Colin G. Miller

	In the clinical trial arena there is a general cry for clinical trial results transparency. This is in part been thrust into the limelight by Dr. Ben Goldacre and the company AllTrials

, although there have been many other voices over recent years requesting that more (all?) data was available for dissemination. Arguably, one of the earliest champions of this space was the US Government when they enacted the requirements for the web site 

 (clincialtrials.gov). Many other regulatory agencies and Governments have joined the “bandwagon” and are now requesting clinical trial registration and trial results transparency for their country, and or region, with perhaps the European Medicines Agency (EMA) leading the charge with EudraCT

	But is this really transparency or a provision of some data to appease the public demand. What is the ultimate downstream benefit, if any, and who pays for these requirements?

	The requirements of double blind clinical trials really came into existence in the 1970’s following the thalidomide babies and other issues and that have been documented comprehensively elsewhere

. Companies developing products, both biopharmaceutical and devices, have generally considered the trials or studies proprietary and therefore confidential. However, the academic world has generally had a need to publish as part of the academic acumen and maintaining funding. Participating in clinical trials was (and is) a valuable source of funding and yet for a couple of decade’s academics were tied from publishing results of the commercially sponsored trials in peer reviewed journals, until such time as the sponsors had the product on the market and could carefully craft the major papers to match the needs from a Medical Affairs and marketing teams.

	Academically sponsored trials often undergo less operational rigor compared to commercially sponsored studies and yet can become the valuable source of data for the further understanding of how a novel product is used and its safety and efficacy in specific cohorts of the population. Again the requirements for publication are the pre-requisite of much of this research however, if the study has a negative outcome or “misses the mark” it is almost impossible to publish. This has been well documented and also the experience of this author.

	The lack of publications and the publication bias of only reporting trials with positive outcomes, has led to a misperception and misunderstanding of many therapeutic interventions and devices. It has also required many more studies to be unnecessarily conducted since many studies are repeats of other compounds etc. The ethics of repeating studies, particularly with a negative outcome, is highly questionable to say the least: Clinical trials are based on the thesis that there is equipoise at the start of the study (the new treatment in this population is not known to be better or worse than the placebo or active treatment arm). If there is sufficient data from prior trials, there is no equipoise and therefore subjects are possibly recruited into studies where there is no benefit and possibly harm.

	With this backdrop and the fact that the clinical trials (mainly positive ones) in the public arena have historically been published in peer-reviewed journals without recourse to the original records, the clamor for greater transparency from the public has been increasing. In 2007 the US Government mandated the requirement that all the results from clinical trials registered on Clincialtrials.gov must be posted in 12 months from last patient last visit

. By the end of 2015, only about one-third of the results that should be posted to Clincialtrials.gov have been posted. While the area of largest discrepancy is the results registered by academic centers, there is also a significant amount of information missing from commercially sponsored studies.

	Clincialtrials.gov is a US government mandated data repository that is run by the National Institute of Health (NIH). This has no direct connection to the US Food and Drug Administration (FDA), however, to submit any form of study to the FDA will require a submission of a Form 8674, which states that compliance with Clincialtrials.gov registration has been accomplished. This requirement extends to ex-U.S. companies that have an interest in possibly bringing a new product into the United States anytime in the future. Obviously this helps drive the global industry to ensure that all commercial trials are registered on Clincialtrials.gov. When it comes to publications, there are now over 1,000 journal editors that have signed on to state that the journal they represent will not publish clinical trial data unless it has been registered and an National Clinical Trial (NCT) number is included in the manuscript or covering letter

. This drives many academics around the world to register the clinical trial to ensure that if they have a worthwhile paper it will be accepted at a major journal for publication upon completion. Interestingly, Phase I studies do not have to be registered on Clincialtrials.gov since they are often in volunteers and Clincialtrials.gov is allowing commercial organizations to keep the early development of new compounds biologics or devices “under wraps” for a longer period of time. However, the journal requirements for an NCT number has driven many Phase I studies to be registered to ensure later publication.

