How can regulators look at the same data and get different results?
The goal of regulatory bodies such as the U.S. Food and Drug Administration (FDA), the European Medicines Agency (EMA), and the African Medicines Agency is always the same: Balance potential benefits and risks to ensure that the drugs that make it to market are both safe and effective for patients in need. Still, a drug that gains FDA approval may give pause to European reviewers, despite having reviewed the same evidence as their American counterparts, and vice versa.
These contrasting conclusions are, in part, due to each agency’s basic structure and processes: Both the FDA and the EMA have distinct interactions with external stakeholders, different methods of endpoint evaluation, and individual comfort levels with risk.
ASH Clinical News spoke with representatives from each agency, as well as independent investigators, to learn how the agencies make their decisions.
The Road to Approval
The road to approval is well-trodden: Once an advanced phase trial (or even an earlier phase study, through accelerated approval mechanisms) is completed and shows encouraging results, the company attempts to secure approval from the relevant regulatory bodies for the drug to enter the market.
In the U.S., that means submitting a new drug application to the FDA’s Center for Drug Evaluation and Research – one of the FDA’s largest centers. The agency then has 60 days to decide whether the application should be filed for review.1 Filed applications are then reviewed by a team of physicians, statisticians, chemists, and pharmacologists, who assess the data to determine the drug’s safety and efficacy for the proposed indication.
Though FDA review is a strict, arduous process, it’s relatively straightforward compared to the complexity of the EMA’s regulatory processes.
“The EMA is the coordinating hub for the network of EU member state–specific regulatory agencies called national competent authorities,” explained Giovanni Tafuri, PhD, a senior heath technology assessment officer at the Italian Medicines Agency, one of the national competent authorities working within the European medicines regulatory network. (Dr. Tafuri offered his own personal views based on his research in this area and noted that they do not reflect the views of the Italian Medicines Agency.)
Similar to the FDA’s use of medical experts as special government employees to serve on advisory committees, the national competent authorities supply thousands of European experts to serve on EMA’s special committees, working parties, and other groups.2 However, Dr. Tafuri noted, “at the FDA, the assessment is more internalized.”
Typically, drug manufacturers seek approval through the EMA’s centralized procedure, which allows applicants to obtain marketing authorization that is valid throughout the EU. This process involves a number of committees, including the Committee for Medicinal Products for Human Use (CHMP).
The CHMP, which is charged with developing a scientific opinion on a drug, comprises an elected chair, one member, and an alternate nominated from each of the 28 EU member states, as well as a member and alternate nominated by Iceland and Norway.3 Member states and the EMA can also nominate as many as five additional members to join the committee. The members (who serve 3-year terms) convene monthly to conduct initial assessments of applications and to review marketing authorizations.
“At the end of this rather complex process of evaluation that involves many, many parties, the committee sends its scientific opinion to the European Commission, which converts the opinion into a decision that allows companies to market the drug in the EU, Norway, and Iceland,” added Francesco Pignatti, MD, head of the EMA’s Oncology, Haematology, and Diagnostics division.
National competent authorities are also responsible for reviewing products that do not pass through the EMA’s centralized procedure, but are instead submitted for review in individual member states.
This is “an infrequently used process,” noted David Bowen, MD, a consultant hematologist at the Leeds Teaching Hospital in the United Kingdom (UK), a recent Seconded National Expert at EMA, and a member of the UK’s National Institute for Health and Care Excellence (NICE) appraisal committee. “This process is probably mostly used for drugs for which the evidence base is quite marginal and also where there may be a specific unmet need in one country.”
Once approved in one member state, the drug can be reviewed under a Mutual Recognition Procedure for consideration of approval in other member states that choose to evaluate the product.
“[It is a] question of where the bar for certainty about the presented evidence is. What is the level of uncertainty that your country or your organization is prepared to accept?”
—David Bowen, MD
Am I Seeing What You’re Seeing?
As an example, when the antibody-targeted chemotherapy agent gemtuzumab ozogamicin (GO) was being reviewed for the treatment of acute myeloid leukemia (AML), the FDA and the EMA handled its approval much differently.4
In the United States, the FDA granted accelerated approval to GO in 2000 but the manufacturer voluntarily withdrew the drug from market in 2010 because it failed to meet post-marketing requirements. (Note: On July 12, 2017, the FDA’s Oncologic Drugs Advisory Committee recommended GO for approval for the treatment of adult patients with previously untreated, CD33-positive AML.)
