FDA’s Issues Draft Guidance on Accelerated Approval: A Substantial Evidentiary and Procedural Overhaul to this High-Profile Pathway

On December 5, 2024, FDA published a new draft guidance on accelerated approval providing a much needed and substantial update to its guidance on the pathway.  FDA’s application and use of accelerated approval has evolved dramatically since it was first developed by the Agency to help address the HIV/AIDS epidemic in the late 1980s.  Since that time, it was formalized in FDA regulations (21 CFR § 314 Subpart H) in 1992, codified in the Food, Drug, & Cosmetic Act by FDAMA (21 USC § 356(c)) in 1997, revised by FDASIA in 2012, and described in guidance, most importantly, in the 2014 Expedited Programs for Serious Conditions – Drugs and Biologics (“2014 Guidance”).  However, it has been recent circumstances, such as, reforms enacted by FDORA in 2022 (as well as FDORA’s requirement to publish new guidance; see HPM’s previous coverage here) and our observations of the evolving ways in which FDA has been applying its accelerated approval authority across a wide-range of areas of drug development that has driven our desire to see FDA provide new guidance on the topic.

The new draft guidance, Expedited Program for Serious Conditions — Accelerated Approval of Drugs and Biologics, is intended to replace much of the 2014 Guidance’s discussion of the topic.  However, the 2014 Guidance cannot be ignored entirely.  FDA’s interpretation of several threshold criteria for eligibility (serious condition, available therapy, unmet medical need) will continue to rely upon the 2014 Guidance.  What will be replaced and is described in more detail in the new guidance is (1) FDA’s overarching view of the pathway’s applicability, (2) interpretation of what constitutes an accelerated approval endpoint, that is, a surrogate or intermediate clinical endpoint, (3) the evidentiary standard for demonstrating that such an endpoint is reasonably like to predict clinical benefit, (4) recommendations and policies regarding confirmatory trials, and (5) the procedures governing the withdrawal of accelerated approval.  Despite including guidance on confirmatory trials and FDA’s authority to require them, the Agency decided to publish an additional guidance focused solely on how it is interpreting the authority to require that such trials “be underway prior to accelerated approval or within a specified time period after the date of accelerated approval.” As such, we will publish a separate blog post focused on confirmatory trials and how FDA is interpreting this key provision of its accelerated approval authority.

Background on Accelerated Approval

The key concept, an ingenuity of FDA itself that has been around since the beginning of accelerated approval, is to allow earlier access to a promising therapy for a serious condition with an unmet medical need based on an endpoint that is reasonably likely to predict clinical benefit.  Accelerated approval endpoints can take one of two forms: (1) a surrogate endpoint that is reasonably likely to predict clinical benefit or (2) a clinical endpoint that can be measured earlier than irreversible morbidity or mortality (i.e., an intermediate clinical endpoint) and that is reasonably likely to predict an effect on irreversible morbidity or mortality or other clinical benefit.  Whether a surrogate or intermediate clinical endpoint (ICE), the endpoint to support accelerated approval is one that can be assessed more rapidly than the ultimate clinical benefit; this is what enables earlier patient access than would be possible while generating the evidence to support a traditional approval, which is typically infeasible.  Such clinical benefit is defined as “a positive therapeutic effect that is clinically meaningful in the context of a given disease,” is supported by a positive benefit-risk profile, and is, generally, a measure of how a patient feels, functions, or survives.

The FDCA requires that FDA also consider the “severity, rarity, or prevalence of the condition and the availability or lack of alternative treatments” when determining whether a program meets criteria for accelerated approval.  It also gives FDA authority to require that the sponsor conduct one or more postapproval studies to verify and describe the predicted effect on clinical benefit, to require pre- and post-approval submission of all promotional materials, and, in certain circumstances, to withdraw approval using expedited procedures for products approved via accelerated approval.

The most recent statutory reforms gave FDA additional authority to set conditions for postapproval studies, such as enrollment targets, the study protocol, and milestones for study conduct and completion, in addition to creating a new obligation for sponsors to submit 180-day progress reports on meeting their postapproval study conditions.  These reforms also gave FDA authority to require that postapproval studies be underway at the time of approval, expanded FDA’s enforcement authority to include failure to comply with the postapproval study conditions and reporting requirements, and laid out the procedures for expedited withdrawal.  Lastly, the reforms enacted by FDORA required that FDA publish guidance on early consultation with FDA to identify novel surrogate or intermediate clinical endpoints, use of novel trials designs for postapproval studies, the expedited withdrawal procedures, and considerations related to the use of surrogate and intermediate clinical endpoints.  FDA’s December 5, 2024 draft guidance on accelerated approvals appears intended to cover each of these topics even if not by dividing them precisely along those lines.

