An examination of the draft position paper from the European Medicines Agency (EMA) and the Committee on Medicinal Products for Human Use (CHMP)
The EMA has drafted a “Reflection Paper on Establishing Efficacy Based on a Single-arm Trial Submitted as Pivotal Evidence in a Marketing Authorization: Considerations on evidence from single-arm trials.” The position paper has been released for consultation by CHMP from 21 April to 30 September 2023.
As many of us are well aware, while single arm trials (SATs) historically have been accepted by the US FDA as primary evidence of efficacy for marketing approval in the setting of accelerated approval (AA) of new oncology agents/indications, their acceptance in other therapeutic areas including rare diseases has been more selective. Of the 4 non-oncology AA approvals in 2022 and 1Q2023, three of the 4 studies were randomized controlled studies with a surrogate endpoint as the primary efficacy parameter. The EU has always been more selective favoring a randomized control trial (RCT) over an SAT. In my experience, requests for scientific advice wherein the trial design presented is an SAT, generally yield recommendations that the sponsor consider a RCT. For those agents that have been approved in the EU based on SATs, the efficacy data tend to be what can be characterized as “remarkable”; e.g., high durable responses in a refractory oncology setting.
In its position paper, EMA outlines their view of what would be acceptable for a pivotal SAT from a design, execution and outcome perspective. Having read the paper, I can say that that the standards have not been lessened. Scientific Advice is recommended to determine the acceptability of a pivotal SAT for the disease in question. Front and center are a need for the applicant to justify why a SAT, rather than an RCT, is an appropriate mechanism to provide clear pivotal evidence of efficacy. It should be noted that this position paper is also applicable to trials that contain more than one arm but do not randomize to a control for a formal comparison.
One of EMA’s concerns with SATs is that they lack a concurrent control arm, randomized allocation to treatment, and blinding of participants, investigators, and outcome assessors; all features of RCTs that are designed to avoid bias. Because external data is needed for a comparison of trial outcome to patients not on trial and thus not treated with the experimental drug or in a concurrent control, the magnitude of effect is more difficult to interpret and less reliable, which results in EMA’s desire to see a greater demonstration of efficacy than one would need for an RCT.
This position paper talks in great detail about sources of bias in SATs and the methods to control for them in order to be sure that the magnitude of the effect seen in a SAT is due to the intervention rather that patient selection or study characteristics. What is of note are the recommendations regarding the planning and execution of the SAT.
One of the main takeaway messages is that we will not be seeing an increase of EMA approvals for SATs that were designed as activity-determining trials and then are ‘converted’ to pivotal efficacy trials based on initial results. This position paper clearly states that for a pivotal SAT to be accepted it must be prospectively designed and executed meticulously. This is very similar to what we have seen in the recent FDA guidance on AA as discussed in our previous summary, FDA Accelerated Approval of Oncology Therapeutics Guidance.
In a SAT meant to be a pivotal efficacy trial, a clear statement as to what would be considered a positive trial is required and the SAP should be completed prior to enrolling the first patient. The choice of the endpoint and the magnitude of the response needed for a positive study must be made on clinical grounds, including the mechanism of action of the agent and the natural course of the disease. The choice of external information or comparator to which the trial results will be compared, and upon which the efficacy expectation is based, are also critically important. These can include knowledge about the natural course of the disease, demonstrating that the endpoint will not change without active treatment, or clinical data, perhaps from another study or publication. Unless there is a clear need to identify a different population, the ITT population would be the primary population for efficacy evaluation.
In planning the study and analyses, care should be taken such that protocol amendments, missing data and changes in analyses are controlled and minimized, as any of these modifications after the study starts have the potential to invalidate the use of the SAT as primary evidence of efficacy.
Subject selection and inclusion/exclusion criteria are also important, to provide reassurance that the effect seen is not the subject of a favorable subject selection process.
While it is nice that we now have a document from EMA that outlines their expectations for a pivotal SAT, the requirements do not lessen the burden to the sponsor in contemplating such an approach. In my mind, one of the benefits of this position paper is that it does a nice job in outlining the issues we should be prepared to address in discussions with EMA, in Scientific Advice discussions, and in study execution.
PROMETRIKA’s experience in direct communication with international regulatory authorities and in execution of SATs positions us to maximize our sponsors’ successful use of pivotal SATs.
