WEBVTT 00:00:00.000 --> 00:00:01.720 Okay, so picture this. You've got a product coming 00:00:01.720 --> 00:00:05.099 off the manufacturing line, right? And it looks 00:00:05.099 --> 00:00:07.440 absolutely perfect. It meets every single specification, 00:00:07.459 --> 00:00:10.699 but then there's a tiny change. Maybe the raw 00:00:10.699 --> 00:00:12.500 materials are a little different or someone tweaks 00:00:12.500 --> 00:00:15.160 machine setting just slightly and boom, the next 00:00:15.160 --> 00:00:17.399 batch is off. It just makes you think, is it 00:00:17.399 --> 00:00:20.140 even possible to make a manufacturing process 00:00:20.140 --> 00:00:23.879 so stable that it can handle those little hiccups, 00:00:23.960 --> 00:00:26.399 those inevitable changes, and still pump out 00:00:26.399 --> 00:00:29.059 high quality products every single time? You've 00:00:29.059 --> 00:00:31.879 just hit on the fundamental question behind quality 00:00:31.879 --> 00:00:34.979 by design, or QBD as it's commonly known. You 00:00:34.979 --> 00:00:36.719 see, it's really a whole different way of thinking 00:00:36.719 --> 00:00:39.420 about pharmaceutical manufacturing. The focus 00:00:39.420 --> 00:00:41.560 is on building that inherent stability into the 00:00:41.560 --> 00:00:43.960 process from the get -go, making sure that it's 00:00:43.960 --> 00:00:46.060 not just reproducible under perfect conditions, 00:00:46.439 --> 00:00:49.399 but truly robust in the face of real -world variability. 00:00:49.820 --> 00:00:51.359 And that's exactly what we're going to unpack 00:00:51.359 --> 00:00:53.950 today. It sounds like it's all about making the 00:00:53.950 --> 00:00:56.729 process as bulletproof as possible. And for our 00:00:56.729 --> 00:00:58.609 listeners who are already deep into the world 00:00:58.609 --> 00:01:00.810 of drug development, we've got a ton of great 00:01:00.810 --> 00:01:02.969 material to dive into. But for this particular 00:01:02.969 --> 00:01:05.230 deep dive, we're zeroing in on the QBD aspects. 00:01:05.590 --> 00:01:08.790 How do you actually create those rock solid, 00:01:09.109 --> 00:01:11.510 dependable manufacturing processes that you're 00:01:11.510 --> 00:01:13.659 talking about? Our mission today is to break 00:01:13.659 --> 00:01:16.200 down the core principles of QBD. We'll explore 00:01:16.200 --> 00:01:18.180 how manufacturers go about identifying those 00:01:18.180 --> 00:01:20.340 critical process parameters, the ones that can 00:01:20.340 --> 00:01:22.500 make or break product quality. And then we'll 00:01:22.500 --> 00:01:24.739 look at risk -based design and how it moves from 00:01:24.739 --> 00:01:27.620 being a theoretical concept to an actual practical 00:01:27.620 --> 00:01:29.819 tool for designing processes that are way less 00:01:29.819 --> 00:01:32.180 likely to veer off course. And to really bring 00:01:32.180 --> 00:01:34.680 this to life, we'll be using real -world examples 00:01:34.680 --> 00:01:37.719 from the OPRD literature. Okay, so let's jump 00:01:37.719 --> 00:01:40.379 right in. What exactly does quality by design 00:01:40.379 --> 00:01:42.500 mean? Is it more than just checking off a list 00:01:42.500 --> 00:01:44.680 of good practices? That's a great question, and 00:01:44.680 --> 00:01:46.780 you're right, it's a lot more than just a checklist. 00:01:47.659 --> 00:01:50.099 QBD is really a systematic approach that starts 00:01:50.099 --> 00:01:52.379 long before you even think about manufacturing. 00:01:52.939 --> 00:01:55.340 You begin with very clearly defined objectives. 