How Should We Really Feel About FDA's Biosimilar Interchangeability Guidance?

By Anna Rose Welch, Editorial & Community Director, Advancing RNA

Chances are, if you were conscious at any point in time within the past few weeks, you may have heard that the FDA released its much-anticipated final guidance on interchangeability. I jest primarily because I’ve spent the past few days reading the seemingly thousands of mainstream media and industry trade publication articles broadly unpacking the changes in the final guidance. And now that the dust has only just started to clear, I’m here to throw my own boulder into the stilling pond. (Who said writing about biosimilars can’t be poetic?) 

Overall, I’ll admit I’ve been slightly puzzled by the coverage surrounding the guidance’s release. That’s not to say the document is not improved — and significantly shorter — than it was in draft form. But I’m still trying to wrap my head around some of the comments about how positively this framework will benefit the biosimilar industry. For instance, one analyst referred to the guidance as “very reasonable,” “rather friendly to biosimilar developers,” and, in some cases, “lenient.” They also made the argument that this final guidance will stimulate more companies to pursue interchangeability.

Now, I would agree the release of this guidance is positive given that any clarity around what the agency is expecting will, of course, help companies plan their development strategies. But on a whole, the final guidance remains pretty close to the draft guidance; it’s still quite a demanding, high bar, and we can’t ignore that companies haven’t been chomping at the bit to pursue interchangeability using the parameters outlined in the draft guidance. I don’t know that the changes made to the final document will sway more companies toward pursuing interchangeability — at least for the molecules we’ve seen thus far. As I discussed in a previous article, the recent FDA public hearing on insulin biosimilar development and interchangeability left me feeling a bit hopeful we could have more action on the interchangeability front, but mostly because of the simplicity of the molecule compared to the Humiras and Enbrels of the world. However, this ultimately will depend on how the FDA chooses to approach interchangeability specifically for insulin.

I may not be a scientist or regulatory affairs professional. But as a former English major, I’m quite skilled in doing comparative document analyses. (That’s basically like analyzing biosimilarity, if you think about it.…) Overall, I’d argue the final guidance is a much more elegant document than the first. (Yes, elegant.) There were a few notable changes to mention. As many have reported, one of these changes was the elimination of the terms “residual uncertainty” and “finger-print-like similarity.” These changes will, hopefully, help eliminate notions that interchangeable biosimilars are a better, more highly similar version of an already approved biosimilar. Though the procedure for switching trials remains unchanged between the draft and final guidance, the FDA eliminated quite a few pages discussing the potential framework for device threshold analyses and potential comparative use human factor studies. Finally, perhaps the biggest change that was hailed as a huge win for the industry was the FDA’s acceptance of EU-sourced reference products, as opposed to only U.S.-sourced reference products.

I’d agree all these changes are positive news for the biosimilar industry and at least suggest a more open-minded approach to interchangeability than may have been presented in the draft guidance. Reading through the two versions of the guidance, I had a few thoughts about some of these changes, which led me to reflect more broadly on their impact on the biosimilar regulatory and development sphere.

Final Vs. Draft: More Supportive Of Analytical Innovation And Tailored Development?

An increasingly hot topic of discussion in this industry is the ability of analytical characterization to provide more accurate understanding of the originator and biosimilar molecules, in turn arguably reducing the need for large Phase 3 clinical trials. Last year, a paper published in BioDrugs by Biosimilar Development editorial board member Francois-Xavier Frapaise highlighted just how advanced analytical technologies have become. This, in turn, sheds light on the structure-function relationship of molecules and provides a more innovative, cost-effective approach forward for biosimilar development. As he argued in the article, an advanced chemistry, manufacturing, and controls package, in addition to a Phase 1 study, should provide developers and regulators with the information they need to demonstrate high similarity. There will also, of course, be post-marketing studies and real-world evidence on these products that can continue to illuminate their safety and efficacy in clinical practice.   

In light of ongoing discussions on this in articles and at industry conferences, I was pleased to note a particular — albeit small — deletion from the draft guidance to the final guidance that, to my knowledge, has yet to get any attention but does show growth on the FDA’s part. The draft guidance stated on page six, “Despite significant improvements in analytical techniques, current analytical methodologies may not detect or characterize all relevant structural and functional differences between the reference product and the proposed interchangeable product.” Though the draft guidance emphasized the use of advanced analytical characterization and its abilities to inform the overall data and information needed to demonstrate interchangeability, the discussion was much less straightforward than the final guidance. Similarly, the sentence highlighted above completely undermines the quality of the data these analytical methodologies can provide in the development of all biosimilars.

In the final guidance, instead of casting doubt on the quality of analytical technologies, the FDA more concisely outlined the impact the structural and functional complexity of a molecule may have on demonstrating interchangeability, as well as (broadly) which molecular features may impact the amount of clinical data needed. But in the slightest drift from the draft guidance, page seven gives a plug to advances being made in the technological sphere and the overall impact this may have on the data required and the clinical studies needed. The guidance states, “Advances in analytics may allow for extended analytical characterization that affect the extent of other data and information needed to support a demonstration of interchangeability and may in certain circumstances lead to a more selective and targeted approach to clinical studies intended to support a demonstration of interchangeability.” Now, similar language to this did appear in the original draft. But I think it’s important to point out that, originally, the FDA was arguing only analytical characterization techniques that led to a “finger-print-like” analysis of the molecule would be impactful on the data requirement needs and clinical studies.

