Yesterday’s FDA approval of Biogen’s Alzheimer’s drug, Aduhelm (aducanumab-avwa), is historic and is of a magnitude that it may be harbinger for future Agency actions, especially in neuropsychiatric conditions. This has been a decision we’ve been closely following. So much so, in November, right after the FDA Peripheral and Central Nervous System Drugs (PCNS) Advisory Committee had met on this therapy, the Pink Sheet quoted HP&M’s Frank Sasinowski in its coverage of the advisory committee as he floated the idea that FDA may turn to consider Accelerated Approval as a path forward on this therapy as the advisory committee had not been asked to nor had offered any opinion on this approval pathway.
Our reading of yesterday’s action is that while this is the direct action of the Director of the Office of Neuroscience, Dr. Billy Dunn, this approval has the full support of the Director of the FDA Center for Drug Evaluation and Research (CDER), Dr. Patrizia Cavazzoni. This is because, not only did the FDA issue a press release, which is typical of any FDA regulatory action of this consequence but, in addition to the conventional press release which quotes Dr. Cavazzoni, she also issued her own personal, fairly extensive public statement explaining and defending this action. Moreover, in her statement, Dr. Cavazzoni observes too that the advisory committee had not discussed the possibility of the Accelerated Approval pathway for this therapy.
While we do not yet have the FDA summary basis for approval, which we expect will contain very illuminating review memos, we can turn to the label and other overarching aspects of the approval for insights. Below we highlight those observations that we believe, in the context of the broader regulatory landscape, set a positive trajectory for future Agency actions. As you will see, ultimately, this Accelerated Approval may have benefited from FDA’s experience in evaluating products under the Emergency Use Authorization (EUA) regulatory framework.
Key Takeaways from the Aduhelm Label: What’s in It and What Isn’t
First, it was noteworthy what was missing from the label. What do we mean? Well, the FDA initially issued draft guidance on how to label drugs approved via the Accelerated Approval (i.e., Subpart H) pathway in 2014, then issued it as final guidance in January 2019. In this guidance, FDA explains that sometimes “[s]imply reporting the endpoint used may convey sufficient information about uncertainty with regard to the limitations of usefulness drug and of uncertainty about anticipated clinical benefits…” Recent Accelerated Approvals we have been involved with, such as the 2017 approval of benznidazole for Chagas and the 2019 approval of Oxbryta for sickle cell disease, have taken this approach as they do not say anything more than identify the surrogate endpoint that was the basis of the approval and noting “continued approval…may be contingent upon verification and description of clinical benefit in confirmatory trials.”
However, the FDA guidance goes on to say in some cases, additional context is needed by specifically inserting a sentence in the indications statement that: no clinical benefit has been established. This sentence was included in the 2016 approval of Exondys 51 for Duchenne muscular dystrophy and the 2018 approval of Andexxa for reversal of anticoagulation therapy.
It is noteworthy that this sentence that no clinical benefit has been established was deemed by FDA as unnecessary for Aduhelm to provide additional context beyond that the approval is based on reduction in amyloid beta plaques. As an aside, we have found this inconsistency between which accelerated approval therapies are saddled with this sentence to be a paradoxical FDA policy as there is but a single standard for surrogate endpoints to be utilized for purposes of every Accelerated Approval (i.e., be “reasonably likely to predict” ultimate clinical benefit). Moreover, in every Accelerated Approval, the clinical benefit must be confirmed in a post-approval clinical study, just as the labels for all Accelerated Approval drugs state.
Second, it was notable that the trials of Aduhelm were conducted in those with early Alzheimer’s disease, and the indication that FDA approved is for anyone and everyone with Alzheimer’s disease. In many development programs, there is strong scientific rationale for generalizing the results (e.g., consistencies in the disease process, drug’s overall benefits and risks) and, as a result, indicating a drug for a population broader than those that were studied in clinical trials (see FDA’s indications and usage labeling guidance here). This approach to labeling encourages sponsors to undertake rationale drug development planning, allowing them to focus on study designs that can most effectively capture treatment benefits (or lack thereof) (e.g., through enrichment – a technique particularly necessary in rare disease settings where numbers are small and heterogeneity is high).
Reflecting on the Use of Accelerated Approval as an Appropriate Tool for Regulatory Flexibility
There is a long and vibrant history of FDA using the Accelerated Approval pathway to help expedite drug development and approval. This goes back to when Sasinowski was at FDA in the 1980’s, he had a hand in aiding others at FDA, like Dr. Bob Temple, to create on the Agency’s own initiative the Accelerated Approval pathway in order to address the raging AIDS crisis. Later, in 1993, in private practice Sasinowski advocated for the first use of Accelerated Approval outside of AIDS and cancer, when Dr. Janet Woodcock approved Betaseron as the first drug for multiple sclerosis. Over the past half of a decade, Sasinowski and Valentine have been involved in more than half of the Accelerated Approvals outside of cancer. Ever since Sasinowski’s 2012 paper, cataloguing FDA’s exercise of flexibility in its approval of drugs for rare diseases (later followed-up in 2015 with an updated analysis by both HP&M’s Valentine and Sasinowski; both papers available here), the word “flexibility” has been used widely, but what it means or how to apply it is still subjective and largely unknown and unknowable except in hindsight such as in these two papers.
