Accelerating Neurodegenerative Clinical Drug Development with AI

0:06
Well, hello and welcome to everyone joining us for this X Talks webinar.

0:10
Today’s talk is entitled Accelerating Neurodegenerative Clinical Drug Development with AI.

0:15
My name is Cory Stanton, and I’ll be your X Talks host for this event.

0:19
Today’s presentation runs for approximately 60 minutes and the webinar includes an interactive Q&A session at the end.

0:25
To make the most of this session, feel free to submit questions and comments for our speakers throughout the presentation and we’ll try to attend to as many as possible.

0:32
If you’re joining us by go to webinar, you can use the questions chat box, which is located in the control panel on the right hand side of your screen.

0:39
And if you require any assistance, please don’t hesitate to contact me by sending a message using this very same chat box.

0:45
And if you’re viewing this presentation from our LinkedIn live event, you can submit your questions using the Comments tab on LinkedIn.

0:51
Please note that all audience members are in listen only mode and that this event will be recorded and made available for streaming to those who registered on xtox.com.

0:59
At this point, I’d like to thank TFS HealthScience who developed the content for this presentation.

1:04
TFS is a full service global contract research organization that supports biotechnology and pharmaceutical companies throughout their entire clinical development journey.

1:13
In partnership with customers, they build solution driven teams working towards a healthier future.

1:18
Bringing together nearly 800 plus professionals, TFS delivers tailored Clinical Research Services in more than 40 countries with flexible clinical development and strategic resourcing solutions across key therapeutic areas, including dermatology, immunology and inflammatory diseases, internal medicine, neuroscience, oncology and hematology, and ophthalmology.

1:38
Now let’s take a moment to meet our panelists.

1:40
For today’s presentation, one of our speakers is Doctor Anne-Marie Nagy.

1:44
Doctor Nagy joined TFS in September 2021 to lead the company’s internal medicine and neurosciences business units.

1:50
Her strategic and operational experience from nearly 25 years at both mid and large sized pharmaceutical and clinical research organizations has benefited biotech clients throughout their clinical development journey.

2:02
She has extensive experience in neuroscience with broad indications such as those in psychiatry and neurology, including rare and gene therapy.

2:09
And also joining us today is Doctor Sara Doan. Doctor Doan is TFS’s Executive Director, Program Strategy and Delivery for Internal Medicine and Neuroscience.

2:17
She brings a robust 27 year background in global clinical research to her role with a particular emphasis on psychiatry.

2:24
Over the past 12 years.

2:25
Her expertise spends a broad range of disorders, including Alzheimer’s, Parkinson’s, narcissistic personality disorder, borderline personality disorder, major depressive disorder, post-traumatic stress disorder, bipolar disorder, migraine and anxiety.

2:39
Recognized for her exceptional ability to cultivate and lead high performing teams, Doctor Doan emphasizes cohesion, morale, open dialogue, and a culture of accountability that fosters innovation and collaboration.

2:50
And now, without any further ado, I’d like to hand the mic over to Sara to kick off the presentation.

2:54
Doctor Doan, floor is yours, right?

2:56
Thank you.

2:57
Welcome everybody and thank you so much for taking the time to join us to have this very, hopefully very in-depth conversation that we have.

3:06
My name is Sara Doan and I’m the Executive Director of Program Strategy and Delivery at TFS.

3:13
This first slide is talking about the agenda.

3:18
We are going to have various parts.

3:20
I will be discussing parts one and two and Doctor Anne-Marie Nagy is going to talk part three and the wrap up.

3:29
In Part 1, we’ll dive into AI.

3:32
We’ll explore what AI really is, AI’s evolution from early sparks to today’s cutting edge technology.

3:40
Then we’ll go into Part 2 where we will navigate through the central nervous system and drug development.

3:46
We’ll talk about current treatment, landscape of neurodegeneration, the drug development drama, and we’ll discuss the risks and costs.

3:57
Then part three, Doctor Nagy will take over and she will discuss how AI impacts CNS drug development and research.

4:05
She’ll talk about the methods for how AI is accelerating drug development, research AI solutions for challenges within drug development in CNS.

