The search for new drugs to combat major diseases is being set back by shoddy animal research, according to work by two teams of scientists.
An Edinburgh University survey of thousands of animal studies found that the majority were not rigorous enough to rule out effects that routinely inflate the benefits of new treatments.
The analysis led by neurologist Malcolm Macleod found that top-tier journals were among the worst offenders for publishing poorly designed animal studies and that two-thirds of animal work from leading UK universities had question marks over its validity.
Illustration: Lance Liu
Animals are used to test experimental compounds before companies decide whether to push ahead with full clinical trials in humans, but flawed animal tests can exaggerate the effects of candidate drugs and lead to trials that end up being a colossal waste of time and money, as well as suffering.
“We could clearly be doing a lot better, and if we do better, our science will become more robust in terms of our results being translatable into new treatments for diseases,” Macleod said.
The scientists pored over animal studies to see if authors had taken basic steps to prevent biases from skewing their results. These involved the random assignment of animals to treatment groups; blinded assessment of how the animals fared; a record of animals removed from the results and a calculation to show that the study had enough animals to be statistically sound.
Studies that fail to rule out biases can often find big effects and so get published in the highest profile journals, but these studies tend to be the most unreliable ones and are often knocked down by better studies done later.
The Edinburgh team first looked at a random collection of 146 animal studies from 1941 to 2012 and found that less than 20 percent of the studies randomly assigned animals to test groups, to reduce the chances of one group, for example, being healthier than the other. Yet in only 3 percent of studies were scientists blinded, meaning that in 97 percent of cases, the authors knew which animals should show a response.
The team went on to look at more than 2,500 animal studies published in leading journals between 1992 and 2012 to see if top-tier publications reported more robust research than lower ranked journals. Again they found the opposite to be true, with less attempt to rule out biases in the top flight journals, according to their report in the journal Plos Biology.
Finally, Macleod’s team looked at more than 1,000 papers from the UK’s top universities: Cambridge, Edinburgh, Imperial, Oxford and University College London to see if the best institutions deserved their reputations. More than two-thirds of the studies did not report even one of the four measures seen as critical for reducing the risk of bias. Only one study, written by Alastair Buchan, dean of medicine at Oxford, included all four.
The problem with animal research is highlighted by a second team from McGill University in Montreal, Canada, which found that flawed animal studies over-estimated the effectiveness of a new kidney cancer drug by up to 45 percent.
The drug, sunitinib, is sold under the name Sutent and targets pathways that cause certain cancers to grow.
Jonathan Kimmelman at McGill said studies reporting little or no cancer effect had simply not been published.
Few animal studies on sunitnib were designed to avoid biases, the scientists wrote in the journal eLife, and most were conducted on one laboratory mouse strain that responded well to the drug: young females with compromised immune systems.
Tests involving other animal types, including mice that had spontaneously developed tumors, showed less of an effect. In many cases it was not even clear how many animals had been tested as the sample size was not reported.
“Only a fraction of drugs that show promise in animals end up proving safe and effective in humans. An important reason is because studies in animals are often not well-designed, and because positive results have a higher chance of being published. They end up skewing what we think we know about the potential of a drug,” Kimmelman said.
“People are motivated to get things published, to get funded, to get promoted and there’s an awful lot of unconscious bias going on way beyond any obvious conflicts of interest,” said Buchan, who has been a researcher for more than 30 years. “Most of what I’ve done has found negative results and it’s very hard to continue getting negative data when everyone else is getting positive data. I worry about the funding agencies not being strict enough, I worry about the investigators not reporting what they need to report, the journals need to be much stricter, and the institutions have to bear some responsibility.”
Buchan has spent much of his career looking for drugs that protect the brain cells against stroke damage.
“We’ve been at this for 30 to 35 years,” he said.
“The reality is that in neuroprotection, we don’t have a drug, and if you told me that in 1980, I would have had to think twice about going into what I thought was a really key field,” he said.
Guidelines developed in the UK and aimed at improving standards of medical research conducted on animals were published in the journal Public Library of Science Biology in 2010.
The Canadian scientists called for new guidance on the design and reporting of pre-clinical cancer studies.
In an article published in Newsweek on Monday last week, President William Lai (賴清德) challenged China to retake territories it lost to Russia in the 19th century rather than invade Taiwan. “If it is really for the sake of territorial integrity, why doesn’t China take back Russia?” Lai asked, referring to territories lost in 1858 and 1860. The territories once made up the two flanks of northern Manchuria. Once ceded to Russia, they became part of the Russian far east. Claims since then have been made that China and Russia settled the disputes in the 1990s through the 2000s and that “China
China has successfully held its Forum on China-Africa Cooperation, with 53 of 55 countries from the African Union (AU) participating. The two countries that did not participate were Eswatini and the Sahrawi Arab Democratic Republic, which have no diplomatic relations with China. Twenty-four leaders were reported to have participated. Despite African countries complaining about summit fatigue, with recent summits held with Russia, Italy, South Korea, the US and Indonesia, as well as Japan next month, they still turned up in large numbers in Beijing. China’s ability to attract most of the African leaders to a summit demonstrates that it is still being
Trips to the Kenting Peninsula in Pingtung County have dredged up a lot of public debate and furor, with many complaints about how expensive and unreasonable lodging is. Some people even call it a tourist “butchering ground.” Many local business owners stake claims to beach areas by setting up parasols and driving away people who do not rent them. The managing authority for the area — Kenting National Park — has long ignored the issue. Ultimately, this has affected the willingness of domestic travelers to go there, causing tourist numbers to plummet. In 2008, Taiwan opened the door to Chinese tourists and in
Taiwan People’s Party (TPP) Chairman Ko Wen-je (柯文哲) on Thursday was handcuffed and escorted by police to the Taipei Detention Center, after the Taipei District Court ordered that he be detained and held incommunicado for suspected corruption during his tenure as Taipei mayor. The ruling reversed an earlier decision by the same court on Monday last week that ordered Ko’s release without bail. That decision was appealed by prosecutors on Wednesday, leading the High Court to conclude that Ko had been “actively involved” in the alleged corruption and it ordered the district court to hold a second detention hearing. Video clips
RSS