00:01:56.060 --> 00:01:58.019 What are we trying to achieve with this drug, 00:01:58.219 --> 00:02:00.599 and what level of quality is non -negotiable 00:02:00.599 --> 00:02:02.969 for patient safety? So you start with the end 00:02:02.969 --> 00:02:05.349 in mind. Exactly. From there, it's all about 00:02:05.349 --> 00:02:07.849 gaining a deep scientific understanding of both 00:02:07.849 --> 00:02:10.650 the drug product itself, particularly its critical 00:02:10.650 --> 00:02:13.669 quality attributes or CQAs, and of course the 00:02:13.669 --> 00:02:16.169 entire process used to manufacture it. The key 00:02:16.169 --> 00:02:18.430 here is that quality is not something you test 00:02:18.430 --> 00:02:21.030 for at the end. It's baked into every single 00:02:21.030 --> 00:02:23.090 step right from the beginning based on sound 00:02:23.090 --> 00:02:26.289 scientific principles and a very proactive assessment 00:02:26.289 --> 00:02:28.710 of potential risks. So it's kind of like designing 00:02:28.710 --> 00:02:30.710 a building to withstand an earthquake rather 00:02:30.710 --> 00:02:32.590 than just hope hoping for the best and inspecting 00:02:32.590 --> 00:02:34.870 it after the fact. I love that analogy. That's 00:02:34.870 --> 00:02:37.289 precisely it. And this proactive way of thinking 00:02:37.289 --> 00:02:39.550 is what leads to those robust and reproducible 00:02:39.550 --> 00:02:41.659 processes we talked about. So now the question 00:02:41.659 --> 00:02:43.979 is, why are these two things, robustness and 00:02:43.979 --> 00:02:46.199 reproducibility, so crucial in pharmaceutical 00:02:46.199 --> 00:02:48.599 manufacturing? Well, when we say a process is 00:02:48.599 --> 00:02:51.800 robust, we're talking about its ability to withstand 00:02:51.800 --> 00:02:54.060 those little variations that always pop up during 00:02:54.060 --> 00:02:56.819 manufacturing. Slight changes in temperature, 00:02:57.379 --> 00:02:59.120 maybe the raw materials from different suppliers 00:02:59.120 --> 00:03:00.479 are slightly different, those kinds of things. 00:03:01.039 --> 00:03:03.219 But the final product quality doesn't suffer. 00:03:03.629 --> 00:03:06.289 and then reproducibility, that's about being 00:03:06.289 --> 00:03:08.629 able to churn out the same high -quality product 00:03:08.629 --> 00:03:11.349 time after time, batch after batch, regardless 00:03:11.349 --> 00:03:13.689 of the scale of production or even where in the 00:03:13.689 --> 00:03:15.909 world it's being manufactured. And for our listeners, 00:03:15.969 --> 00:03:18.250 this is about understanding that a well -designed, 00:03:18.310 --> 00:03:21.270 robust process is the foundation of consistent 00:03:21.270 --> 00:03:24.460 medicine quality and supply. You nailed it. And 00:03:24.460 --> 00:03:26.960 that consistency is what directly impacts patient 00:03:26.960 --> 00:03:29.020 safety and how well the drug actually works. 00:03:29.680 --> 00:03:32.000 So let's dig a little deeper into how manufacturers 00:03:32.000 --> 00:03:34.159 figure out what those critical process parameters 00:03:34.159 --> 00:03:36.560 or CPPs really are. What are they exactly and 00:03:36.560 --> 00:03:38.580 how do you go about identifying the really important 00:03:38.580 --> 00:03:41.800 ones? As far as I understand it, CPPs are the 00:03:41.800 --> 00:03:44.300 key variables in the manufacturing process. The 00:03:44.300 --> 00:03:46.889 ones that have a direct impact on the CQAs of 00:03:46.889 --> 00:03:49.629 the final drug product. So things like the temperature 00:03:49.629 --> 00:03:52.430 at a specific stage in the reaction, the mixing 00:03:52.430 --> 00:03:55.669 speed, how long a drying step takes, those are 00:03:55.669 --> 00:03:59.169 all potential CPPs, right? You got it. And understanding 00:03:59.169 --> 00:04:01.449 how those CPPs are linked to the CQA, that's 00:04:01.449 --> 00:04:04.469 like the holy grail of QBD, you absolutely need 00:04:04.469 --> 00:04:07.129 to pinpoint which process