by Katy Spink, DHC’s COO and Managing Partner
There’s no denying that approval of aducanumab (AduhelmTM)1 is the biggest biotech/pharma topic of the quarter. It’s inexplicable on the surface, especially relative to other recent decisions, isn’t it? While I share the concerns that have been voiced elsewhere on this subject (three examples among dozens: here, here, and here), the topic I’ve found myself musing on most frequently in the last few weeks is how to view this decision in light of recent regulatory trends within the cell and gene therapy field.
In a previous post, I observed that there have been a large number of FDA-driven delays and rejections of C> products of late, and in last quarter’s inaugural issue of the Quarter Horse I sat down with DHC’s regulatory Master Practice Expert to discuss (among other things) this trend. Since the aducanumab decision three weeks ago, I’ve been thinking through parallels (or lack thereof) in our field, trying to reconcile FDA’s rigorous enforcement of CMC and clinical standards for C> products for diseases with substantial unmet medical needs (leading to delay and rejection following delay and rejection) while simultaneously apparently relaxing those standards in the case of aducanumab.
Does it make sense to precisely enforce FDA’s historical standards for decisions when considering a cell and gene product while reconsidering them for aducanumab? In considering this, let’s look at two cases of delayed and rejected C> products indicated for tremendous unmet needs: those of Mesoblast’s remestemcel-L (RyoncilTM) for pediatric steroid refractory aGVHD, and Iovance’s lifileucel for metastatic melanoma.
In the case of Mesoblast, the FDA reversed the AdCom’s recommendation. This was consistent with FDA’s historical posture of being among the strictest regulators in the world, requiring very robust and statistically valid demonstrations of clinical efficacy. Sure, fine, nothing to see here with respect to precedent.
With aducanumab, though, this pattern reversed itself: the AdCom noted insufficient efficacy data and FDA overturned them with approval of a stunningly broad label.
Those are two seemingly opposite cases of FDA overturning AdCom, aren’t they?
If we’re looking for any sort of inconsistencies in the AdCom’s Mesoblast position that could justify not adhering to their recommendation (not that FDA is obligated to so do; they historically follow those recommendations only ~80% of the time), we can point out that the recommendation was not unanimous and that the transcript shows several members on the fence and voting what accounted to a very weak yes. In other words, the AdCom’s recommendation was less enthusiastic than it may have appeared on the surface. That could be one factor in play.
As I’ve observed previously, FDA’s decision not to approve remestemcel-L also appeared to be at least partially grounded in concerns about Mesoblast’s CMC package, especially in relationship to product characterization. Now, here’s where things get really interesting. The AdCom that was convened for remestemcel-L was actually a CDER AdCom (the Oncologic Drugs Advisory Committee). While there was a morning session of the ODAC dedicated to discussion of product characterization and potency that included guest members with greater expertise on this topic, it is important to note that such concerns were not considered in the final question the AdCom voted on (which related solely to strength of the clinical data), nor are they within the bailiwick of the voting AdCom members, who are primarily expert oncologists and statisticians.
All that is to say that FDA’s overruling here doesn’t appear to be as simple as a binary disagreement on the strength of the clinical data. FDA was looking at a fuller picture that included CMC issues.
Thinking about Mesoblast’s CMC issues is a natural lead-in to the generous handful of other C&G therapies that have experienced late-stage delays due to CMC issues.
Take, for example, Iovance. It’s hard to argue that the level of unmet need in PD-1 Refractory Metastatic Melanoma is less than that of Alzheimer’s disease. And yet, despite very strong evidence of clinical efficacy, BLA filing for lifileucel has been delayed by more than a year due to FDA’s concerns regarding lifileucel’s potency assay(s). Iovance hasn’t yet even been able to make its case to the AdCom, and that’s all due to CMC issues.
Concerns about CMC issues makes sense; I’m certainly very much in favor of strong product characterization. But, should FDA be uncompromising regarding the requirement to tie proven efficacy to an in vitro assay in one case while at the exact same time an admittedly different branch is making a significant compromise on clinical efficacy on the basis of substantial unmet need?
The most logical way to explain this incongruity is to consider it from the CMC angle, especially because it ties in with what we know about patterns in C>. Consider Scott Gottlieb’s quote from his May 2018 remarks to ARM’s annual board meeting:
“In contrast to traditional drug review, where 80 percent of the review is focused on the clinical portion of that process, and maybe 20 percent is focused on the product issues, I’d say that this general principal is almost completely inverted when it comes to cell and gene therapy.”
But is it reasonable to have those CMC issues constitute what is functionally a blocking position to even filing a BLA, while concerns about the robustness of clinical efficacy data can be considered by a third-party body like an AdCom? What if we took this opportunity to explicitly acknowledge this difference between manufacturing of traditional biologics and manufacturing of C>, and as a result create more flexible avenues for review of C> CMC issues in the same ways that we’ve had available to us for many years when it comes to demonstrations of clinical efficacy? For example:
- Given that C> review failings are overwhelmingly CMC-related, why not convene separate complementary C> CMC AdComs to weigh in on those questions? The historical make-up of FDA AdComs as clinical expert review boards made sense for traditional therapeutics when 80% of review issues were clinical, but as referenced in the Gottlieb quote, this is not the case for C>. Why not just adjust the AdCom process accordingly?
- And what about a version of the “Accelerated Approval” track for CMC, in which there may be an option to allow for post-approval commitments on the CMC front in the face of stunning efficacy data in an area of significant unmet need?
This would be the CMC-version of the post-approval clinical trials required for any Accelerated Approval today and actually, such commitments would be much more achievable than the ones expected now.2 Furthermore, for therapies with high levels of unmet need where clinical accelerated approval is a possibility, having a similar path available for CMC would better align development timelines across those two important aspects of product development.
Whatever we may think about the degree of wisdom demonstrated by FDA’s flexibility in its clinical efficacy standards for aducanumab, I am sure we can agree that unmet patient need is not restricted to Alzheimer’s disease alone, and that in the future, there may be appropriate opportunities to demonstrate thoughtful and measured flexibility in other ways for products addressing significant unmet medical needs.
1 For those of you who’ve been hiding under a rock, FDA went against recommendations of AdCom and approved aducanumab for Alzheimer’s patients despite significant controversy over the strength of Biogen’s clinical efficacy data. With a stunningly wide swath of on-label indications that aren’t limited to those at the earlier stages of the disease, to boot!
2 It’s hard for studies to convert after an Accelerated Approval (from the clinical side of things) because of the difficulties of dealing with randomized trial enrollment. What patient wants to take the risk of being in a placebo arm of the trial, when they can already get access to the drug through their doctor?