Developing A Combination Product: Early-Stage Formulation And Device Considerations At GSK
Successful development of a safe and effective combination product depends on careful consideration of factors impacting the alignment between drug formulation and delivery device. In this episode of In Combination, host Tom von Gunden combines the perspectives of three functional area leads who work collaboratively during early-stage product development and device selection at GSK: Kris Angamuthu, formulation lead; Dominick DeGrazio, early device lead; and Sindhuja Kuchibhatla, senior manager for device innovation. Together, the trio illuminates common scenarios for cross-functional interaction, using the emergence of a new combination product as a representative illustration.
Interested in joining Tom on an episode? Contact him today.
Episode Transcript
Tom von Gunden, Chief Editor, Drug Delivery Leader:
Welcome to another episode of In Combination, the series in which I talk with folks in the biopharma industry who come at common topics from different angles of vision based on their roles in an organization or the kind of organization that they are in. My name is Tom von Gunden, Chief Editor and Community Director at Drug Delivery Leader and your host for this series.
Our topic today is early-stage combination product development. And for that conversation, I have folks who come from different functional roles in early-stage development but also work interactively and collaboratively, at times. And that's what we're going to hear about today.
So, the folks from GSK joining me today are Sindhuja Kuchibhatla.
Sindhuja Kuchibhatla, Senior Manager for Device Innovation, GSK:
Hi, Tom. Thanks for having me.
And she is, let me make sure I get this right: Senior Manager for Device Innovation.
Kuchibhatla: Correct. Yes.
Thanks for joining me. Also joining me is Dominick DeGrazio. Dominick is Early Device Lead at GSK.
Welcome, Dominick.
Dominick DeGrazio, Early Device Lead, GSK:
Tom, thanks for having me.
And also, Kris Angamuthu, who is a formulation lead at GSK.
Welcome, Chris.
Kris Angamuthu, Formulation Lead, GSK:
Hi, Tom, glad to be here.
To set the context and understand where you folks sit within the organization, and particularly in the work on early-stage combination product development, if you could each take a quick run at what your functional area contributes to combination product development and what your particular role is, just to give us a little bit of background before I start to ask you about how it all works in concert across your areas. So, I'll let any one of you who believes it's the right place to start, [to] start.
Angamuthu: So, maybe I will go first, because everything starts with the formulation, especially for the combination products. So, what I focus on at GSK is to develop formulations and manufacturing process[es] for early-stage clinical development and late-stage clinical development leading to commercialization. I specialize in sterile formulation development for new chemical entities.
Great. Thank you, Kris. Perhaps Dominick, maybe next.
DeGrazio: Absolutely. So, as an early device lead, I sit within the Early Engagement and Device-ability group. The intent of our group and my role is really to take a drug product, take several inputs based off of the Target Product Profile [TPP], and de-risk that program and that project that's interested in being formally developed as a drug-device combination product.
So, there are certain risks that would be encountered, and we want to de-risk those certain entities prior to it actually being formally developed in the late stage. And so, that's really the intent of our group at the early stage.
Great. Thank you, Dominick. Sindhuja, what does your group do and where does it come into play?
Kuchibhatla: Yes, absolutely. So, we're more focused on the technology side of it. We're taking a lot of the pipeline and portfolio needs that Dominick and Kris are monitoring and trying to pair them with the appropriate technologies. So, as we look at what's out there in the landscape and in the market, we're trying to assess [a range from] very novel technologies to very mature technologies that have been tried-and-true.
So, we're assessing them for platform-ability. We're testing them for manufacturability. And then one of the key core areas that our team really focuses on is de-risking and assessing the feasibility of these technologies at an early stage. So, really trying to make sure that we identify the right platform for the right asset.
All right. Great. Thank you all for setting the context.
So, thinking about a common scenario ― or what I assume would be a common scenario ― perhaps you have a newer, complex formulation that you're going to have to determine how that's actually going to get delivered to patients. And, assuming for the sake of this conversation that it would be a scenario where ultimately the decision is to have it become a combination product delivery, what goes into that when the program is identified and launched? What are some of the key needs and components that go into it? And how do you each contribute to how you collaborate with each other along the way? Are there common components that you look at together across your group? So, there's some that are unique to your groups. And let's just trace how it typically unfolds.
