Directive and, therefore, have to be supported by animal toxicity studies Therefore, micro-dosing cannot replace animal testing, and administration of chemicals or pharmaceuticals to humans in low doses to study pharmacokinetics and toxicokinetics (biokinetics) (Amberg et al., 1999; Monster et al., 1976) has been used for a long time in research. Recent developments in analytical chemistry such as LC/MS-MS or accelerator mass spectrometry have only refined microdose studies due to more simple sample workup and higher sensitivity. Micro-dosing in early human studies only investigates pharmacokinetics and is performed after administration of very low single doses (max. of 100-fold below the pharmacologically active dose in animals). As a prerequisite for performing microdose experiments in humans, single dose toxicity data in an appropriate animal model are needed to ensure that the microdose given to humans can be considered a safe dose. Thus, toxic effects are not expected in humans and a toxicity profile cannot be established. Toxicity in animals is the relevant endpoint in all safety testing and this can thus not be studied with micro-dosing. However, compounds with an unfavourable human pharmacokinetic profile are not further developed and in that sense, the use of animals in toxicity testing may be reduced due to earlier termination of an unpromising compound. On the other hand, if a compound shows a favourable human pharmacokinetic profile in micro-dosing, all standard animal safety tests are needed for further clinical development, so that micro-dosing in humans can also result in an increase in the number of animals used for a specific compound (single dose toxicity study plus standard tests) (EMEA, 1994).

The US National Academy of Sciences has recently issued a report on “Toxicity testing in the 21st century”. The report discusses a “vision” to reduce the need for animal testing based on a combination of in vitro testing, “omics”-technologies applied to in vitro systems, and physiologically-based pharmacokinetic modelling within the next decades. Animal testing should only be used when unclear results are obtained or specific concerns are present. However, it needs to be noted that the mandate of the NAS committee was restricted to environmental chemicals where daily human doses are much lower then those used in therapy with pharmaceuticals. Therefore, the conclusions of this report cannot be applied to pharmaceutical safety testing at the present time.

Source & ©: SCHER,  The need for non-human primates in biomedical research, production and testing of products and devices (2009),
Section 3.2 Section on “Safety testing of pharmaceuticals”, p.22-23.

3.2 Are there alternatives to the use of primates in research on infectious diseases?

The SCHER opinion states:

Infectious diseases

In infectious disease research and vaccine development, there are no ideal small-animal models for studying HIV infection. A model for investigating immunogenicity is the Trimera model, where a human immune system is introduced into wild-type mice. However, conclusions are limited since the viability of the human cell transplant is short and so the Trimera model is mainly used for the investigation of short-term immunity and rapid screening of candidate HIV vaccines (Ayash-Rashkovsky et al., 2005).

Generating a genetically modified mouse permissive to infection with HIV is difficult, since all species-specific factors needed for complete HIV replication in mice have not been identified. Recently, a mouse model (HIV/MuLV) has been established based on the infection by HIV-1 enveloped by a mouse retrovirus envelope. Since the mouse immune system is intact, studies of HIV candidate vaccines and adjuvants can be made over longer time periods compared with other models (Boberg et al., 2008). Similar to the Trimera model, HIV/MuLV-challenge system is primarily useful for screening candidate vaccines, but further testing of such vaccines requires studies in NHPs.

It has been claimed that the failure to achieve protection against infection in clinical studies of HIV vaccines invalidates the use of NHPs in preclinical vaccine studies (Gordon and Langley, 2008). However, over the years, new knowledge about the virus and how the immune system interacts with it has been gradually collected both in humans and in NHPs. As a result, new and better animal models have continuously been developed. In parallel, based in part on observations in these studies, in vitro techniques have given us a better understanding about fundamental reactions at a cellular level in the immune system. To date, in vitro systems only show that a given formulation can induce an immune response in human cells and cannot demonstrate that this immune response protects the host against viral infection. Therefore, NHPs cannot be replaced in this area of research at present and in vitro methods can only be regarded as complementary techniques.

Many alternative methods, without the use of NHPs, have also been used in search for a malaria vaccine. After publication of the genome of Plasmodium falciparum, the most important human malaria parasite (Gardner et al., 2002), genomics-related technologies, recombinant DNA and cell engineering have increased the knowledge about genes and pathways involved in human malarial infections.However, no in vitro system today mimics the complex biology of the human malaria parasites and the interaction with the immune system of the host. Infection of mice with sporozoites of Plasmodium berghei or Plasmodium yoelii to evaluate liver-stage protection by candidate malaria vaccines has not translated readily to effective malaria vaccines in humans. Thus, mice can be used to dissect basic parameters required for immunity, but may not represent preclinical vaccine models (Schmidt et al., 2008). While NHPs also have their limits in the studies of protective immunity, they will likely be needed to develop human malaria vaccines in the future.

