An eight-year study by researchers in the US has found that fewer than half of the most influential cancer biology experiments of the last ten years, including some involving animals, are reproducible.
The Reproducibility Project: Cancer Biology (RP:CB) initiative, led by the Center for Open Science (COS), US, aimed to replicate all experiments published in 53 of the leading cancer papers over the last decade, but found that only five of these could be fully reproduced.
Four of these five studies involved animals, but overall research using animals was less likely to be reproducible.
On animal research, the authors highlighted the difficulties in producing reliable data due to the small effects that are recorded in animal research, or changes in the environment of the animals, such as the impact that variations of gut bacteria in the animals in different facilities might have on the results.
However, failure to replicate the study in the RP:CB project did not necessarily mean that the overall conclusions of the papers were incorrect – Science reported that two animal studies that weren’t reproduced by RP:CB have led to promising early clinical results for new drugs and treatments..
“[The paper] has huge implications for the success of these things moving up the pipeline into the clinic,” said Tim Errington, project leader at COS, in Science magazine.