In 2022, the AgencyIQ team spent days (and days, and days…) observing and analyzing key issues for over a score of products that came before the FDA’s advisory committees. Though the agency doesn’t have to follow the advice of its committees, it usually does. And the meetings provide an opportunity to see where the FDA is eager to leap into the fray, and where it’s proceeding with caution in drug development. Here are 5 key observations from our year of listening in on this fascinating process.
Takeaway 1: Pay attention to dosing strategies
- Quick background: Historically, dose-finding studies for cancer treatments have been based on determining the maximum tolerated dose (MTD). MTD correlates the tolerability and safety of a treatment at the highest dose that can both be tolerated and treat the cancer. There are several limitations to the MTD approach. Initially developed for systemic chemotherapies, dose-finding based on MTD is generally not applicable to new treatments that are more targeted in their effects, such as immunotherapies and new molecular targeted agents (MTAs). In addition, the MTD approach ignores target interactions and off-target toxicities, evaluates too few patients at each dose, and offers only a short observation period to observe dose-limiting toxicities (DLTs).
- What we saw in 2022: The FDA is now taking a “more isn’t better” approach. Though many of the concepts underlying this approach have been circulating in various formats, signs of the FDA’s hew, harder line began to surface during several Oncologic Drugs Advisory Committee (ODAC) meetings.
- In a hearing for Oncopeptide’s Pepaxto (melphalan flufenamide), FDA reviewers raised the issue that the sponsor had only conducted limited dose exploration and used the MTD without exploring alternative doses, thus not effectively establishing the true optimal dosage. [ Read AgencyIQ’s analysis of the Pepaxto ODAC meeting here.]
- Similarly, Spectrum Pharmaceutical’s poziotinib faced criticism related to the sponsor’s dose-finding approach from Richard Pazdur, director of FDA’s Oncology Center of Excellence (OCE), remarked that “proceeding with a drug development program when you don’t have a well-founded dose is literally building a house on quicksand.” [ Read AgencyIQ’s analysis of the poziotinib ODAC meeting here.]
- Secura also saw Copiktra (duvelisib) come under fire for dosing considerations at another ODAC meeting, both for safety considerations and for not adequately exploring whether lower doses would be effective. Almost half of those taking duvelisib in the pivotal trial had to reduce their dose and about 80% of patients had Grade 3 or higher treatment-emergent adverse events while on duvelisib. Although pharmacokinetic and pharmacodynamic data suggested a lower dose was efficacious, that lower dose was not explored further by the sponsor, earning Secura harsh criticism from the FDA. [ Read AgencyIQ’s analysis of the Copiktra ODAC meeting here.]
- More context for the FDA’s new dose optimization scrutiny: Duvelisib is a member of a class of oncology treatments called PI3K inhibitors; a May 2022 ODAC meeting focused on lessons learned about this drug class. A key point from that meeting was the overall inadequacy of dose exploration and optimization for the class, a finding borne out in the specific case of duvelisib. The webinar also highlighted the importance of considering the patient experience in evaluating oncology dosing. The real-world implications of dosing established by a MTD strategy include frequent dose reductions and interruptions, and even treatment discontinuations, because of adverse events, noted FDA presenters. In turn, this means that patients may be receiving sub-therapeutic cancer therapy because of tolerability problems. Traditional severity grading may not fully capture how burdensome treatments are. However, a fuller picture of the patient experience with cancer treatment dosing can come with increased use of patient-reported outcomes (PRO) such as those tracked in the PRO-Common Terminology Criteria for Adverse Events (PRO-CTCAE) developed and managed by the National Cancer Institute, according to meeting presenters. [ Read AgencyIQ’s analysis of the May 2022 OCE meeting here.]
- Project Optimus, established in 2021, formalizes the FDA’s ambition to reform outdated dose optimization practices. Goals of the project include disseminating the FDA’s “expectations for dose-finding and dose optimization” and making it easier for developers to meet with oncology review divisions on the topic of dose optimization. The project also seeks to inject more evaluation of a range of doses earlier in the development program in an approach “that leverages nonclinical and clinical data.”
- What to watch for: Given OCE Director Pazdur’s forthright approach to this issue in several 2022 ODAC meetings, oncology sponsors who don’t include a coherent dosing strategy (or strategies) in their development programs can expect to have their feet held to the fire. Although dose optimization is of obvious concern in oncology drug development, sponsors could expect the FDA to scrutinize dose-finding strategies for certain other products where dose-dependent toxicities or drug-drug interactions, for example, could affect tolerability and safety.
