170 - Emerging Trends in Personalized Development (S12E5)

Welcome to the deep dive. You know, for the longest
time, medicine was about finding treatments that
worked, well, for most people. Sort of a broad
approach. Exactly. But things are changing fast.
Today, we're diving into a real revolution happening
in pharmaceuticals. personalized development.
It's a fascinating shift. We're moving beyond
that one size fits all model. And the cutting
edge now is all about tailoring therapies specifically
to you, the individual patient. That's right.
We've gathered quite a bit of information looking
at how new strategies are really taking hold.
Things like biomarker driven trials and importantly
patient centric design. These aren't just buzzwords
anymore, are they? Not at all. They're becoming
central to how we actually develop and even deliver
new medicines. It's more than just tweaking old
drugs. It feels like a, well, a fundamental change
in thinking. And that's exactly what we want
to unpack for you today. We're aiming to get
past the high -level concepts and understand
the practical ways this personalization is actually
happening. Yeah. How is all that individual patient
information being used? What does it mean on
the ground for treatment? Precisely. So let's
get into it. Where do we start? Well, a really
foundational piece of this puzzle is the rise
of biomarker driven trials. Okay, biomarkers.
What does that mean in this context? So what's
really transformative here is that we're moving
towards selecting people for clinical studies
based on specific biological characteristics.
Their individual makeup. Ah, I see. So instead
of just enrolling anyone with, say, condition
X. Exactly. You target those who share particular
molecular signatures or biomarkers. The idea
is you're finding the patients who are most likely
to actually benefit from that specific treatment
being tested. Right. So you're not casting such
a wide net anymore. You're using these biological
clues, these markers, to hone in. Precisely.
And this is having a huge impact already. Can
you give an example? Oncology is a great place
to look. Think about certain lung cancers. If
we can identify specific mutations, like EGFR,
for instance, then we can use highly targeted
therapies. And these often lead to significantly
better outcomes than, say, traditional chemotherapy
for those specific patients. Because the drug
is designed to work best in people with that
exact marker. Exactly. The trial can clearly
show the drug works really well for that group
because you focused on them from the start. That
makes a lot of sense. A really clear example
of how a marker unlocks better treatment. OK,
so that's biomarkers. You also mentioned patient
-centric design. Yes, another really key trend.
And this sounds like what it is, making the whole
drug development process much more focused on
the patient's actual experience. So putting their
needs and maybe preferences right at the center.
Absolutely. It goes beyond just, does the drug
work? It really asks you to consider things like
the burden of the treatment itself. You mean
like how difficult it is to take side effects?
Yeah, exactly. How does it impact their daily
life? What's the effect on their overall quality
of life, not just the disease symptoms? So it's
about the whole picture, ease of use, tolerability,
how it fits into someone's actual life. Correct.
And these considerations are genuinely starting
to shape things like how trials are designed,
what outcomes are even measured, sometimes even
the drugs formulation or how it's delivered.
Interesting. It feels like a much more holistic
view. It is. Now, if you think about what powers
both these biomarker strategies and this patient
-centric approach, it's really the ability to
integrate and analyze huge amounts of patient
data. Ah, the data piece. That seems crucial.
Where's all this individual data coming from?
Well, it's exploding, really. We're getting genetic
information, detailed electronic health records,
lifestyle factors, even data from wearable trackers
and things. Wow, that's a lot of different streams
of information. How do you make sense of it all?
And that's where it gets really sophisticated.
Artificial intelligence and machine learning
are becoming absolutely invaluable tools for
sifting through all this complexity. AI in drug
development, it really does feel like we're stepping
into a new phase. What can AI actually do with
that data? What kinds of patterns can it find?
The real strength of AI here is finding these
incredibly intricate patterns in massive data
sets, patterns that, frankly, humans might just
miss. OK. So as some of the literature points
out, Different types of neural networks are being
used. You have deep neural networks, or DNNs,
which are great at learning complex nonlinear
relationships, perfect for predicting drug responses.