Japan approved GO as an orphan drug in 2005 and opted to extend its approval in 2010. The EMA, however, failed to approve the drug in 2008 because of a lack of randomized clinical trial data.
What drove the regulatory bodies to reach such varied decisions?
“Internationally, we all agree on the value of certain endpoints. We all prefer to see differences in overall survival (OS),” said Dr. Pignatti. The divergence occurs when those data are not available. Then, agencies turn to other efficacy endpoints, such as progression-free survival (PFS) or – occasionally in the case of hematologic malignancies – response rates.
In a recent study of the EMA and FDA decision processes for anti-cancer drugs, Dr. Tafuri and Francesco Trotta, PhD, a senior pharmaco-epidemiologist at the Department of Epidemiology of Lazio Regional Health Service in Italy, interviewed regulators from each agency to determine the formal and informal factors that ultimately drove decision-making.5
“Most respondents attributed the divergence in decision-making to a different evaluation of clinical endpoints,” said Dr. Trotta.
He added that EMA regulators tend to identify PFS as a clinical benefit, whereas FDA respondents are more likely to consider it a surrogate endpoint to be confirmed by OS.
The study also revealed that the agencies handle risk differently. “Unlike the EMA, the FDA has a prevailing attitude to take risks in order to guarantee quicker access to new treatments,” Dr. Trotta said.
Dr. Bowen agreed, explaining that it is probably more common for the FDA to approve a drug, whereas the EMA might err on the side of caution.
“That comes down to the question of where the bar for certainty about the presented evidence is. What is the level of uncertainty that your country or your organization is prepared to accept for a drug to be licensed?” he said, noting that it is difficult to define the exact criteria either organization uses to make decisions.
Another important distinction is that the datasets the two agencies base their decisions on aren’t exactly the same. The FDA examines raw data (the actual clinical data collected from a study), and the EMA relies on an analysis of data provided by the company seeking drug approval.
“This gives the FDA the capability to initiate its own independent analysis, which we don’t have,” said Dr. Pignatti. “It also ensures that companies are diligent in analyzing their data, which is more difficult for us to check.”
Although neither agency designs clinical trials for investigational products, they do provide guidance for drug sponsors initiating such studies. Manufacturers can even get advice from both organizations at the same time.
The FDA seems to play a more active role in shaping the clinical trial pathway, though. Before phased clinical trials can begin, drug sponsors are required to submit an investigational new drug application to the FDA that outlines initial animal testing results, the drug’s composition and manufacturing, and the plan for testing the drug in humans.
Time to Approval
As one would expect, each review method results in different timelines to approval, with the EMA typically taking longer to reach a decision. The advantage of the EMA system – the large diversity of opinions included in the process – puts the agency at a disadvantage when it comes to efficiency.
“In the European [example], pharmaceutical legislation has been about harmonizing the requirements among the countries,” Dr. Pignatti said. “This type of set-up ensures that all countries have an equal say in these decisions.”
It also means that the FDA boasts faster time to approval. Bringing safe, effective, innovative drugs to market more quickly has long been a stated goal of the FDA leadership. In 2002, legislators amended the 1992 Prescription Drug User Fee Act (PDUFA) to establish specific goals for both standard and priority reviews, requiring that standard reviews be completed within 10 months and priority reviews be completed within six months.6
According to the 2016 PDUFA Performance Report, 95 percent of the drug applications in the 2015 fiscal year were acted on within the set goal.7
The EMA has a maximum of 210 days to carry out its assessment, according to the agency’s 2016 annual report, but its time to approval is nearly universally slower than the FDA’s.8
“The fundamental reason for this difference is related to the process by which the EMA comes to its decision,” Dr. Bowen said, which includes multiple rounds of review.
