In a major policy shift, the Department of Health and Human Services issued a Notice of Proposed Rulemaking (NPRM) requiring researchers to submit results from trials of investigational compounds that fail to meet clinical or safety endpoints – findings that often remain hidden under current practice. The proposed regulations would make nearly all clinical trials subject to mandatory reporting requirements under Title VIII of the Food and Drug Administration Amendments Act of 2007 (FDAAA) by clarifying that researchers must register trials, including early stage and expanded access studies, and submit summary results to ClinicalTrials.gov, in a standard format, no matter what the findings. The publicly accessible database is operated by the National Library of Medicine (NLM), part of the National Institutes of Health (NIH).
A primary goal of the proposal is to reduce unnecessary duplication of trials, particularly when results suggest a candidate may be unsafe or ineffective, by providing greater and timelier transparency to researchers and third parties.
Francis Collins, NIH director, and numerous insiders at the NIH long have pushed for greater transparency in reporting clinical trials outcomes. In recent months, increased pressure from Collins and his team to reform the clinical trials process broadly drove the timing of policy change, according to Kathy Hudson, deputy director for science, outreach and policy at the NIH.
"We want to extract as much information and knowledge as we can from every penny that we're investing in clinical trials, and do so in the most effective and ethical way," Hudson told Medical Device Daily's sister publication BioWorld Today.
Deborah Zarin, director of ClinicalTrials.gov in the NLM's Lister Hill National Center for Biomedical Communications, confirmed that the process was driven "at the highest levels of the NIH."
The HHS proposed rule would clarify the requirements about public sharing of summary data from clinical trials of drugs and devices regulated by the FDA, directing researchers to include baseline characteristics of participants, primary and secondary outcome results, and information about adverse events. With few exceptions, the proposed rule would require researchers to submit results to ClinicalTrials.gov within one year after collection of primary outcomes data are completed.
The HHS NPRM was published in Friday's Federal Register, with comments due in 90 days.
The NIH took the HHS proposal a step further by releasing a draft NIH policy on the same day to apply the data reporting requirements to all interventional clinical trials that it funds. Hudson said the agency is working with NIH-supported researchers and institutions to ensure the responsibilities in the proposed policy on dissemination of NIH-funded clinical trial information are understood and to remove any unanticipated obstacles.
"There's a whole series of reforms that we're developing that focus on clinical trials and human subject research," Hudson explained. "This is a realty important centerpiece of that."
The combination of the NPRM and the NIH policy will ensure that most trials conducted at academic medical centers, will fall under at least one of the programs, Zarin added.
"As a result, it will, I hope, induce those institutions to develop processes and procedures to both facilitate and support investigators but also to incentivize investigators to do a good job of both registering and results reporting," she said.
'We don't want to repeat our mistakes'
Currently, ClinicalTrials.gov contains registration information for more than 178,000 clinical trials and summary results for more than 15,000, but these numbers include trials that are not subject to FDAAA, including phase I trials of drugs and biologics and small feasibility studies of medical devices. The proposed rule, developed by NIH in coordination with the FDA, specifies how data collected and analyzed in a clinical trial would be required to be submitted to ClinicalTrials.gov. to meet requirements established by FDAAA.
"The law was written in a way that permitted the Secretary to make some decisions through the rulemaking process," Hudson explained. "The law laid down a floor of requiring registration and results reporting for trials of approved, FDA-regulated products."
However, the law "left open the door" but did not specifically require the HHS Secretary to include products that were not eventually approved by the FDA.
"That's vital," Hudson insisted, "because those are the trials that have a higher likelihood of having failed, and we don't want to repeat our mistakes."
Trials that pass muster with the FDA meet a higher threshold of scrutiny that gives an incomplete picture of the overall development process for drugs and devices, she added. The proposed rule would ensure that all information, "not just the good stuff," gets out into the public eye. "If nobody ever sees the data, we never learn from it," Hudson said.
Notable changes from current practice outlined in the proposed rule include:
• A streamlined approach to determine which trials are subject to the proposed regulations and who is responsible for submitting required information.
• Expansion of the set of trials that are subject to summary results reporting to include trials of products that are not approved.
• Additional data elements that must be provided at the time of registration, or not later than 21 days after the first participant is enrolled, and at the time results are submitted, generally within 12 months of the study's completion.
• Clarified procedures for delaying submission of results when studying an unapproved, unlicensed, or uncleared product or a new use of a previously approved, licensed or cleared product and for requesting extensions to the deadline for submitting results.
• More rapid updating of several data elements to help ensure that users of ClinicalTrials.gov have access to accurate, up-to-date information about important aspects of a clinical trial.
• Procedures for timely corrections to any errors discovered by the responsible party or by the agency as it processes submissions prior to posting.
The changes are designed to provide improved access to clinical trial information by researchers, third parties such as journal editors and readers, and patients and family members, Hudson said. The proposed rule does not affect requirements for the design or conduct of trials or for the data that must be collected during trials.
NIH seeking to 'make this as easy as possible'
Zarin predicted few immediate changes in the look of the ClinicalTrials.gov site, but she said the rate of submissions could more than double once the new rule and NIH policy are in place.
"It will become more imperative for people who want to find all of the publicly available evidence about an intervention or clinical situation to look in ClinicalTrials.gov as well as to look in PubMed," Zarin told BioWorld Today. At any given time, she said, an estimated 50% to 60% of the results entries posted on ClinicalTrials.gov do not have associated publications, so the website is the only "searchable, accessible place to find results for those studies."
For systematic reviewers, ClinicalTrials.gov is an essential complement to PubMed when conducting a rigorous search of scientific evidence to answer a given hypothesis, "and as we get more and more studies, that will become even more important," Zarin said.
Entries in ClinicalTrials.gov often are more complete than published journal data, she added, in terms of basic facts about pre-specified outcome measures, results and adverse events.
The NIH is improving the usability of the data submission site and staffing up to offer one-on-one assistance to investigators who will be entering results for the first time. In the long run, higher quality submissions will likely prove less burdensome to investigators because they'll improve the overall quality of trial design.
"What we hear from our heaviest users, who oversee large groups of studies at pharmaceutical companies or data coordinating centers at academic institutions, is that, when they realize what they have to do to submit results, it circles back and improves the writing of the protocol for the next study," Zarin said. "It improves tracking of certain items that they know they're going to need to report. That kind of feedback loop, with the higher volume of trials that institutions need to report, might lead to efficiencies and, we like to believe, actual improvements in methodology."
The NIH hopes to have a final rule in place by the end of 2015.