Draft Guidance - Clinical Trial Considerations to Support Accelerated Approval of Oncology Therapeutics
This guidance issued in March 2023 has been the subject of much talk. While in my opinion, it does not provide any new thoughts on the part of FDA regarding how they handle requests for approval using the accelerated approval (AA) pathway, it does put in writing the policies they have implemented in recent reviews and communications with sponsors. This should not be a surprise to us in the industry as most new guidances and initiatives actually codify existing policies which have been used in the review and approval of products so that sponsors can understand what principles and approaches are acceptable. Maybe I am showing my age, but I remember when the AA pathway was first implemented; it was modeled on the review and approval of azidothymidine (AZT) or zidovudine, the first product for the treatment of Acquired Immune Deficiency Syndrome (AIDS), in 1987. Since that time, it has been used extensively in oncology and to a lesser extent in other therapeutic areas.
There are two overarching themes in the guidance; single arm studies as the basis of approval and the confirmatory study. While FDA has been requesting that confirmatory studies be ongoing at the time of submission of marketing applications for AA for quite some time, this request is now in writing with this guidance, and as a result, the interpretation of this requirement is no longer questioned. It is noteworthy that late last year, one sponsor (ADC Therapeutics; camidanlumab tesirine, see link) reported that, “The FDA provided strong guidance that for it to consider an accelerated approval path, a randomized confirmatory Phase 3 study must be well underway and ideally fully enrolled at the time of any BLA filing.”
FDA begins the guidance discussing single arm studies with response rate as the primary endpoint. Some of the limitations of these trials include:
- Small safety databases
- Inability to interpret time-to-event efficacy endpoints such as survival
- Low magnitude response rates
- Difficulty in determining individual contribution of each component in combination regimens
- Challenges in comparing study results to historical trials
They conclude that for all of these reasons, a randomized trial is preferred; either a single trial with one endpoint for AA and another for full approval (2 in 1 study) or 2 separate studies. If 2 separate trials are proposed, the confirmatory trial should be well underway, if not fully enrolled, at the time of submission. While the choice of one or two trials is up the sponsor, the best approach will depend on whether the indication for AA and full approval are the same or if the AA indication is in a later stage population. The 2 in 1 trial approach is one that has been used for many years and is also favored by FDA as it provides further assurance that the confirmatory data will be generated in a timely manner. FDA also looks closely at the recruitment for the confirmatory endpoint and may request that the initial analysis for AA not be started until the trial is fully accrued. This is a request I clearly remember from projects I worked on in the past where we used the 2 in 1 design.
Most 2 in 1 trials in oncology will use an endpoint such as response rate or progression free survival (PFS) for the initial AA filing and measures of clinical benefit (such as overall survival (OS) or, in some cases, PFS) as the confirmatory endpoint.
Other trial considerations are outlined in the guidance. Some relevant specifics are outlined below:
- Sample size should achieve adequate power for clinically meaningful and statistically significant improvement in both the AA and clinical benefit endpoints (e.g., safety and duration of response). Trials with response as the endpoint, should be sized sufficiently to adequately determine response duration and safety.
- Analysis of response-based AA endpoints should be conducted on a pre-specified number of initially randomized patients. Analyses of efficacy to support AA should be avoided until the trial is close to or fully enrolled to mitigate potential challenges in accrual if AA is granted.
- Measures should be in place to prevent circumstances that may jeopardize the trial results or trial integrity. For example, blinding of data for the endpoint supporting verification of clinical benefit should be maintained until the endpoint’s protocol-specified analysis time point is reached to ensure a robust assessment of this endpoint.
One needs to keep in mind that FDA is concerned with safety AND efficacy, especially as more programs include combinations of agents rather than single-agent trials. Simply showing an improvement in a few outcomes versus standard of care may not be sufficient, and FDA will be looking at both benefit and potential for harm from treatment on the investigational arm. A prime example is where one sees a larger response rate in the treatment group compared to the control but an interim analysis of OS does not result in a trend in favor of the treatment arm. Even if not statistically significant, this may be cause for concern. The agent should help the patient live better, longer, or both; shrinking the tumor but not extending survival and potentially shortening it is not a benefit.
For single arm studies oncology trials, overall response rate (ORR) based on RECIST is the most common efficacy measure and what FDA generally expects to see. FDA does leave the door open using new or modified methods with the caveat that these must be supported by a strong rationale and discussed with and agreed to by FDA in advance. Another item of note which has been mentioned before in various forums but is clearly outlined in the guidance is that the product should provide a significant efficacy advantage over available therapy. As a result, we need to pre-specify the historical trials that will serve as the basis for this assessment and explain why they were chosen. In addition, the magnitude and duration of response must be considered clinically significant. If response is the endpoint, the main efficacy evaluation should include a blinded review with the response assessments including only complete and partial responses; there is a clear message that for AA approval decisions based on response, stable disease is not an appropriate measure. The guidance also requests that sample size and related considerations be pre-specified so that modifying studies to increase the number of patients in order to have a single arm to be used to demonstrate clinical benefit based on response becomes the exception rather than the rule.
Why is all of this important? In 2021, FDA requested that select oncology applications be withdrawn when the confirmatory studies were not positive or treatments for the disease in question had advanced such that a confirmatory study was not feasible. While sponsors voluntarily withdrew select applications, there were also Oncologic Drugs Advisory Committee (ODAC) meetings held to review the data and provide input to FDA on the continued need for these indications. This is discussed in more detail in one of my previous blogs (link).
In October and December 2022, members of the Division of Oncology Drug Products at FDA published an article (link) in NEJM and a special communication (link) in JAMA focusing on AA. Much of the information included in this guidance is also outlined in the October 2022 publication. In the December 2022 special communication, oncology approvals from January 1, 2002 to December 31, 2021 based on single-arm trials were reviewed with a focus on the status of confirmatory studies. Of the 116 AA approvals based on single-arm trials, 45 (38%) fulfilled their postmarketing requirements to verify clinical benefit, 61 (52%) were pending verification and 10 (9%) were withdrawn. In looking further at FDA’s list of ongoing cancer drug approvals, 15 (13%) of the applications pending verifications have passed the original project completion date agreed to upon approval.
This does not mean that FDA is cutting back on AA approvals for oncology products. In 2022, there were 10 approvals and an additional 4 in the first quarter of 2023. From their perspective, they are providing more guidance to industry, so there are clear expectations for the application for AA and there is greater assurance that the confirmatory study will be completed in a timely manner. This is further corroborated by comments made by Dr. Marks, the head of FDA’s Center for Biologics Evaluations and Research (CBER) where one of his initiatives is to clarify the use of AA for gene therapy (link). As of June 2023, two cell and gene therapy products have been approved by FDA using the AA pathway; SKYSONA (elivaldogene autotemcel) suspension approved in September 2022 to slow the progression of neurologic dysfunction in boys 4-17 years of age with early, active cerebral adrenoleukodystrophy (CALD) and Tecartus (brexucabtagene autoleucel) suspension approved in July 2020 for the treatment of adult patients with relapsed or refractory mantle cell lymphoma (MCL). Though not an oncology drug, FDA also approved delandistrogene moxeparvovec (Elevidys), the first gene therapy for Duchenne muscular dystrophy, via the AA in June 2023.
All of this is very important to those of us working on oncology products and other serious and/or life-threatening illnesses where biomarkers and other endpoints are being considered and evaluated as potential efficacy measures to support AA.