DeGrazio: I think it would all start with the formulation, ultimately, but typically will ultimately be driven by the TPP, the target product profile. So, with a given indication, for an asset, it will have this TPP, which includes a list of various things, including route of administration, dose frequency, potential dose volumes, and number of injections that would potentially be administered, [as well as] whether this is supposed to be for at-home use or to be administered solely in the clinic.
And so, we would like to take that information within the TPP and design around that, design what the product in the end would potentially look like to accommodate and meet those initial high-level TPP input requirements. But again, it would all start with the formulation as well because, depending upon what the TPP is, if we're saying that this has to be for a subcutaneous administration, ideally, they would want this to be in a liquid dosage form.
So, formulators have to develop this to be in a stable liquid formulation, to begin with, and at the appropriate concentrations. So, but that's a general starting point. It all starts from that TPP at least.
Kris? Sindhuja? Anything to add to Dominick's description there about the starting point?
Angamuthu: Yes. Just reiterating, the TPP, the quality target product profile, is the real starting point for the formulation. So, from the formulation standpoint, the drug concentrations and the dose volume required for the administration are determined, and the formulation screening and development work begin with an intent to have a stable formulation that's safe and efficacious.
And then as a follow-on for the combination product development, the aspects, like identifying the suitable primary pack container for device administration while still retaining the quality target profile of the product, is the real starting point.
Kuchibhatla: Just to build off of that, I was just going to say that TPP really allows us to open up a whole new set of criteria that we can start then diving into the primary pack, like Kris said, in the device.
So, really, what that list of requirements from the project team allows us to do is develop a new set of requirements from a technical standpoint for the device and the primary pack and be able to then create a set of criteria that we’re assessing the landscape for. So, for example, Dominick mentioned that the number of injections is important. If we're driving towards one injection, in many cases we're then stemming off of what sort of springs, what sort of devices, what sort of volume capability do we need?
And then in terms of the route of administration, sometimes that defines a lot of our needle attributes, and certain patient populations need certain injection depths and all. So, it gives us a good set of criteria that's aligned across clinical, commercial, regulatory, a lot of different functions, and primary drug product that we can then go into and start opening up the device conversations for.
And then once those device conversations really kick in or engage in full effect, what starts to happen once the focus really starts to turn toward, Okay, how are we actually going to put this into something that gets into patients
Kuchibhatla: So initially, it starts with a full landscape search of tech scouting and trying to assess all of the different devices that could meet this broad set of requirements that we've now established across the business.
But then pretty quickly, when we start to get into conversations around assessing each technical specification or each technical feature of the device, we're then bringing in Dominick's team and Kris's team to try and evaluate and start setting up some physical assessments to take this through another round of downselection.
So, we're starting with maybe 20, 30, maybe 60 technologies and trying to get down to maybe 10 that we can bring into the lab and assess further. Dominick, you can maybe talk about some of the tests that we do when we get to that point. But it really starts kind of a back-and-forth engagement between our groups.
DeGrazio: Yes, exactly. I think that the device-ability team will obviously work with the drug product team very closely to obtain the drug product, understand the various formulation characteristics and properties, understand elements of the manufacturing process that could potentially impact the behavior of the formulation and drug product, and then couple that in a way that's also device-agnostic.
So, by evaluating primary packaging, there are only a handful of primary packaging systems that could be utilized and integrated into, say, the full drug-device combination product you’re specifically talking about --say, an autoinjector, for example. You primarily have your prefilled syringes or your cartridges, right? So, in the absence of having an actual formally selected device, and when we're downselecting the array of devices that Sindhuja is looking at, prior to that we would be able to put drug product formulation on stability in the various primary packaging systems ― whether it be a cartridge or a prefilled syringe -- to start gathering data from at least a compatibility standpoint and also from a functionality standpoint. Agnostic, again, of the device.
Once we have an understanding of that general high-level compatibility and functionality information, we would be able to couple it with more specific information from the devices during landscaping that would be downselected. And then we could begin some sort of feasibility testing of the drug product in the associated primary packaging systems, with certain elements of the selected devices that we would be looking at, to begin to intertwine the two.
If we utilize, say, an autoinjector toolbox or simulator, where we don't even have the device, but we have a simulator that still behaves in a way where we can replicate certain spring forces, coupled with certain needle gauges. It's more of a plug-and-play, in a way, to at least evaluate the behavior of the drug product in the given primary packaging system with certain characteristics that are equivalent to that in certain device options that that we would be looking at.