factors and within 00:04:07.129 --> 00:04:09.949 what range of operation have a significant effect 00:04:09.949 --> 00:04:12.490 on the quality of the drug. That knowledge is 00:04:12.490 --> 00:04:15.199 the bedrock of a well -controlled process. Sounds 00:04:15.199 --> 00:04:17.500 pretty complex. So how do manufacturers actually 00:04:17.500 --> 00:04:19.920 determine which parameters are the critical ones? 00:04:20.240 --> 00:04:22.519 It can't be as simple as just listing out every 00:04:22.519 --> 00:04:24.360 single step in variable, right? Oh, absolutely 00:04:24.360 --> 00:04:27.240 not. It requires a systematic and very scientific 00:04:27.240 --> 00:04:29.920 approach. Typically, identifying CPPs involves 00:04:29.920 --> 00:04:32.139 a mix of existing knowledge about the process, 00:04:32.420 --> 00:04:34.139 along with some good old -fashioned experimental 00:04:34.139 --> 00:04:36.579 studies. Scientists might use techniques like 00:04:36.579 --> 00:04:39.100 design of experiments or DOE to deliberately 00:04:39.100 --> 00:04:41.259 change multiple process parameters at the same 00:04:41.259 --> 00:04:43.300 time, and then they analyze the effects on the 00:04:43.300 --> 00:04:45.740 CQAs. So they're deliberately introducing variability. 00:04:46.339 --> 00:04:48.759 Exactly. This allows them to use statistical 00:04:48.759 --> 00:04:51.279 analysis to determine which factors are the heavy 00:04:51.279 --> 00:04:53.860 hitters, the ones that really influence the quality 00:04:53.860 --> 00:04:56.560 attributes. It's all about building a data -driven 00:04:56.560 --> 00:04:58.560 understanding of what's going on in the process. 00:04:58.839 --> 00:05:00.139 And this is where it gets really interesting, 00:05:00.360 --> 00:05:04.379 because one of the OPRD papers you shared, OPRD2012d2 00:05:04.379 --> 00:05:07.519 .pdf, had a perfect example of this in action. 00:05:08.000 --> 00:05:10.240 They were scaling up a reaction, and at a small 00:05:10.240 --> 00:05:12.800 lab scale, everything was smooth sailing. Right. 00:05:12.920 --> 00:05:15.160 But when they moved to a much larger 100 -liter 00:05:15.160 --> 00:05:18.540 reactor, they ran into this gelling problem at 00:05:18.540 --> 00:05:20.800 a certain stage, something they'd never encountered 00:05:20.800 --> 00:05:23.459 before. That's a classic example of how scaling 00:05:23.459 --> 00:05:25.800 up can suddenly reveal CPPs that you didn't even 00:05:25.800 --> 00:05:28.399 know existed. In that particular case, it was 00:05:28.399 --> 00:05:30.579 the longer time it took to reach and maintain 00:05:30.579 --> 00:05:32.740 the target temperature in that bigger vessel, 00:05:32.779 --> 00:05:34.980 along with potential differences and mixing at 00:05:34.980 --> 00:05:37.259 that scale. Those turned out to be the culprits 00:05:37.259 --> 00:05:39.579 behind the unexpected gelling issue. And I remember 00:05:39.579 --> 00:05:41.600 they ended up with a decent yield for that batch, 00:05:42.100 --> 00:05:43.899 but it was only through some manual intervention. 00:05:43.959 --> 00:05:46.769 They had to basically babysit the process. Exactly, 00:05:46.769 --> 00:05:48.970 and that highlighted a gap in their understanding 00:05:48.970 --> 00:05:51.910 of the process. Even though they managed to save 00:05:51.910 --> 00:05:54.629 that batch, the gelling was a clear sign that 00:05:54.629 --> 00:05:56.889 they hadn't identified all the key factors that 00:05:56.889 --> 00:05:59.089 controlled the process. It was a wake -up call 00:05:59.089 --> 00:06:01.209 to investigate further and really understand 00:06:01.209 --> 00:06:03.509 what was critical for preventing that gelling 00:06:03.509 --> 00:06:05.550 from happening consistently, especially at those 00:06:05.550 --> 00:06:07.730 larger scales. So even though they got a decent 00:06:07.730 --> 00:06:10.709 yield that time, it wasn't a reliable or robust 00:06:10.709 --> 00:06:13.069 process. It was a bit of a fluke. Precisely. 