4:15
She’ll review a case study and discuss the challenges and ethical considerations that are to be taken within AI.

4:24
We’ll then have a wrap up with the key takeaways and we’ll talk about the future of AI as it pertains to CNS drug research.

4:34
As Corey mentioned, well as time permits will follow with the Q&A session from the audience, so please feel free to ask any and all questions.

4:43
We will address what we can as time permits, and anything subsequent to that we’ll have in follow up.

4:51
To keep things fun and interactive.

4:53
We’ll be featuring an audience full with a few questions along the way, so stay tuned for that.

5:00
Let’s explore what artificial intelligence or AI is.

5:06
Artificial intelligence, termed AI for short, is giving computers a blame brain.

5:12
Essentially, this technology lets them mimic human smarts.

5:18
Think of it as a teaching machine to think and learn a bit like we do.

5:24
Who knows, maybe one day they’ll even figure out how to remember where we left our keys.

5:32
I don’t think that’ll be the case, but we’ll see.

5:36
Types of AI systems.

5:38
There’s machine learning.

5:39
Machine learning is creating algorithms to make predictions or decisions based on data.

5:46
This approach not only aids in early identification, but it also allows more personalized patient care, improving outcomes and potentially reducing healthcare costs.

5:59
Then we have what’s known as deep learning.

6:02
This is advanced programming where complex patterns, image, speech recognition are implemented.

6:10
This has shown great promise, particularly in the analysis of electronic health records and diagnosing imaging.

6:20
Finally, generative AI.

6:22
What is this?

6:23
It’s producing new content, like existing examples.

6:28
Think of ramping the same old dad jokes, keeping it familiar but never the same and way, way funnier.

6:36
Essentially, AI is capable of gathering and using knowledge, perceiving the environment, making decisions, automating actions, and more.

6:47
As we progress to the next slide, we’ll talk about some of the varying areas of how AI has developed, starting with the 1950s, where Alan Turing created what’s known as the Turing Test to check if a computer can think like a human.

7:07
Essentially, you can chat with the computer and not easily tell it’s a machine.

7:12
Well, then it passed the test.

7:15
Then when was AI officially turned?

7:18
In 1956, you had John McCarthy.

7:21
He attended Dartmouth Conference, and that’s where the term AI becomes the official name.

7:30
And you have 1966 or we have Joe Wisenberg who created what’s known as the Eliza program.

7:38
Eliza, it engages in human dialogue or chat box to hold conversations with humans.

7:47
Followed by 1997 where you have the Deep Blue.

7:51
It was a chess computer program that was designed and it beat a world chess champion.

7:59
Many other areas where we have the creation and the groundwork being laid for quick evolving field of AI, further leading us into 2011 where iPhone created Siri’s support.

8:16
It’s always ready to help, but sometimes you ask for pizza and you get a history lesson instead.

8:22
So we’ve had some areas where we’ve had to fix it along the way.

8:26
In 2014, Amazon launched Alexa.

8:29
We still use it to today, and it’s being amplified.

8:33
It’s blossoming into your own virtual personal assistant.

8:37
Then even up to now, 2020 and beyond, you have ChatGPT.

8:42
Oh, everybody uses this.

8:44
It’s the AI that is always ready to assist you.

8:48
Ask it a question, even if your biggest question is, should I wear socks with my sandals?

8:54
But it’s advancing and it’s becoming quite helpful as we progress.

8:59
To Slide 6, let’s navigate through the CNS drug development.

9:06
OK, here’s the fun part.

9:08
Let’s go into our first full question.

9:11
And there’s no right or wrong answers.

9:14
We just want to hear and feel what people are thinking and where they’re at.

9:18
So if I could give it back to Corey so that you can launch the first full question for us.

9:23
Yes, that’s right.

9:24
Thank you, Sara.

9:25
We do have an interactive poll question for audience members to respond to in real time appearing on your screen.

9:30
Now.

9:30
You can vote on the question by clicking on any of the answers and hitting the submit button.

