a discussion with Don Fink
Q: To kick things off, would you please remind us of why it’s even possible to consider providing proposed guidance to the FDA?
A: Absolutely. On the FDA website there are what’s called Instructions for Submitting Drafts of Proposed Guidance Documents Electronically. There’s even guidance for how to submit a draft guide; so, guidance on how to write guidance. And this opportunity for sharing ideas is open to everyone. Anyone who wants to submit proposed draft guidance can do so. That’s an incredibly helpful proposition that has been historically underutilized. The FDA is open to hearing suggestions for guidance provided from external sources; we just have to offer it.
Q: And the difference between guidance and something more prescriptive is…
A: This is an important point for everyone working in this field to understand: guidance is exactly what it sounds like. It provides recommendations that reflect the agency’s current thinking on a topic. It gives developers something to aim towards rather than being an enforceable hard-and-fast rule. It’s always possible to out-perform recommendations provided in any particular guidance while nevertheless experiencing other problems which might lead to a rejection or delay in the approval of a product license application, or the reverse could happen: missing specific guidance recommendations but still having the necessary clinical data and overall product characterization needed to lead to an approval. That description of the term “guidance” remains true whether the “draft of proposed guidance” comes from us or any other non-government entity or is a final guidance document from the FDA itself. It’s a guiderail.
Q: What makes Dark Horse a good fit for submitting this proposed guidance?
A: Look, if anyone is in a position to help give the FDA material assistance that could help advance the setting of these guiderails, it makes sense that it would be the consultancy with the most experience and the deepest bench of clients in the fields of cell and gene therapies. And, unlike if such a document were to be submitted by a sponsor with particular self-interests, we don’t have a potential conflict of interest to address. Whatever we can do to strengthen the field is a win for product developers and we happen to have a wide-angle view of the development process due to our broad client base and depth of technical expertise. These assets make it easier for DHC to identify necessary next steps related to gene therapy product development, and we view submission of a proposed draft guidance document as a support that DHC is unusually well-equipped to provide.
The guidance recommendation we’re putting together isn’t multi-faceted or unnecessarily broad. It’s clean and to the point: as simple as FDA guidance documents get. We’re going to recommend setting a limit for empty capsids present in the final product.
Q: What makes this the time to take this step?
A: We started to discuss this internally after the FDA-convened AdComm meeting held in September 2021, focused on the topic of toxicity risks associated with AAV for Gene Therapies. (We shared our initial take on the meeting here: https://darkhorseconsultinggroup.com/aavs-the-search-for-actionable-advice-and-or-paths-forward/)
General industry reaction to the meeting indicated that yes, it was widely agreed that guidance around the use of AAV, particularly at high doses, would be valuable to the field. (Very specifically for products given by intravenous infusion: that’s important. AAV therapies that by their nature can be delivered in a targeted fashion—injected directly into the back of the eye, for example—typically require lower doses relative to systemic administration.)
AAV, as a vector modality, has constraints, especially those associated with product impurities. Empty AAV capsids represent a type of identified impurity that is unable to contribute to a direct therapeutic benefit and may instead actually be contributing something of concern, namely immunogenicity. So, there’s a section in existing FDA guidance that describes CMC Information for Gene Therapy INDs. It addresses the topic of impurities including those that are related to AAV products. The guidance says such impurities a) should be measured and b) may be reported as a ratio of…you guessed it, full-to-empty capsids. So, there is existing FDA guidance that already acknowledges the reality of these impurities and specifies that your product should be monitored and characterized for impurities…and then it stops there without providing any specific recommendations pertaining to the setting of limits for AAV impurities.
So the backdrop for what we are intending to address in our proposed draft guidance already exists, right? And, of course, it was highlighted in the FDA AdComm meeting on AAV, which was called by the agency to discuss in a public forum the concerns surrounding adverse events (including subject deaths) related to systemic high-dose AAV administration.
There was a related discussion about a statement that was made by one of the FDA staff gene therapy CMC reviewers back a while ago that drew our attention.
Q: Would you please give us more detail on that?
A: Sure…this was a meeting back in late 2020: a workshop on systemic immunogenicity considerations for high-dose AAV. A question was posed to the panel regarding whether or not sponsors should be required to substantially remove AAV empty capsids from investigational products for doses that exceed a certain threshold (in other words, high-dose AAV). Zenobia Taraporewala, from the Office of Tissue and Advanced Therapies at CBER, said (and this quote is taken directly from a publicly available recording of the meeting, link below):
“Our [FDA] take on it is you want to limit or shoot for as little empty capsids as possible, because it is obvious that this is an antigenic mode that is an impurity and many a times you will have product that will be 75% of impurity and just 25% of the product carries the vector genome which is doing the job, so to speak, or delivering the therapeutic effect. So, I think it would be generous on my part to say that having anything more than 50% empties is really a non-starter for many programs that use high doses.”