Takeaway 2: The FDA is getting serious about reining in accelerated approvals
- Quick background: The FDA can grant an accelerated approval to drugs intended to treat serious or life-threatening diseases for which adequate therapies are lacking, based on “adequate and well-controlled” clinical trials showing the drug has a significant effect on either a surrogate endpoint “that is reasonably likely to predict clinical benefit,” or an intermediate clinical endpoint “that can be measured earlier than irreversible morbidity or mortality,” as Jacqueline Corrigan-Curay, CDER’s principal deputy center director, explained in a March 2022 update on the FDA’s accelerated approval program. However, these approvals need to be confirmed by a postapproval study, or risk having approval withdrawn (or, more frequently, be voluntarily withdrawn upon the FDA’s request). [ Read AgencyIQ’s meeting recap, including more details about accelerated approval requirements, here.]
- In 2022, the FDA continued its scrutiny of products that have received accelerated approval, some of which may have failed to confirm a favorable benefit-risk profile in confirmatory trials, either because the trials failed or because they weren’t completed. The FDA is also casting a sharp eye on products seeking the accelerated approval pathway that lack a sufficient data package to support approval. What started out at the OCE under Project Confirm for oncology-only accelerated approval products has now also spread to infectious disease products that have received accelerated approval as well.
- Makena, a product with a complicated history, faced a 3-day hearing before an advisory committee in October 2022, after several rounds of the formal dispute resolution request (FDRR) process. Despite the product’s lack of substantial evidence of effectiveness to prevent preterm birth, even with a narrow indication, the FDA had difficulty removing the product from the market. This has led the agency to advocate for stricter measures at the time of approval such as the confirmatory trials being underway. Yet the bigger question remains: how to balance unmet need while gathering a sufficient data package to support an application and the use of this particular pathway. [ Read AgencyIQ’s analysis of the Makena hearing here.]
- FDA officials voiced similar concerns at the ODAC meeting for poziotinib. In addition to concerns over the lack of statistically persuasive data, there were concerns about enrollment to the confirmatory trial and the trial’s delayed start. Stemming from these and other concerns, FDA’s position at the poziotinib meeting was that confirmatory overseas trials are not the answer to this challenge. OCE staff, including Pazdur, co-authored a 2022 paper showing that when confirmatory trials are ongoing at the time of accelerated approval, the confirmatory data are made available significantly more quickly than when trial enrollment hasn’t commenced when accelerated approval is granted. Additionally, OCE pointed out ongoing issues with sponsors who agree to confirmatory studies but are unable to conduct the study due to lack of patient enrollment, later asking the FDA to change the study design. [ Read AgencyIQ’s analysis of the poziotinib ODAC meeting here.]
- A conundrum: The accelerated approval itself may hinder enrollment in confirmatory trials. When a product is already commercially available under the accelerated approval pathway, it may be difficult to enroll patients in a randomized trial with a placebo arm – and, depending on the condition being treated, a placebo arm may be required to show benefit in a “hard” endpoint such as overall survival. Similar conditions may arise when a drug has been made extensively available under so-called compassionate use programs.
- More context and updates: New provisions in 2023’s omnibus bill may give the FDA additional leverage to achieve withdrawal of products whose confirmatory trials have not confirmed safety, effectiveness, or both. In the new legislation, FDA is now required to specify the conditions for a post-approval study (or studies), “which may include enrollment targets, the study protocol, and milestones, including the target date of study completion.” Previously, FDA generally only published information about the high-level requirements of the postmarketing required study (more commonly known as a PMR). The agency now also has explicit authority to require that a confirmatory study be underway at the time of accelerated approval. Other provisions may help the FDA expedite the process of withdrawing approval, though AgencyIQ analysts are dubious that much will come of these updates. The price of non-performance is also going up, with the FDA receiving authority to levy much steeper penalties against companies that are not staying on track with PMRs. A new Accelerated Approval Council would include a who’s who of FDA leaders – the directors of CDER, CBER, OCE, Office of New Drugs, Office of Orphan Products Development, Office of Tissues and Advanced Therapies, Office of Medical Policy, and at least three directors of review divisions (including one within the Office of Neuroscience). The council would “ensure the consistent and appropriate use of accelerated approval across the FDA.” [ Read AgencyIQ’s full analysis of the FY2023 omnibus here.]