So, it is not surprising there have been questions about the future of Accelerated Approval making their way through the industry in the last year. While historically it seemed that the barrier to use of the pathway was a seemingly high evidentiary bar for demonstrating that a surrogate endpoint is “reasonably likely to predict” ultimate clinical benefit, recent concerns expressed by senior CDER officials indicated the barrier to its use was instead fears of whether the Agency could be sure that post-approval confirmatory studies would be conducted in a timely manner and could be designed in a way that would reliably produce interpretable data. The basis for this concern materialized in October 2020 when CDER proposed withdrawal of approval for Makena after completion of its confirmatory study. Then further, in April of this year, when FDA’s Oncologic Drug Advisory Committee met to review several oncology indications granted Accelerated Approval over the last five years in what appeared to be part of a broader evaluation of Accelerated Approvals in oncology. The question was raised: does FDA still view Accelerated Approval as a realistic tool for expediting drug development?
Yesterday’s approval of Aduhelm helped answer that. Both Dr. Dunn, in his memo to the PCNS Advisory Committee members, and Dr. Cavazzoni, in her statement, emphasized the utility of Accelerated Approval in devastating conditions where the needs for treatments are urgent. This is certainly the case in Alzheimer’s disease – and as FDA articulates in its December 2019 draft guidance on substantial evidence of effectiveness, is exactly the context in many rare diseases.
In fact, it is in that document where FDA most extensively describes its flexibility in drug review. For example, it describes the three major ways in which this substantial evidence of effectiveness standard may be met, the latter two which reflect a demonstration of flexibility via approval based on a single study: (1) two positive adequate and well-controlled studies, (2) one positive adequate and well-controlled study with confirmatory evidence, or (3) one adequate and well-controlled study with statistically very persuasive evidence. FDA also describes how FDA will apply flexibility in applying this standard to different data sets on a case by case basis where either the condition is rare or is serious or there is a great unmet medical need. To us, through this approval action, it seems that FDA is showing that the Agency it intends to include the Accelerated Approval pathway in its armamentarium of ways in which to exercise flexibility. The evidence on the surrogate endpoint would still have to meet the substantial evidence standard by one of those three ways, but the Accelerated Approval pathway is clearly a type of flexibility that FDA has available to it in its discretion.
Closing Thoughts: A Non-COVID-Related Approval Based on Review Experience During the COVID-19 Pandemic
So how did we get here? There is no question the last year has been unlike any other for all of us, but FDA has been steeped in “new” too. From PPE to diagnostics, treatments to vaccines, the review staff have been applying a new statutory standard: that which supports Emergency Use Authorization (EUA) of a product for COVID. While this is not new (see, e.g., EUAs during the Ebola epidemic), the COVID-19 pandemic required all-hands-on-deck.
Yesterday’s action by FDA gives the world a window on one way in which the Agency may be turning to Accelerated Approval; FDA has a regulatory tool, “Accelerated Approval”, that does, in some way, allow reviewers to approach the types of flexibility seen in FDA’s EUA authority where reviewers look to see that “known and potential benefits” outweigh the “known and potential risks”. FDA always considers protecting against “known and unknown risks” when evaluating every new drug, so the risk evaluation side of this EUA equation would be one familiar to FDA reviewers (see, e.g., the Aduhelm warning for Amyloid-Related Imaging Abnormalities (ARIA)), and is not unique to Accelerated Approval.
Instead, what stands out is the reliance on an unvalidated surrogate that is merely “reasonably likely to predict” clinical benefit (i.e., it is a surrogate that is not so well established that it would support a full, or traditional, approval). This is, in fact, accounting for “potential benefits”. So, maybe yesterday’s action is in some way a hidden outgrowth of the Agency’s new experience with and comfort from exercise of its EUA authority in these COVID times. It may well be that FDA’s COVID experiences have revealed to FDA that it could employ that “potential benefit” part of the EUA standard in the context of Accelerated Approval….such an epiphany could come from being hidden in plain sight all this time. Perhaps someday we will look back and count today’s action as the harbinger of those unexpected findings and this epiphany that led to a new way to position Accelerated Approval as a finding the potential benefits of a therapy outweighed its known and potential risks. Onward!