Part of why this drew our attention was because it provided a starting point of relative agreement.
Q: And bringing us back to the current endeavor…
A: Yes, so then, I was looking through cell therapy guidance, and noted that it addresses the viability of cellular material…how nonviable cells are essentially agreed to be not a good thing and something one should limit in the product. We started discussing the relationship (which is really quite direct) between this and the FTE capsid ratio.
The goal is to benchmark…to make a recommendation to establish a release criterion for AAV vector capsid impurity. We want to encourage the official setting of a reasonable, achievable impurity limit that will emphasize the area at which one maximizes efficacy while minimizing adverse side effects.
TL;DR: Impurities exist. They’re not good. Empty AAV capsids are a type of impurity and we should look at ways to remove them. In making that effort, we should strive to get to a reasonable ratio of those impurities relative to useful capsids that are going to deliver the product in question.
Q: You’ve mentioned in past discussions that this is analogous to guidance already in place in cell therapies. Would you please circle back to that for a moment?
A: Right, so, first, what are capsids? They’re a delivery mechanism, like a container ship, that exist to carry the therapy in question as cargo. Empty capsids do not contain the cargo. In many ways, they’re analogous to nonviable dead cells and in cell-based products, it’s agreed that dead cells don’t contribute a benefit and in fact, can pose a significant safety risk when given via infusion.
Q: Let me interrupt you quickly to ask for an example of those empty capsid safety risks?
A: The most common example is with the liver. That’s a key site of toxicity concern in certain contexts, and the reason is that a large number of capsids traffic to the liver. That can be dangerous if you have certain liver susceptibilities underlying your disease. Minimizing empty capsids trafficking to the liver minimizes the risk, especially because dosages are based on vector genome concentration (i.e., full capsids), but currently, the number of empties (or total capsids) in that dosage can vary widely.
Q: OK, so back to the similarities to the minimum release criterion for cellular product viability…
A: Yes, so it’s already established in writing as FDA guidance describing CMC Information for Somatic Cell Therapy IND Applications. With respect to the presence of nonviable cells, the minimum criterion for cell therapies is set at greater than or equal to 70% viable cells. In other words, if we flip and look at that criterion from the other direction, the recommendation is to minimize nonviable dead cells so that there are 30% or less of them in the total dose given to a patient. If we were to directly translate that to empty capsids in gene therapies, given the similitude between empty capsids and dead cells, that would equal a maximum release criterion of less than or equal to 30% empty capsids.
Q: And, again, this is a recommendation, not a hard line, right?
A: Right. If you as a developer find yourself unable to achieve such a level, you are free to submit data that support using a gene therapy product that has an acceptance criterion for empty capsids that is higher than the recommended maximum level. Safety data, for example, can be generated demonstrating that the presence of empty capsids exceeding the level recommended doesn’t affect patient safety, because that’s the primary goal here: ensuring patient safety first, followed by demonstration of therapeutic efficacy. This is why the deepest possible characterization of the product being developed is a best practice.
And/or there are scenarios where this may not apply, such as those examples of direct delivery at lower overall doses to target a discrete anatomic location, right? That’s the classic example of using AAVs as the vector modality, but not in the high doses generally required for broad distribution throughout the body.
A: So, the issue you’re hoping to address with the proposed guidance is already a recognized issue in the field of AAV, and it is analogous to guidance already in place for cell therapies.
B: Yes. FDA has called out this AAV issue as one of concern from its own perspective and the beauty of the solution is that this concern can be managed through the chemistry, manufacturing and controls (CMC) component of producing an AAV gene therapy product. We want to have measurements for identified critical quality attributes of the product as part of characterization that enables predictions to be made about both product safety and effectiveness. This is clearly an example of a case in which an attribute of the product (its ratio of empirical capital) is a) a measurable attribute and b) one that speaks directly to the topic of patient safety. This is essentially the low-hanging fruit of the guidance world.
We hope that our proposed guidance can support the FDA by providing an impetus to officially update existing guidance for something that’s already known to be an issue and even has a precedent corollary in cell therapy guidance.
Q: Is there anything else you’re hoping to include?
A: Well, there are some established methods out there for controlling or containing the FTE ratio, and in some cases those methods may even enrich product purity. We believe it could be helpful to acknowledge that some manufacturing issues can contribute to a higher percentage of empty capsids in an AAV gene therapy product and that there some steps that can be taken to remedy this situation leading to a lower percentage of empty capsids in the final product.
Anything of this nature we will include as supplemental information in an appendix to the proposed guidelines. Considering the technical understanding of our team (of Dark Horse scientists, as well as a highly qualified member of the academic scientific community who shall remain unnamed for the moment), why not inform the field of practical techniques available that can save precious time for those who wish to follow the guidance?
Q: And when might we expect to see this guidance submitted to FDA?
A: There’s no time like the present! We’re aiming for early 2022. Watch this space, as they say.