The pharmaceutical company Pfizer spent more than $33 million on television advertisements for Lyrica, its blockbuster drug used to treat diabetic nerve pain and fibromyalgia, in September 2017. The drug consistently tops the lists of the most expensive marketing campaigns — the company spent $23.1 million in March 2018, and $19.5 million in April 2018. The ad spots are ubiquitous, showing pain sidelining people from time with family and friends, until Lyrica helps ease their symptoms.
The commercials end with the familiar roll call of potential side effects and risks of the drug: dizziness, swollen hands, weight gain. One of the side effects is blurred vision — a complaint echoed on message boards online.
When the Food and Drug Administration approved Lyrica in 2004, it did so on the condition that Pfizer continue to study the drug — and specifically, conduct a study of Lyrica’s effect on vision. That study was supposed to be completed in 2009. But according to the FDA’s own public database, nearly 10 years after the deadline, it remains delayed, with no robust data on the relationship between the drug and eyesight available.
It is not at all uncommon for drugs like Lyrica to make their way into medicine cabinets across the country before drug manufacturers or the agency that regulates them have all of the information about them. The FDA can approve a drug with caveats, like requests for additional research on niggling issues about safety and efficacy the approval process couldn’t nail down. Often, the federal agency asks drug companies to conduct longitudinal studies, sometimes stretching up to 10 years, to understand the impact of long-term use of the drug. The arrangement — known as post-marketing research — allows drugs to get to market faster and more cheaply, while still ensuring that all safety questions eventually get answered.
But the success of that system hinges on the FDA holding the companies accountable for completing the required studies. Otherwise, the requests are toothless, the data non-existent, and the risks still unknown. And according to experts, internal government reviews, and an analysis of publicly available data on post-marketing trials, the FDA has a poor track record when it comes to following up. Required studies are often delayed, sometimes for years, while the drugs in question remain on the market all the while. Even though the FDA has the authority to penalize companies for falling behind on their required research, pharmaceutical industry advisers say that the agency has never forced the withdrawal of a drug, or even fined a company, because it failed to perform a post-marketing commitment.
“It’s never happened,” says David Clissold, an attorney in Washington, D.C. who advises pharmaceutical companies on regulatory issues.
The promise and peril of post-marketing research dates back decades. When the Austrian researchers Oleh Hornykiewicz and Walther Birkmayer injected the compound levodopa into patients suffering from Parkinson’s disease in 1961, it represented the first major medical breakthrough in the treatment of the disorder, which, until then, was treated with rest, stress reduction, and dietary supplements.
A few years prior to the levodopa breakthrough, scientists discovered that an absence of dopamine in the striatum of the brain causes the symptoms associated with Parkinson’s, which range from tremors and involuntary movement to impaired speech. Levodopa, or l-dopa, was ultimately shown to increase dopamine levels and ease those impairments.
Though the drug was incredibly effective, the agency responsible for approving it for commercial use — the FDA — was in a bind. L-dopa helped return physical function to suffering patients, but it wasn’t a cure. Parkinson’s patients would presumably need to be taking the drug for years, even decades, but none of the research had looked to see if long-term use was dangerous. What’s more, the pharmaceutical company sponsoring the drug hadn’t even completed toxicity studies in animals, let alone humans. They knew the drug worked, but a crystal ball could have better predicted the side effects than those scientists.
Despite its potential dangers, doctors, patients, and their families were clamoring for the drug. So the FDA crafted a compromise: It would approve the drug and allow patients to take it, but only if the pharmaceutical company that manufactured it agreed to do more studies once it was already on the market. So-called post-approval studies had been done before, but usually on the manufacturer’s own accord, out of a desire to better understand their drug for themselves. L-dopa marked the first time the FDA asked for such research itself.
Today, post-marketing research is used in the majority of new drug approvals. In 2015, for example, approximately 80 percent of the 45 novel drugs approved by the FDA hit pharmacy shelves with some sort of post-marketing study requirements. On average, each new drug was saddled with about seven additional studies. That’s more than double the number of studies from nearly a decade ago, when new legislation, the Food and Drug Administration Amendments Act of 2007 (FDAAA), first gave the agency more power to compel companies to perform such studies, and to punish those companies if they fail to comply.
Given this legislation, which went on the books in 2007, some advocates for the post-marketing process argue that the increase is to be expected. “[The] FDA has a lot more authority now to require those studies, and they are exercising it,” says Clissold.
A spokesperson for the FDA agreed. “Because the FDA has received its [post-marketing requirement] authorities (statutory and regulatory) over time, it is not unexpected that more [post-marketing requirements] are issued now than were issued 20 years ago,” FDA spokesperson Lauren Smith Dyer wrote in an email message. Data in the FDA’s 2016 annual Federal Register report on the Performance of Drug and Biologics Firms in Conducting Post-marketing Requirements and Commitments supports her claim. The number of these so-called PMRs has indeed remained fairly constant from 2009 to 2014, since the FDA received broader authority, according to that report. There was a sharp jump in the number of post-marketing trials required by the FDA immediately after 2007 and the number has remained at this new normal ever since.