Once a dossier for a new drug has been submitted, the EMA allocates rapporteurs (French for “reporters”) from two of its member states to complete a primary assessment. The dossier then goes back to the CHMP. At that stage, all member states have a chance to comment on the initial dossier or pose questions about the report. This list of questions, often long, is then returned to the drug’s sponsor.9
“The company then has the option to ‘stop the clock’ and take its time to properly consider all of these questions, so that adds a bit of time onto the standard duration of EMA approval,” Dr. Bowen said.
The EMA also has accelerated pathways to drug approvals, but they are still slower than the FDA’s. The EMA’s priority review process can shorten the typical approval time from just over 12 months to just under 10, but the timeframe is not guaranteed.
Even in scenarios in which drug sponsors are seeking approval without randomized controlled trial (RCT) data available, the FDA is quicker to reach a decision. From 1999 to 2014, the FDA granted 60 unique indications for drugs without RCT results, whereas the EMA approved 44.10 Of the 44 applications that were made to both agencies that had comparable review packages, the FDA granted 43 approvals, but the EMA approved just 35. In addition, the FDA took an average of 8.7 months to review the products, compared with 15.5 months by the EMA.
“Unlike the EMA, the FDA has a prevailing attitude to take risks in order to guarantee quicker access to new treatments.”
—Francesco Trotta, PhD
A similarity between the agencies is that pricing is not a factor for determining whether to approve a drug.
“Neither a drug’s potential price nor the price of a similar drug product plays a role in the approval process,” Jeremy Kahn, a trade press officer and spokesman for the FDA, told ASH Clinical News. “The FDA does not have the authority to regulate drug prices.” Instead, in the U.S., manufacturers, distributors, and group-purchasing organizations determine drug prices.
In Europe, pricing and reimbursement decisions about an investigational drug are made only after the EMA has granted its approval. Then, each EU member state determines its own pricing or reimbursement plan.
For instance, in the UK, NICE establishes prices by running cost-effectiveness analyses on certain drugs, including all newly licensed cancer drugs. “At NICE, our objective is to start the review process during the EMA marketing authorization assessment process, on the assumption that the drug is going to get a license,” said Dr. Bowen.
“This means that there is not a long time between the point of licensing and the point at which a decision is made about whether the drug is cost-effective for use in the National Health Service (NHS) and patients may or may not have routine access to the drug.”
Once a license is granted by the EMA, NICE aims to decide within three months whether the drug is cost-effective enough for use by the NHS. To reach this decision, the NICE appraisal committee relies on academic information, models, and discussion in what Dr. Bowen called a “challenging, rigorous” process.
“At the end of the day, there’s a set figure, and if the drug falls below that figure, it is deemed cost-effective and will be routinely commissioned by the NHS for use,” he explained. “If it falls above that figure, it won’t be commissioned and there is no mechanism for a patient to be treated with the drug within the NHS because the drug is not deemed to represent effective use of NHS resources, compared with other pressures on the fixed NHS budget.”
Other member states rely on alternative methods from the NICE process. For instance, French and German agencies use comparative-effectiveness analyses to decide whether a drug should receive marketing authorization. Reviewers examine the novel drug and its associated data, then score it against the current standard treatment to determine an appropriate price. They can then enter negotiations with the drug company to set a market price.
“The heterogeneity of systems and criteria used by health technology assessment and reimbursement bodies among the EU member states often leads to serious discrepancies in pricing, reimbursement, and patient access across countries,” Dr. Tafuri noted.
Seeking Public Input
Mr. Kahn indicated that interacting with patients, caregivers, and advocates has long been a priority of the FDA. “In this tradition, the FDA intends to enhance future patient engagement by providing a more transparent, accessible, and robust experience for patient communities,” he said, adding that the agency is even considering adding a new Office of Patient Affairs.
Drs. Tafuri and Trotta also highlight the high value the FDA places on seeking public input. “All FDA respondents considered patient input ‘highly valuable,’ though they still thought the FDA ought to make regulatory decisions based on independent scientific grounds,” they wrote. “FDA respondents stressed that their agency has been increasing its transparency and interaction with the outside world over time, integrating the perspectives of patients, physicians, and health-care system specialists at all levels.”5
Some FDA regulators interviewed for the study identified public input as a leading factor contributing to the divergence in decision-making between the EMA and FDA.
