Despite the ongoing epidemic of misuse, overuse, and diversion of opioids, the US Food and Drug Administration (FDA) has set a low bar for approval of these medications over the past 20 years, new research suggests.
Results of a cross-sectional study reveal that between 1997 and 2018, the majority of approvals of opioids for the treatment of chronic pain were based on pivotal trials that lacked critical safety and efficacy data.
The study results also show that the FDA did not require manufacturers to collect safety data on tolerance, withdrawal, overdose, misuse, and diversion in any rigorous fashion.
In addition, during the study period, 17 of the 39 new drug applications (NDAs) (only one was an innovator product, known as a new molecular entity) for chronic pain were approved with an “enriched enrollment randomized withdrawal” (EERW) trial design. Such a design, in this case, allowed manufacturers to exclude 32% to 43% of the initially enrolled patients from the double-blind treatment phase.
“The question for regulators, policymakers, and others is, how did we get to a point where these approvals took place based on trials that were by design unlikely to yield some of the most important information about safety and efficacy that patients and clinicians would care about?,” study investigator G. Caleb Alexander, MD, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, told Medscape Medical News.
The study was published online September 29 in the Annals of Internal Medicine.
“Cooking the Books”
Little is known about the evidence required by the FDA for new approvals of opioid analgesics.
To characterize the quality of safety and efficacy data in NDAs for opioid analgesics approved by the FDA between 1997 and 2018, the investigators conducted the cross-sectional analysis using data from ClinicalTrials.gov, FDA reviews, and peer-reviewed publications regarding phase 3 pivotal trials.
The investigators examined the key characteristics of each NDA, including the number, size, and duration of pivotal trials, trial control groups, use of EERW, and systematically measured safety outcomes.
Results showed that most of the 48 NDAs evaluated were for new dosage forms (52.1%) or new formulations (18.8%). Only one (2.1%) was for a new molecular entity.
Of 39 NDAs approved for the treatment of chronic pain, only 21 products were supported by at least one pivotal trial. The mean duration of these 28 trials was 84 days, and they enrolled a median of 299 patients.
Results showed that for 17 of the 39 opioids approved for chronic pain, pivotal trials had an EERW design. For the latest period ― 2012 to 2018 ― trials of all eight of the approved opioids used the EERW method.
This EERW design allows the manufacturer to assess efficacy “among a subset of patients most likely to respond and least likely to have adverse effects, reducing generalizability to real-world settings,” the investigators note.
They call on the FDA to stop relying on this type of trial to assess opioid efficacy.
In an August 2020 article published in the AMA Journal of Ethics, Andrew Kolodny, MD, pointed out the pitfalls of the EERW approach. In such a study, all participants are made physiologically dependent on the opioid in a 4- to 6-week open-label phase. Only those who tolerate the drug and find it helpful are included in the randomized study. Kolodny is co-director of opioid policy research at the Heller School for Social Policy and Management, Brandeis University, Waltham, Massachusetts.
“Critics of EERW have correctly described this methodology as ‘cooking the books,’ ” Kolodny writes.
He notes that the agency’s decision to rely on EERW trials for opioids was “based on discussions at private meetings between FDA officials and pharmaceutical company executives hosted by an organization called Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT).” The 2013 meetings were reported in an article published in The Washington Post.
Little Sign of Change
Among NDAs for chronic pain, the investigators found that eight (20.5%) included pooled safety reviews that reported systematic assessment of diversion. Seven (17.9%) reported systematic measurement of nonmedical use, and 15 (38.5%) assessed incident tolerance.
The study revealed that 8 of 9 products that were approved for acute pain were supported by at least one pivotal trial. The median duration of these 19 trials was 1 day, and they enrolled a median of 329 patients.
The investigators note that the findings “underscore the evidence gaps that have limited clinicians’ and patients’ understanding and appreciation of the inherent risks of prescription opioid analgesics.”
Alexander, who has been an FDA advisory committee chairman and currently serves as a consultant to plaintiffs who are suing opioid manufacturers in federal multidistrict litigation, said the study “is a story about missed opportunities to improve the safety and to improve the regulatory review of these products.”
Coinvestigator Peter Lurie, MD, who was an official at the FDA from 2009 to 2017, said, “There’s not a lot of signs that things are changing” at the agency.
The study shows that the FDA has “accepted what the companies have been presenting,” said Lurie, who is president of the Center for Science in the Public Interest.
The FDA “absolutely has the authority” to require manufacturers to undertake more rigorous trials, but agency culture keeps it from making such demands, especially if doing so means a new applicant might have to conduct trials that weren’t previously required, Lurie told Medscape Medical News.
“FDA is pretty rigorous about trying to establish a level playing field. That’s a virtuous thing, but it becomes problematic when that prevents change,” said Lurie.
The most recent FDA guidance to manufacturers, issued in 2019, does not provide advice on criteria for endpoints, study duration, or which populations are most likely to benefit from opioid treatment. The agency also does not require drug manufacturers to formally collect data on safety, tolerance, overdose symptoms, or constipation.
The guidance does suggest that the agency would likely take into account public health considerations when evaluating opioids, such as the risk to the overall population for overdose and diversion.
Commenting on the study for Medscape Medical News, Kolodny said that as far as he is aware, “this is the first scientific publication in a peer-reviewed journal demonstrating clearly the problems with FDA’s opioid approval process.”
The article offers “overwhelming evidence that they are improperly approving the most dangerous medications ― medications that killed more people than any other medication on the market,” added Kolodny, who is also president of Physicians for Responsible Opioid Prescribing.
Asked to respond to the study findings, FDA spokesperson Charles Kohler said the agency “does not comment on specific studies but evaluates them as part of the body of evidence to further our understanding about a particular issue and assist in our mission to protect public health.”
Ann Intern Med. Published online September 29, 2020. Abstract