- What to watch for: Again, many of these policy shifts will be felt most deeply by oncology sponsors. But accelerated approval is commonly sought in other areas where serious conditions still have large unmet need, such as rare pediatric diseases and neurodegenerative diseases. Sponsors hoping for accelerated approval based on surrogate endpoints should carve out space for discussions with the FDA regarding confirmatory trials early in the game, and should not bank on single-arm trials, foreign-only trials, or vague plans for post-approval enrollment satisfying the agency, given public pressure for curbs in this area – and the agency’s new legislative ammunition.
Takeaway 3: Approvals that leverage foreign-only data are mostly nixed, though some exceptions might remain
- Quick background: In an early 2022 ODAC meeting for the anti-PD-1 inhibitor sintilimab, the FDA strongly expressed that was not inclined to premise approvals on single-country data drawn from a non-diverse study population – an approach that limits generalizability of the data to the U.S. population. At this meeting, and in an article published nearly contemporaneously with the meeting, FDA officials ticked off a host of concerns.
- Additional points raised by the agency in this ODAC meeting included the fact that the study was conducted in China, a country for which data integrity concerns are a recognized issue and standards of clinical care differ from those in the U.S. These concerns persist even if the drug is not first-in-class or is intended to fill an unmet medical need, the FDA made clear in its presentations and briefing documents. Nested within the issue of generalizability is another big problem: lack of diversity for single-country clinical trials, especially those run in such a racially and ethnically homogenous country as China. A single-country trial exhibits “lack of diversity by design,” said Pazdur during the sintilimab meeting. The FDA has been consistent in its calls for trials to be representative of the U.S. population in its pre-BLA meetings, he added. [ Read AgencyIQ’s analysis of the sintilimab meeting here.]
- Sintilimab was the warning shot for several other applications pending at the FDA that rely on data solely from China. As noted in an article published in the New England Journal of Medicine (NEJM) in December 2021 by OCE Director Richard Pazdur and Deputy Director Julia Beaver, many new checkpoint inhibitor trials are conducted “exclusively or predominantly in China.” At the time of the ODAC meeting for sintilimab, the FDA tallied at least 25 oncology applications based on Chinese pivotal trials that were either headed for submission or under review by the FDA.
- That meeting also clarified that better generalizability is just one reason FDA favors multi-regional trials, and favors bringing China into that fold. Rather than continue the trend of enrolling Chinese patients by running single-country trials, Pazdur said he’d rather see China in more MRCTs. “We want to bring China into the multiregional arena here,” he said. “We feel that we would all benefit from having China participate fully in multi-regional trials…. The world would be a better place from having all countries participate in these multi-regional trials. We don’t want to pit one country against the other…. We are all going to be stronger with a global regulatory environment,” he went on.
- At another ODAC meeting, the FDA has also made clear that conducting confirmatory trials solely overseas due to lack of enrollment or any other reason is not acceptable. At the 2022 poziotinib meeting referenced earlier, Harpreet Singh, director of FDA’s Division of Oncology 2, called such an approach “antithetical” to the oncology community’s “commitment to diversity and having a patient population that’s reflective of the US patient population.”
- In contrast, at the Psychopharmacologic Drugs Advisory Committee meeting for Acadia’s Nuplazid (pimavanserin), FDA’s Billy Dunn, who leads the Office of Neuroscience, offered a different perspective, given the lack of diversity in the pivotal trial for Nuplazid. At the time, Dunn said “we are able to base approvals and considerations data on foreign data. We do need to work with the sponsor to understand the applicability of the foreign data to the domestic population […] We do consider data from foreign and ex-U.S. sources; as long as there’s no reason to believe that those data are inapplicable to our population, we can rely on them for our regulatory action.” [ Read AgencyIQ’s analysis of the Nuplazid meeting here.]
- Nuplazid eventually received its second complete response letter for the proposed indication (hallucinations and delusions associated with Alzheimer’s disease psychosis) that brought the drug before the advisory committee – but concerns were more based on the overall data package than on foreign-only data.
- What to watch for: Dunn’s remarks during the Nuplazid meeting highlight the diversity of responses still being seen across the FDA on this question, and hint at the possibility of some regulatory flexibility regarding foreign-only trials. However, sintilimab’s sponsors received scathing criticism from OCE’s Beaver, Singh, and other FDA officials premised in part on their lack of communication with the agency about their development program. Sponsors seeking approval in any therapeutic area who hope to rely entirely or primarily on foreign-only data, and on single-country data in particular, should signal their intention to the FDA early on and be prepared for less than positive feedback.