9:34
And the question we have for you is which phase of neurodegenerative drug development best describes your current focus and your options are A) discovery or preclinical, B) early clinical trials, C) late stage clinical trials, D) post approval or simply not applicable.

9:52
This doesn’t apply to you.

9:53
Again, we’re just looking to get a sense of our audience members who are here today.

9:57
And which phase of neurodegenerative drug development best describes your current, your main focus?

10:02
And the options are discovery, preclinical, early clinical trials, late-stage clinical trials, post approval or the simply doesn’t apply to you.

10:10
Thank you to everyone who has voted so far.

10:12
I’m going to leave the poll open a little while longer to allow us to get a majority of our audience vote in.

10:16
If you haven’t voted yet, make sure you do so now and we’ll get a sense of who we’re delivering this presentation to today.

10:23
And which phase of neurodegenerative drug development best describes your current focus.

10:26
It does look like we have a majority of votes.

10:28
I’m going to go ahead and close the poll and let’s take a quick look at those results.

10:33
So we can characterize our audience here.

10:35
We have a pretty overwhelming majority, 50%, so half our audience coming in at early clinical trials, but some responses as well, 20% in the discovery preclinical phase, 25% in the late stage clinical trial phase.

10:48
No one here voting for post approval and just 5% coming in at this doesn’t apply to them.

10:53
So half our audience today seems to be focused on early clinical trials.

10:57
Sara, back to you.

10:59
Thank you.

11:00
And that’s interesting- TFS is centered around phase one through phase three mostly.

11:06
So it sounds like we have an audience that’s like-minded with how we are functioning.

11:10
So this, this is good.

11:12
Again, if you have any questions, please put them in the messaging and we’ll address as many questions as we can at the end.

11:20
OK.

11:22
As we are taking a look at the treatment landscape in neurodegeneration, what does it currently look like?

11:29
What does modern daytime look like?

11:32
Obviously, as you can see by what we have below, there’s been a notable shortage of disease modifying treatments for neurodegenerative diseases.

11:41
After two decades of hitting dead ends, in June of 2021, we were finally able to have some good news.

11:49
The FDA approved the first new Alzheimer’s treatment that was brought to us by Biogen.

11:56
Then Biogen later released another in 2023 and Lily soon proceeded thereafter in 2024.

12:04
So as you can see, we’ve had a lull and then we’ve picked up and we hope to be progressing forward with this more and more as we come more in the years to come.

12:17
Let’s go to the next slide.

12:20
In here, we want to talk about the major challenges within CNS drug development.

12:27
The unique psychology of CNS disease creates significant challenges in the clinical drug development process.

12:35
We have high failure rates due to intricate brain function.

12:40
We have longer development timelines.

12:44
We have lengthier regulatory review due to potential for severe side effects.

12:51
We have limited benefits to disease progression, offering symptomatic relief rather than modifying the disease.

13:00
So as you can see, there’s a hurdle within CNS whereas other therapeutic areas don’t have as much.

13:09
We’ll talk about this in our next coming slides as well to follow, but keep that in mind as we continue our discussion.

13:18
Let’s progress to the next in better understanding the risks of drug development within CNS.

13:26
Many CNS disorders have poorly understood mechanisms.

13:31
Our understanding of which underlying psychology and mechanism of action in neurodegeneration is limited, which complicates the development even further.

13:44
If you have Parkinson’s, Alzheimer’s, multiple sclerosis present in different ways for each patient, they have a very different very complex presentation.

14:00
This makes it challenging to apply a one-size-fits-all treatment approach.

14:06
If we consider Alzheimer’s disease, multiple classes of drugs are undergoing phase three trials where an effective treatment continues to remain challenging.

14:19
Let’s look at the risks of CNS and drug development in the next slide.

14:25
We have what’s known as crossing the blood brain barrier (or BBB for short) drug delivery in blood vessels in the brain are equipped with these tight junctions that form the blood brain barrier, a selective barrier without any pores.

14:46
The barrier limits the entry of numerous therapeutic agents complicating drug delivery to the central nervous system.