00:06:13.370 --> 00:06:15.529 That unexpected event was what prompted them 00:06:15.399 --> 00:06:17.800 to explore different ways to run the reaction, 00:06:18.079 --> 00:06:21.019 like using continuous flow conditions. The idea 00:06:21.019 --> 00:06:23.680 there was that the more efficient heat transfer 00:06:23.680 --> 00:06:25.819 and mixing in a continuous flow reactor could 00:06:25.819 --> 00:06:27.740 potentially solve the temperature control and 00:06:27.740 --> 00:06:29.740 hold time issues that were causing the gelling 00:06:29.740 --> 00:06:32.459 problem in the larger batches. This is a great 00:06:32.459 --> 00:06:34.920 example of using process understanding to design 00:06:34.920 --> 00:06:37.800 a more robust and reliable system. And this brings 00:06:37.800 --> 00:06:40.120 us right to the concept of risk -based design, 00:06:40.139 --> 00:06:42.300 doesn't it? Because that gelling problem they 00:06:42.300 --> 00:06:45.100 had at scale, that was a real risk to their ability 00:06:45.100 --> 00:06:47.959 to consistently manufacture the drug. Absolutely. 00:06:48.319 --> 00:06:51.459 Risk -based design is a cornerstone of QBD. It 00:06:51.459 --> 00:06:53.720 involves systematically identifying potential 00:06:53.720 --> 00:06:55.779 sources of variability in your manufacturing 00:06:55.779 --> 00:06:58.500 process and then figuring out how likely they 00:06:58.500 --> 00:07:00.800 are to occur and how bad the impact would be 00:07:00.800 --> 00:07:04.240 if they did. By conducting thorough risk assessments, 00:07:04.579 --> 00:07:06.680 manufacturers can figure out which CPPs need 00:07:06.680 --> 00:07:09.399 the tightest control strategies and the most 00:07:09.399 --> 00:07:11.540 in -depth understanding during development. So 00:07:11.540 --> 00:07:13.519 it's like a pre -mortem for the manufacturing 00:07:13.519 --> 00:07:16.199 process. What could go wrong and how bad would 00:07:16.199 --> 00:07:18.649 it be? You got it. And this proactive approach 00:07:18.649 --> 00:07:20.949 is all about preventing problems before they 00:07:20.949 --> 00:07:23.029 even have a chance to show up. And it makes a 00:07:23.029 --> 00:07:24.810 lot of sense because in pharmaceuticals the stakes 00:07:24.810 --> 00:07:27.009 are high. They are indeed. And this risk -based 00:07:27.009 --> 00:07:29.529 approach is what ultimately leads to the establishment 00:07:29.529 --> 00:07:32.129 of the design space. Think of the design space 00:07:32.129 --> 00:07:34.689 as a scientifically defined safe zone for your 00:07:34.689 --> 00:07:37.589 process. It lays out the acceptable ranges for 00:07:37.589 --> 00:07:39.730 your critical input variables and parameters, 00:07:40.129 --> 00:07:41.990 the ones that have been proven to consistently 00:07:41.990 --> 00:07:44.329 deliver a high quality product. So it's like 00:07:44.329 --> 00:07:46.389 having a set of boundaries for your process. 00:07:46.889 --> 00:07:48.389 And if you stay within those boundaries, you 00:07:48.389 --> 00:07:51.110 know you're going to be okay. Exactly. And here's 00:07:51.110 --> 00:07:53.230 the really powerful thing about the design space. 