EAE is a good model of MS

Baker D, Amor S. Experimental autoimmune encephalomyelitis is a good model of multiple sclerosis if used wisely. Mult Scler Relat Disord. 2014 Sep;3(5):555-64. doi: 10.1016/j.msard.2014.05.002.

Although multiple sclerosis is a uniquely human disease, many pathological features can be induced in experimental autoimmune encephalomyelitis (EAE) models following induction of central nervous system-directed autoimmunity. Whilst it is an imperfect set of models, EAE can be used to identify pathogenic mechanisms and therapeutics. However, the failure to translate many treatments from EAE into human benefit has led some to question the validity of the EAE model. Whilst differences in biology between humans and other species may account for this, it is suggested here that the failure to translate may be considerably influenced by human activity. Basic science contributes to failings in aspects of experimental design and over-interpretation of results and lack of transparency and reproducibility of the studies. Importantly issues in trial design by neurologists and other actions of the pharmaceutical industry destine therapeutics to failure and terminate basic science projects. However animal, particularly mechanism-orientated, studies have increasingly identified useful treatments and provided mechanistic ideas on which most hypothesis-led clinical research is based. Without EAE and other animal studies, clinical investigations will continue to be “look-see” exercises, which will most likely provide more misses than hits and will fail the people with MS that they aim to serve.

This paper has just appeared on pubmed

We were asked to do a counterpiece to
Behan PO, Chaudhuri A. EAE is not a useful model for demyelinating disease. Mult Scler Relat Disord. 2014;3(5):565-74, 

However, rather than read a rant against EAE and then try and defend a position on which we may not be interested in defending or write a paper about how they don’t know what they are taliking about, we wrote our own unrelated paper . 

This does not try to defend the indefensible but indicates that both basic scientists could conspire to bring about a drug’s down fall. Notably, Neurologists play a part in this process. 

Whilst it is a clinical pass time to beat up the animal work..the failure of animal studies to deliver human treatments is such a common theme, it can’t always be the animals fault. 

Do Neurologist play any part in the downfall of treatments…I think almost certainly…..yes.

It’s always …..easy to make excuses in hindsight.


About the author



  • EAE is nearing EDSS. It's like the dead parrot sketch on Monty Python. EAE allowed lots of students to gain their PhDs. It has done very little for humans with MS. I don't think it's fair to blame neuros. Quite simple, EAE is not a good model for MS. 50 years of EAE research has not delivered the cause or cure. It could never do this. Researchers will defend it to the hilt – they have to as it's their lively hood. Hopefully as better treatments become available we can say goodbye to EAE. It has served PhD students well. It has done nothing for those with progressive disease. As the viral basis of MS is revealed, we will see why EAE failed.

    • The best selling MS drug comes directly from EAE……blinkered views are the problem…..What's yours?

      I dont defend to the hilt your need to read the posts.
      If you read the link you can see that EAEers are not beyond reproach and for many of them the mechanism is what is being investigated not a treatment.

      However you are geting your wish…..the parrot may well be dead soon. For immune modulating DMT its been dead in the UK for the past ten years or so.

      As you are keen to defend neuros for trying, dont forget to vote for them.

    • Geez, no comment. You realise this bench to bedside business has nothing to do with EAE…politics…getting closer…

  • Sounds like Fawltey Towers. "Don't mention the falibility of neuros"
    I mentioned it once will icgetvaway with it ….obviously not

    Dont burst any bubbles

  • What are your thoughts on the concept of meta-analysis of animal studies or multi-centre animal studies? Are these potential ways to get a big enough pre-clinical dataset to be more sure that a compound will translate well to human use?

    • meta analysis has a problem because the negative data seldom gets published so the published data is skewed. However it can be done as is the case of vesterinen et al 2010. However without a critcal eye you can lump the crud in with the good. The vesterinen meta analysis was done with a view of finding neuroprotives which cant be done with the data as the data is about immune modalators
      So without a critical eye it is rubbish in rubbish out.

      Next on could say if we dont compare like with like are we going to get sensible data. 70% of treatments used to occur before disease is induced in EAE. This never happens in MS. Most MS drugs give drug in remission but most eae experiments thiz is never done so the models rather lack external validity so you are largely doing a meta analysis of cherry picked data. Should we have prrregistration of animal studies to identify the negative data. No but a journal of negative results may help but unless the experiments have quality control in them then they are not worth reporting. Eg the score of some control groups are say three and then we find data with. conttol group of one meaning a limp tail then it iz probably a rubbish experiment meaning the data is dubious. anx should be repeated. Get quality control and alot of rubbish unrepeatablestuff would go

  • next multicentre studies. This is not as easy as one would think especialliy if it is across countries. A group of stroke researchers have been investigating the mechanisms of doing a multinational study it is hard. There is the issue of ethics different countries want different things. It is not harmomised so one size does not fit all and then there iz the issue of protocols. This could be like difficult because each lab knows best. You could argue that we should do what top labs do but many of them have the quality control issue in their published work so this may not be a good place go . Contract research organisations are after your money and dont always do a good job.

    However if you had an independent lab test immunemodulators before they go into human trials then many ideas would fall on the sword because the animal data was not reproducible. The point is to care that you have a solid model system and then you have to happy that some drugs dont work and
    you are not desparate for a result. Labs whose data is reprodcible would be good choices to do this.

By MouseDoctor



Recent Posts

Recent Comments