So that's what we would do, at least at a high level from an early device-ability space, before further downselecting, narrowing down to get even more specific to the form device that we're really looking at potentially industrializing.
Gotcha. And then, Kris, when this landscaping and downselecting and narrowing of options work is underway, how does that touch you, or your group, and the thinking and work that goes on there? What does that landscaping process mean to your folks on the formulation side?
Angamuthu: That's a really interesting question. Now, when the device and primary pack landscaping is set out, it brings its own challenges to the program, meaning if it's a simple vial product, there are not many volume constraints in the given primary pack.
The dose volume administration is limited by the route of administration itself. But for device products, the primary pack and then the dose volume restrictions all could have an influence on the formulation design. For example, the concentration in the product: that's important input from the landscape assessment to let the formulators know, what are the feasible device options and what are the primary pack options that would enable these device options? [What] would provide feedback to the formulators on what would be the most optimal formulation design to enable the device and the primary pack option?
So early engagement between me, Dominick, and Sindhuja to really understand the landscape and the device options, and put that in context with the program TPP, would allow us to identify early on if formulation tweaks, or the way a formulation is put together to enable the whole device-ability ambitions. Those are critical.
You describe, really, the early collaboration or interaction among the three of you. How steady is that? How often are you or your groups in conversation with one another along the way?
Kuchibhatla: I was going to say our evaluations, by the way, Tom, are happening within days, within weeks. We're not spending months on these evaluations, which I think is very critical because we're just looking for [a] very initial read to help us not focus on evaluating 20 [or] 30 technologies at a time.
We're trying to really narrow our focus on that, so we're in constant communication. And I think that's what's been really critical about being able to work so fast is that we are quickly able to take the formulation parameters and then test with the device. But then, when we look at some of these newer devices, like we're working with very, very high springs [or] motors, sometimes, or things like that, that have physical impact on how the formulation is delivered. And we very quickly need Dominick and Chris's input on those things before we go too far in assessing those devices.
Dominick, did you want to add something there?
DeGrazio: Yeah, I was going to say that, early on, there's typically a lot of communication and pinging going back and forth, especially between the drug product team and myself, and Sindhuja, simply because the drug product team, and depending on the specific program, because each program has their own team and each team works differently. It isn't standardized or harmonized, per se.
So, there could be many different formulations that are being evaluated at a given time, or they're doing formulation development, so they're looking at 16 different formulations. And we need to evaluate, how do they go about evaluating or selecting out of those 16 formulations what is to be utilized? Now, typically the formulators think, we need to get something in Phase 1 very quickly, and that's often in a vial presentation.
But there are times the behavior of the product in the vial is much different from that in the syringe or cartridge. And so, we need to evaluate that. All of the various permutations of the different formulations have to be evaluated. And so, there's constant communication in terms of what could be a lead formulation, what could be a backup, or what could be a backup to that, or an alternative to account for a different scenario.
And so, the different array of formulation options are constantly being evaluated. And because some of the things and problems that might arise are not really seen. They are going to be seen, per se, at “t0,” right? They happen over time.
And so, when those things do tend to happen over time, then there are additional changes that are made. So, then you'd have to go back to the formulator and provide that additional input. And maybe that will change what's considered the lead option of the drug product. So, there is constant communication based off of the data that we generate, as well. And that's used as input back to them [the formulation team] And then, again, we will also align with the device innovation team from a feasibility perspective.
Because not only will we be testing the formulations in the primary packaging systems to be used as a way to select the formal formulation during the development phase, we'll also want to test these in, say, a plug-and-play manner -- with say, as I mentioned before, a simulator -- to try out different spring forces or different needle gauges if we're looking at a certain type of autoinjector. So, all of that is used as input to go back to the formulation and see what can potentially be changed or altered, and vice versa.
Kuchibhatla: One thing to add is that, apart from just assessing the devices as they are, we're also trying to assess the manufacturability of them. So, when we open up the process and the evaluation or the assessment of that device to its entire supply chain, then we really need to bring in all of our team members.
And it's like an all-hands-on-deck assessment of making sure that the packaging meets our needs, that filling is compatible with our current filling network. And then the assembly, not just the device, but even within the primary pack, is suitable. So, those come with their own set of requirements that we don't necessarily bring in right away.