00:07:53.269 --> 00:07:55.329 If you operate within those established boundaries 00:07:55.329 --> 00:07:58.310 during manufacturing, it's not considered a change 00:07:58.310 --> 00:08:00.920 to the approved process. This gives me manufactures 00:08:00.920 --> 00:08:04.180 a lot of flexibility and encourages them to continuously 00:08:04.180 --> 00:08:06.860 improve their processes without having to constantly 00:08:06.860 --> 00:08:09.319 go back for regulatory approval for every minor 00:08:09.319 --> 00:08:11.860 tweak. That sounds like a huge win in terms of 00:08:11.860 --> 00:08:14.220 efficiency and being able to adapt to new information 00:08:14.220 --> 00:08:16.660 or improvements. It absolutely is. And going 00:08:16.660 --> 00:08:18.660 back to that OPRD example we talked about earlier, 00:08:18.839 --> 00:08:22.560 OPRD2012D2 .pdf, the whole experience with the 00:08:22.560 --> 00:08:24.740 gelling issue at scale that directly informed 00:08:24.740 --> 00:08:27.420 how they approached risk -based design for future 00:08:27.420 --> 00:08:29.759 batches of that product. They learned the hard 00:08:29.759 --> 00:08:33.019 way that temperature and hold time were much 00:08:33.019 --> 00:08:35.399 riskier parameters at larger scales than they 00:08:35.399 --> 00:08:37.360 had initially thought based on their small -scale 00:08:37.360 --> 00:08:39.419 experiments. And that wasn't the only lesson. 00:08:39.279 --> 00:08:41.679 than they learned from that OPRD paper. They 00:08:41.679 --> 00:08:43.360 also ran into some trouble during the methyl 00:08:43.360 --> 00:08:45.740 ester hydrolysis step. I think they were using 00:08:45.740 --> 00:08:48.080 standard alkaline hydrolysis conditions and ended 00:08:48.080 --> 00:08:50.559 up with some unwanted side reactions that affected 00:08:50.559 --> 00:08:53.120 the purity of their product. That's right. The 00:08:53.120 --> 00:08:55.240 hydrolysis of the methyl ester turned out to 00:08:55.240 --> 00:08:58.379 be another critical step where the initial seemingly 00:08:58.379 --> 00:09:00.820 straightforward approach didn't quite translate 00:09:00.820 --> 00:09:03.600 well to the larger scale. They discovered that 00:09:03.600 --> 00:09:05.899 those standard alkaline conditions were causing 00:09:05.899 --> 00:09:08.799 ring opening and epimerization which led to the 00:09:08.799 --> 00:09:11.100 formation of impurities that they didn't want. 00:09:11.820 --> 00:09:13.919 This really highlighted the fact that the reaction 00:09:13.919 --> 00:09:16.039 conditions things like the choice of base, the 00:09:16.039 --> 00:09:18.159 temperature, and how long the reaction ran for 00:09:18.159 --> 00:09:21.000 those were all critical process parameters that 00:09:21.000 --> 00:09:22.960 had a big impact not only on the yield of the 00:09:22.960 --> 00:09:25.440 product, but also on its purity. They ended up 00:09:25.440 --> 00:09:27.500 having to investigate exactly who hit the nail 00:09:27.500 --> 00:09:29.399 on the head. And it wasn't just about those main 00:09:29.399 --> 00:09:31.480 reactions either. There were also challenges 00:09:31.480 --> 00:09:34.100 during the scale up of the patent procedure specifically 00:09:34.100 --> 00:09:36.559 related to the sulfonyl chloride hydrolysis. 00:09:37.019 --> 00:09:39.259 It seems like the amount of water used as a regent 00:09:39.259 --> 00:09:41.279 and the longer heating times needed at a larger 00:09:41.279 --> 00:09:44.019 scale led to the formation of a sulfonic acid 00:09:44.019 --> 00:09:46.899 byproduct. So even something as seemingly simple 00:09:46.899 --> 00:09:49.279 as the amount of water became a critical factor. 00:09:49.299 --> 00:09:51.610 That's right. The conditions that worked beautifully 00:09:51.610 --> 00:09:54.250 at a smaller laboratory scale around 13 grams 00:09:54.250 --> 00:09:56.570 didn't provide the same level of control when 00:09:56.570 --> 00:09:59.529 they scaled the reaction up. They observed a 00:09:59.529 --> 00:10:01.750 clear link between the increased amount of water 00:10:01.750 --> 00:10:04.309 relative to the starting material at larger scales, 00:10:04.789 --> 00:10:07.029 the longer heating times needed to get the reaction 00:10:07.029 --> 00:10:09.350 to finish, and the formation of that unwanted 00:10:09.350 --> 00:10:12.570 byproduct. This just goes to show that even parameters 00:10:12.570 --> 00:10:14.950 you might think are well defined can behave differently 00:10:14.950 --> 00:10:18.090 at different scales, which is why thorough investigation 00:10:18.090 --> 00:10:20.570 and control throughout development are so important. 