John P. Gonzalez, PhD, CMPP, Solanum Medical Communications, Macclesfield, UK
Calls for urgent improvements in the design, registration, reporting, and transparency of preclinical studies have recently arisen (1). In an open letter to the USA Today newspaper published June 20, 2018, over 500 researchers, including four Nobel Laureates, called for increased openness about the important role of animals in health research (2). “We need better animal research, better reported” was the headline of a BMJ editorial published earlier this year (3). As with all biomedical publishing, preclinical publications facilitate transparency and the sharing of data and ideas with other scientists, helping to reduce unnecessary use of laboratory animals and avoid duplication of studies that have been conducted elsewhere but were not disclosed.
For a variety of reasons, a lot of preclinical research is never published and some of that which is published has been described as “not fit for purpose,” despite the science being sound (4). Preclinical research is designed to be predictive of the potential efficacy and safety of a molecule, informing the decision to take a molecule into the clinic for administration to humans. However, success rates for compounds transitioning to the clinic are generally accepted to be very low, and the total prevalence of irreproducible preclinical research has been estimated to exceed 50% (5).
Accuracy of Reporting in Preclinical Publications
A comprehensive survey of publications describing animal research reported that a number of the studies omitted the hypothesis or objective of the study, as well as the number and characteristics of the animals used (6). Most of the papers surveyed did not use randomization or blinding. In addition, only 70% of the publications that had used statistical methods described their methods and presented the results with a measure of error or variability. The survey exposes various issues around the accurate and comprehensive reporting of methods and results in preclinical publications, highlighting the problem that researchers face in replicating designs and reproducing results from animal studies in which information is incomplete or absent.
In a study analyzing 109 investigator brochures presented for ethics review for phase I and II trials, the content of 708 preclinical efficacy and safety studies contained in them was examined (7). For 89% of the studies, no reference to a published report was provided, and only 6% of these reported an outcome demonstrating no effect. Furthermore, the majority of the investigator brochures (82%) only reported studies with positive findings.
The Importance of Preclinical Research in Clinical Study Planning
In a recent case, the MVA85A tuberculosis vaccine failed to improve the prevention of tuberculosis in a large study of South African infants despite positive preclinical results in four animal species (8). It was revealed that other unsupportive preclinical research existed and was dismissed as failed or irrelevant by the researchers who were seeking funding and approval for human trials.
Similarly, an announcement that the STRIDER (sildenafil therapy in dismal prognosis early onset fetal growth restriction) trial intervention had resulted in 11 infant deaths due to lung-related problems has raised questions regarding the preclinical data that was assessed prior to the commencement of the study (9). It is suggested that a comprehensive search of the animal literature would have revealed a 2009 study in sheep showing adverse effects of sildenafil, including hypotension, reduced fetal oxygen supply, and further fetal growth reduction. Consideration of the results from this study alone could have halted the STRIDER trial in infants.
These examples highlight the patient impact and ethical issues surrounding the use of selective animal data when decisions are being made to take a compound into the clinic.
Standards and Guidelines in Preclinical Publishing
There are well-established ethical and regulatory standards requiring that research be published, regardless of study outcome (positive or negative), when human subjects have participated (10,11). To date, such standards and requirements have not been routinely applied to preclinical research, but improvements are coming to make preclinical research more transparent. For instance, there are calls for systematic reviews of animal studies to be performed before commencement of a Phase I study and their subsequent integration into the Cochrane framework for clinical studies (1). Elsewhere, an international register of preclinical trial protocols has been set up as a comprehensive listing of animal studies (12). Protocols would be registered at study inception to increase transparency, which would help to avoid duplication and reduce the risk of reporting bias by enabling comparison of the completed study with that outlined in the protocol.