14:57
To cross the blood brain barrier, CNS drugs are often modifications which then can reduce efficacy and increase toxicity.

15:10
Let’s go on to the next slide and let’s discuss the cost.

15:16
So we’ve already talked about how difficult it is to even manage how to tackle these neurodegenerative disorders crossing the blood brain barrier or BBB.

15:29
And then on top of that, you have cost, you have development costs.

15:34
That’s the cost of developing the drugs for the central nervous system.

15:38
They’re high, they’re notably higher at a rate of about 30% more than non-CNS drugs.

15:48
Then you have an increase in cost, largely due to the complexity of CNS, which requires extensive research through clinical trials.

15:58
There’s specialized expertise to tackle the unique challenges of the brain and the nervous system.

16:05
Then what we have market risks.

16:08
CNS drugs are faced with a much lower chance of reaching the market as compared to non-CNS drugs.

16:18
This is specifically the likelihood of ACNS drug making it into the market being at about 50% less than other drugs.

16:29
This is a significant difference in points to the obstacles involved in developing these medications.

16:38
Even after gaining approval, CNS drugs encounter serious market challenges.

16:45
These include tough competition from other treatments.

16:50
They include pressure on pricing.

16:53
There’s difficulties with reimbursement from insurance companies.

16:58
Such issues can greatly affect the success and availability of the CNS drugs, making the path from development to market particularly challenging.

17:10
So with all of this, what do we do next?

17:14
Do we stop?

17:15
Do we continue?

17:16
How do we progress?

17:17
How do we use what we feel as a very plausible way to advance clinical research within neurodegeneration?

17:27
What are the next steps?

17:30
I’m going to hand it over to my colleague, Dr. Anne-Marie Nagy, to discuss how AI is accelerating CNS development.

17:39
She’ll present a case study and more.

17:42
Doctor Anne-Marie.

17:46
Thank you.

17:46
Doctor Doan.

17:49
So in this section of the presentation, I will speak about how AI can change the clinical research landscape in CNS drug development.

18:01
Maybe before moving to the next slide.

18:06
Let’s have this full question and see where do you come from.

18:11
So please Corey, if you want to pull the questions.

18:16
Yes, that’s right.

18:17
We do have our 2nd and final poll question for our audience today.

18:21
This time the question we have for you is how would you describe your organization’s use of AI in clinical research?

18:28
And your options are A) not using AI yet, B) exploring AI applications, C) actively piloting AI solutions, D) fully implemented AI tools, or E) simply not applicable to you.

18:40
And again, we’re looking to just get a sense of, as we’re, you know, talking about AI today, how you would describe your organization’s use of AI in clinical research.

18:49
Are you not using AI yet?

18:50
Are you exploring AI applications?

18:53
Are you actively piloting AI solutions?

18:55
Are you using fully implemented AI tools?

18:57
Or does this not apply to you?

18:59
I do think our audience members who have voted already will leave the poll open a little while longer so we can get a majority audience vote in.

19:05
If you haven’t voted yet, make sure to hit an option and submit it now.

19:10
We’re just looking to see how you would best describe your organization’s use of AI in clinical research.

19:15
We do have a majority audience votes.

19:16
I’m going to go ahead and close the poll and let’s take a quick look at those results.

19:20
So we have an interesting division of percentages here.

19:24
Our majority here to 42% not using AI yet.

19:28
Our next highest percentage vote 37% at exploring AI applications, 16% are actively piloting AI solutions.

19:36
We have no votes for fully implemented AI tools and 5% came in at this doesn’t apply.

19:41
So between the 42% and the 37% seem to be very much in that in that early realm of either not using AI yet or just exploring AI applications.

19:50
So you know, certainly a long way to go here I think and something to give us to think about for the future of AI and clinical research.

19:57
Anne-Marie, back to you.

19:59
Yeah, thank you.

20:00
So interesting.

20:02
So yes, AI has a great potential to offer solutions to many of these challenges that Doctor Doan highlighted here in CNS drug development.

20:11
So AI can speed up at various stages of drug development from discovery to clinical trials by analyzing large data sets and improve our decision making processes.

