0:00
OK, thank you for the nice introduction.
0:03
I know it's a lunchtime, everybody's looking for lunch.
0:06
So I'm not going to make it very difficult.
0:10
Let me start.
0:12
So the topic is integrated formulation strategy for soluble enhancement of poorly soluble drugs from early screening to scalable oral delivery.
0:24
The topic is very big, so I cannot talk about everything in 20 minutes, but I'll give you some showcase about our approach, you know how we do I'll give you a focused introduction in early stage strategy development strategy.
0:43
So who are we?
0:48
Ardena is a global CDMO and CRO.
1:07
We have footprints in Europe and in US.
1:09
In Europe we are in Sweden, Belgium, Netherlands, Spain. And Ardena has acquired a new site in New Jersey this year in February, and we cover drug substance manufacturing, development drug products development and manufacturing.
1:32
We have analytical development and validation capabilities as well as we also provide full CMC support for IND NDA filing.
1:44
So with that I will go to my topic.
1:49
I'm going to talk about industry wide challenges that we face with poorly soluble drugs.
1:57
And you know, how do we rapidly formulate molecules, if we make formulations?
2:06
This is a very important point, specially for drug discovery companies where we have lots of limitations.
2:13
And you know, I'm going to discuss to give you some case studies of challenging molecules.
2:21
I'm going to focus on one of our enabling technologies, that’s Hot Melt Extrusion and at the end I will give you a summary.
2:29
So what are the problems to face?
2:32
You know, we know 70% of NTEs exhibit poor solubility.
2:40
This is this trend is going to be go, you know more in that we are going that direction more and more molecules are becoming more lipophilic, they are becoming larger, and they are becoming more complex in structure.
2:55
So the properties are usually classified in BCS Class 2 and class 4 type of molecules and you know bioavailability is a major problem.
3:06
So this creates problems during clinical development.
3:11
The bottlenecks are poor solubility leads to low variable absorption food effects are issues with these molecules and we have unpredictable PK profiles.
3:26
So this brings out again with problems with our first in human studies where we need, we require a lot of escalation, and you know the formulation development using this time becomes a little bit challenging.
3:42
We need to have we go through a lot of rework. At early stage development.
3:47
Again, the other problems we have is API is limited.
3:52
The physical chemical properties of the API are not well known.
3:58
We have low yield when we are manufacturing the APIs, and you know we have there's no sufficient characterization.
4:06
So this requires a fast track development approach.
4:11
That's what I'm going to discuss next.
4:14
So manufacturing scalability issues are also that come, you know the enabling technologies at early stage you start with small scale and our formulation is really scalable.
4:26
Do we need to change a lot?
4:28
This is one question you need to ask.
4:30
Regulator is another issue.
4:31
It's related with that reproducibility is very important and cost and sustainability consideration has to be taken consideration also when we are talking about enabling technologies.
4:43
So the first step of formulation development for poorly soluble drugs is understanding the DCS classification.
4:50
Molecules can be classified at DCS Class 1 where whereas I have high solubility, high permeability.
4:57
Then we have the second class of molecules that have poor solubility but high permeability and we have the DCS Class 3 molecules that have good solubility but poor permeability and therefore they don't have both good properties.
5:18
BCS Class 2 is classified into 2A and 2B.
5:24
The 2A is resolved by particle size reduction.
5:31
You increase the surface area, you decrease the particle size.
5:35
Usually you get the dissolution rate faster and the exposure is achieved just by changing the particle size.
5:42
If you go to the DCS class 2B group, you need enabling technologies where you need to convert the API into amorphous form.
5:51
For example, you destroy the crystal structure, and you make metastable forms amorphous solid dispersion.
6:01
And in that case you may need also that's the next slide.
6:07
You may need to add some other additives like lipophilic components, surfactants and so on.
6:18
So at Ardena we have for amorphous solid dispersion, 2 technologies, that's spray dry technology and hot melt extrusion technology.
6:26
So this gives us a lot of potential to bring molecules into first in human.
6:36
So what we do is at early stage, we don't know which technology works for your molecule.
6:42
So we cover, we check the hot melt extrusion, the spray dry technology the suspension, the liquid formulation, all these things will be evaluated parallel with little amount of API and other fast pace.
7:07
So how can we make it fast?
7:11
We need to have processes to make sure that we minimise experimenters in order to make sure that we don't waste a lot of material and a lot of time in conducting experiments.
7:24
So the in silico ASD modelling will help us to identify a polymer.
7:31
This helps us to see the API chemical structure, functional groups and understand the thermodynamic properties of these molecules with a library of ASD polymers.
7:44
How do they interact?
7:45
We have estimation of theoretical polymer API interaction.
7:51
The Gibb’s energy is a good information.