00:10:20.789 --> 00:10:22.669 It really highlights the fact that what works 00:10:22.669 --> 00:10:25.230 perfectly in a small flask in the lab doesn't 00:10:25.230 --> 00:10:28.250 always translate smoothly to industrial scale 00:10:28.250 --> 00:10:31.049 production. You often uncover new challenges 00:10:31.049 --> 00:10:33.070 and realize that parameters you might have considered 00:10:33.070 --> 00:10:35.490 insignificant can become absolutely critical 00:10:35.490 --> 00:10:37.169 when you're dealing with much larger quantities. 00:10:37.309 --> 00:10:39.750 Couldn't have said it better myself. And the 00:10:39.750 --> 00:10:42.330 OPRD literature is a goldmine of these real -world 00:10:42.330 --> 00:10:44.970 examples that really bring the concepts of QBD 00:10:44.970 --> 00:10:47.980 to life. It's not just theoretical jargon. It's 00:10:47.980 --> 00:10:51.100 about solving real manufacturing problems and 00:10:51.100 --> 00:10:53.100 creating more reliable and efficient processes 00:10:53.100 --> 00:10:55.500 that ultimately benefit patients. It definitely 00:10:55.500 --> 00:10:57.399 makes it much more relatable and understandable. 00:10:57.700 --> 00:11:00.480 Absolutely. And speaking of real -world examples, 00:11:00.679 --> 00:11:02.820 there's a great one in the OPRD literature about 00:11:02.820 --> 00:11:06.379 the production of an R alcohol described in OPRD2014 00:11:06.379 --> 00:11:09.440 .bodypdf. When they were scaling up their solvent 00:11:09.440 --> 00:11:11.879 extraction process, they ran into a major problem 00:11:11.879 --> 00:11:14.059 with emulsions that just wouldn't separate. Emulsions, 00:11:14.120 --> 00:11:16.379 those are the bane of so many chemists' existence. 00:11:16.639 --> 00:11:18.860 I can only imagine the frustration. Tell me about 00:11:18.860 --> 00:11:21.669 it. Anyway, through their investigations, they 00:11:21.669 --> 00:11:24.070 started to suspect that residual protein from 00:11:24.070 --> 00:11:26.850 the fermentation broth was acting as an emulsifying 00:11:26.850 --> 00:11:29.970 agent. So to fix this and make their extraction 00:11:29.970 --> 00:11:33.149 process more robust, they introduced an ultrafiltration 00:11:33.149 --> 00:11:35.590 step to remove those proteins before extraction. 00:11:36.000 --> 00:11:38.539 This is a perfect example of how an unexpected 00:11:38.539 --> 00:11:41.600 problem during scale -up led to a deeper understanding 00:11:41.600 --> 00:11:44.059 of what was going on and the implementation of 00:11:44.059 --> 00:11:46.519 a targeted risk -mitigating solution that was 00:11:46.519 --> 00:11:49.120 completely in line with QBD principles. So the 00:11:49.120 --> 00:11:51.159 emotion formation wasn't even on their radar 00:11:51.159 --> 00:11:53.700 as a major risk at the smaller scale. But when 00:11:53.700 --> 00:11:55.659 they scaled up, it became a real threat to their 00:11:55.659 --> 00:11:57.960 ability to consistently and efficiently extract 00:11:57.960 --> 00:12:00.740 the product. And their solution, adding that 00:12:00.740 --> 00:12:03.220 ultrafiltration step that was pure QBD thinking. 