20:25
So in the next couple of slides I will present you several examples and several ways on how AI is driving these changes in this field.

20:34
But by no mean I would pretend being exhaustive of course.

20:38
So let’s start with the early stage.

20:40
So at early stage, AI can accelerate drug discovery by identifying novel target more quickly and efficiently by uncovering previously unknown pathways or molecular interactions.

20:55
But thanks to AI, they can screen thousands of compounds rapidly and more efficiently than using the traditional methods, identifying the most promising candidates for further development.

21:10
AI using by predicting potential adverse events or how the drugs will interact with disease target, we can better predict the efficacy and the toxicity of these drug candidates.

21:26
So this approach definitely will help us with identifying safety concern early on in the development process, reducing the risk of late stage failures and improving obviously the patient’s safety.

21:40
Another way AI can accelerate drug discovery is by predicting the peaky sort of pharmacokinetic and pharmacodynamic profile of drug candidates.

21:52
So the AI systems can stimulate how the CNS drugs will be absorbed, distributed, metabolized and excreted in the body and will help us with optimizing the dosing regiments, the delivery methods or ensuring that the drug reached the intended targets in the brain while minimizing local or systemic side effects.

22:17
So obviously now you understand that such models are particularly relevant for CNS drugs that needs to cross the blood brain barrier effectively.

22:29
And then a few words about biomarkers.

22:31
So as you know, there is a high focus on finding the Holy Grail in the holy biomarkers that will help us to correlate the biomarkers with the clinical efficacy of a drug in a clinical trial.

22:46
So there is a high pressure from regulators and from all the stakeholders involved with the clinical trials to involve biomarkers.

22:56
So AI will definitely help us identify novel biomarkers that ultimately can be used for early diagnosis, monitoring the disease progression and or has to the endpoint.

23:10
For now, all the study endpoints like in Alzheimer’s or Parkinson’s are basically subjective endpoints using subjective rating scales or questionnaires.

23:20
So these biomarkers will definitely help us to provide qualitative data at the end of the day.

23:31
In the next slide, we will see so when we are moving to from drug discovery to the clinical trial itself.

23:41
So AI can also make clinical trials more efficient by optimizing protocols, improving the study design and helping identifying the right patient population that we need to include.

23:56
So AI can analyze the genetics, biomarkers or clinical data to stratify patients more accurately into highly specific subgroups based on the disease stage, treatment increasing so the progression rate or response to a therapy.

24:13
So AI will definitely improve trial outcome by decreasing the variability of the patient cohort.

24:20
It will also help us implement adaptive client clinical trials where the study design can be modified based on interim outcomes.

24:29
For example, AI can analyze ongoing data to adjust the dosage, to modify the patient group, to adjust some inclusion criteria or the sample size, or even halt the trials early in the treatment if the treatment shows significant efficacy or potential arm.

24:49
So ultimately AI will make the trials more efficient and they will reduce the time to market.

24:57
Integrating data coming from real world evidence such as clinical records, imaging data, genetic information or patient reported outcomes will also better inform on the study endpoint, which to the endpoint to include and on study design, decreasing the risk of failures.

25:20
In the next slide, I will speak about how AI will positively impact patient engagement by enhancing patient recruitment and retention.

25:32
So I mentioned AI can analyze electronic medical records of the patient.

25:36
And by doing this, we should be able to better locate where the patients are located in a map and who are those patients who we need to enroll in a clinical trial with some specific criteria.

25:53
So it will help us to recruit more accurately a better subgroup of patient population by using the electronic medical recourse or any source data that are existing outside.

26:06
AI algorithms can also analyze imaging such as MRI PET scan to detect early signs of Alzheimer’s or Parkinson’s disease, often before clinical symptoms appear.

26:20
So early detection is crucial for enrolling patients in trials at the stage where interventions might be the most effective.

26:29
And finally, AI driven tools like wearable devices on mobile apps can monitor the patient’s symptoms and adherence to trial protocol in real time.

26:40
On top, such mobile apps that track symptoms and provide personalized feedback will empower patients to take an active role in managing their disease or their participation in a clinical trial.