7:55
Basically the lower, the more negative the Gibb’s energy, the higher the interaction between the API and the polymer.
8:04
So you take the structure, you know what type of interaction do they have.
8:08
This is any silico modelling, and we evaluate several polymers based on that.
8:13
We say, OK, these polymers have the highest potential to work with this molecule.
8:19
Sometimes we don't even have the chemical structure.
8:22
What do we do?
8:23
We can take some solubility experiments and by taking by doing that solute experiment we can do some predictive modelling and identify which polymer would work with that API.
8:42
So the next step was once you select the polymer, the potential polymers, it may not be one, it may be a couple of them or three, four, depending on the how to interact with the polymers.
8:53
Or sometimes you may end up only with one.
8:56
The next step is to check it in the lab.
9:01
So for spray drying we can use solvent screening method, high throughput micro evaporation through the wells.
9:10
This with this method, we would be able to evaluate several formulations with little amount of API, with less than 200 milligrams API, we'd be able to evaluate up to 14 experiments.
9:25
And this takes up to the time down.
9:29
And once we do that, we go to the next level that is the spray drying, the small scale spray drying process.
9:36
And we evaluate that and we make formulations and we bring this to animal PK studies.
9:45
So this has to happen in a very short period of time within a few weeks.
9:53
Parallel to that, we conduct also hot melt extrusion screening.
9:59
Basically what we do here is we start with a thermal characterization with DSC, and we try to understand, you know the drug loading, and we select the polymer also.
10:15
Once we have good understanding using that thermal characterization process, we take to it to one level higher.
10:23
That is a vacuum compression moulding process which is actually assisted by cryogenic milling.
10:31
The vacuum compression moulding, what it does is it takes a polymer in the API, and you put it under vacuum and you heat it up and you try to convert it into amorphous form.
10:45
Usually if you take only that step, that's not going to be to mimic our hot melt extrusion process, which has actually a mixing capability.
10:54
So in order to mimic that, we introduce it a cryogenic milling before that. You take the API and the polymer and cryogenically mill it and do the compression moulding.
11:07
We have been able to get a good data or comparable data when it comes to the processing temperature for the on the HME.
11:16
You know, you pretty much come very close sometimes even on the spot.
11:21
By using this, you don't need to try several temperature profiles to do the hot melt extrusion.
11:26
So that helps a lot.
11:29
And this DSC discs are also very important to evaluate.
11:33
You can put this these discs, you know in 4075 open dish and try to see the physical stability.
11:43
The same material can be used for impurity studies and OK, these discs can also be used for SSKD super saturated kinetic dissolution studies.
11:57
Even though this method is not exactly the same that like the hot melt extrusion region process, it can give you also good idea about the exposure.
12:06
If you mill it and fill it in capsule and give it for animal PK studies.
12:11
This is not exactly the same.
12:13
HME is more efficient that gives you the information.
12:17
But this is directional.
12:18
At this stage we want to see directional information.
12:23
So the key here is also not to waste a lot of material.
12:27
So with little amount of material, for example, the first on the on this stage we can with less than 200 milligrams, we can check up to 12 experiments.
12:38
If we had done this on the hot melt extrusion, even the smallest extruder for 11 millimetre extruder you did for each batch 30 gram.
12:46
And if you go even by the smallest drug loading, 5% drug loading, imagine how much material you would need.
12:52
So we have with this technology you'd be able to see the feasibility with less than a gram of API.
13:03
I'll give you some case studies here for rapid HME formulation screening.
13:12
This is a molecule that we used for animal studies.
13:19
We did you know, we performed you know thermal miscibility screening micro prototype development of VCM using you know less than 150 milligram API for proof of concept prototype development via large scale HME.
13:35
You know, we have been able to identify combination of polymer carrier solubilisers, components of suitable drug loading, demonstrate ASD formulation via XRPD, DSC and non-sync dissolution.
13:51
And of course HME prototypes could be prepared later on once we you know we narrow down on 11 millimetre extruder and they can be filled, milled and filled into capsules and sent for animal studies.
14:07
So this is a rapid screening process for animal studies.
14:13
The first step which is actually very important tool, we call it a phase diagram.
14:19
You take a couple of pointers using the DSC and you take a couple of samples of drug loading, and you determine the processing window.
14:32
Basically here this molecule has we cannot process above 200°C.
14:38
For this particular example, actually the limitation was not the API but the polymer that selected and we have been able without doing any experiment that would be able to go up to 25% drug loading.
14:51
If you go outside that, this process may not work.
14:56
So we took for example, you know 10% drug loading, 20% drug loading and they were amorphous.
15:02
We did SSKD studies and this was prepared for animal PK studies.