Takeaway 4: Single-arm trials continue to cause sponsor headaches
- Quick background: As noted above, earlier in 2022, the FDA put four approved phosphatidylinositol 3-kinase (PI3K) inhibitors products before an ODAC hearing due to mounting safety concerns and excess deaths shown in confirmatory trials of approved products. The committee heard a litany of safety concerns for a class of oncology drugs that, though effective, may have toxicities that significantly impact overall survival. Single-arm trials will not capture critical time-to-event data, said the FDA, who also raised issues of dose-finding in a drug class with a narrow therapeutic window.
- The discussions at this ODAC meeting exposed flaws in single-arm trials. Without a comparator arm, it is hard to characterize safety, since it can’t be established whether treatment-associated adverse events are due to the drug or underlying disease. Efficacy assessments are also less robust when historical controls or cross-population controls are used as comparators in single-arm trials. In addition, the follow-up for the investigation arm is short and time-to-event endpoints are not interpretable.
- The committee agreed with the FDA on these points; Lingering concerns also centered around the inability to ascertain overall survival from single-arm studies, though others acknowledged the difficulty in using such a long-term endpoint for indolent cancers. Still, “Randomized data are always valuable,” noted PCNS member Jorge Nieva, of USC’s Keck School of Medicine.
- In a completely different arena, single-arm trial data caused headaches for Y-mAbs Therapeutics, who sought approval for omburtumab, a radiolabeled monoclonal antibody proposed to treat pediatric neuroblastoma. Here, the sponsor used external control data in its analysis of one of its two single-arm trials, comparing overall survival data against an ex-US historic control. Despite the difficulties of running a randomized trial for this rare pediatric condition, FDA reviewers still found a host of issues with the data package stemming from discrepancies between the external control group and the single arm that received study drug – and the committee was in unanimous agreement. [ Read AgencyIQ’s analysis of the ombertumab meeting here.]
- What to watch for: The FDA’s intense scrutiny of PI3K inhibitors, including devoting an entire ODAC meeting to issues with the drug class and associated development programs that run a single study arm, make clear that sponsors of these products should plan to steer entirely clear of single-arm trials. And ongoing concerns about the inability of single-arm trials to capture overall survival and tease out important patient experience data extend beyond this drug class. Statistical analysis is also often complicated by referencing external data to shore up single-arm studies, as seen with omburtumab. Though single-arm trials may be inevitable for some conditions and in some populations, sponsors should be prepared with a sheaf of data to support this approach.
Takeaway 5: The FDA is still willing to exercise considerable regulatory flexibility – sometimes
- Quick background: In 2022, several products were presented to the various advisory committees to discuss if there was substantial evidence of effectiveness and in some cases, whether the FDA should exercise regulatory flexibility. In almost all of the hearings where regulatory flexibility was at issue, the committee sided with the FDA’s view and voted against approval. However, 2022 saw some notable exceptions – in one case, for an ALS drug where FDA was under public (and Congressional) scrutiny. In the developing policy area of COVID-19 vaccine approvals, the agency came under fire for fast and flexible decision-making – but in fact, followed exactly the road map it had laid out for itself relatively early in the pandemic.
- For example, in November 2022, FDA’s ODAC unanimously rejected the radiolabeled monoclonal antibody, omburtamab. to treat pediatric neuroblastoma patients with metastases, a rare condition. The committee cited a lack of sufficient data, although the condition remains with no approved therapies, as we just discussed. The sponsor, Y-mAbs, submitted a package with the external control data and other issues that resulted in a statistical analysis fraught with confounders. The fact that FDA reviewers had significant disagreements requiring adjudication across the data package, combined with significant safety issues, left the committee disinclined to endorse omburtamab despite acknowledged unmet need. [ Read AgencyIQ’s analysis of the ombertumab meeting here.
- More recently still, FDA’s Cardiovascular and Renal Drugs Advisory Committee (CRDAC) rejected Cytokinetics’ omecamtiv mecarbil, a novel heart failure drug due in part to a lack of substantial evidence of effectiveness, although the drug was intended to fill an unmet medical need in a condition that still has significant morbidity and mortality even with standard of care treatment. Safety was also at play for CRDAC members – an ongoing theme when committees are asked whether the agency should employ flexibility.