26:56
So moving to the next slide, I would like to speak about this exciting recent development that showcase how AI, specifically machine learning based video analysis, could revolutionize the way we detect early symptoms of Parkinson’s disease.

27:20
So as a reminder, Parkinson disease is marked by motor symptoms such as bradykinesia, slowed movement as well as tremor and rigidity.

27:30
So the study was conducted by researchers at the University of Florida this year.

27:35
So it’s a very recent study and they showed how machine learning based computer analysis of video recordings can detect movement dysfunctions associated with early stage Parkinson’s faster than the human eye.

27:51
So the results are shown on the right side of the slides and are very promising.

27:58
So the machine learning algorithms achieve detection accuracy of 79% using fingertips videos, 75% with hand movements videos, and 79% from leg agility videos.

28:13
When combining all these three tasks, the accuracy increased to an impressive 86%.

28:22
Perhaps more strikingly, this method correctly classified 97% of Parkinson’s patient and 73% of those without the disease.

28:35
So definitely these findings show the potential of AI in enhancing our ability to diagnose and treat neurological disorders, potentially leading to early intervention.

28:49
Now let’s have a look to the challenges associated concomitantly to the use of AI in drug development.

29:05
So in this section, I will discuss, I will not be exhausted either, but I want to show you that these AI tools and methodology will bring also a lot of challenges that all the stakeholders involved in a clinical trial will have to address.

29:26
So from a sponsor perspective, they will have to navigate the complex landscape of AI regulation in healthcare.

29:34
They will have to ensure that the AI system will generate data quality integrity.

29:41
So the decision-making using AI tools or AI data will be as good as the data will be good.

29:50
And they will have to address responsibility in case of an AI related errors or adverse.

29:56
Even so they will have to discuss with regulators and more guidelines will be needed concomitantly to the use of these AI from the investigator perspective.

30:12
So to the next slide, they will have to ensure the scientific validity of using AI.

30:19
So they will have to ensure AI-driven research methods meets rigorous scientific standards.

30:26
They will have to ensure that they will be able to integrate AI within the existing practice as well and that they will have enforced the right skills and expertise to effectively use and interpret AI data in CNS research.

30:46
And finally, from the patient perspective, obviously the biggest concern that we need to address will be the privacy and the data protection.

30:56
So as you know, AI is collecting a huge amount of sensitive neurological data.

31:03
So the patient will have to understand also how AI will be used in their treatment or during the participate, during their participation in a clinical trial.

31:13
So this has to be addressed during the consenting process and we will have also to address any worries or apprehensions using AI system that might not accurately represent diverse patients with categories or relying on AI for clinical health decision.

31:38
So this is ending our presentation.

31:42
So as a wrap up and the main key take away, AI is not a new concept.

31:53
It has evolved exponentially over the last 20 years.

31:56
So it’s time to consider AI offerings and capabilities.

32:02
It has a lot to offer to mitigate all the challenges that currently exist in CNS drug development.

32:10
But concomitantly, we also will have to address the associated challenges.

32:19
So that’s ending our presentation.

32:22
So and I will open the floor for any Q&A.

32:29
Yes, that’s right.

32:30
Thank you very much, Anne-Marie and Sara, I’m just going to bring by webcam back here so we can all come back into the full.

32:37
Really appreciate your insights from that presentation.

32:39
As you mentioned, we are now ready to begin the Q and A portion of the event.

32:42
And as we do so, I invite the audience to continue sending questions or comments.

32:46
Just a reminder, if you’re attending via go to webinar, please use the questions chat box.

32:50
And for those of you watching on LinkedIn live events, feel free to post your questions using the comments tab.

32:55
We already have received some audience questions and we’ll see how many we can hit in the time that we have left.

33:00
But I’ll go ahead and kick off with the first one.

33:02
And this one’s probably best suited I think for Sara to answer. This audience member was asking about how does AI driven patient stratification differ from traditional methods in terms of identifying participants for these neurodegenerative trials.

33:18
OK, good.

33:19
Good question.

