Thursday, 31 December 2015

EAE is rubbish again for finding drug responses

de Bruin NM, Schmitz K, Schiffmann S, Tafferner N, Schmidt M, Jordan H, Häußler A, Tegeder I, Geisslinger G, Parnham MJ. Multiple rodent models and behavioral measures reveal unexpected responses to FTY720 and DMF in experimental autoimmune encephalomyelitis. Behav Brain Res. 2015 Dec 12. pii: S0166-4328(15)30315-6.

Experimental autoimmune encephalomyelitis (EAE) is a widely-used rodent model for multiple sclerosis (MS), but a single model can hardly capture all features of MS. We investigated whether behavioral parameters in addition to clinical motor function scores could be used to assess treatment efficacy during score-free intervals in the relapsing-remitting EAE model in SJL/J mice. We studied the effects of the clinical reference compounds FTY720 (fingolimod, 0.5mg/kg/day) and dimethyl fumarate (DMF, 20-30mg/kg/day) on clinical scores in several rodent EAE models in order to generate efficacy profiles. SJL/J mice with relapsing-remitting EAE were studied using behavioral tests, including rotarod, gait analysis, locomotor activity and grip strength. 

Prophylactic treatment with FTY720 prevented clinical scores in three of the four EAE rodent models: Dark Agouti (DA) and Lewis rats and C57BL/6J mice. Neither prophylactic nor late-therapeutic treatment with FTY720 reduced clinical scores or reversed deficits in the rotarod test in SJL/J mice, but we observed effects on motor functions and sociability in the absence of clinical scores. Prophylactic treatment with FTY720 improved the gait of SJL/J mice. DMF was tested in three EAE models and did not improve clinical scores at the dose used. These data indicate that improvements in behavioral deficits can occur in absence of clinical scores, which indicate subtle drug effects and may have translational value for human MS.

So we use animal models as a tool to predict response of drugs in MS  so in this study they take Dimethyl fumarate and FTY720 and find that FTY720 (fingolimod) inhibits disease in EAE Dark Agouti (DA) and Lewis rats and C57BL/6J mice but not SJL mice, unless you do subtle behavioural tests. On the other hand DMF did not have an effect.

So first thing you have to say is that we know that fingolimod and slow release dimethyl fumarate work in MS and so if they don't work in EAE, then the models have no validity for their purpose and are of no use.

So I hear many of you say....yeah. EAE is rubbish and is useless!!!!

Yes we have to deal with this critisism everyday and this paper adds to this view.  Therefore it is appropriate to comment.

However positive responses of both DMF and FTY720 in EAE would have both been part of the drug response to developing these drugs, so why the negative data?.

What about EAE in SJL mice and fingolimod why the failure? Is the SJL mouse strainrubbish to study drug effects?.....

Whilst I personally think it is pretty rubbish...not because it does not respond to FTY720, but because it gets too sick and its front and hind legs are affected (we would  cull these animals as they can't get to food and water) and doesn't relapse with high enough frequency to be useful for drug testing (OK this is a personal preference and a little dig). What have other found?

Easy there are loads of studies showing FTY720 works in EAE

Webb M, Tham CS, Lin FF, Lariosa-Willingham K, Yu N, Hale J, Mandala S, Chun J, Rao TS.Sphingosine 1-phosphate receptor agonists attenuate relapsing-remitting experimental autoimmune encephalitis in SJL mice. J Neuroimmunol. 2004;153(1-2):108-21.

I bet I could find loads of examples as contract research organisations that do contract drug testing and use fingolimod (FTY720) as the positive control, because it works in MS.

If you go to Novaritis and ask if it works in SJL mice, they would probably say, we like to see positive responses in more than one strain/species but you have to use the human equivalent dose to get an idea of whether it will work in humans. In this study they use 0.5mg/kg but the human equivalent dose is I believe 3mg/kg and this certainly works in ABH mice, so was enough used early enough?

Immunosuppression with FTY720 is insufficient to prevent secondary progressive neurodegeneration in experimental autoimmune encephalomyelitis. Al-Izki S, Pryce G, Jackson SJ, Giovannoni G, Baker D. Mult Scler. 2011;17:939-48

Now to DMF. In the human they us only a small dose of slow release material but it is used 2 times a day because it has terrible pharmacokinetics even  with the slow release formulation and without it it would be many more timesadayand if you do not use the slow release formulation as made by Biogen it is even more rubbish and so to get any response you have to use bucket loads of the stuff. At once a day dose if you look at work done by Biogen they use 100-200mg/kg dimethyl fumarate to get any form of inhibition because of the poor pharmacokinetics of a drug, so if you pluck a dose say 30mg/kg out of the air it may not work. 

Indeed, if you use lower doses that 100mg/kg/day you have to dose animals at least twice a day, which is a pain in the bum when you have to do weekend dosing. But indeed that is how it can work and we have done it in C57BL/6 mice recently.

So this shows the importance of doing a dose-response in animal experiments. Yes, it uses more animals but maybe it helps you get more sensible answers. So first you can read and find what companies who make the drug and others have used and there are ways to examine dose response without doing them in EAE.

However, too often we see one dose used in EAE experiments to give the perfect response. This no doubt leads to the over-estimation of success of a particulary drug as they have given more and more until it worked, generally we use ten times dilutions.

However when you get negative results, you need to find out why

So we move into 2016 and no doubt we will get some great animal work but this will be peppered with the usual guff that plagues basic science. 

So we will no doubt have some more MD rants in the future. 

But if referees actually read and implimented suggestions based on the ARRIVE guidelines there would be alot less unreproducible guff that we get to read. 

Baker D, Amor S.Publication guidelines for refereeing and reporting on animal use in experimental autoimmune encephalomyelitis. J Neuroimmunol. 2012; 242:78-83.


  1. I do not see problems with EAE, but I see problems in the drugs themselves ...

  2. Are the DMF results in this paper, more or less, not expected? The large human trials have shown DMF to have a weak effect on human (MS) progression. (One weak but statistically significant. In another the effect was so weak that it wasn't statistically significant). Have I misunderstood?

    1. I think you are talking brain atrophy in MS. In EAE they are really looking at the immune response...This has a downstream effect on nerve content. However you point it that DMF is not the best DMT is taken, but this is not why they did the study their view is that it should work in all models


Please note that all comments are moderated and any personal or marketing-related submissions will not be shown.