00:12:03.450 --> 00:12:05.789 identify the problem, understand the root cause, 00:12:05.970 --> 00:12:07.929 and implement a control strategy to minimize 00:12:07.929 --> 00:12:09.929 the risk and improve the robustness of the whole 00:12:09.929 --> 00:12:13.360 process. Exactly, and in that same OPRD paper, 00:12:13.559 --> 00:12:17.179 2014g .pdf, they also talk about their work on 00:12:17.179 --> 00:12:19.419 optimizing the catalyst loading and reaction 00:12:19.419 --> 00:12:21.679 conditions for producing an optically active 00:12:21.679 --> 00:12:24.100 intermediate. It wasn't just about getting the 00:12:24.100 --> 00:12:25.899 desired chemical transformation to happen, it 00:12:25.899 --> 00:12:27.759 was about fine -tuning the reaction parameters 00:12:27.759 --> 00:12:30.259 to get the highest possible yield and the desired 00:12:30.259 --> 00:12:32.639 optical purity and doing it in the most efficient 00:12:32.639 --> 00:12:36.029 and reproducible way possible. That proactive 00:12:36.029 --> 00:12:38.669 optimization of key parameters is another major 00:12:38.669 --> 00:12:41.230 element of the QBD philosophy, designing for 00:12:41.230 --> 00:12:43.610 optimal performance and consistent quality right 00:12:43.610 --> 00:12:46.509 from the start. It's so helpful to see how these 00:12:46.509 --> 00:12:49.230 real -world examples from the OTRD literature 00:12:49.230 --> 00:12:52.090 really illustrate the practical side of QBD. 00:12:52.269 --> 00:12:54.409 It shows that it's not just an abstract concept. 00:12:54.509 --> 00:12:56.730 It's about tackling real -world manufacturing 00:12:56.730 --> 00:12:58.950 challenges and building better, more reliable 00:12:58.950 --> 00:13:01.850 processes that ultimately benefit patients. Absolutely. 00:13:01.929 --> 00:13:03.929 And while we focused on these specific examples 00:13:03.929 --> 00:13:06.090 from OPRD, it's important to remember that the 00:13:06.090 --> 00:13:08.330 principles of QBD are tightly connected to the 00:13:08.330 --> 00:13:10.730 broader regulatory landscape that governs pharmaceutical 00:13:10.730 --> 00:13:12.870 development and manufacturing. So it's not just 00:13:12.870 --> 00:13:15.190 a nice to have. It's woven into the fabric of 00:13:15.190 --> 00:13:17.370 how drugs are developed and regulated. Exactly. 00:13:17.750 --> 00:13:20.129 For instance, the emphasis on thoroughly understanding 00:13:20.129 --> 00:13:21.990 and controlling your process parameters. and 00:13:21.990 --> 00:13:24.730 how they impact product quality that aligns perfectly 00:13:24.730 --> 00:13:27.210 with the regulatory requirements for having well 00:13:27.210 --> 00:13:29.309 -defined specifications and robust analytical 00:13:29.309 --> 00:13:31.889 methods. It's all outlined in guidelines like 00:13:31.889 --> 00:13:37.909 21 CFR 314 .50D3. You simply can't ensure quality 00:13:37.909 --> 00:13:40.230 without being able to accurately measure and 00:13:40.230 --> 00:13:43.149 control those critical attributes. And that proactive 00:13:43.149 --> 00:13:45.750 mindset of preventing contamination and ensuring 00:13:45.750 --> 00:13:48.529 consistent product quality through robust process 00:13:48.529 --> 00:13:50.690 design that's completely in line with the principles 00:13:50.830 --> 00:13:53.710 of Good Manufacturing Practice or GMP, which 00:13:53.710 --> 00:13:56.470 are detailed in regulations like 21 CFR Part 00:13:56.470 --> 00:13:59.210 211. A well -designed process that's informed 00:13:59.210 --> 00:14:01.970 by QBD is the foundation of a clean and controlled 00:14:01.970 --> 00:14:04.309 manufacturing environment, which, as our excipient 00:14:04.309 --> 00:14:06.509 development source also points out, is essential 00:14:06.509 --> 00:14:09.029 for preventing contamination in facilities. Exactly. 00:14:09.169 --> 00:14:10.750 And the fundamental principle behind it all, 00:14:11.149 --> 00:14:13.009 understanding and controlling process variables. 