33:21
AI driven patient stratification offers several advantages over traditional methods in identifying participants.

33:31
For neurodegenerative trials, there’s data integration.

33:36
AI can analyze diverse data sets where traditional methods cannot.

33:43
We rely on limited criteria and we overlook variables.

33:48
Then we have predictive modeling.

33:52
AI algorithms can identify patterns and correlations that might not be standard with traditional methods.

34:02
Doing this we can allow for more precise predictions, disease progression, so a finer pinpoint.

34:10
Then we have personalization.

34:14
AI can help identify subgroups of patient databases.

34:19
Traditional methods typically used broader classification.

34:24
We can do efficacy, efficiency, we can reduce bias.

34:32
As Anne-Marie was alluding to, there’s plenty of tests within Alzheimer’s and within Parkinson’s where it is really radar based and instead we would be able to use AI to assess and pinpoint the information and the details better.

34:51
So overall AI approaches are actually a supplement to what we’re doing.

34:57
It’s not anything where we would have disadvantages, rather it’s a add to, if you will, to the traditional methods that we have that would help expedite, that would help make efficient and quite honestly, it would help to bring patients to light sooner and faster than our traditional methods would.

35:22
Thanks, Corey.

35:23
Great, excellent.

35:24
Thank you so much for that.

35:25
Sara, we have another question here from the audience and I guess I think this one I’ll direct towards Anne-Marie.

35:32
This audience member is asking about what role could AI play in overcoming the challenges of crossing the blood brain barrier in drug delivery.

35:41
I know we talked about that a little bit during the presentation.

35:43
Maybe you could expand a bit more on how AI could help in those, you know, the BBB as Sara call it, challenges in drug delivery.

35:51
Oh, yes, certainly.

35:52
That’s really an interesting one.

35:53
So yeah, yeah, I have touched base a little bit in my presentation about this question.

35:58
AI will help optimize the drug design and creating drug interaction in the pharmacokinetic or pharmacodynamic of the drug.

36:08
It can also identify novel daily free mechanism.

36:11
So a couple of examples, I’m thinking of AI can assist in optimizing existing drug candidate by suggesting modification to improve the blood brain barrier permeability.

36:24
Or it can help design innovative nano carriers like liposone nanoparticles or exosomes for drug delivery that are optimized for blood brain barrier crossing over.

36:40
Or as I mentioned in my presentation, we can think about also better predicting the timing, the dosage or the delivery method by simulating the pharmacokinetics and the pharmacodynamics model of drugs in the brain.

36:56
So there are many, many more examples on how we can optimize drug candidates compared to the traditional methods, but it will for sure accelerate the discovery of new target.

37:13
Great, excellent.

37:14
Thank you so much for that.

37:16
Anne-Marie.

37:18
Another question here that’s come in from the audience this member was asking about, you know, we mentioned how biomarkers can be predicted by AI.

37:29
Do we have a list yet?

37:30
Do we know which specific biomarkers, you know, how, how do we, how has had that information been delivered at this point?

37:37
No, it will be.

37:38
So we know actually there are many biomarkers that are currently on the research in Parkinson’s and in Alzheimer’s.

37:47
The biggest issue that we have with biomarkers today is the lack of correlation of these biomarkers of with the clinics.

37:59
So there is no correlation with a clinical efficacy.

38:02
So we still have to use primary endpoints.

38:06
So that is where AI will accelerate to help us to define how the biomarkers are involved in the mechanism of the pathophysiology of the of the disease.

38:19
Because for now, maybe we are we are studying the right biomarkers, but we are not able to correlate them to the clinic.

38:28
So by understanding better the physiopathology of the disease, it will allow us to better identify novel, and potentially completely novel biomarkers, by better understanding the underlying mechanism of the pathology of the, of the disease.

38:50
Great, excellent.

38:50
Thank you for that.

38:52
Anne-Marie, I think we’ve got time here to just hit one more question and Sara, I think I’ll direct this one towards you and maybe Anne-Marie may want to chime in as well.

39:00
This audience member was asking about, you know, in CNS trials, placebo response is often, you know, it’s a great problem when it comes to evaluation.

