Biases in Animal Studies may Differ from those in Clinical Trials
The analysis of 63 animal studies of statins, was led by Lisa Bero, PhD, UCSF professor clinical pharmacy.
In previous studies, Bero determined that drug-company-sponsored clinical trials were associated with publication of outcomes that favor the sponsor. Bero’s work has been cited as part of policy reform efforts that have led many journal publishers, agencies and institutions to require researchers to disclose funding sources and possible conflicts of interest when presenting their research.
The impetus for the current study, Bero said, was to explore whether or not industry-funded animal studies also would be likely to yield more positive outcomes for the companies’ drug candidates.
But in their analysis the researchers found the opposite: Results of animal studies that had industry sponsorship were less likely to measure a benefit for statins in slowing or preventing arterial disease. Of the studies that disclosed funding, 9 of 19 industry-sponsored studies had results that favored statins, in comparison to 18 out of 28 studies that favored statins among studies not funded by industry.
The explanation may be, said Bero, that “the interests of the pharmaceutical industry might be best served by underestimating efficacy prior to clinical trials, and overestimating efficacy in clinical trials. By underestimating efficacy in preclinical studies, the pharmaceutical industry could reduce the money spent on clinical trials that did not lead to marketable products.”
“Because demonstrating drug efficacy in human studies is linked to drug company profits, drug companies may have more incentive to publish favorable efficacy findings of human drug studies than animal studies.”
However, the reason for the opposite findings obtained in analyzing animal and human studies merits additional investigation, Bero said. Selective reporting of study outcomes might play a role, she suggested.
Conclusions of all the studies tended to be favorable in Bero’s PLoS Biology analysis. While the industry-sponsored animal studies had somewhat less favorable results, they nonetheless were more likely to present conclusions that favored the statin even when data were less favorable. This result highlights the role of “spin” in communicating research findings, Bero said.
The UCSF researchers also found methodological problems to be common, both in non-industry and industry-sponsored studies. Furthermore, Bero found that harmful side effects were not investigated.
“Not a single animal study we looked at assessed adverse events following the statin intervention,” Bero said. “As toxicity data from animal studies must be submitted to drug regulatory authorities before a compound can proceed to testing in humans, it is surprising that so little data on harm appear in the published scientific literature.”
In about half the studies analyzed, it appeared that animals were not assigned to treatment or placebo arms of the study randomly, a requirement of high-quality clinical trials. Furthermore, in about half the animal studies analyzed animals were identifiable to the person assigning treatment, a violation of the practice of “blinding.”
Criteria for including or excluding animals from studies often were not included in published reports, the UCSF researchers found, and many studies also failed to account properly for changes in the assigned treatment arm that occurred during the course of treatment.
Most of the industry and non industry studies analyzed in Bero’s PLoS Biology report were done using rabbits and mice. To gauge atherosclerosis, targeted by statins, researchers quantified blood vessel qualities such as number of damaged blood vessels, blood-vessel diameter, plaque severity, blockage to coronary and other arteries, and plaque rupture.
The study was funded by the National Institute of Environmental Health Sciences.