FDA’s Overview of When Accelerated Approval is Appropriate

The new draft guidance explains, that there are two circumstances in which the accelerated approval authority would most clearly apply: (1) when the course of the disease is long or (2) the clinical event that is relevant for demonstrating benefit occurs infrequently.  While these examples could be interpreted to mean that the key consideration is the amount of time needed to accumulate enough evidence of benefit, FDA’s authority does not limit use of accelerated approval only to such circumstances. The new guidance states, for instance, that “accelerated approval may be considered where an effect on a surrogate endpoint could be shown in a smaller number of patients” than would be required to show an effect on a clinical outcome. To us, FDA’s application of accelerated approval in numerous other instances, such as benznidazole for Chagas disease, Oxbryta for sickle cell disease, Qalsody for SOD1 ALS or Kebilidi for AADC deficiency, demonstrate that, while accelerated approval determinations are highly circumstance specific, the authority can apply to a wide-range of development pathways and disease areas.

Moreover, it was also interesting to see how FDA discussed the concept of unmet medical need in the limited fashion that it did in this new guidance. As noted earlier in this post, the new guidance is intended to replace previous guidance in most respects except that FDA’s interpretation of what constitutes a serious condition, available therapy, and an unmet medical need will continue to rely upon the 2014 Guidance. However, while FDA reiterated its interpretation that the statutory requirement to take “into account . . . the availability or lack of alternative therapies” and the regulatory requirement that the drug provides “a meaningful therapeutic benefit over existing treatments” means that accelerated approval is only available to therapies that address an unmet medical need, it chose to highlight one telling example of how a sponsor might satisfy this criteria even when there are available therapies – when the new therapy has similar efficacy but a different mechanism of action compared to the available therapies.

This nod to innovation stands as something of a reminder to sponsors to not ignore disease areas simply because treatments exist and that expedited pathways are still available to those willing to pursue new therapeutic modalities and approaches.  This may also help to derisk parallel product development amongst multiple products for the same condition, where another product in development may gain traditional approval or another product already on the market under accelerated approval may confirm clinical benefit prior to completion of clinical trials to support accelerated approval (e.g., in Duchenne Muscular Dystrophy).  It will be important for FDA to communicate to sponsors its views on whether a product would address an unmet medical need on mechanistic grounds to help ensure greater certainty for investment in the development of therapies for these serious conditions.

On the other hand, sponsors must be aware of several limitations articulated by FDA that may prevent or limit the use of accelerated approval.  First, patients may be exposed to safety risks from a drug that subsequently fails to demonstrate clinical benefit.  Second, given the potential for reliance upon smaller or shorter clinical trials (than would be typical for traditional approval), there may be less information at the time of approval about rare or delayed adverse events.  Third, accelerated approval should not be considered if the completion of an adequate and well-controlled postapproval clinical trial to verify and describe the clinical benefit is infeasible.  Each of these risks and considerations will influence FDA’s decision-making with respect to accelerated approval.  Moreover, the infeasibility of conducting an adequate and well-controlled postapproval clinical trial becomes ever more important because FDA has authority to withdraw an accelerated approval using expedited procedures if the sponsor either fails to conduct the study with due diligence and according the conditions stipulated by FDA, or that study fails to verify the clinical benefit of the product.

FDA’s recommendation to sponsors for how to handle such risks will sound like something of a common refrain (and for good reason) at this point – communicate with the Agency early and often during development.  In particular, the guidance states that sponsors should discuss (1) potential eligibility, (2) proposed surrogate or intermediate clinical endpoints, (3) clinical trial designs, and (4) the planning and conduct of confirmatory trials.  Notably, FDA in this guidance was quite keen to convey their concern that sponsors not overlook the importance of the postapproval confirmatory studies, reemphasizing in a footnote following these recommendations that the “accelerated approval pathway will not be an option for every serious disease with an unmet medical need, particularly when evidence is insufficient to support use of a surrogate endpoint or intermediate clinical endpoint, or when an adequate and well-controlled confirmatory trial would be infeasible” (emphasis added).

