How are pharma companies in Europe doing on clinical trial transparency?

European pharma lobby group EFPIA recently published a survey of its member companies focusing on trial transparency. This blog provides an overview of the key findings, with some comments.

Pharma critics routinely slam the sector for opacity. Industry argues that its practices have significantly improved, and that such criticisms are no longer valid. So what does the survey show?

What is the new survey about?

EFPIA’s survey focuses on the Principles for Responsible Clinical Trial Data Sharing.

This 2014 document sets out voluntary commitments to transparency made by major pharma companies belonging to EFPIA and/or its American counterpart, PhRMA. The Principles themselves focus heavily on sharing Individual Patient Data (IPD) and Clinical Study Reports (CSRs), and the survey reflects that.

The survey was run during July-September 2020, and all 34 EPFIA member companies responded, as did one smaller pharma company. EFPIA then encouraged companies with policy gaps to up their game:

“After receiving the initial response, members whose answers revealed a level of non-compliance in one or several Principles were contacted individually to clarify the requirements and clarify potential misunderstandings. The data shown in this report include results including agreed policy amendments… After the survey, clarifications have led individual EFPIA member companies to put policies and corrective measures in place to further improve their compliance with the Principles.”

Protocols, CSR synopses, and IPD

Protocol sharing. All companies reported sharing study protocols from clinical trials for medicines and indications with marketing authorisation in EU and US.

CSR synopses. All companies reporting sharing CSR synopses after EU and US market approval. (The survey question did not cover sharing of full CSRs.) Going beyond their formal commitments, just over half of companies reported additionally making CSR synopses available of studies with investigational products that have not yet been approved in any country, but it is unclear whether that means synopses of all studies, or only of some studies. While 80% of companies reported making redactions to the synopses they share, this may be less problematic than it sounds as such redactions are often minimal in scope.

IPD sharing. With one exception, companies reported sharing (anonymised) patient-level data from clinical trials in patients for medicines and indications with market authorisation in EU and US.

TranspariMED comment:

The commitment to IPD sharing is highly welcome, as voluntary disclosure by companies is currently the only realistic way that independent researchers can access these data. Equally, the commitment to sharing full protocols – even if only after marketing authorisation – is good news for the research community. Assuming that companies live up to their commitments in practice, something that future research should be able to verify, this glass is significantly more than half full.

Reporting results on trial registries

Registry reporting. Companies reported “100% compliance with the legal requirements” of the American ClinicalTrials.gov and European EudraCT trial registries.

TranspariMED comment:

Large European and American pharma companies have an exceptionally strong track record of making clinical trial results public on trial registries if and when that is required by law. The EU Trials Tracker and the FDAAA Trials Tracker, both of which are regularly updated, clearly show this. This is huge progress compared to only five years ago.

However, the WHO recommends that the results of all trials should be made public on trial registries, even when there is no legal requirement to do so. The industry Principles do not go that far, and the EFPIA survey did not cover this issue.

(Footnote: A limitation of the FDAAA tracker is that it only shows trials from 2017 onwards. It therefore misses trials conducted 2007-2017 that are also subject to U.S. reporting laws. So we cannot be sure about companies’ reporting performance for those older trials, but unless proven otherwise, it seems reasonable to assume that large pharma players are broadly compliant.)

Publication in journals

The industry’s 2014 Principles say the following about publication in journals:

“All company-sponsored clinical trials should be considered for publication in the scientific literature irrespective of whether the results of the sponsors’ clinical trials are positive or negative. At a minimum, results from all phase 3 clinical trials and any clinical trial results of significant medical importance [sic] should be submitted for publication. This commitment also pertains to investigational medicines whose development programs have been discontinued.”

All companies stated that they have “committed to publication of clinical trial results in the scientific literature irrespective of whether the results of the clinical trials are positive or negative,” and with one single exception all companies also replied that they had made a commitment “to publish all phase 3 trial results and all trial data of significant medical importance”.

In addition, 80% of companies stated that they “encourage [sic] publication of clinical trial data and results regardless of clinical phase and marketing authorization status.”

TranspariMED comment:

Under Article 36 of the Declaration of Helsinki, there is an ethical obligation to make the results of all clinical trials public. WHO best practices also require publication of all trial results.

Industry’s current voluntary commitments fall significantly short of these benchmarks, essentially covering only “all phase 3 clinical trials” and setting no timeframe for publication.

At the same time, studies consistently show that the results of industry trials are more likely to be made public than those of academic trials.

(Footnote: Many academic institutions fail to keep trial registry data up to date, and the people doing this type of meta-research usually fail to validate their findings with the sponsors who ran the trials. So academic trials that never got off the ground in the first place might get miscounted as “unreported”, making industry vs academia comparisons in the existing literature somewhat unreliable. On balance, while industry’s record is far from perfect, and it’s hard to be certain, it appears that overall industry has a better publication record than academia.)

Where do we go from here?

Improvements in clinical trial transparency over the past decades have been stunning. Remember that 25 years ago, trial registries did not even exist.

Due to laws, regulations, rising public expectations, and efforts by large industry players to safeguard their reputation, we have come a very long way. More broadly, the internet has enabled a degree of transparency that would have been literally impossible in the past.

Thanks to laws and regulations setting out strict rules for companies wanting to bring medicines onto the market, the pharma industry has arguably come further than academia in terms of trial transparency. It has certainly come lightyears further than the medical device industry, which in Europe remains woefully under-regulated.

While it is completely unrealistic to expect industry to act against its own interests and those of its shareholders – there is no substitute for effective regulation – those interests do include reputation management. A voluntary commitment by EFPIA and PhRMA members to make all Phase 2-4 trial results public on trial registries within a year would go a long way in that regard.

In parallel, we need to urge legislators to make trial reporting in line with WHO best practices a legal requirement for all trials, no matter who is running them – as the UK seems likely to do this year.

New legislation is the only way to set binding minimum transparency standards for all players, including for medical device manufacturers and academic institutions.

After all, from a patient perspective, it really does not matter who ran the clinical trial that your future survival might hinge on. What matters is that the results are made public, rapidly and in full.