Still, critics of this system worry that the practice places too much risk on ordinary consumers. “Mandatory post-market study authority has, at some level, been a substitute for doing studies before the drug has been approved,” says Sidney Wolfe, senior adviser in the Health Research Group at Public Citizen. He argues that the increasing use of the post-marketing arrangement means that fewer and less-extensive clinical trials are conducted before a drug gets to market.
Wolfe and other critics also suggest that the FDA seems to be paying scant attention to whether or not drug companies are actually performing the studies they promise to do. An analysis of information available in the public FDA post-marketing trial database supports this claim. According to the database, last updated on July 31, 2018, the agency issued 280 post-marketing commitments in 2011 and 2012. Of those, 103 were scheduled to be completed by January 2018, but only about a third — 37 total — have been marked as fulfilled.
AAfter a potential new drug is discovered in the lab, a pharmaceutical company tests the drugs on bacteria and animals to gather preliminary data on safety and efficacy. Next come clinical trials in humans — in groups of gradually increasing size — to determine if the drug is safe and if it works in people. With all that data in hand, the company then submits its drug for approval to the FDA. In the process of approving a drug, the FDA can require a pharmaceutical company to conduct additional studies once the drug is on the market. In 2015, for example, it issued a total of almost 250 study requirements.
Post-marketing studies like these are, in many cases, necessary and important because testing done by pharmaceutical companies can’t capture all the potential issues with a drug. “When you test the drug initially, it’s in very few people and in a shorter term,” says David Gortler, a former FDA medical officer and current associate professor at Georgetown. “Often, [in real life] people take the drug for a longer term.” A side effect or adverse event might therefore occur in only one in thousands or hundreds of thousands of patients — too small a number to register in earlier and smaller clinical trials. That’s why successive research trials overseen by the FDA tend to narrow in on the subpopulations who might be at risk. The elderly, for example, are typically excluded from early stage clinical studies, so approval of geriatric drugs and associated post-marketing trials are often essential for getting much-needed drugs into older hands.
In the process of approving a drug, the FDA can make a pharmaceutical company conduct additional studies once the drug is on the market. Sometimes these studies (post-marketing “commitments”) are the result of the FDA asking the company to agree to do these studies; other times, the studies (post-marketing “requirements”) are dictated by law. In these latter cases, the FDA can force the company to comply with the mandate for further research, whether companies want to or not.
It sounds like a winning arrangement, but not everyone thinks the system benefits patients — and even those who do think the post-marketing system is in need of refinement. “It’s always a balance,” says Ken Kaitin, director of the Tufts Center for the Study of Drug Development. On one extreme, one might say “‘Let’s keep all drugs in development until we know absolutely everything,’” Kaitin says. The other extreme, he suggested, is to put a drug on the market with only cursory testing and let patients decide for themselves if they want to take it.
The best option, according to Kaitin, is somewhere in the middle. “We have to use the best information on safety and efficacy of the drug, and if it looks like it will provide benefit and not cause harm, approve it with the condition that more studies are done,” he says.
In the middle of the AIDS epidemic, the FDA relied heavily on post-marketing requirements and so-called accelerated approvals so that experimental drugs could get into the hands of patients sooner. Though the side effects may have been severe, many patients, their doctors, and, after years of lobbying and public protest, the FDA, believed that the certainty of death meant these patients should be allowed to try to combat their illness as soon and as aggressively as possible, says pharmaceutical expert Larry Sasich.
The upside of post-marketing studies is calculated on the assumption that the FDA is actually keeping careful track of these post-marketing commitments — including making sure that the companies perform the studies that they promise to do and penalizing them if they fail to meet their commitments. But an examination of agency records suggests that’s not the case, meaning that cost-benefit analysis of post-marketing studies is profoundly skewed. Research shows that the FDA has issued two formal letters and one untitled letter — letters that note an infraction without rising to the formal level of a warning letter — using their authority under the FDAAA about failures to fulfill a post-marketing requirement. Any other letters that the FDA may have sent to drug companies under FDAAA, however, are not publicly available.
Merck Pharmaceuticals was the recipient of the first such letter in 2012 when the FDA reprimanded the company for its failure to complete a required post-marketing study on the risk of pancreatitis associated with its type II diabetes drug Januvia. Januvia was of special concern because similar chemical concoctions were determined to increase the risk of acute pancreatitis among patients taking the drug, as well as non-fatal side effects like severe muscular and joint pain.
To deal with the fallout, Merck sent a new protocol to the FDA and ultimately executed the study. The company submitted the results to the FDA in December 2012 and February 2013, according to a statement from a Merck representative.
In the second case, the FDA issued an untitled letter to the company Shionogi, Inc. over failure to comply with post-marketing requirements for the painkiller Rybix. Claiming it was too expensive to do the study, it elected to pull the drug from the market on its own. In the third case, the FDA issued a warning letter in June to Perrigo for not complying with the post-marketing requirements for its testosterone gel.
The FDA’s reluctance to use its authority is no surprise, according to Kaitin. “When [the FDA’s new enforcement powers] first came out I thought, that is never going to happen,” he says. “The culture of the agency over the last few decades has been more towards a partner with industry to make important drugs available, and not to be an adversary. For them to have a policing role creates more of an adversary relationship with industry.”