For many years, the EMA overlooked patient involvement, but a recent push by the agency opened up the discussion to patients and other external stakeholders. “For several years now, we have had a framework for patient interaction,” said Dr. Pignatti. “Patients are not members of the committee, but we do call on them as experts in advisory committee meetings. We also have working groups of patients to whom we can address communication questions, for instance.”
Patient opinion is now also considered when the CHMP is making decisions, though the meetings remain closed to the public.
“The opportunity for patients to voice their views during different stages of the EMA regulatory processes are much greater than they have been in the past,” Dr. Tafuri noted. However, EMA regulators in his study expressed mixed opinions on the value of holding public hearings and incorporating patient advocacy groups in the process.
“Even respondents who agreed in principle to the establishment of public hearings in Europe were afraid that [adding] steps to the decision-making process could further slow … the approval of new anti-cancer medicines,” the authors wrote.5
Despite their differences, both agencies have a close relationship and rely on each other during the approval process. “One of our most valuable collaborators is the EMA,” said Mr. Kahn.
Since 2004, the FDA and EMA have collaborated in groups called “clusters.” “We have formed [clusters] that focus on treatments for children, establish effective measures for the development and use of biosimilar medications, evaluate new treatments for patients with cancer, set standards to help develop medicines personalized to a patient’s genetic makeup, and much more,” he said.
Dr. Pignatti said six organizations (the FDA, the EMA, Health Canada, the Australian Government Department of Health, the Pharmaceuticals and Medical Devices Agency in Japan, and Swissmedic) also participate in monthly, confidential calls to discuss methodology and ongoing evaluations. Dr. Tafuri explained that those interactions are typically part of an effort to “exchange information, not necessarily to find convergence on decisions.” This scheduled communication also helps to foster global collaboration, which will become a larger part of drug review and approval in the coming years.
As drugs become more complex and therapies become more individualized, collaboration will be essential to providing safe and effective medications across the globe, Dr. Tafuri said. “In a globalized world, international cooperation among different agencies – none excluded – is key.”—By Jill Sederstrom
- U.S. Food and Drug Administration. Drug Approval Process. Accessed May 16, 2017, from https://www.fda.gov/downloads/drugs/resourcesforyou/consumers/ucm284393.pdf.
- European Medicines Agency. How We Work. Accessed May 16, 2017, from http://www.ema.europa.eu/ema/index.jsp?curl=pages/about_us/general/general_content_000125.jsp&mid=WC0b01ac0580028a46.
- European Medicines Agency. CHMP Members. Accessed May 16, 2017, from http://www.ema.europa.eu/ema/index.jsp?curl=pages/contacts/2010/02/people_listing_000002.jsp&mid=WC0b01ac0580028c7c.
- Tanimoto T, Tsubokura M, Mori J, et al. Differences in drug approval processes of 3 regulatory agencies: a case study of gemtuzumab ozogamicin. Invest New Drugs. 2013;31:473-8.
- Tafuri G, Stolk P, Trotta F, et al. How do the EMA and FDA decide which anticancer drugs make it to the market? A comparative qualitative study on decision makers’ views. Ann Oncol. 2014;25:265-9.
- U.S. Food and Drug Administration. Frequently Asked Questions about the FDA Drug Approval Process. Accessed May 17, 2017, from https://www.fda.gov/drugs/resourcesforyou/specialfeatures/ucm279676.htm.
- U.S. Food and Drug Administration. FY 2016 PDUFA Performance Report. Accessed May 21, 2017, from https://www.fda.gov/AboutFDA/ReportsManualsForms/Reports/UserFeeReports/PerformanceReports/ucm548126.htm.
- European Medicines Agency. Annual Report 2016. Accessed May 17, 2017, from http://www.ema.europa.eu/docs/en_GB/document_library/Annual_report/2017/05/WC500227334.pdf.
- European Medicines Agency. Marketing Authorisation: The Evaluation Process. Accessed May 18, 2017, from http://www.ema.europa.eu/docs/en_GB/document_library/Presentation/2009/10/WC500004235.pdf.
- Hatswell AJ, Baio G, Berlin JA, et al. Regulatory approval of pharmaceuticals without a randomised controlled study: analysis of EMA and FDA approvals 1999-2014. BMJ Open. 2016;6:e011666.