Mettler-Toledo International Inc. (MTD) Earnings Call Transcript & Summary

So 5:00 here in Switzerland. I think most of you are attending from Americas. I see some names still from Scandinavia and from Europe. So we typically will start a little bit later, so to give everyone a chance to handle all the difficulties to log in and to connect the sound, et cetera. So I will mute all lines now so there will be no interruption. And I will keep it muted during the presentation, so I will not be able to answer your questions during the webinar. But you can always use the chat to write in your questions, and I will try to answer them as good as I can. Yes. If you need a certificate of participation, you can send me an e-mail with your name and your company. And I think that's all that we need to -- maybe also the country, that's all that we need to make that certificate for you so that it's valuable. So 1.5 more minutes, I will play this beautiful music again. So yes, obviously, it was good to wait a couple of minutes because some people could make it still. This is Christoph from the front desk, announcing Christoph, the presenter. A little bit later, Christoph and Bob McDowall. You have noticed I muted the lines. Questions will be addressed at the end of this webinar. We will take a little bit more than an hour. I hope that is fine for you. That is good for you. There is a lot of information in this webinar. So it's better that we start right from the beginning. Welcome from Switzerland to the webinar about method validation. From the subtitle. You can see that we will mainly speak about methods, which are used mainly in regulated GMP laboratories for the pharmaceutical industries. However, the new guidance can also be applied to ISO 17025, the international quality standard for testing and calibration labs, which also include food applications. We will not specifically make that difference. Before we start the presentation, we show you the legal disclaimer. The original title of today's webcast is Validation and Maintenance of cGMP Methods. How to Handle the Dynamics of Requirements and Firmware Changes. Actually, we will talk about a lot more, as you can see in today's agenda. Those of you who already have seen one of our webinars know that we always start with background information, terms and abbreviations. Next we will speak about technical aspects of the method development, we will use a titration application as a sample. Then we will speak about setting the scene with the data integrity model, how to prepare your company most appropriate. We will give you an outline of which regulations you should prepare and how to interpret them. Then we will explain the difference between procedure and method to continue with the milestones of the method validation and how long each step should take, just an estimate. There have been new guidance documents drafted in Summer 2020, and we will discuss how to interpret those. We will speak about typical mistakes. Pitfalls, we see quite often in pharmaceutical labs, those which may lead to death by compliance situations. Who is on the presenters end of this webcast? It will be Bob McDowall and me, Christoph Jansen, who will present. I will mainly ask the questions to the expert, Bob McDowall. I'm a chemist working for more than 10 years from Mettler-Toledo as a global technical key account manager for our analytical business unit. Bob, can you quickly introduce yourself?

Hi, I'm Bob McDowall. I'm a Director of RD McDowall Limited based in London in the U.K. And I've got nearly 50 years experience, 6 years as a forensic toxicologist, 15 years working in the pharmaceutical industry for 2 major pharmaceutical companies and 28 years-or-so as a self-employed consultant. I'm also a trained auditor and also do a lot of writing and training. And one of the current areas of interest is both data integrity and data governance where I've written a book on -- or 2 books, I should say, on validation of chromatography data systems and also data integrity and data governance for the regulated laboratories.

This is not the first webinar with Bob McDowall. We have a series of webinars that we still webcast, most of them around data integrity where we covered many aspects of it except for method validation, which is our focus today. You may consult our website for the dates and registrations. First, like always, we like to explain the most important and commonly used terms and abbreviations for better understanding. Terms like AIQ or Analytical Instrument Qualification might be quite familiar to you, as well as URS or User Requirement Specification; DQ, design qualification; IQ installation qualification; OQ operational qualification; and PQ performance qualification. Bob, can you please explain ICH and what is important to know about that?