But, once we have an idea of where these physical assessments are going and we have some data, we can layer on some of the manufacturing requirements and try to see how much of that is also being met. Because ultimately, when we're proposing a device to be ready for clinic, we have to make sure that the manufacturing train that comes along with it is also ready for the clinic. So, we have to make sure that those are also in sight.
DeGrazio: There has to be a certain level of maturity of the devices that we are evaluating before formalizing the team to look towards industrializing one or another, or the potential options for devices where we get the other cross-functional stakeholders involved in greater discussion. I just wanted to add on to that.
Great. Thank you. And the question that was coming to my mind was: We've framed this conversation around early-stage development. And so, maybe this concept won't have any resonance or meaning for you, but is there a point at which something no longer is “early stage”? And if not, what triggers that shift? And to what extent do any of the three of you have involvement beyond what we might describe as early stage?
Kuchibhatla: I guess to start, something moving to late stage really depends on that selection and maybe the risk assessment or the risk profile that that device carries. So, it really is making sure that whatever risks we are able to take into late stage are acceptable. So, all teams across the board, not just device, need to be comfortable with that risk profile and accept that it does meet those program needs again.
So, it kind of goes full circle back to when we initially started with the TPP and making sure that all of their criteria are met. But in terms of what that handoff looks like, it's definitely not a hard pass [of] the ball to the other side of the court. It's kind of like a very ramp-down and ramp-up for the late-stage teams. And there's a lot of overlap.
So, we've kind of made sure that that happens in conjunction, where their team members are sitting on the early-stage meetings for a long period of time. And early-stage team members sit on the late-stage meetings for a significant amount of time at the start of the project to make sure that it's not really a handoff but a very seamless transition.
DeGrazio: Yeah, I would say it's definitely not cut-and-dried, especially for, say, programs that are looking to be developed in a drug-device combination product that isn't considered to be platform. Well, it's an established platform versus something that's novel or new. So, I think for things that we've been doing: GSK has been developing certain autoinjectors -- say it's a 1 mL auto injector -- for 10 [to] 15 years. And the project is known to be, okay, it's going to be paired with this one autoinjector. I think that's a bit of a different situation because it's platform; It's been done for years. I think the transition is not nearly as long and drawn out, per se.
It is more like, here, we give you this design brief that summarizes all the early development work that's been done, the risks. And then here you go, you can take it from a late stage and then continue the development in formal industrialization. But for things that are new and novel, where there isn't really a process that's well-established, where you still have kinks to figure out, there is a lot more time involved with making the transition much smoother.
Despite early-[stage] folks like me having to be part of conversations for late phase, because a lot of the de-risking that we do has to incorporate elements that would be considered from a late phase earlier on. We have to look at those things. And so, in the late phase, individuals have to be understanding of some of the things that we're doing in an early phase.
They're aware of the data that's generated, the risks that they'll also have to account for in the late-stage development. So, I think that depends on the different scenarios, for sure.
Kris, anything from your perspective on that sort of trajectory of timelines and staging?
Angamuthu: Again, it's all going back to, how much of the risk is acceptable at the early stage, and how much of it can really be transferred to the late-stage development?
Okay. And, for any of you, are there any common scenarios or illustrative points that would say, when something moves more quickly, these are the reasons it can happen? And are there any points or scenarios where you would say, these are common ways that something might slow? Are there any things you can point to [to] say, if these things happen, it tends to go faster? If these things arise, it tends to slow down?
Angamuthu: If there is a platform device and packaging solution readily available and a formulation [that] is fully compatible across the board with this primary pack and device, and then [if] the TPP allows the seamless integration of the formulation, primary pack, and the device to move forward the development. All of this plays a key role. It's a complex and dynamic interaction between these individual counterparts. And all three individual counterparts need to chime in really well to have a seamless development and path forward. That could really shorten or lengthen the development journey.
And also, the maturity of the device and the compatibility of the formulation with the primary components that the device in itself determines the length of the journey for development.
In short, if [there are] not many risks, and then the device and the primary pack [are] at a very high maturity level, enabling a true platform option for the formulation, that could enable the shortest development journey.
Gotcha. Dominick, Sindhuja, do you have anything to add to that perspective on speed and pace?
Kuchibhatla: Yeah, I guess two things for me are, one, the flexibility with our vendors. A lot of the device manufacturing companies that we work with, when they're flexible and they understand how early we are and how complex some of the requirements that we're working with are, it tends to work very well.