39:10
Is there an AI based method to handle this issue?

39:12
Can we bring that into the fold?

39:14
Yeah, absolutely.

39:15
There’s several methods being explored with AI and placebo responses.

39:20
In the CNS, we talked about predictive modeling, how AI can analyze the historical data to identify patient characteristics.

39:32
We have adaptive trial designs where AI can optimize trials in real time based on interim results.

39:39
So we don’t have to wait until the entire study duration.

39:44
We have the machine learning algorithms that we talked about.

39:47
They can identify broader biomarkers or other indicators that can correlate with placebo responses.

39:56
We could do simulation models.

39:58
AI can simulate various trial scenarios and predict the different impacts of the placebo effects that can happen.

40:08
There’s a multitude of different ways, but absolutely putting in how the placebo variants can go into effect.

40:17
These are the approaches that we would use to enhance the accuracy and reliability within the central nervous system of the trials, minimizing any influence within placebo responses.

40:30
Great, excellent.

40:31
Fascinating on that one.

40:32
Anne-Marie, did you have anything to add or thought on that one?

40:35
No, it was perfectly summarized by Doctor Doan.

40:40
But what I would like to add here is about the placebo.

40:44
It’s really a concerning issue the placebo response, in the neurodegenerative diseases like Parkinson’s or Alzheimer’s because there is such an unmet need to have good treatment available for the patient.

41:03
So doctors don’t show that there are actually a couple of treatments that are now finally approved, but these treatments will only decrease the progression of the disease.

41:18
And for now, this decline of the disease progression has an effect up to 18 months only, right.

41:28
So these are not probably the one that will cure the disease.

41:32
They will just slow down the progression so far.

41:34
So we still have a lot to do in order to find the treatment, right.

41:40
So the placebo response is something that definitely is one of the reasons why the clinical trials are failing.

41:54
So the AI with all the AI methodology and data integration and having this data in real time.

42:02
So being able to stop this trial early on or to adjust the trial in order to mitigate the placebo response on time rather than the end of the at the end of the study or at the end of the clinical development plan is really critical, you know, for the drug development in CNS.

42:21
So that’s what I, I wanted really to, to highlight here that this passive response is actually not allowing to prove to find the right drug treatment because of all it will imply in the clinical trial.

42:41
Excellent.

42:41
Thank you for, for that insight on it.

42:42
I think it’s probably a good place for us to stop as, as we’re sort of running out of time here.

42:47
And it leaves us on the note of, you know, we’re, we’re probably just on the cusp of what AI can do.

42:52
And, and we’re going to, it’s going to be very interesting to see how this all evolves in the future.

42:56
I do thank you both for your answers to those questions.

42:59
I thank the audience for sending in your questions.

43:01
If you have any further questions or if you’re someone who’s watching the recording on this webinar, you can get in touch with the team at TFS using the contact info up on screen.

43:08
You’ve got 2 links for you there to note down.

43:10
There’s the info@tfscro.com e-mail address and you can also just visit tfscro.com and explore the website there and the team will have a chance to follow up with you after the webinar.

43:20
I do thank everyone once again for participating today.

43:22
If you are attending by go to webinar, I’ve added a link to the chat box where you’ll be able to view the recording of this event and you can share this link with your colleagues so that they may register to view it as well.

43:31
You will also be receiving a follow up e-mail from X Talks and after we wrap things up here, a survey window will appear on your screen.

43:36
Your participation in the survey is always appreciated as it helps us to improve our webinars.

43:40
For those of you watching on LinkedIn Live, we encourage you to create an account on xtalks.com and register for this event, which will not only give you access to today’s recording, but to our entire library of free recording, including other webinars, editorial articles and and much more.

43:54
Now please join me once again and thanking our speakers for their excellent presentation.

43:57
Doctor Nagy, Doctor Doan, thank you so much for your time here today.

44:02
We really appreciate it.

44:03
We hope our audience found this informative and we wish you all a great rest of your day.

44:07
Take care everyone, Thank you.

44:10
Bye.