ICH used to be called the International Conference on Harmonization. It changed its name to International Council on harmonization and the long title is, as you can see on the screen, International Council for Harmonization of Technical Requirements for the Pharmaceuticals for Human use, which is why we call it ICH. They have a number of strands there that representatives from initially the European, U.S. and Japanese regulators and pharmaceutical industry work together to produce a series of standards. And the ones that are of most interest to an analytical audience today are ICH Q2(R1) which is on analytical validation. This is currently undergoing revision and will be ICH Q2(R2) probably sometime in 2022. There is also in parallel a new guidance being developed, Q14, on analytical procedure development. That is being worked up, as I say, in parallel and should come out at the same time for public comment. APLM is Analytical Procedure Life Cycle Management. This has been ongoing now for probably just over 10 years. And it is a series of initiatives, both from the industry, but also from the United States pharmacopeia. You'll see underneath the APLM, there are a number of terms like analytical target profile, analytical control strategy, target measurement uncertainty, total analytical error and procedure performance qualification. These are all part of a draft USP General Chapter 1220 on analytical procedure life cycle management. A draft of 1220 was issued in the summer of 2020 and is currently undergoing revision in light of industry comments. But it is a new way of approaching analytical method and procedure development, validation and use.

Okay. From reading some material, I stumbled upon the term analytical method and analytical procedure. Bob, can you please explain the difference?

The key thing here is terminology. And one of the big problems is that sometimes, we mix things up. When you look at the pharmacopeias, they refer to analytical procedures. This typically is the whole of the process from sampling to final generation of the final reports. Whereas an analytical method is a subset of an analytical procedure and is really around the preparation of the sample and the instrumental analysis. And it's important that sometimes even experts talk about a method when they mean a procedure and talk about a procedure when they actually mean a method. It's important to have that understanding in place and actually challenge people to make certain what actually is meant by that phraseology.

And we are talking about methods in this webcast, right?

Correct.

The purpose of quality control is assuring constant quality in a defined range of a product by scientific, reproducible and accurate tests. In regulated pharma quality control or GMP, for patient safety, this means to follow certain rules. The rules are spelled in rather general statements that need a lot of guidance for interpretation and also experience. In our webcast, the flexibility of compliance, we discuss that. One of the most important guidance documents is the U.S. pharmacopeia or USP. Remember that chapters are numbered and the number has a big relevance. USP general chapters between 1 and 999 are mandatory. They are the law. Chapters between 1,000 and 1,999 are informational or strong guidance. If you have good and scientifically sound arguments, you may do it different. Typically, most of the scientific methods for checking the quality of a product are supported by instrument suppliers. They deliver hardware, instruments, software and, in many cases, application support, guidance and training. We have a large collection of predefined methods, pre-installed that are at least a technical start.

Well, one of the classic ones will be on the balance for USP 41 repeatability and minimum weight -- or accuracy, I should say. So you have that built into your balance set.

And you mentioned it, where possible, many of our suppliers pre-installed methods are based on USP methods. Which scientific method is the best for an assessment, depends on various aspects. When looking into modern pharma QC labs, one could get the impression the solution is often HPLC or other chromatography methods, which is often correct, especially for the final product. Whatever method was in the registration defined has to be used in the quality control in the defined way. However, we see R&D labs in pharma companies who consider more and more the other well-established methods, such as melting point, density, refractive index, UV/Vis spectroscopy and sometimes even in a combination to identify a product and its purity. Those methods might be faster and better cost effective. Many of these methods are described in pharmacopeias. Bob, any comment to this?

Pharmacopeial methods are the legal requirement. If you change it to HPLC, you have to cross verify. And if there's a problem with the HPLC method, the pharmacopeial method is the legal standard and takes precedence over it. And it's interesting to note that the United States pharmacopeia is actually legally above the FDA's GMP regulations. And that's in the Food, Drug and Cosmetic Act.

Do you always have to validate a method, for instance, identify a substance, such as cholesterol in a pharmaceutical quality test for incoming goods even for USP methods?

No. If it's a pharmacopeial method, it is deemed validated as written, okay? You have to verify that the way that you use that validated pharmacopeial method works. So it's a verification. It's less than developing a method de novo. So what you would do is verify that for cholesterol 147 to 150, you would have a fineCifan standard at 135 and Acetanilide 160 as the USP standards, and you would check to see that everything works with that method. Because the pharmacopeia says you want a standard top and bottom. And if it doesn't meet, then you do a mixed melting point. So the verification of a pharmacopeial method for something like melting point should be relatively quick. If you are developing a new method for my super duper blockbuster drug, then you are doing a method validation because there's not a pharmacopeial method for this. And so that will follow ICH-Q2(R1), APLM in 1220, something along those lines. So it's a full validation as opposed to a pharmacopeial method verification. There is a difference.

