As scientists, it’s in our nature to care about the quality of our research. We care deeply. But cutting through the noise and complexity of biological research to deliver ‘quality’ findings is easier said than done.
When I was first introduced to the concept of Quality by Design, I quickly realized that it isn’t something that you can tick off a list. It’s more than that: It’s a way of thinking and acting to deliver quality results, every single time.
Read on to learn:
What is Quality by Design (QbD)?
ICH’s official definition: The ICH defines QbD as “a systematic approach to development that begins with predefined objectives and emphasizes product and process understanding and process control, based on sound science and quality risk management.”
In other words: Quality by Design (QbD) systematically identifies, explains and manages sources of variability that affect the quality attributes of a process.
What are quality attributes?
They are the features of a product or process that actually impact quality. This distinction is important – as processes always vary in some way, but not all variation affects quality.
What are sources of variability?
Sources of variability are factors – control settings or other inputs to the process – that affect these quality attributes. Since not all output variation is important, not all input changes matter, either, since the variation they produce may not affect any quality attributes.
Common examples are the composition of input materials, environmental factors like temperature and human factors such as operator.
A key characteristic of a well-controlled process is consistency – and all these factors can lead to problems with maintaining it. That’s why QbD is particularly important to embed into processes that you use over and over again.
What is "quality" in the pharmaceutical industry?
Essentially, quality means freedom from defects. And “quality” can mean different things in different pharmaceutical contexts:
Imagine two teams of scientists are investigating two new drugs:
- One is studying oral formulation and considering different combinations of excipients
- The other is developing an assay for understanding their drug’s mechanism of action
What is deemed “acceptable” – or in other words, used as a marker of quality – will differ for each team.
The team studying oral formulation, for instance, would probably consider effects on solubility critical. Meanwhile, inaccuracies in measuring the degree of competitive inhibition would be key for the other team.
In other words, “quality” in the pharmaceutical industry is the degree to which a process, product or outcome meets specified requirements.
Examples of requirements might be:
- An assay is reliable (within specified limits of tolerance) under usual conditions
- An assay is reliable (within specified limits of tolerance) across a range of conditions
- An assay shows a strong signal with minimal noise
- A manufacturing process is reliable (within specified limits of tolerance) at scale
Jargon-busting interlude: 4 QbD acronyms worth knowing
For those of you who’ve already done some reading around the subject, QbD may seem almost esoteric – an alphabet soup of acronyms that sound more like Shyriiwook, the language of Wookiees, than English.
Here are four acronyms we’d recommend learning:
- Quality Target Product Profile (QTPP)
ICH definition: “A prospective summary of the quality characteristics of a drug product that ideally will be achieved to ensure the desired quality, taking into account safety and efficacy”.
In other words: It’s the specification of quality that you want your product to meet. To develop a QTPP, you need to understand the factors (temperature, concentration of compound A, etc.) that influence the product’s attributes (quality, safety, and efficacy), and how ( i.e., increasing concentration of Compound A beyond an identified concentration limit decreases quality in a specific way).
- Critical Quality Attribute (CQA)
ICH definition: “A physical, chemical, biological, or microbiological property or characteristic that should be within an appropriate limit, range, or distribution to ensure the desired product quality”.
In other words: CQA is analogous to multiple responses in design of experiments (DOE), i.e., the signals that you’ll measure to tell you about your process or product. If you are using DOE to assess an assay, you might want the response to be within an appropriate range.
- Critical Process Parameter (CPP)
ICH definition: “A process parameter whose variability has an impact on a CQA and therefore should be monitored or controlled to ensure the process produces the desired quality”.
In other words: CPPs are equivalent to factors in DOE. For example, you explore the changes in temperature, pH, cooling time, reagents, buffers, and salt to understand which affect your CQAs (responses).
Worth noting: You can control CPPs. But there may be many more CPPs than you realize. Which means you’d need a systematic approach to identifying, evaluating and monitoring them. During QbD, you’d identify potentially high-risk parameters. You’d then design and conduct experiments, using DOE when appropriate, to see if the parameter is critical to meeting the QTPP. You can then develop a plan to monitor or control the critical parameter, perhaps using process analytical technology (PAT). This allows real-time monitoring of a previously defined CPP.
- Critical Material Attribute (CMA)
Definition: A raw material whose variability has an impact on a CQA – and should be monitored or controlled to ensure the process produces the desired quality.
In other words: Like CPPs, CMAs are also equivalent to factors in DOE. But CMAs, unlike CPPs, are outside of your control.
For example: Some immunotherapy treatments, like chimeric antigen receptor (CAR) T-cell therapy, collect and use a patient’s lymphocytes to treat certain cancers. However, the quality of the source material (in this case, T cells) can vary widely between patients.
Why QbD is so important in experimentation
Regulatory authorities including the Food and Drug Administration (FDA) and the ICH do place quality at the heart of biological research.
But more importantly, by identifying the uncontrollable sources of variability, QbD helps reduce the risk that you will need to go back and redo your process, like during scale-up and manufacturing stages.
So even if the regulatory authorities aren’t peering over your shoulder, QbD helps you change how you approach your experiments for the better.
Let me break it down with a very simple, yet extremely powerful idea: Strip away the regulations and jargon, and imagine that assessing processes for your experiments is like searching for peaks in a mountain range.
You might want to find the highest peak you can – in other words, you want to maximize how much product your process can produce. But what you’re likely to find is that the highest peak is also very sharp and narrow. You can’t build anything on it – it’s impractical and unstable.
QbD is about changing your criteria. You’re not just looking for how productive your process is, but how stable it is. So instead of searching for the highest peak, you’re looking for a region of the range that’s almost as high, but also has a usably flat area. Now, you’re still very high up, but you can do something with your finding, which is likely to last a bit longer.
When to embed Quality by Design into your processes
QbD techniques are powerful tools for understanding and controlling your process, but to get the full benefit you need to account for the real-world variations your process will experience while you are developing it. Which means that you should start thinking about quality as early as possible in the process.
This ensures that from the first tentative steps into the experimental space, to when the first product leaves the laboratory (and beyond), QbD can equip you with the tools to deliver quality results.
The moral of the story? Don't leave quality 'til last
QbD isn’t something to do just because the regulatory authorities tell you to.
Embedding QbD early on can help avoid costly mistakes later in the developmental process. It can also inform ‘go/no go’ decisions, make research more efficient and reduce risks – and ultimately, make research less expensive.
