Hacker News Reader

Comment by bwb | Hacker News Reader

Only 8% of experiments on mice work on humans… friendly reminder.

A review [0] found that only 37% of research on mice translate in humans. But as someone interested in neurodegenerative diseases I can only recall that there were ~2500 clinical trials done on Alzheimer because of interesting results in mice and ... not a single one was effective in humans.

For less investigated diseases such as ALS or Parkinson it's about 700 trials without any success.

An excellent article discussing why mice studies are not applicable to humans, at least for ALS, is this one: https://www.nature.com/articles/nrneurol.2015.65

[0] http://jamanetwork.com/journals/jama/article-abstract/203577

Do you have a reference on this? It'd be quite interesting to read further on why studies using mice are often poor predictors of human reactions to exposure.

I read it years ago in a medical journal talking about the insanity of websites that post cutting edge research when it is only the early result.

Sorry not sure where it is located, but you could Google it.

You have locations like this: https://theconversation.com/of-mice-and-men-why-animal-trial...

It showed that of the most-cited animal studies in prestigious scientific journals, such as Nature and Cell, only 37% were replicated in subsequent human randomised trials and 18% were contradicted in human trials. It is safe to assume that less-cited animal studies in lesser journals would have an even lower strike rate.

Not sure why that is higher, but it is phrased weird.

That's actually quite a lot.

ya that is what I thought too, although I think news organizations should stop reporting on any "medical breakthroughs" involving animals.

Would also like some source/info on this, seems suspicious, like something that anti-cruelty organizations came up with and people are mindlessly believing and repeating...

Google on this one, as I read it years ago in a medical journal.

8% seems good to me, as that is the point of testing... but the point of the article was to avoid the hype of these early tests.

Source? It would be nice to scope such vague statements. Some kind of pharmaceutical mechanism of action ~totally transpose to the human while some other do not at all, it's heavily topic dependent. For example mitochondrial targeted antioxidant are specy agnostic.

Google, not sure how that is vague :), it is a specific number I read in a medical journal years ago.

Reply on news.ycombinator.com