In a surprise move, the US Food and Drug Administration (FDA) has released a second draft guidance document regarding the distribution of scientific and medical publications, this time focusing on publications intended to convey risk information about drugs and biological products.
The guidance document is the second so-called "reprint" guidance released this year.
In February 2014, FDA released a similar guidance document, Distributing Scientific and Medical Publications on Unapproved New Uses - Recommended Practices, which lays out FDA's expectations on how companies can distribute literature regarding their products, including off-label use. In short, literature should be from reputable sources, not have been supported by the company, be scientifically sound and from an "adequate and well-controlled clinical investigation," and be accompanied by a bibliography containing all other known studies on the drug (including contradictory ones).
The guidance, however, has come under fire from some groups, including the Washington Legal Foundation (WLF), which have said FDA's guidance violates an existing court injunction as well as companies' First Amendment rights to free speech.
Both guidances are surprising—literally—because they were not included among the Center for Drug Evaluation and Research's (CDER) 2014 guidance agenda, which typically outlines all documents the agency is considering for release during the calendar year.
On Safety and Efficacy
The newest draft guidance, Distributing Scientific and Medical Publications on Risk Information for Approved Prescription Drugs and Biological Products—Recommended Practices, specifically concerns safety and risk information, FDA noted.
As explained by FDA:
"FDA recognizes that the safety profile of a drug evolves throughout its lifecycle as the extent of exposure to the product increases and that it can be helpful for health care practitioners to receive significant new risk information about an approved product in a timely manner. FDA anticipates that the earliest distribution of new risk information will generally involve distribution of recently published studies, as opposed to textbooks or clinical practice guidelines. Accordingly, FDA is providing guidance for firms that choose to distribute new risk information in the form of a reprint or digital copy of a published study."
While companies often extoll the benefits of drugs to medical professionals, they're often just as concerned about communicating the safety profile of their drug as well. As FDA notes in its guidance, that can include notifying medical professionals that a study had found a drug to have fewer adverse events than previously believed, or that the events are less serious than previously known.
Differing Standards between Safety and Efficacy
But effectiveness and safety are treated very differently by FDA, which has differing standards governing how information on each comes to be included in a drug's label.
For example, while efficacy data is generated from "adequate and well-controlled studies," safety data is instead generated from a wide range of data, including preclinical data.
"The [safety] assessment is based largely on observations of adverse events from studies intended to assess effectiveness and includes all patients exposed to the investigational drug during its development," FDA wrote in the draft guidance. Those observations are then turned into a "coherent assessment" of the risks and their association frequency and severity, FDA explained.
"The threshold for inclusion of an adverse event in approved human drug labeling is a determination that 'there is some basis to believe there is a causal relationship between the drug and the occurrence of the adverse event,'” FDA added, citing 21 CFR 201.57(c)(7).
FDA on Data Quality
FDA's guidance document is split up into two sections, one reflecting premarket/postmarket data quality concerns, and the other on the criteria for determining whether new safety data may be distributed to medical professionals.
For premarket scientific and medical literature reports, FDA says it will give greater credence to data that comes from placebo-controlled and dose-response studies, and particularly those involving large numbers of patients.
In addition, FDA said it will also weight more heavily adverse event reports that happen in tandem with one another, and in particular events which rarely occur spontaneously in populations not exposed to drugs. Conversely, adverse event reports that occur in either limited number or in non-drug-exposed populations "provide more ambiguous evidence of a causal relationship," regulators said.
Once approved, a drug's label may be updated based on data from postmarketing pharmacovigilance activities, including new spontaneous reports (such as through MedWatch) or epidemiological studies run by the manufacturer or others.
But the meat of the guidance comes in the last two pages, when FDA outlines its criteria for what is appropriate to distribute to medical professionals.
Before distributing safety data, companies should "Carefully consider the reliability and persuasiveness of the data," FDA writes. This will depend heavily on the data supporting the side effect, as well as the side effect itself, and in some cases "even one occurrence of an adverse event" is cause of concern, whereas in others it might simply be statistical noise better attributed to outside factors.
In other cases, if existing data has already shown a strong correlation between a drug and an effect, the evidence needed to overturn that consensus must be even stronger, FDA said.
The agency also said it "does not intend to object to the distribution of new risk information that rebuts, mitigates, or refines risk information in the approved labeling, and is distributed by a firm in the form of a reprint or digital copy of a published study," with the added condition that the study must meet pre-defined criteria for suitability, similar to those contained within its February 2014 guidance.
Key criteria include:
- The study must be methodologically sound, well designed, and acknowledge its own biases and/or shortcomings.
- For data intended to reverse an existing consensus or understanding, FDA says the data must be "at least as persuasive as the data sources that underlie the existing risk assessment of causality, severity and/or incidence."
- A study's conclusions should be a "fair characterization of all relevant information in the safety database, including contrary or otherwise inconsistent findings."
- The study or analysis should be "published in an independent, peer-reviewed journal."
- The distributed copy of the study or analysis should include a coversheet disclosing the data's shortcomings, that the data may be inconsistent with the FDA-approved label, that FDA has not yet reviewed the data, and any conflicts of interests between the company and the study.
- The reprint or copy should be accompanied by the FDA-approved labeling, be separate from any promotional materials, and all statements made by a company representative in reference to the material should be "constituent with its content" and the content of the disclosure cover sheet.
The guidance has not yet been published in the Federal Register for comment.
Distributing Scientific and Medical Publications on Risk Information for Approved Prescription Drugs and Biological Products—Recommended Practices