And it opens up a lot of different opportunities for us to engage with them, not just on the particular device configuration that they have today, but for any device configuration that we could potentially work with them on.
And then, two, when we're doing early de-risking assessments: I think the breadth of the assessment seems to help us out more than going really in-depth in one area. So, I think, for early-stage assessments, making sure we cast a wider net and making sure we look at all of the different functions that are to be receiving this product downstream, we really consider all the considerations across the board. So yes, I would say flexibility with our vendors and the breadth of the assessment. How much we can open that up really helps us move fast.
DeGrazio: And to add on to that, I would also say that, during the development process, another thing that we tend to leverage if possible is modeling. That's another key element that will enable us to move faster where, say, we might not have the ability to obtain a bunch of drug product material. I think [that], at an early stage, you're limited by the amount of material that you could have to perform some of the de-risking feasibility studies.
And so, if you can get a little bit of material but are able to model some of the characteristics that can be used as inputs, or some of the device de-risking aspects, that can also help save a lot of the time and testing and waiting that can occur in the absence of having material to conduct the actual physical experimentation. So, that that's another big thing that we try to leverage, at an early stage. I think that's probably one of the bigger things.
Great. Well, as I move toward the close here, I have one more question ― more of an invitation, I guess, in terms of what to talk about: Using specific examples that you have in mind, or common scenarios, any final words-to-the-wise or lessons learned that you want to impart to our audience based on your experiences?
Angamuthu: So, for a combination product, the clinical phase of development, during which devices are introduced to the clinic ― that can sometimes have an impact on the timeline. For example, I have seen some commercial examples for Phase 1 clinical introduction [where] the formulation was introduced in a vial format, and the device identification and development mix right at or after [the] Phase 1 clinical program. And [then] the formulation is introduced in Phase 3 in the device. To me, that is a riskier proposition because the formulation introduced at Phase 1 in the vial format may not provide full information on the compatibility between the formulation of the primary pack and the device integration to be introduced at a later time or imported at Phase 3.
So, the early collaboration between the device groups, primary pack groups, or the formulation groups is key to the overall success of the program because this would enable the matrix team to identify the best solution for that medicine with respect to the device and primary pack. And then the formulation designs lower the risks, or minimize the risks, for the program overall. So, the early collaboration is the key.
Gotcha.
DeGrazio: Specifically, with a lot of formulators, [with] a lot of people in [the] early space in general: they're just looking to get into the clinic as fast as possible. And oftentimes, the end result of what you want to attain ― to commercialize in the market ― is not top-of-mind. And I think a lot of the problems that occur later on could be resolved, or at least mitigated to a certain degree, by thinking of the end goal that you have in mind at the very beginning.
So, that way you can design it right the first time, rather than having to make changes later on, in later development, where it becomes much more difficult and much more time-consuming. So, trying to develop the product and develop the formulation, the right dosage form, and whatnot, with what is intended for potential marketing. In the end, upfront should be top-of-mind, because what you do in the very beginning, you can have a lot of unforeseen challenges later.
If you're able to do a lot of high-throughput screening, have an array of different formulations that could be evaluated to accommodate different potential attributes based upon what you want in the end --different packaging systems, different conditions that could be seen later on -- if you can try to de-risk those things earlier on and design a product that is right from the start, I think that would help with some of the issues that are commonly encountered later on in development.
Gotcha. Good advice. Thank you for that. And Sindhuja, any final thoughts from you?
Kuchibhatla: Yeah, I think, for me, I echo what Kris and Domonick were saying about starting early and trying to maintain the collaboration across the teams. But it's also, trying to be very flexible, whether it's the people working on it, whether it's the criteria that we're working off of.
As we've started to try and get ourselves from a device standpoint earlier and earlier in the process, more things are unknown, and more things are likely to change by the time we get to selecting a device or prioritizing a device. So, it really is important for us to make sure that we are keeping things open, but starting a lot of these processes at risk and starting them ahead of time. So, yes, just openness and flexibility, whether it's with the vendor or is with our own team, trying even with the devices that we're working on or even with the formulation.
Well, thank you for that. And Sindhuja, Dominick, Kris: I want to thank you all for joining me on this episode of In Combination and sharing your perspectives on the early-stage development process. And to our audience, I want to thank you for joining us. We'll see you here next time.