- However, in perhaps the most extraordinary example of the FDA’s willingness to bend over backward in exercising flexibility, the agency called back its Peripheral and Central Nervous System (PCNS) advisory committee for a second meeting for Amylyx’s Relyvrio (AMX0035) for amyotrophic lateral sclerosis (ALS), which the agency eventually endorsed in September. Relyvrio, then called AMX0035, had received an initial narrow thumbs down from PCNS in March 2022. At that time, the committee cited issues included statistical questions, a small sample size, irregularities in study conduct, and other issues (including primary reliance on a single study) in reaching its 6-4 vote against sufficient evidence to support an efficacy claim for AMX-0035. It’s notable that although the FDA review team was critical of the data package, the agency itself had asked Amylyx to submit the package for approval. [ Read AgencyIQ’s analysis of that advisory committee meeting here.]
- But that wasn’t the end of the story. In July 2022, the FDA scheduled a highly unusual second PCNS meeting to re-review AMX0035, after the sponsor submitted new analyses of previously submitted data (but no new data). At that meeting, though Tristan Massie, the FDA statistical reviewer, still found that the data “do not provide a statistically persuasive effect on mortality,” other FDA leaders hammered home the agency’s freedom to exercise regulatory flexibility when warranted. Dunn, the ONS head, told the committee that “The serious nature of ALS contextualizes our consideration of the strength of evidence being presented in support of the effectiveness of AMX0035.” Further, Teresa Buracchio, director of OND’s Division of Neurology 1, told the committee that “confirmatory evidence can be just about anything,” including data from natural history studies. [ Read AgencyIQ’s analysis of the second advisory committee meeting here.]
- In the end, the FDA may have felt pressure to reconvene the committee to have some external endorsement of its Relyvrio approval. By centering the day’s discussions on regulatory flexibility, the FDA shifted the emphasis away from strict attention to the success or failure of individual components of Amylyx’s development program. Said Dunn,“We recognize that the committee are not regulators.” He drilled in further on his characterization of the FDA’s role in such edge cases: “Effectiveness is a regulatory decision that we as regulators must make and it’s a qualitative decision that requires the exercise of scientific judgment.”
- ALS advocacy groups have exerted coherent and targeted pressure on the FDA and elected officials to advance therapies for the fatal, rapidly progressive neurodegenerative disorder, as our POLITICO colleagues have reported in detail. It’s unknown to what extent these external factors prompted the FDA to grant Relyvrio accelerated approval; Billy Dunn extracted an extraordinary – but largely unenforceable – commitment from Amylyx leadership to withdraw their product should the confirmatory trials fail.
- The totality of the FDA’s approach to Relyvrio plays out in the shadow of the accelerated approval of Biogen’s Aduhelm (aducanumab) for Alzheimer’s disease – a much-analyzed and -maligned action that was the subject of a recent Congressional report that detailed extensive irregularities in the agency’s actions, including off-the-books communications with the sponsor. [ Read AgencyIQ’s recap of the Congressional review at the top of this article.]
- Shifting gears completely, the FDA has been criticized for flexibililty in another arena — its approach to authorizing some rounds of Covid-19 boosters, most notably the bivalent mRNA boosters that the agency, together with the Centers for Disease Control and Prevention, authorized in anticipation of a fall-winter surge of cases. Some critics faulted the FDA for not conducting human clinical trials of the effectiveness and safety of the reconfigured vaccines and instead relying on in vitro immunogenicity data and a well-established safety profile.
- However, this immunobridging strategy was exactly what the FDA has intended to do all along. As laid out previously by the FDA, and as discussed at the April 2022 VRBPAC meeting, the agency generally intends to use a strategy to authorize or approve strain-adjusted vaccines that aligns with its practice for influenza vaccines, where blood samples from vaccinated individuals are used to test antibody production when the samples are exposed to the virus in question, or a lab-manufactured stand-in. In this case, the agency has followed a well-established model for vaccine authorization and approval. Although the FDA may appear to have veered from its lane with these actions, it’s actually following a well-established roadmap. An upcoming VRBPAC meeting will give agency officials an opportunity to lay out future plans for its regulation of vaccines for the mutable – and still deadly – virus. [ Read AgencyIQ’s preview of that meeting here.]
- What to watch for: As with the related area of accelerated approvals, the FDA’s application of regulatory flexibility still seems somewhat divergent across centers. The broad concept of flexibility that’s baked into approvals, especially accelerated approvals, gives the agency “wiggle room” that may be useful with serious conditions (and well-organized advocacy groups), as with Relyvrio. However, the FDA is flexing increasing muscle in areas where recent therapeutic advances make drugs with concerning toxicities look more and more hazardous, as with the PI3K inhibitors.
Featuring previous research by Alexander Gaffney and Kedest Tadesse
To contact the author of this analysis, please email Kari Oakes.
To contact the editor of this analysis, please email Alexander Gaffney.