15:08
We have been able to provide this with 2 1/2 weeks, which is really very fast.
15:15
And this is done actually parallel to the screening of the other technologies.
15:20
So we select one and this is the one that goes forward.
15:27
So how does it correlate?
15:28
You know, if you see this example, we have said number one is borderline, number two is a good spot to make amorphous solid dispersion, number three is at outside the window, number four is at the borderline.
15:44
So we duplicated it in the by extrusion and you can see number one as expected gave us amorphous solid dispersion.
15:54
The picks that you see here are picks that come from the solubilizer.
15:58
We have added in this formulation solubilizer.
16:01
These peaks are not crystalline peaks and number one as expected, gave us amorphous or a good amorphous and number 2 you can see here, we have started to see some peaks here, some residual crystallinity, number 3 which is where it's outside the window, it gave us full crystallinity, number 4 the same thing which was borderline.
16:24
So instead of doing experiments, extrusion for all of this, just by doing a few milligrams experiments, you'll be able to predict the drug loading and the temperature in the processing temperature.
16:37
So this is another issue.
16:40
You know there is a need, when the API melting point is very high, hot melt exclusion is not applicable.
16:49
But actually it's not true.
16:51
If you find the right polymer and find the right processing condition, you'll be able to do that.
16:57
For example, this molecule has a melting point around 301°C and we have been able to select a formulation using PVPVA or solar blast to make amorphous solid dispersion.
18:15
So we have to be able to select the formulation, you know to process this material at 200° C.
18:22
So key here is drug leading has to be kept a little bit lower.
18:27
The reason is by this experiment you are not melting the API, but you are dissolving the drug substance into the hot polymer.
18:36
So that's dissolving.
18:37
So we should be able to be within the solubility of the drug in the polymer.
18:43
So we took also aggressive screw configuration, and you know, we have to control the screw speed and the feed rate in order to make sure we have the right specific mechanical energy.
18:56
So by this we have been able to make amorphous formulations with process temperature about 235°C in this particular case after we scale it up.
19:10
So instead of processing it at 301°C, we have been able to process at 235°C.
19:18
Another issue is thermal decomposition.
19:20
This molecule as you can see decomposes at smelt.
19:26
So how can we process this?
19:28
So these have extensive thermal screening studies have been done to evaluate melting point depression miscibility with various HME polymers, optimise drug loading to allow drug dissolution below the melting temperature.
19:42
So by selecting appropriate temperature, no appropriate temperature, screw profile and so on we have been able actually to process this material 72°C below the melting point.
19:54
That means we avoided the decomposition temperature.
19:58
So we have been able to get a good amorphous solid dispersion.
20:04
This is for APIs with high crystallisation propensity.
20:07
That means some molecules you can get amorphous solid dispersion, but they recrystallise fast.
20:12
So here also we have been able to select a good formulation.
20:20
With relatively appropriate reselected drug loading and appropriate screw design, screw profile processing temperature and you know, we have been able to make amorphous solid dispersion.
20:32
Actually we compared this also with SDD and we did some PK studies.
20:39
The HME formulation was actually you know better than the SDD formulation.
20:44
So crystallisation propensity may be considered as a problem, but by taking the right approach would be able to handle this problem too.
20:57
This is temperature sensitivity of API you know here HME is also blamed for degrading the products when you are processing it at high temperature.
21:08
So not necessarily.
21:09
Again it's not only the temperature but the processing conditions.
21:14
The shear is key here.
21:16
Now this data shows us in the first page we have a process; the process was not optimised.
21:26
By that we have been able to see lots of impurities on this.
21:41
So key here is to decrease the specific mechanical energy.
21:45
So by decreasing the specific mechanical energy, we have been able to get no impurities.
21:53
Key here is when we are trying to take down the impurity profile, the crystalline amorphous conversion has to be also taken into consideration.
22:03
It's a sweet spot, a balance, you know you make it, you know if the specific mechanical energy is too low, you may get approved a pure material but with crystallinity.
22:15
If you do it the other way round, you may have a good amorphous solid dispersion but impurity.
22:19
So find the sweet spot, that's also possible.
22:24
I think with this I will summarise.
22:27
I think I may be close on time.
22:32
The summary is fast track, API sparing and cost effective formulation and enabling technology evaluations are necessary for early stage development and solutions for challenging molecules to make ASD by HME has to be known.
22:48
Good understanding is very important.
22:50
So we have to have systemic respect, scalable formulation development which I have not touched actually there are lots of discussions we can take further if I have the time.
23:00
But we from the beginning we have to have good understanding of the drug substance, the processes, and the scalability has to be taken into consideration.
23:09
So we have to have end to end alignment from clinical to commercial manufacturing.
23:14
Thank you so much.
