179 - Preparing for the Unknown (S12E14)

Ever feel like developing a new drug isn't quite
a straight line? More like, I don't know, navigating
a maze blindfolded? Huh, yeah, or on a moving
train, maybe. There's just so much that can shift
under your feet. That's exactly it, that inherent
uncertainty in pharma development. It really
is baked in. You've got the science itself, which
can always surprise you. Then there's the regulatory
side, manufacturing the market. Right, always
changing. A rigid plan just doesn't cut it? Not
in that kind of dynamic environment. No way.
And that's what we're really getting into today.
Strategic risk management and flexible design.
Basically, how do you plan for the unexpected?
Exactly. Thinking ahead, building in adaptability.
We've been looking at quite a range of material,
preclinical stuff, trial design, real world manufacturing
fixes. All to figure out how proactive planning
really makes a difference. Yeah, our mission
here isn't just to say, oh, problems happen.
It's to give You, the listener, some solid ideas
about anticipating those bumps and building agility
right into the process. Moving beyond just reacting
when things go wrong. Precisely. Looking ahead,
trying to see around the corner. Okay, so let's
untack that uncertainty first. Where does it
all stem from? What are those big sort of initial
hurdles? Well, number one is the science. Always
the science. You can get completely unexpected
biological responses. Like the drug just doesn't
do what you thought it would. Or it does something
else entirely. Think about older pre -clinical
models like sarcoma 37 and rodents. Sometimes
those models themselves could trigger an immune
response in the animal. Ah, so the model reaction
wasn't purely about the drugs effect. Potentially,
yes. It could muddy the waters, you know. It
highlights how even our basic tools for prediction
have limitations we need to understand, relying
on just one model. Risk. Okay, so scientific
surprises. What else keeps developers up at night?
Regulations. Definitely regulations. They aren't
static, they evolve. Think about the EU's increasing
focus on risk regulation. Right, the goalposts
can move mid -game. They really can. And now
with new tech, like AI showing up in medical
devices, the rulebook is still being written
in real time. You have to be adaptable. Trying
to hit a moving target. And what about actually
making the drug? Especially lots of it. Oh yeah,
manufacturing. That's another big one. Yeah.
Scaling up from a tiny lab batch to commercial
quantities. That's a huge leap. Things change
when you scale. Everything changes. Consistency,
purity. Keeping that dialed in is tough. We've
seen examples where tiny tweaks in mixing or
crystallization have major downstream effects.
It's complex. And then you throw in market dynamics,
competition. Exactly. The whole landscape can
shift while you're still in development. So yeah,
a static follow the steps plan. It's just not
sufficient. OK. So if rigid plans fail, what's
the alternative? How do you manage all this?
Well, it comes down to being proactive. Strategic
risk management is key. And a huge part of that
is contingency planning. Contingency planning.
Basically having a plan B ready to go. Plan B,
plan C, exactly. It means thinking through potential
setbacks before they happen and defining how
you'll respond. Like if your talk studies show
something weird. Exactly. Unexpected toxicology
findings. Maybe a trial fails to show efficacy.
Manufacturing delays. What's your alternative
approach? What resources do you need? Who makes
the call? So you're not scrambling in the moment.
You've already sketched out the detour. You've
sketched the detour, yeah. And crucially, those
plans need to be flexible, too. You can't predict
the exact nature of the problem. Right. A small
issue needs a different response than a total
showstopper. Completely different playbook, which
leads to another tool. scenario analysis scenario
analysis okay painting pictures of the future
kind of yeah you create and analyze different
plausible futures best case worst case maybe
a couple of most likely versions based on what
you know now how does that help day -to -day
well for each scenario you think about the impact
on efficacy safety the regulatory path market
access the whole picture okay understanding those
potential impacts helps you decide where to put
your resources now If a likely scenario involves,
say, a regulatory delay in Europe, maybe you
invest more in the U .S. pathway in parallel
now, just in case. So you're stress testing your
strategy against different possible realities.
That's a good way to put it. Stress testing,
identifying weak spots, and shoring them up early.
Okay, that makes sense. Now you also mentioned
flexible design. How is that different from contingency
planning? It's about building adaptability into
the process itself right from the start, rather
than just having backup plans for the process.
Engineering it to be more bendy, less brittle.