Even when legally required to send a letter to pharmaceutical companies that aren’t holding up their end of the bargain, the FDA rarely does so. A certain subset of post-marketing studies are dictated by a law called PREA — the Pediatric Research Equity Act — that ensures that many drugs get tested on children, not just adults. When a drug company fails to perform a required pediatric trial on time, if there’s not a valid reason for the delay, the agency is required by law to issue a noncompliance letter.
Before PREA and other accompanying legislation, most drugs for children were prescribed “off-label,” where doctors used adult information to scientifically guestimate if and how much of a certain drug to give a child. But children are not small adults, and that extrapolation sometimes leads to devastating consequences — like the antibiotics ciprofloxacin and chloramphenicol, which have both been found to be toxic to children. The introduction of PREA was intended to close that “guesstimation” gap. However, in keeping with the data on post-marketing studies at large, evidence suggests pharmaceutical companies are often late or non-compliant with their PREA obligations, too.
Of the 140 PREA-mandated trials in the FDA publicly available database (last updated July 31, 2018) that were scheduled to be completed by January 2018, only 42 — or 30 percent — are marked as fulfilled. Thirty-eight of the trials are noted as delayed in the database, representing a total of 27 different drugs. The FDA has issued non-compliance letters for just 15 of those drugs. In one case, the company Pfizer was required by PREA laws to complete two trials of its anti-fungal drug Vfend by 2003. Fifteen years later, the status of the post-marketing research remains “delayed,” but no non-compliance letter has been issued. That drug also remains on the market.
“The PREA PMR/PMC database does not necessarily reflect the most up-to-date information,” wrote FDA spokesperson Dyer in response to an inquiry about the delayed studies. “The FDA issues non-compliance letters to sponsors who have failed to submit their pediatric assessments required under PREA by the final due date, have failed to seek or obtain a deferral or deferral extension, or have failed to request approval for a required pediatric formulation. The current listing of FDA non-compliance letters is up-to-date.”
Even if the FDA were inclined to follow up on post-marketing commitments, it’s far from clear if the agency could do so effectively. Internal government watchdogs have been noting for years that the FDA has failed not just to enforce post-marketing commitments, but also simply to keep track of the promises that were made. In a report published in June of 2006, the Office of the Inspector General of the Department of Health and Human Services took a comprehensive look at the way the FDA managed post-marketing trials, and found a system in disarray. Though pharmaceutical companies are required to submit yearly reports on the status of their post-marketing studies, nearly a third were missing or incomplete at the time of the analysis. Worse yet, the FDA didn’t know which companies were doing the required work and which companies weren’t. The “FDA cannot readily identify whether or how timely post-marketing study commitments are progressing toward completion,” the OIG wrote.
Ten years later, a follow-up study found more of the same. “For PMRs to be effective, FDA must be able to fully account for them. However, FDA’s system for overseeing PMRs is not yet up to the task,” the report said. Elizabeth Havener, who worked as the lead analyst for the OIG’s 2016 follow-up report, remarked on the FDA’s consistent inconsistency. “We found some problems in managing data and work processes that we first identified in 2006, and 10 years later, we find these problems persist,” she said.
The FDA, for its part, argues that both the agency and its system are working as they should. According to Dyer, the FDA takes action when a company is non-compliant, and when they don’t have a valid reason for that non-compliance. “It is the FDA’s practice to give firms an opportunity to take voluntary and prompt corrective action before the Agency initiates an enforcement action,” she wrote in an email.
However, Kaitin says the FDA doesn’t follow up for several reasons. On a purely practical level, the FDA doesn’t have the manpower to pursue each of these drugs. What’s more, Kaitin suggests the FDA might not believe that the studies are actually providing critical information in the first place. “The FDA, if they require these studies now, they should be required to follow up and assess whether the results give useful information and better understanding of the drug,” he said.
Critics say that the absence of many post-marketing studies is particularly troubling given that the post-marketing data that does exist has a big impact. Nearly all the post-marketing requirements that were fulfilled in 2014 led the FDA to request the sponsor to make changes to the safety or efficacy indications on the drug’s label, according to the OIG’s 2016 analysis. Analyzing the post-marketing studies available shows that the majority are not fulfilled by the date set by the FDA, and the OIG suggests that the agency is not equipped to adequately keep track of them. This suggests that most drugs remain on pharmacy shelves with their original — and potentially incomplete, incorrect, or misleading — labels.
As a result, most general physicians prescribing medications probably do not know that the drugs they give patients are still in trials, according to Wolfe. “There’s nothing in the labeling about that. There doesn’t have to be,” he says.
Other experts agree. Sasich says that though doctors might think they write evidence-based prescriptions, that might not be the case. “If you ask physicians where they get the data they make prescribing decisions on and they’ll tell you it’s the scientific data, but I don’t think that’s what’s happening,” he says. “The thing that comes to mind is that physicians are making prescribing decisions, and patients are accepting drugs, on the basis of promotion, instead of evidence.”
This article was produced by students in the Science, Health & Environmental Reporting Program at the NYU school of journalism.