Accelerated Approval Endpoints and Evidentiary Criteria to Support Accelerated Approval

As noted above, the endpoints for accelerated approval can fall into one of two categories – a surrogate endpoint or an intermediate clinical endpoint, either of which must be reasonably likely to predict clinical benefit to support an accelerated approval. Moreover, an application (NDA or BLA) based upon a surrogate or intermediate clinical endpoint seeking accelerated approval must still meet the substantial evidence of effectiveness standard and contain sufficient information to demonstrate that the drug is safe for use under the conditions of the proposed labeling. However, as the new guidance helps to explain the additional burden of accelerated approval is to provide “adequate evidence” that the endpoint is reasonably likely to predict clinical benefit and goes on to provide a number of factors for sponsors to consider.

For those somewhat familiar with accelerated approval and surrogate endpoints, you will find that the guidance reiterates a now well-established framework. Surrogates are generally biomarkers, such as laboratory, radiographic, imaging, physical, or other measures that are thought to predict benefit but are not inherently measures of clinical benefit. Surrogates fall into one of three categories in terms the strength of the evidence to support their predictive capacity:

1. validated surrogates (e., one that is known to predict clinical benefit and could support traditional approval);
2. reasonably likely surrogates (e., one that is reasonably likely to predict clinical benefit and could support accelerated approval); and
3. biomarkers that do not qualify as either (e., markers for which there is insufficient evidence to demonstrate is predictive capacity or simply lack predictive capacity based upon the available data).

In contrast, FDA articulated its most detailed guidance to date on the use of; intermediate clinical endpoints (ICE). Beyond reiterating the key criteria, a clinical endpoint that can be measured earlier than irreversible morbidity or mortality (IMM) and is reasonably likely to predict an effect on IMM or other clinical benefit, the new guidance states that there is a “threshold” consideration before relying upon an ICE. FDA will consider “whether the demonstrated therapeutic effect on the [ICE] alone would be a basis for traditional approval” because accelerated approvals based on an ICE will be considered “only when it is critical to confirm the effects on IMM or other clinical benefit” (emphasis added). The new guidance provides two examples of when an ICE may be appropriate:

1. following demonstration of “a short-term benefit in a chronic disease [but] a longer duration of effect is necessary to demonstrate a clinically meaningful benefit,” and the short-term benefit is reasonably likely to predict “a longer duration of effect;” and
2. the ICE demonstrates clinical benefit on a “less serious or earlier symptom of serious disease, but the benefit observed is anticipated to predict a favorable disease outcome.”

Notably, the first example provides clarity about a long-standing point of confusion regarding accelerated approval using an ICE – that is, the same clinical measure (captured at two different timepoints) can serve as both the ICE (accelerated approval endpoint) and as the measure of ultimate clinical benefit (traditional approval endpoint). While examples of this approach are uncommon, it stands out to us that FDA included this example in its guidance. Of note, the 2004 accelerated approval of natalizumab (Tysabri) for relapsing-remitting multiple sclerosis was based on the Kurtzke Expanded Disability Status Scale (EDSS) score at one-year; two-year EDSS data were used to verify and describe the clinical benefit for conversion to traditional approval. Further, the recent approval of Kebilidi for AADC is likely to follow this approach, suggesting that FDA may be gaining some comfort with type of ICE-clinical benefit relationship.

Regardless of the type of endpoint, surrogate or ICE, the critical consideration for FDA and sponsors will be whether there is “adequate evidence” that the endpoint is reasonably likely to predict clinical benefit. The new guidance goes on to describe that such determinations will (of course) “be a matter of judgment that will depend on the biological plausibility of the relationship between the disease, the endpoint, and the desired effect, and the empirical evidence to support that relationship.” Such empirical evidence may come from a variety of sources, such as epidemiological, pathophysiological, therapeutic and pharmacologic, but it is not limited to such sources should new innovative methods or tools provide supportive empirical evidence.

