Preclinical randomised controlled trials could tackle high attrition rates in drug development

Researchers have come up with a potential solution to the problem of experimental drugs failing in clinical trials, despite promising results from preclinical, animal studies.

Arthur Liesz of the Munich Cluster for Systems Neurology (SyNergy), Munich, Germany, and colleagues performed a preclinical randomised controlled multicentre trial (pRCT) to test a potential stroke therapy using a design and rigour usually reserved for clinical randomised control trials.

The therapy, an antibody that stops circulating immune cells from crossing into the brain where they could cause inflammation and damage after a stroke, has had mixed results in trials. Some studies have found that the anti-CD49d antibody improves outcomes after a stroke, while another study reported no significant benefit.

The researchers, from six independent European research centres, investigated the efficacy of anti-CD49d antibodies in two mouse models of stroke – one representing a relatively mild stroke, the other a relatively severe stroke. The results pooled from the six centres revealed that treatment with anti-CD49d antibodies was effective in the mild stroke model, but not in the severe stroke model.

“Using two stroke models was a critical part of the design of our study,” says Liesz. “Clinical stroke is a highly variable disease and one drug might not be appropriate to cure all of them.”

Only a small fraction of the results from preclinical studies are replicated in the clinic, wasting time, money and scientific endeavour, not to mention risking potentially harmful clinical trials.

Animal studies are not subject to the same exacting requirements as clinical trials, and yet the results from these studies are used to support subsequent clinical trials. A separate study investigating previous animal studies has reported the inadequate reporting of data, statistical flaws, and missing cross-validation of data from independent study centres. “Virtually all preclinical studies performed by individual research groups are underpowered and often fail to adequately control for bias,” notes Liesz, whose findings were published in Science Translational Medicine
^[1]
on 5 August 2015.

“We have seen that preclinical studies with the same rigorous design as clinical trials are indeed feasible,” says Liesz.

Michael Tymianski from the Toronto Western Hospital Research Institute, Toronto, Canada, is impressed. “The approach of conducting multicentre preclinical trials of a therapy sets a new standard for life sciences research, which is currently plagued by irreproducibility of a substantial proportion of preclinical data,” he writes in a commentary piece^[2]
in the same issue of Science Translational Medicine.

Nevertheless, Tymianski, who notes that more than 1,000 experimental stroke treatments have not translated successfully to the clinic, does not believe Liesz and colleagues’ findings will solve the problem completely.

“It is difficult to believe that all of these failures resulted only from irreproducible or even false results,” he writes. “It is an important piece of the larger unsolved translational puzzle in stroke research, but there are other key pieces that must be defined before the roadmap to success is clear.”

Liesz says their findings have implications for other disease areas. “In view of the frustrating experience of numerous failed clinical trials and the extremely low success rate of translational research from bench to bedside — in nearly all biomedical research areas, not only stroke research — I would argue that a confirmative preclinical study design such as a pRCT should be mandatory before moving forward to conduct a clinical trial,” he concludes.