Full article reprinted from "The Gray Sheet" - September 29, 2008
As key portions of the government's new device clinical trial registry requirements went into effect last week, Medtronic clinical research director Lisa Griffin Vincent offered advice on complying with some of the thornier aspects of the law. Find out what she had to say...
Full article reprinted from "The Gray Sheet" - September 29, 2008
As key portions of the government's new device clinical trial registry requirements went into effect last week, Medtronic clinical research director Lisa Griffin Vincent offered advice on complying with some of the thornier aspects of the law.
"It's a challenging law to interpret," Vincent said.
Stiff penalties for non-compliance, including criminal sanctions and civil money penalties up to $10,000 per day, make it imperative that device companies understand their obligations under the trial registry program, she and other speakers said at the AdvaMed 2008 MedTech Conference in Washington, D.C. last week.
The program for registering device clinical trials and reporting their results was created in last year's Food and Drug Administration Amendments Act (FDAAA). Congress' intent was to make full information about clinical trials available to the public, but regulators and industry alike have spent the last year scrambling to figure out exactly how to fulfill their obligations under the statute (1"The Gray Sheet" June 23, 2008, p. 3).
Registration Hazard: Estimated Trial Finish Date
Beginning Sept. 27, sponsors or principal investigators of all device clinical trials falling under the law are required to submit information on their studies to the National Institutes of Health's Clinical Trials Data Bank (see sidebar: "2Clinical Trial Registry Data Bank Key Deadlines").
This includes initial registration information, including the trial's purpose, start and finish dates, endpoints and eligibility criteria, as well as basic results information once the trial is completed.
Companies have to submit the registration information within 21 days after the first patient is enrolled in a trial. However, unless a company indicates it wants this information made public immediately, NIH will keep it hidden until after the device is approved or cleared for marketing by FDA, at which point NIH will post the information on its ClinicalTrials.gov Web site. Posting will occur within 30 days after FDA's decision.
Companies should pay especially close attention to the estimated trial completion date they give NIH in their registration submission, Vincent said. It is this date that will be used to set the deadline for submitting the eventual study results to the data bank.
The statute states that results have to be submitted within a year of either the estimated or actual completion date, whichever comes first. It defines trial completion as the date on which data for the primary outcome was collected on the last patient.
Anticipating that date "is very difficult, particularly in complex trials," said Vincent.
"Because this date triggers when your results will be reported, you need to go in and update this as often as you can when you reasonably think that date is going to be delayed."
Sponsors should also go back to the registration Web site when the study is finished to insert the actual completion date, Vincent recommended.
Pointing to a potentially more serious problem, she noted that the law's definition of a trial's completion date differs from traditional understanding, and could undermine trial integrity. Typically, she said, sponsors and researchers think of the completion date as when the database has been "locked down," and all information has been gathered and checked for all outcomes (see sidebar: "3What Trials Does The Law Apply To?").
"In many trials, you might even have multiple primary outcomes, two or three oftentimes - an efficacy and a safety - and then you might have some secondary outcomes" that require the trial to continue for many months after the completion date defined in the law, she said.
The issue becomes particularly troublesome in the case of a blinded study that needs to continue for secondary and other endpoints, "including perhaps long-term safety," Vincent said. "That can jeopardize the integrity of the overall study results," she observed, "and that isn't in the best interests of the public or the patient."
Device companies have been working to explain the issues surrounding the definition of the completion date to NIH and FDA regulators, Vincent said. "I'm still hopeful we might be able to come to some different agreement."
Coordinating Conflicting Registration Obligations
Coordinating FDAAA compliance with adherence to other important clinical trial registration policies will be another major challenge for trial sponsors, Vincent said.
For example, while FDAAA exempts feasibility and observational trials, the International Committee of Medical Journal Editors requires that such studies be registered in a database such as ClinicalTrials.gov as a condition for article publication.
"If you hope to publish and not jeopardize the potential to publish, you'll want to pay attention to the ICMJE's more conservative requirements," Vincent said.
Further, where FDAAA gives sponsors 21 days after the first patient enrollment to register the trial, ICMJE requires registration prior to the first enrollment. For all practical purposes, therefore, device companies need to register their trials in the NIH data bank in accordance with the ICMJE rules.
"What we've seen is that the journal editors are holding fast to this rule," said Vincent. "So if you don't register prior to that first enrollment, you still can submit your article for publication, and it's up to the individual journal editor. But we've seen a number of examples where they've just automatically turned it down; they won't even review it for potential publication."
In a particularly tricky aspect of compliance, FDAAA requires delayed public disclosure of registration information until after a new device is approved by FDA, while ICMJE forbids such delay.
The way the data bank has been designed, it is up to companies to first check "yes" if the trial being registered is for a new, never-approved or -cleared device, which will automatically trigger delayed disclosure. Then, when the device is approved, the firms must go back and check "no" to have the registration information released to the public.
The dilemma is that by checking "yes" and thereby providing truthful information under the act, a company loses its chance to have trial results accepted for publication in a journal.
Medtronic and other firms have informed NIH and FDA about this Catch-22 and are hoping the issue can be ironed out as implementation of the data bank requirements continues.
In the meantime, companies that hope to publish may want to not check the box at all at initial registration, and simply leave it blank, Goodwin Procter attorney Mark Heller suggested.
But he added that advice from NIH has so far been unclear. "It's a concern," he said.
- Mary Houghton
Sign up for your 30-day, risk-free trial of "The Gray Sheet" today.
"The Gray Sheet" gives you 51 issues per year filled with useful articles that will help you meet your business and regulatory objectives.
.jpg)