The guidance goes on to warn that “pharmacologic activity alone” cannot provide adequate evidence and that clinical data “should be provided.” FDA’s use of “should” here is interesting because the guidance also stated that in the context of certain rare disease development programs (e.g., gene therapies) “where there is (sic) data supporting a relationship between the therapeutic target and the surrogate” clinical data may not be necessary. FDA could, for instance, decide that the “totality of the evidence” (which may include compelling nonclinical data) may be sufficient to determine that a surrogate endpoint is reasonably likely to predict clinical benefit.  This may represent an important difference in the application of accelerated approval between CDER and CBER, as many therapeutic approaches regulated by the drug center have similarly targeted mechanisms to traditional gene therapy but are not mentioned (e.g., antisense oligonucleotides, small interfering RNA).

Irrespective of disease area, FDA will consider “all relevant evidence and may consult external experts” or convene an advisory committee when deciding whether the available evidence supports the relationship between the accelerated approval endpoint and clinical benefit. FDA considers the strongest supportive evidence to be data from interventional studies showing that “the extent of change” in the surrogate correlates with “the extent of improvement” in the measure of clinical benefit. However, the guidance acknowledges that such data may be unavailable in certain settings, such as for rare diseases. For such cases, other sources of information, including nonclinical animal data, epidemiological data, and other clinical data will be considered “to determine if the convergence of evidence supports” the endpoint as reasonably likely to predict clinical benefit.

Finally, FDA provided several factors it considers important to consider:

1. How well understood is the relationship between the pathophysiology of the disease and the surrogate endpoint.
2. How reliable and consistent is the epidemiological evidence supporting the correlation between the surrogate and the clinical outcome.
3. Has the predictive relationship between the surrogate and clinical benefit been demonstrated in an interventional clinical trial previously, and whether the intervention was in the same or a closely related pharmacologic class.
4. Does the relationship between the surrogate and the clinical outcome require achieving a certain magnitude or duration of effect on the surrogate to predict a clinical benefit (or is otherwise necessary to support a favorable benefit-risk assessment).

A Few Concluding Thoughts

As we noted in a few places throughout this post, FDA’s new guidance frequently acknowledges that some flexibility or adjustments to the evidentiary bar may be necessary and warranted in the context of rare diseases. Repeatedly, the guidance stressed that accelerated approval determinations are case-by-case and circumstance-specific. Our experiences and observations of this space further underscore each of these points. However, we were pleased to see FDA acknowledge, explicitly in the context of rare diseases and implicitly in the context of common diseases, that deciding whether the evidence supports a determination that an endpoint is reasonably likely to predict clinical benefit can be based on a totality of evidence approach. In the context of common diseases, FDA may require more and stronger interventional data showing “the extent of change” on the accelerated approval endpoint correlates with “extent of change” in the measure of clinical benefit but nonetheless appears to acknowledge in this new guidance that other information can supplement such data in a meaningful way.

We were also excited to see FDA provide more explicit discussion of ICEs than had been included in previous guidance. That being said, the guidance does leave some ambiguity with respect to what extent FDA’s discussion of the evidentiary criteria applies solely to surrogates or can reasonably be extrapolated to ICEs as well. This portion of the guidance begins by discussing both surrogates and ICEs collectively or simply refers to “an endpoint that is reasonably likely to predict.” However, as the guidance turned to more specific factors and detailed discussion, the focus changed to surrogate endpoints. While we are disappointed by this potential ambiguity, it is our experience that much of FDA’s discussion of the evidentiary criteria can apply to both surrogates and ICEs, with the important caveat that each determination is circumstance-specific.

FDA’s new accelerated approval guidance is a must read. The discussion provides a wide-ranging and much needed update on FDA policies and interpretation of its accelerated approval authorities. Moreover, given its rich and exciting content, it is a topic we expect to return to time and again.

Want to Hear More about Accelerated Approval?

This blog post focused largely on the selection and justification of suitable endpoints for accelerated approval, as well as a discussion of the broader context where accelerated approval would be appropriate.  Undergirding the program is timely initiation and completion of confirmatory studies, which serves to limit the extent of patient exposure to drugs that may not ultimately demonstrate clinical benefit.  FDA intends to use its new statutory authorities under FDORA to set conditions for accelerated approval related to confirmatory studies, including that such studies may be required to be “underway” prior to approval.  Failure to meet these conditions may prevent approval or result in withdrawal of the accelerated approval.  Shortly after publishing the draft guidance that is the subject of this blog post, FDA published a second draft guidance that provides further context to considerations for determining whether a confirmatory trial is “underway” prior to approval, which we will cover in a subsequent blog post.