Why aren’t there any good medical treatments for a stroke?

I meet Russell at Australia’s oldest pub, the Hope and Anchor Tavern in Hobart, which has been taken over by gaggles of scientists for the launch of this year’s Beaker Street Festival. Russell, from the University of Tasmania, sports glowing steampunk goggles and a lab coat covered in fake blood.

“There have been thousands of medicines that looked like they improved strokes in animals and yet, when researchers would go to test them in humans, they suddenly would not show the same improvements,” Russell tells me.

One analysis of this problem found that 69 per cent of experimental studies (i.e. animal trials) into stroke treatments were “positive”, meaning the tested drug seemed to work. In the next phase of study, early clinical research in people, only 32 per cent were positive. In phase three clinical trials, the positive hit-rate had dwindled to a measly 6 per cent.

Russell’s PhD focuses on whether the way we use animal models to mimic stroke is partly to blame for this trend.

One method is called middle cerebral artery occlusion (MCAO), which involves inserting a thread into an anaesthetised rat’s neck and up an artery into the brain to mimic a blockage.

“But it’s not like a human stroke in many ways because that animal is under surgery, it’s asleep, it’s got a physical blockage with thread instead of a blood clot blockage, which would have lots of different chemical markers,” Russell says.

Aside from being an imprecise imitation of a stroke, the method is difficult, prone to error and takes years of training to perform.

“Think about how small a rat is, and you’re inserting a sewing thread into their brain,” says Russell. “It’s a good way to mimic brain damage, but a bad way to mimic a blood clot.”

Pink dye and laser beams

There’s an easier way, however, to summon an actual blood clot in a study animal’s brain by using a hot-pink photosensitive dye called Rose Bengal.

“You can inject an animal with Rose Bengal dye, and then you can shine a laser light onto their brain at the specific point where you want a blood clot to form, and the dye in their bloodstream will react with the light and cause a blood clot,” Russell says.

The method forms a blood clot spontaneously, which gets closer to how real strokes arise in humans. But there’s a catch.

Russell ran a meta analysis of more than 400 studies using the Rose Bengal method in mice and rats. One group of stroke treatments seemed to show promise in these studies: antioxidant treatments, such as edaravone. The next logical step would be to test these antioxidant treatments in humans.

“But my research is trying to get people to stop and think a little bit more about which animal study is best for their specific research question,” Russell says. “If you look at how the Rose Bengal dye reacts, the light oxidises the dye, that’s what causes the blood clot to form.

‘If [research improves, we’ll have] a better understanding of stroke. That [is] a good thing, whether or not we find a magical treatment.’

Stroke researcher Dr Ash Russell

“So if you put an antioxidant into an animal and then try to oxidise the dye, it’s likely that the antioxidant is stopping the initial blood clot from forming.”

These interventions, it turns out, may be treating the model of stroke rather than stroke itself, one of the many reasons why so much promise in animal trials has not come to fruition in humans.

Data in the dark

There’s another issue with stroke research: the file drawer problem.

Studies reporting positive results are more likely to be published, an issue also known as publication bias. “New wonder drug prevents strokes in rats” is a far sexier paper than “We tested this drug in rats and it didn’t work”.

One PLOS Biology analysis found that publication bias meant the efficacy of some stroke drugs was being overestimated by a third because data from experiments that found the drugs had little or no effect weren’t seeing the light of day. The results were sitting at the bottom of the metaphorical “file drawer”.

That’s not only a scientific problem, but also an ethical one.

Firstly, it means drugs that don’t actually work could be making their way to human trials.

Secondly, most scientists try to study lab animals only when absolutely necessary. If experiments are carried out on rats but the resulting data isn’t published, that testing never contributes to the sum of human knowledge. The sacrifice of those animal lives is wasted.

To help address this, the authors of the PLOS Biology analysis suggest the creation of a central register of animal experiments and their results, allowing researchers to analyse both published and unpublished data so no experiments go to waste.

Russell says that with more scientists scrutinising better ways to conduct stroke science, the field is improving.

“If the quality of research keeps improving, we’re going to keep having a better understanding of stroke,” they say. “That can only be a good thing, whether or not we find some magical treatment that happens to work.”

Angus Dalton travelled to Hobart as a guest of Beaker Street Festival.

Examine, a free weekly newsletter covering science with a sceptical, evidence-based eye, is sent every Tuesday. You’re reading an excerpt – sign up to get the whole newsletter in your inbox.