Further, a number of guidelines are available on the EQUATOR Network to assist authors and medical writers with the reporting and publication of preclinical research. The main guidelines are listed in Table 1 below.
|Table 1: A selection of the key guidelines supporting the publication and reporting of preclinical and animal research from the EQUATOR Network website.|
|Animals in Research: Reporting In Vivo Experiments (ARRIVE)||www.nc3rs.org.uk/arrive-guidelines|
|Collaborative Approach to Meta-Analysis and Review of Animal Data from Experimental Studies (CAMARADES)||www.dcn.ed.ac.uk/camarades/default.htm
|A gold standard publication checklist to improve the quality of animal studies, to fully integrate the Three Rs, and to make systematic reviews more feasible||www.ncbi.nlm.nih.gov/pubmed/20507187
|A call for transparent reporting to optimize the predictive value of preclinical research||www.ncbi.nlm.nih.gov/pubmed/23060188
|A protocol format for the preparation, registration and publication of systematic reviews of animal intervention studies||https://onlinelibrary.wiley.com/doi/epdf/10.1002/ebm2.7
|Systematic reviews and meta-analysis of preclinical studies: why perform them and how to appraise them critically||www.ncbi.nlm.nih.gov/pubmed/24549183
What role for medical publication professionals?
The calls for better reporting of preclinical research present an opportunity for the medical publication profession. Given the questions over transparency and quality of some of the outputs described above, there is a clear need to improve standards in preclinical publishing. This could be facilitated through formal publication planning and delivery support in a similar way that is provided in the clinical arena. As an anecdotal observation, many publication professionals come from a research background and are familiar with basic research; they also have the necessary skills to evaluate, report, write, and systematically review research data from any biomedical source and at any stage of the development process.
Some of the reasons that preclinical research is not being published are shown in Figure 1 below.
- The time and resource barriers could be addressed by using medical writing services; but, as many companies do not currently make financial provision for preclinical publications support, funding would need to be proactively included in the budget cycle.
- A common mindset exists that preclinical publications can only be written by the researchers themselves, which would need to be altered to introduce the use of medical writing services. This mindset could be challenged by showcasing examples of pivotal clinical trials that have been successfully published in top-tier medical journals with assistance from medical writers and sharing information about the type of support that can be provided by medical writers.
- If the barrier to publishing is that journals are not interested in negative studies, there are journals that publish negative studies, such as BMJ Open Science and PLOS ONE. BMJ Open Science will publish all sound experiments irrespective of their findings, and the journal emphasizes that understanding what doesn’t work as well as what does work provides clear insights into the utility of preclinical evidence. Similarly, PLOS ONE will accept research papers reporting negative results and replication studies, which it states are all vital parts of the scientific record.
The trigger for change in the requirement to publish all preclinical data may only come through regulatory changes. If there was a mandate that no molecule goes into the clinic without registration of all studies and disclosure of results, the landscape for development of preclinical publications would likely change rather quickly.
As a call to action, publication professionals and department heads should consider approaching preclinical departments to offer the planning, budgeting, and writing skills of medical writers, particularly those with a preclinical research background, to assist in the writing and publishing of their research. Another option is to offer education of staff in preclinical departments on relevant aspects of GPP3 and ICMJE Recommendations, and perhaps medical writing tips, as a low-resource, collegial approach to improving publication practices across a company.
- Ritskes-Hoitinga M, Wever K. Improving the conduct, reporting, and appraisal of animal research. BMJ 2018; 360:j4935
- Godlee F. We need better animal research, better reported. BMJ 2018; 360:k124
- Kilkenny C, Browne WJ, Cuthill IC, Emerson M, Altman DG. Improving Bioscience Research Reporting: The ARRIVE Guidelines for Reporting Animal Research. PLoS Biology. 2010;8(6):e1000412. doi:10.1371/journal.pbio.1000412.
- Freedman LP, Cockburn IM, Simcoe TS. The Economics of Reproducibility in Preclinical Research. PLoS Biol 2015; 13(6): e1002165.
- Kilkenny C, Parsons N, Kadyszewski E, et al. Survey of the Quality of Experimental Design, Statistical Analysis and Reporting of Research Using Animals. McLeod M, ed. PLoS ONE. 2009;4(11):e7824. doi:10.1371/journal.pone.0007824.
- Wieschowski S, Chin W, Federico C, et al. Preclinical efficacy studies in investigator brochures: Do they enable risk–benefit assessment? PLOS Biology, 2018; 16 (4): e2004879 DOI: 10.1371/journal.pbio.2004879
- Cohen D. Oxford TB vaccine study calls into question selective use of animal data. BMJ 2018; 360:j5845.
- Symonds ME, Budge H. Comprehensive literature search for animal studies may have saved STRIDER trial. BMJ 2018;362:k4007
- Battisti W, Wager E, Baltzer L, et al. Good Publication Practice for Communicating Company-Sponsored Medical Research: GPP3. Ann Intern Med. 2015;163:461-464.