157 - The Role of Quality by Design (QbD) in Pharmaceutical Development and Manufacturing (S11E7)

Welcome to the deep dive. When you think about
the medicines we take, how do we actually know,
like really know, that every single pill or dose
is going to work properly and be safe? Yeah,
it's a huge question. And the answer is, well,
it's complex. It goes far beyond just checking
the final product off the line. Right. It's not
just pass fail at the very end. We're looking
into something called quality by design today,
QBD. It sounds like a smarter way to build quality
in from the start. That's a great way to put
it. Instead of just hoping the cake turns out
right and maybe testing a slice, QBD is like
deeply understanding the recipe, the oven, the
ingredients, everything, so you're confident
before you even bake it. Okay. So our mission
today is to unpack how this QBD approach actually
gets used. We'll be focusing quite a bit on a
tool called Design of Experiments, or DOEI. DOEI,
got it. And while our sources don't spell out
the FDA's exact stance on QBD and specific applications,
we can certainly talk about how the whole philosophy
aligns with what regulators are looking for.
Safety, efficacy, consistency. Makes sense. And
we've gathered quite a range of materials for
this. Scientific papers, regulatory info, some
case studies should give us a good picture. Definitely.
So let's dive in. At its heart, what is QBD?
What are the core ideas? OK, so fundamentally,
QBD is a a systematic approach to drug development.
There's this key guideline, ICHQ8, that defines
it really well. ICHQ8. Yeah, it's kind of a big
deal in the pharma world. It calls QBD a systematic
approach to development that begins with predefined
objective and emphasizes product and process
understanding and process control based on sound
science and quality risk management. Whoa, okay,
that's a mouthful. So breaking that down, it
starts with knowing what you want to achieve.
Exactly, predefined objectives. And the big emphasis
is on understanding, understanding the product,
understanding the process used to make it, and
controlling that process. It's all built on science
and managing risks. So it's less about following
a fixed recipe blindly and more about knowing
why the recipe works. Precisely. You have to
grasp the links between how the drug actually
performs, its quality features, and the nitty
-gritty of the manufacturing steps. It's not
just, does it work? Right. But why does it work?
And how do we keep it working reliably? Okay,
that leads us nicely into this design of experiments
thing. DOE. That's the tool for gaining that
understanding right. Spot on, DOE is absolutely
central to QBD. Think of it as super efficient
experimentation, instead of the old way, you
know, changing one thing, seeing what happens,
changing another. One factor at a time, yeah.
Slow, very slow, and you miss interactions. DOE
lets you systematically vary multiple inputs
process parameters, like temperature or mixing
speed, maybe attributes of the raw materials
all at the same time, in a structured way. Okay.
And the goal is? The goal is to map out how these
changes affect the really important characteristics
of the drug, the critical quality attributes,
or CQAs. CQAs. So that's like, it's purity, how
strong the dose is, how quickly it dissolves,
stuff like that. You've got it. Purity, potency,
stability, dissolution rate, all the key things
that define quality and performance. And we actually
have a great example from our sources involving
a drug called Torstrepib. Torstrepib. OK, what
do they do? They used DOE pretty extensively
during its development. These experiments were
crucial for identifying the CQAs for both the
active pharmaceutical ingredient, the API. The
main drug substance. Right. And the final drug
product, like the tablet, the DOE helped them
understand the relationship between the process
steps and those critical quality aspects. So
what kind of things did they figure out using
DOE for Torster PIP? Well, one key outcome was
establishing the design space. That's a term
from the ICH guidelines again. Design space sounds
like the operating. Exactly. It's the range of
process parameters and material attributes where
you've proven that quality is assured. It's not
just one single setting, but a defined multi
-dimensional space. For Torstrepib, Dewey helped
them see how different factors affected impurity
levels. Impurities are always a big concern,
I imagine. Huge. So they ran experiments comparing,
for instance, using sodium carbonate versus sodium
phosphate to control a specific impurity during
a reaction step. And did one work better? I heard
there was a bit of a surprise there. There was.
Initially, sodium carbonate looked better for
controlling that impurity, but the experiments
revealed something unexpected. If they let the
reaction run too long with the sodium carbonate
or maybe use too much, the whole reaction could
just stall, stop working. Really? Why? It turned
out there is a secondary reaction happening.
When the ratio of reactants hit a certain point
because of the side reaction, it basically used
up one of the essential starting materials too
quickly. Wow. So the DOE didn't just confirm
things, it uncovered hidden complexities. Absolutely.