	The data that is presented or “published” on Clincialtrials.gov is a very stylized set of requirements in an XML format. The published results have to match the primary and secondary objectives that are detailed in the trial registration documentation. This requires careful determination of protocol design to ensure that only the key primary and secondary objectives are registered and the matching data at the end of the study can be clearly defined and identified. Unlike most of the other repositories out there the managers of Clincialtrials.gov review the incoming results and will “scrub them” to ensure they make sense and match the primary and secondary end points.

	Historically, when a commercial trial was completed, the sponsor of the trial would clean the data, lock the database and present this to the FDA (or other regulatory agencies) and maintain very stringent control over the database, including who had access to it and what went into the public domain in the form of scientific publication’s. With the clamor for data transparency, this is no longer an option and the challenge is to have a clean data set that is available for “public dissemination” in 12 months from last patient last visit. The tie between the FDA and Clincialtrials.gov is stronger in this area, since the Clincialtrials.gov will allow a delay of results posting on the Clincialtrials.gov site if there is an application going into the agency.

	The term “publication” has taken on a second meaning in this field of data transparency: results have to be “published” on websites such as Clincialtrials.gov. This terminology does provide some confusion since, in the academic field, this clearly means the data has been presented in a peer-reviewed journal and therefore published in manuscript format. Many journals are now requesting the data sets (or condensed data sets) are available on the Internet for further public review and scrutiny.

	Just in these two paragraphs it can be seen that there are potentially three data sets, in different formats, available for evaluation and further dissemination: 1. Regulatory agency 2. Clincialtrials.gov and 3. Journal publications. Add to this the high likelihood that the data is going to other regulatory agencies and other country or regional databases (EudraCT and WHO databases), then the data maybe in a plethora of places.

	The requirement for data publication and transparency now takes on a new challenge to the highly regulated industry. This is compounded with the possibility that within a global trial with sites from many countries, there is starting to be the requirement for the data from that specific country to be available either to the ethics committee or Institutional review board or country regulators.

	So called “whistleblowers,” with the newfound power of big data analytics, can mine these disparate data sets and start identifying differences. Over time, the biopharmaceutical and devices industry will have the tools to manage this, but in the meantime there will be a period of possible non-uniformity in the data sets depending on the breakdown of the multiple datasets in the public domain.

	Data transparency is certainly increasing the burden to the biopharmaceutical industry in the way they manage “their” data and product launches. The extra cost of data transparency has to be factored in to product development and with increasing pricing pressure on from governments on keeping healthcare costs down, this becomes a further challenge to an industry that is having to rapidly change from the previous two decades of existence.

	Ultimately, the value proposition has to go back to the challenges discussed in the beginning of this paper: preventing multiple trials being run on the same cohorts of patients. Whether this is directly paid for by the sponsoring company, or indirectly through a reduction in investigator run studies in Phase IIIb and IV or the Phase IV post marketing studies will have to be evaluated over time. Certainly from an ethical perspective this is the appropriate approach and theoretically should ensure a reduction in the number of wasted studies. This is probably greatest in the area of negative studies where the results should be published and prevent other negative studies from being conducted.

	Clinical trial data transparency is here to stay and the requirements to “publish” the data are still being developed. While this is a major challenge for the biopharmaceutical and devices industry over time, it should provide a platform to ensure greater ethical clarity in our development of healthcare products and provide more confidence to the consumer about the products they are using. (At the time of writing, there is a serious trust deficit in the minds of the patient and the biopharmaceutical industry). This disruptive influence of data transparency in the status quo of product development may have much longer implications to the healthcare process and approach to the way patients are treated both in the information and the medications they receive.

	Colin G Miller, PhD, FICR, FIPEM, CSci, is CEO of 

https://eudract.ema.europa.eu/results-web/

http://www.fda.gov/AboutFDA/WhatWeDo/History/Overviews/ucm304485.htm

https://clinicaltrials.gov/ct2/about-site/history

http://www.icmje.org/recommendations/browse/publishing-and-editorial-issues/clinical-trial-registration.html

Articles

The disruptive influence of data transparency in the status quo of product development may have much longer implications to the healthcare process, and information for patients.

Colin G. Miller

Articles

Trial Transparency: PR Hype or Added Value?