Then you have things like convolutional neural
networks, CNNs, which are really good at analyzing
medical images, maybe finding diagnostic markers
and scans. I see. Image analysis. And recurrent
neural networks, RNNs, they can handle sequential
data, like a patient's health history over time,
to maybe understand how a disease progresses.
And feed for. networks. FFNs are often the sort
of fundamental building blocks for many of these
more complex systems, so different tools for
different analytical jobs. So AI isn't just one
thing, it's a whole toolkit for making sense
of this complex individual biology and predicting
responses. Precisely. And these insights, they
directly feed into how we customize treatment
strategies. By understanding a patient's unique
profile, we can tailor therapies much more precisely.
Okay, let's talk about that customization. How
does it actually change what happens to a patient?
Well, it can influence several key things. First,
as we said, picking the right drug based on their
biomarkers. But second, and this is really important,
it allows for optimizing the dose. The dosage?
Why is that so variable? Well, this gets into
something called pharmacokinetic and pharmacodynamic
variability, PK and PD. Big terms, but crucial
concepts. OK, PK and PD variability. Break that
down for us. Why does the same dose affect people
differently? So think of pharmacokinetics, the
PK, as what your body does to the drug. How it
absorbs it, distributes it around, metabolizes
it, gets rid of it. Right. The drug's journey
through me. Exactly. And pharmacodynamics, the
PD, is the flip side. What the drug does to your
body. Its effects. Both the good ones and the
side effects. OK. Body on drug. Drug on body.
Got it. and these processes. They are influenced
by loads of factors that differ between individuals.
So my journey and the drug's impact could be
quite different from yours, even with the same
pill. Absolutely. For instance, a drug's basic
properties, like how well it dissolves in water
versus fat, really impacts how it's absorbed.
And that interacts differently with individual
body compositions. Makes sense. And then there's
protein binding. Drugs often latch on to proteins
in your blood, like albumin or AGP. How much
they bind can vary, and that changes how much
free drug is actually available to do its job.
Interesting. So less free drug might mean less
effect. Potentially, yes. And maybe the biggest
variable is metabolism, how our bodies break
down drugs. This involves enzymes, especially
the cytochrome P450 family, but also others like
aldehyde oxidase, xanthine oxidase. There's huge
variation here between people. And that variation
is often genetic, right? Often, yes. So you might
have genes that make you break down a certain
drug super fast, meaning you don't get enough
effect at a standard dose. or you might break
it down very slowly, leading to buildup and potential
toxicity. Wow. So understanding my specific metabolism
for a drug could lead to a much better dose for
me. That's exactly the goal. More precise dosing,
choosing drugs that fit your metabolic profile,
aiming for better results and fewer problems.
This level of detail is really... Quite amazing.
The potential seems enormous. Can we talk about
some other specific ways this customization happens?
Sure. We touched on using genetic info to predict
response or side effect risk. That's a direct
application of biomarker strategies. Right. Another
maybe more established example is adjusting doses
based on body surface area or BSA. Body surface
area. Yeah. It's common, especially in cancer
chemotherapy. It's a way to account for the fact
that body size influences how a drug is distributed
and cleared from the body. So instead of just
one standard dose, the calculation takes your
size into account. Correct. It's a step towards
personalization, recognizing that one size doesn't
fit all. And this whole drive is changing clinical
trials too. How so? If you're testing personalized
approaches, the trials themselves must need to
adapt. They absolutely do. We're seeing more
sophisticated designs emerge. Like what? Well,
things like biomarker stratified trials. Here,
you divide patients into groups based on whether
they have a specific marker or not. Then you
can see how well the treatment works in each
specific subgroup. Okay, so testing within defined
populations. Right. And then there are also strategy
designs. These actually compare the overall outcome
of using a biomarker guided treatment plan versus
just using the standard non -personalized approach.