00:14:13.549 --> 00:14:15.529 So you consistently produce a product that meets 00:14:15.529 --> 00:14:18.230 its predetermined quality attributes that's also 00:14:18.230 --> 00:14:20.710 at the heart of process validation, which is 00:14:20.710 --> 00:14:22.850 a critical aspect discussed in the Pharmaceutical 00:14:22.850 --> 00:14:25.450 Master Validation Plan source. You need to be 00:14:25.450 --> 00:14:28.019 able to prove through data and scientific understanding 00:14:28.019 --> 00:14:30.679 that your designed process consistently performs 00:14:30.679 --> 00:14:33.100 as intended under different operating conditions. 00:14:33.720 --> 00:14:36.100 QBD provides the scientific basis for a successful 00:14:36.100 --> 00:14:39.059 validation program. So to sum it all up, we've 00:14:39.059 --> 00:14:41.320 really explored the essence of quality by design. 00:14:41.470 --> 00:14:44.049 It's about intentionally building quality into 00:14:44.049 --> 00:14:46.629 every step of the manufacturing process, from 00:14:46.629 --> 00:14:49.549 the initial design to the final product. It involves 00:14:49.549 --> 00:14:51.570 deeply understanding the product and how it's 00:14:51.570 --> 00:14:54.049 made rigorously identifying those critical process 00:14:54.049 --> 00:14:56.990 parameters and using a risk -based approach to 00:14:56.990 --> 00:14:59.370 design a manufacturing system that's both robust 00:14:59.370 --> 00:15:02.289 and consistently reproducible. It goes beyond 00:15:02.289 --> 00:15:04.750 just checking off boxes. It's a fundamental commitment 00:15:04.750 --> 00:15:07.429 to quality in every aspect of the process. I 00:15:07.429 --> 00:15:11.049 completely agree. And by grasping these principles 00:15:11.049 --> 00:15:13.629 our listeners gain a much deeper understanding 00:15:13.629 --> 00:15:16.110 of what goes into developing and manufacturing 00:15:16.110 --> 00:15:18.669 pharmaceutical products. They can now appreciate 00:15:18.669 --> 00:15:21.269 that ensuring the quality of a medicine is about 00:15:21.269 --> 00:15:23.509 so much more than just the final test results. 00:15:23.850 --> 00:15:26.730 It's embedded in the very DNA of the manufacturing 00:15:26.730 --> 00:15:28.789 process. It really makes you think about the 00:15:28.789 --> 00:15:31.090 future of pharmaceutical manufacturing. With 00:15:31.090 --> 00:15:33.730 increasingly complex drug formulations and constantly 00:15:33.730 --> 00:15:36.190 evolving technologies, how will the principles 00:15:36.190 --> 00:15:39.809 of QBD adapt and evolve? How can we leverage 00:15:39.809 --> 00:15:41.850 advancements like continuous manufacturing, which 00:15:41.850 --> 00:15:44.450 we touched on earlier, and the incredible potential 00:15:44.450 --> 00:15:47.230 of advanced analytical techniques to take quality 00:15:47.230 --> 00:15:50.029 by design to the next level and ensure that everyone 00:15:50.029 --> 00:15:52.529 has access to high -quality medicines for years 00:15:52.529 --> 00:15:54.850 to come? It's a fascinating and critically important 00:15:54.850 --> 00:15:57.190 area to consider. I couldn't agree more. It's 00:15:57.190 --> 00:15:58.990 an exciting time to be involved in this field. 00:15:59.200 --> 00:16:01.120 Well, on that note, thank you so much for joining 00:16:01.120 --> 00:16:04.240 us for this deep dive into quality by design. 00:16:04.720 --> 00:16:06.559 It's been a pleasure exploring these concepts 00:16:06.559 --> 00:16:09.039 with you. And to our listeners, we hope this 00:16:09.039 --> 00:16:11.580 is giving you a deeper understanding of the complexities 00:16:11.580 --> 00:16:14.700 and the incredible importance of ensuring quality 00:16:14.700 --> 00:16:17.139 in pharmaceutical manufacturing. It's been a 00:16:17.139 --> 00:16:18.840 pleasure being here. Thank you for having me. 00:16:18.980 --> 00:16:21.259 Until next time, stay curious and keep exploring.