A cell culture system with a transfected human hepatoma cell line has been developed (Bartenschlager and Lohmann, 2001) that allows research on hepatitis C anti-viral treatment without the use of NHPs. However, for development of vaccines, it is still necessary to test the efficacy of candidate vaccines in chimpanzees.

Source & ©: SCHER,  The need for non-human primates in biomedical research, production and testing of products and devices (2009),
Section 3.2 Section on “Infectious diseases”, p.23-24.

3.3 Are there alternatives to the use of primates in research on the human brain?

The SCHER opinion states:


Magnetic resonance imaging (MRI) is now routinely used for diagnostics in humans. Functional MRI (fMRI) is also widely used in most fields of neuroscience. It is often considered as an alternative to replace research involving NHP. MRI measures changes in vascular parameters and relies on a link between neural activity and vascular variations (blood flow, oxygenation levels). It is thus only an indirect measure of neural activity. Moreover, fMRI is limited because our knowledge on neurovascular coupling and its underlying mechanisms is incomplete (Weatherhall, 2006). Recently, these issues became highly relevant in the context of how to correctly interpret the signals used for functional brain imaging (Vanzetta, 2006). In addition, the temporal and spatial precision of fMRI is low, measuring variations in blood oxygenation levels in the order of seconds, far from the millisecond range at which neural cells process information. Also, fMRI and other neuroimaging techniques (EEG, MEG) give large-scale functional views by being able to record activity or activations from the entire brain. Because of these characteristics fMRI studies cannot replace studies collected with invasive microelectrode techniques; these are much more precise anatomically and temporally. In summary, fMRI and microelectrode studies in NHP are complementary. In fact, the use of fMRI and of other haemodynamic-based functional brain imaging with NHP is of fundamental interest to bridge the vast knowledge acquired in NHP research with established techniques and data acquired in humans, as well as to understand the neural bases of what is measured using fMRI and to validate the technique itself (Logothetis, 2008; Vanzetta and Grinvald, 2008).

The very promising diffusion imaging techniques (DWI, DTI) use magnetic resonance imaging for the non-invasive detection and tracing of neural fibres. However, while the potential of DTI to study connectivity in normal and diseased human brains is highly significant, this technique still needs to be validated and evaluated by histological studies. Current shortcomings are lack of anatomical precision, the inability to evaluate the direction of fibres, and the fact that algorithms interpreting signals acquired with diffusion imaging are based on untested hypotheses (Tuch et al., 2005).

Computer modelling is rapidly improving and is expected to reach significant importance in the domain of robotics and the development of machines based on neural knowledge. Many laboratories integrate computer modelling in their research; it is important to acknowledge, however, that most available models still have poor prediction rates due to limitations in our present understanding of basic brain physiology.

Theoretical approaches to simulate brain anatomy and function depend on empirical data for verification. Neurophysiological and neuroanatomical investigations are the main sources for the development of biologically plausible computational models and neural network models require detailed knowledge on the characteristics of individual neurons, their connectivity and pharmacology. These are acquired with invasive neurophysiological methods. Moreover, knowledge of how the brain works in healthy individuals is important to better understand the pathophysiology of disease.

The Blue Brain project is the first comprehensive attempt to reverse-engineer the mammalian brain, in order to understand brain function and dysfunction through detailed simulations. The first phase was reached in 2007 with a complete modelling of a unique rat cortical column of 10,000 neurons that required the full computational power of a supercomputer. A realistic model of a primate brain will have to contain up to 100 billion neurons. While this approach is necessary for the advancement of neuroscience, there is no foreseeable time when an artificial model of a primate brain will be feasible.

Source & ©: SCHER,  The need for non-human primates in biomedical research, production and testing of products and devices (2009),
Section 3.2 Section on “Neurosciencess”, p.24.

3.4 Are there alternatives to the use of primates in research on organ transplants?

The SCHER opinion states:


In vitro models may be useful in the initial investigations into the presence or absence of antibodies and receptors and rodent models of transplantation and xenotransplantation are usually simpler than primates, especially with regard to the measures required to control the immune response. However, they cannot replace long term studies of function in animals including NHPs.

Source & ©: SCHER,  The need for non-human primates in biomedical research, production and testing of products and devices (2009),
Section 3.2 Section on “Xenotransplantation”, p.25.

Themes covered
Publications A-Z

Get involved!

This summary is free and ad-free, as is all of our content. You can help us remain free and independant as well as to develop new ways to communicate science by becoming a Patron!