How Hopeful Should We Be?

This is always the most fun question to address in this industry, and, quite often, I feel we leave the discussion more frustrated than buoyed. Do I feel as though the subtle changes in language leave this guidance more open to a case-by-case evaluation of interchangeability — Absolutely. However, will advances in analytics lead to changes in how regulators approach switching studies? My answer remains no, and that feeling is informed by a discussion which occurred at the Biosimilar Medicines Conference in Amsterdam a few months ago. A panel of regulators took to the stage to discuss the ongoing industry push for tailored clinical development and the role of clinical trial data (and specifically what kinds of data) in resolving our least favorite thing ever in this industry — residual uncertainty.

The goal of this panel was to explore which questions are left unanswered —and if those questions can be left unanswered — following an analytical excavation, if you will, of a biosimilar molecule. There were a few case study references in which the clinical results raised questions about how much trust can be placed on comparability data. As one regulator argued, these case studies — specifically related to biosimilar Herceptin — revealed the difficulty in knowing how a specific quality attribute can impact the clinical performance of a drug.

I plan to fully unpack many other questions that were discussed throughout the panel in a future article. But it was this statement about the uncertainty over how analytical and functional testing can foretell clinical outcomes that stuck with me as I read through the FDA’s interchangeability guidance. There was certainly emphasis throughout the guidance on the important role analytical techniques can play in determining the amount of information needed to attain interchangeability. As both the draft and final guidance stipulated, a product that is typically used only once likely will not require a switching study to be deemed interchangeable. Directly following this statement only in the final guidance, the FDA included this statement: “For products intended to be administered more than once, sponsors are encouraged to meet with FDA to discuss the planned development approach, including any proposed justification of why data from a switching study is not needed.” Arguably this leaves the door open for a company to persuade the FDA, either using rhetorical fireworks or analytical data, to reduce the burden of a switching study. But given the ongoing regulatory challenges and questions pervading the non-interchangeable biosimilar approval process, it is unlikely we’d see any movement toward more “targeted approaches to clinical trials” (as mentioned in the final interchangeability guidance).

But what’s also important to consider is whether the industry itself has embraced the step-wise approach regulators have touted for years, enabling companies to put their best data forward when it counts. For instance, an interesting line of questioning about the European Medicines Agency’s tailored development pilot program revealed that, of the four pilot applications received, the data provided has been too immature for the regulators to provide meaningful suggestions on tailoring the overall development process. I acknowledge that this pilot program for approving a biosimilar in the EU will differ greatly from companies most likely approaching interchangeability after having their biosimilar approved in the U.S. But this anecdote underscores the ongoing impatience the industry is facing to see movement within the regulatory sphere. As one regulator acknowledged, developers want to see their product approved and launched as soon as possible, and regulators would like to see fewer clinical trials involved in biosimilar development. But this also depends on the quality of the data being submitted in the first place to get the most valuable regulatory recommendations for moving forward with development.

Now, an argument could be made — and was made at the recent FDA public hearing — for progress in terms of interchangeability for insulin biosimilars. Actually, insulin was one of the molecules that several of the regulators at the conference in Amsterdam felt was a promising candidate for tailored development approaches in general (not related to the FDA’s interchangeability standard). But once again, it all comes down to the quality of the data manufacturers bring forth and when they bring that data to the regulators.

Comparator Requirements Change: But Is This Really Progress?

As I mentioned above, one of the biggest changes made in the guidance document is the FDA’s move to permit EU-sourced reference products in interchangeability switching studies. From a cost and sourcing standpoint, this is, of course, a valuable change to the guidance. But I do think if you look at the text of the guidance, this permission is shrouded in wording that could translate into excessive regulatory caution for those who choose a foreign reference product.

As the final guidance specifies, in order to use a non-U.S.-licensed comparator product in interchangeability switching studies, the sponsor will be required to provide a scientific “bridge” between the EU-sourced reference product and the U.S.-sourced reference product. However, “The type and extent of bridging data needed to justify the use of a non-U.S.-licensed comparator in a switching study may be different or more extensive than is needed in other contexts.” (You can find the reasons why the agency feels this way on page 17 of the guidance.)

Those of you who have been particularly in tune with the regulatory landscape of late may have felt a slight shiver down your back, as I did, upon reading these expectations. There has been an impassioned push from trade organizations and manufacturers over the past year to encourage regulatory agencies to waive the bridging studies requirements for a noninterchangeable biosimilar approval for scientific, ethical, and cost reasons. In fact, this was a common talking point in the industry’s public comments responding to the FDA’s Biosimilar Action Plan released last summer.

So, while on the surface, it sounds like a win to use an EU-licensed reference product, the language in the guidance is not altogether encouraging of such a decision, nor does it suggest we’ll see movement in the future in terms of reducing or removing bridging studies requirements for biosimilars.

As always, please feel free to reach out and share your thoughts either as a comment at the bottom of this article or via email (always off-the-record): Anna.Welch@jamesonpublishing.com.