That's the power of it. Now, interestingly, this
stalling didn't actually ruin the batch in their
case. They had a safety net built in. They had
what's called a process acceptance range, or
PAR. Basically, they knew they could remove any
leftover starting material in a later purification
step. But finding that stalling phenomenon, that
gave them much deeper process understanding.
I bet. You wouldn't find that just by testing
the end product. No way. The DOE results also
let them rank which variables had the biggest
impact on getting rid of impurities. And when
they looked at the crystallization step later
on, DOE helped define the precise boundaries
for that part of the design space, too. So really
mapping out the safe and effective zone for manufacturing.
That's it. And this enhanced understanding you
get from QBD powered by DOE, it directly leads
to better control over those critical quality
attributes. Like keeping those impurities consistently
low in the Tors Krapiv example. Exactly. Optimizing
the process based on that DOE data meant they
could confidently keep impurities within safe
acceptable limits batch after batch. And does
this apply to other CQAs too, like how well the
drug dissolves? Yeah. Same principle, absolutely.
While the sources don't detail a specific Torster
Pibb DOE for dissolution, understanding how formulation
factors like particle size or binders in a tablet
and process parameters affect drug release is
a classic QBD -DOE application. You control those
inputs to ensure the drug dissolves predictably.
Sometimes the dissolution rate itself is a specification,
isn't it? It often is. It can be a key quality
standard directly linked to how the drug will
be absorbed in the body. It really emphasizes
that QBD isn't just about observing correlations
if we do XY happens. It's about the Y. Precisely.
That unexpected reaction stalling with Horstropiv
is a perfect example. It forces a deeper dive
into the underlying chemistry and physics of
the process. So how does all this fit with the
regulators? Like the FDA, you mentioned our sources
are a bit light on specifics here. They are on
like direct quotes about QBD and NDAs. But we
can definitely infer the alignment. The FDA's
core mission, above all, is ensuring drugs are
safe and effective for patients. Right. That's
job number one. And the principles of QBD, understanding
your product, understanding your process, controlling
your process based on science and risk, they
feed directly into that mission. It's exactly
the kind of proactive approach regulators want
to see. The FDA looks closely at the manufacturing
part. Oh, absolutely. They scrutinize the manufacturing
process description, the controls in place. They
conduct pre -approval inspections, P .A .Is on
site. P .A .Is. Pre -approval inspections. They
go to the manufacturing facility to verify the
information in the application is accurate and
crucially, that the site is complying with current
good manufacturing practices, CGMPs. CGMPs, right.
baseline rules for quality manufacturing. Exactly.
And even earlier, during the investigational
new drug phase, the IND phase... Before human
trials start. Correct. Even then, regulators
focus on ensuring the manufacturing process for
the trial drug is reproducible and that the drug
is well characterized. They pay close attention
to impurities and potential contaminants right
from the start. QBD provides a framework to generate
that characterization data systematically. So
QBD helps build the evidence package that gives
the FDA confidence. That's a good way to think
about it. It demonstrates a deep level of understanding
and control, which inherently reduces risk. It
aligns perfectly with the goals of CGMP. You
can almost see QBD as the scientific engine that
helps you design and implement truly robust GMPs.
It helps you figure out which parts of the GMPs
are most critical for your specific product and
process. Exactly. It tells you where to focus
your control strategy for maximum impact on quality.
Okay, this has been really insightful. Let's
try and sum up the key takeaways from this deep
dive on QBD. Sure. I think the main point is
that QBD represents a shift towards a proactive,
science -driven way of ensuring pharmaceutical
quality. It's not an afterthought. It's built
in. Using tools like design of experiments to
really dig deep into how manufacturing processes
work. Right. To gain that fundamental process
understanding. And that understanding is what
allows for effective control over those critical
quality attributes, the things that really matter
for the patient. And while we didn't have specific
FDA mandates in our sources today, the whole
QDD philosophy seems perfectly in tune with regulatory
expectations for safe, effective, and consistently
produced medicines. Definitely. It's about building
quality, safety, and efficacy in from the very
beginning based on solid science and risk assessment.
It's a framework for developing a robust, reliable
manufacturing process. Which ultimately means
better, more dependable medications for all of
us. Makes you appreciate the science that goes
into ensuring quality. It really does, and it
leads to a final thought, perhaps. As pharmaceutical
manufacturing gets ever more complex, think biologics,
cell therapies. Yeah, cutting -edge stuff. How
will QBD itself need to evolve? How can the systematic
approach adapt to meet the challenges of these
new modalities and technologies? That's something
worth pondering. A fascinating question for the
future. For listeners wanting to dig even deeper,
checking out those ICH guidelines, particularly
Q8, and looking for more QBD case studies would
be a great next step. Absolutely. There's a wealth
of information out there once you start Lookout.