To see if the personalization actually adds value
overall. Exactly. These designs are really important
for proving the benefit. But as you can imagine,
they add complexity. I bet. Finding enough patients
with specific markers, analyzing smaller groups,
it sounds challenging. It definitely is. Recruitment
can be harder, and the statistical analysis needs
to be very careful. Now, thinking about the real
world science behind this, we often talk about
publications like Organic Process Research and
Development, OPRND, focusing on the how -to of
making drugs. How does that kind of work connect
to personalized medicine? That's a great point.
While OPRND might not be publishing studies on
specific personalized therapies being discovered,
the work they focus on is absolutely critical
for personalized medicine. How so? Think about
it. OPRND is all about the detailed chemistry,
manufacturing, controlling the quality of drug
substances and products. Right. Yeah, getting
the drug made consistently and well. Exactly.
Now, imagine you figured out a very precise personalized
dose for someone. That precision is useless if
the drug formulation isn't optimized so that
exact dose is actually absorbed properly by that
individual. Ah, I see. So things like optimizing
solubility, how the drug dissolves, which is
a big topic in OQR &D and formulation science,
become even more crucial when the dose is tailored.
Absolutely paramount. You need to ensure that
specific personalized amount gets into the body
effectively and predictably. The fundamental
process understanding from OPRD underpins that.
So the making it part has to be incredibly reliable
for the using it personally part to work. Couldn't
have said it better. And another angle is analytical
methods, developing super sensitive, accurate
ways to measure things. Like measuring the biomarkers
in a patient sample or the drug level. Precisely.
That's vital for identifying the right patients.
and for monitoring if the personalized treatment
is actually working as intended, maintaining
the right drug concentration. That development
and validation of analytical methods is core
OPR &D type work. Okay, so even if it's not the
headline discovery, that rigorous process science
is the essential foundation enabling personalized
approaches. It's indispensable, absolutely. That
really connects the dots. So, okay, the potential
is huge, the science is advancing. But what are
the big hurdles still in the way? Well, there
are definitely significant challenges. One major
one is continuing to develop and roll out robust,
easy -to -use diagnostic tools. The tools to
actually measure those biomarkers reliably and
efficiently in everyday practice. Exactly. We
need those to be widely available and accurate
to guide treatment decisions effectively. Makes
sense. What else? And then there are the logistical
and, frankly, ethical considerations around collecting
and using all that sensitive patient data. Privacy,
security, consent, these are big issues. Yeah,
handling all that personal health information
responsibly is critical. Absolutely. And as we
mentioned, designing and running the clinical
trials for personalized medicines is still complex.
It requires new statistical thinking, operational
adjustments. So diagnostics, data ethics and
logistics and trial complexities seem like key
areas needing ongoing work. Definitely. But looking
forward, the trajectory is pretty exciting. What
do you see coming down the pike? Well, we can
definitely expect AI and data science tools to
get even more sophisticated, refining our ability
to predict responses, identify subtle patterns,
and tailor therapies with even greater precision.
More powerful analytical engines. Right. And
the basic research continues identifying novel
biomarkers, understanding disease mechanisms
at a deeper molecular level. All of that feeds
back into creating new personalized targets and
strategies. The whole field seems incredibly
dynamic. The idea of treatments becoming so precisely
matched to us as individuals. It's quite something.
It really is. So just to sort of recap our conversation
today. We've really seen these key trends emerging
strongly in personalized development. Right.
The move towards biomarker driven trials. The
increasing focus on. Patient -centric design,
thinking about the whole experience. And underpinning
it all, the integration of vast amounts of patient
data, analyzed smartly, to customized treatment
strategies. And the ultimate goal, the potential
benefit for you listening, is clear. Therapies
that are hopefully more effective, more targeted,
because they actually take your unique biology
into account. Fewer side effects? better outcomes
potentially. That's the promise. Which leads
us to maybe a final thought for you to chew on
as we get better and better at understanding
and addressing the unique biological signature
of each person. What does that ultimately mean
for health care? What does it mean when treatments
can become truly as individual as we are?