When the moment strikes, researchers like Marian McDonagh, PharmD, associate director of the evidence-based practice center at the Oregon Health and Science University, Portland, turn away from the clean, neatly packaged world of published literature and toward the gray — that is, gray literature. Gray literature refers to unpublished data, such as that found in white papers, reports, and government documents.
A particularly valuable source of unpublished data are the US Food and Drug Administration (FDA) regulatory documents publically posted on the [email protected] website. Commonly referred to as FDA action packages, these documents can collectively run several hundred pages and include various sections like medical review, statistical review, pharmacology review, and risk assessment and risk mitigation review.
And to some, the level of detail is distracting. To others, however, the detail is just right.
“That’s part of where we find the value; [it’s] that there’s so much detail in there,” Dr McDonagh told Cancer Therapy Advisor. In fact, she published a study showing that these FDA documents can be a useful source of unpublished data when conducting a systematic review.1 “You can spend many hours looking at these [documents] and not find anything that’s truly stand-out special, but when you do, it’s usually pretty big,” she said.
Review of these documents has repeatedly revealed a crucial detail missing or downplayed in the published clinical trial study. For example, the published clinical trial study for platelet inhibitor prasugrel mentioned a link between only colonic neoplasms and prasugrel use, yet an FDA briefing document revealed multiple types of cancers to be linked to prasugrel use.2 This reality of reporting bias in the published literature is not new or uncommon.3
Huseyin Naci, PhD, MHS, assistant professor of health policy, London School of Economics and Political Science, United Kingdom, once used FDA documents to identify preapproval trials that led to an accelerated approval — most of which were for cancer indications — and then reviewed the confirmatory trial requirements.4