Exactly. And adaptive clinical trials are a prime
example. Right. We've talked about those. Remind
us how they work. The core idea is allowing pre
-planned changes during the trial based on the
data coming in. Pre -planned being the key phrase
there. Crucial. It's not making it up as you
go. You might adjust the dose, maybe refine the
patient group you're enrolling, or even drop
a treatment arm that's clearly not working. Compared
to traditional trials where everything is fixed
upfront. Right. Think about phase three trials.
Are you trying to show superiority, non -inferiority
equivalence? An adaptive design might let you
shift focus slightly based on early results.
Making it more efficient. Often, yes. Potentially
using fewer patients, getting clearer answers
faster. Really valuable. Especially in rare diseases
where patients are hard to find. So you learn
and adjust as you go within a defined framework.
Precisely. And that flexibility concept. It applies
just as much to manufacturing. Flexible manufacturing.
How does that work in such a tightly controlled
environment? Well, one powerful tool is modeling
and simulation. Using things like computational
fluid dynamics. Simulating the process digitally.
Exactly. You can model how things might work
at scale, anticipate mixing issues, test process
changes virtually before you commit expensive
resources in the real plant. It's like a dry
run. A digital twin for your process almost.
Sort of, yeah. It helps you anticipate and troubleshoot.
And alongside that is just deep process understanding,
robust process development. Knowing your process
inside and out. Thoroughly. Knowing the critical
parameters, how they interact, that knowledge
gives you the ability to make adjustments intelligently
if you need to, like if a raw material batch
very slightly. And we see this in practice, like
in the OPRND literature. Oh, absolutely. Great
examples there. Take GW641597X. The initial process
wasn't ideal. What did they do? They systematically
experimented, changed solvents, tweaked how things
were added, found a better way to purify the
product through crystallization. Big improvements
in yield and safety. So they didn't just stick
with version 1 .0? Not at all. They were flexible,
willing to rethink and optimize. That adaptability
was key. Any other examples jump out? Yeah, the
synthesis of sulfonamide -13. They used design
of experiments, DOE. A structured way to test
variables. Exactly. Systematically changing temperature,
time, amounts to map out how everything affected
the outcome. This let them find the sweet spot
for yield and purity. Giving them that detailed
map you mentioned earlier? Precisely. It gives
you the control and the knowledge to adapt if
things drift. And one more, facing a tough reaction,
an olefin metathesis macrocyclization. Sounds
complex, and it is. What happened there? The
first catalyst didn't work well, so instead of
giving up, they did extensive catalyst screening.
Tested lots of different options until they found
one that'd do the job efficiently. So exploring
alternatives when Plan A hits a wall? It's that
willingness to explore. To not be locked into
one approach, that's flexibility in action. These
examples really bring it home. It's not just
theory. Absolutely not. And remember, regulators
expect you to manage these changes properly.
ICH guidelines like Q9, Q10, Q11 provide the
framework for quality risk management and controlling
changes, including impurities. So flexibility,
yes, but within a controlled, documented system.
Always. Rigor and adaptability have to go hand
in hand. OK, so looking ahead now, the landscape
keeps changing. What's next in preparing for
the unknown? Well, the pace isn't slowing down,
is it? Think about AI and machine learning starting
to permeate development. Huge potential there,
but also new uncertainties, right? Especially
on the regulatory side. Exactly. So preparing
for the future means a real commitment to continuous
learning, staying agile. We need to understand
these new tools and be ready to adapt our strategies
as they evolve. It's not a one and done adjustment.
Not at all. It's an ongoing mindset. And underpinning
all of this, good data, robust data collection,
solid analysis like we've discussed regarding
clinical trial fundamentals. Quality data fuels
informed decisions. It's the bedrock. You can't
manage risks or adapt effectively without reliable
data guiding you. So wrapping this up, we've
really covered the inescapable uncertainties
and how strategies like contingency planning
and scenario analysis help manage risks. And
the power of building in flexibility from the
start. Adaptive trials, agile manufacturing,
learning from those real -world chemistry examples.
It seems the core message is about being proactive,
building resilience into the system. That's it.
Anticipating challenges, embracing adaptability.
It really increases the odds of getting innovative
medicines through that maze and to patients.
It turns uncertainty from just a problem into
maybe an opportunity for smarter development.
Well put. It fosters resilience. So here's something
to think about as we finish. We see AI... advanced
manufacturing, personalized medicine, all starting
to converge. How will that combination force
us to rethink preparing for the unknown in drug
development? What new kinds of risk management,
what new forms of flexibility will we need as
these fields blend together? And maybe consider
how do these ideas of planning and adapting apply
even beyond pharma in whatever field you work
in or are interested in? How do you prepare for
your unknowns?