Heinz-Dieter Isengard once said only the sample knows the true result. No method provides absolutely true results. There is always an error involved or a number of errors in a workflow. Weighing, dilutions, pipetting, temperature measurements, atmospheric influences, et cetera. Even purity variations of standards, chemicals or titrants may contribute to deviations from the true result. They all contribute to the final measurement uncertainty and should be considered. Other than random systematic influences to the uncertainty, blunders or transcription errors are not included in the measurement uncertainty calculations, although they may have a big impact on the deviation of the true value. By using electronic records in a seamless workflow, those type of errors should be minimized. The measurement uncertainty, value determination is mandatory from regulations, such as ISO/IEC 17025 from 2005 to not always talk about pharma regulations. These are mathematical algorithms to calculate the uncertainty after performing a number of single measurements. For determination of the measurement uncertainty of titration, we offer a service called Newpack, where you do some guided measurements, you send us the results, and we calculate the uncertainty and print a comprehensive report for you. This service only exists for titration. Okay. Let me say a few words about method development. Chris Burgess is one of the gurus in this topic once said about the importance of method development, method validation is merely the confirmation of good method development, which underlines the advantages of a clever development of the method. Actually, method development is, in most cases, not so terribly difficult and can be made in a couple of days. With a well thought design space, you will keep some flexibility for later improvements, which we will discuss later in this webcast. If we take a titration method as an example, there are a few things to consider. You need a good reaction, find a proper sample size and find a way to confirm the accuracy, make sure your chemical reaction is good enough. Your sample has to react fast and stoichiometric with the titrant. Optimize your sample size to balance the accuracy with optimal time and chemical consumption. Take more sample and the accuracy is better because your relative error becomes smaller. But the titration will take longer and will use more titrant. Find the optimum. To assess the accuracy use commercially available standards with defined characteristics. You'll need high purity. This also applies to all chemicals in the process. Use chemicals that are stable enough, for instance, against oxygen or carbon dioxide, if not take actions to protect them. There is a lot more to say about this and more details can be looked up in that guide that you can download from our website. Don't underestimate proper method development. In titration, larger sample sizes lead to more result precision, but this costs more time and titrants or chemicals, as mentioned before. When optimizing a method, the optimal sample size should be evaluated. Thus, part of the method development is finding the proper sample size, find the right balance. Please check if the result has linear response, as you can see in the left chart, or nonlinear response to the sample size, as shown in the right chart. When you develop a method, make sure the expected result is above the limit of quantification or LOQ. Of course, if you always measure higher than the limit of quantification, you do not need to determine it. Make a number of measurements and evaluate the relative standard deviation at certain concentrations. Repeat the measurements of each concentration to get an RSD for it, Relative Standard Deviation. As a rule of thumb, in a titration method, the sample size is typically good as long as the RSD or Relative Standard Deviation, is not more than 0.3%.

It's interesting dealing with quantitative bioanalysis the LOQ was defined at about plus or minus 10%. We're sometimes down in the nanogram, sub-nanogram region. And sometimes, we may have a dynamic range of somewhere between 10,000 nanograms to a few nanograms. So you've got a big dynamic range to handle.

As an instrument supplier, we quite often observe in the market that methods are pushed to the technical limits, limits that only work under ideal conditions. Those methods sometimes are not robust enough. They are not rugged enough. What if the pH analyte volume or conditioning of the sensor has variations? This often causes headaches in routine applications because the conditions are not always ideal. Bob, do you have any recommendation for the method development with this respect?

Yes, certainly. I think the key thing here is that many laboratories rush method development. And they rush it and don't fully understand the impact of these changes in any of the parameters or any of the materials that they're using. And ruggedness is addressed, albeit in a different name within the draft of USP 1220. And it's talking about an analytical control strategy. And here's what you try and identify early in the method development process. The various parameters, either with the instrument or with the preparation and analysis that impact the ruggedness or the robustness of the method. Once you've identified them, you then need to understand if you vary these components, what the impact is on the method. And if they need to be controlled very carefully, what have you got to do to control that so that when you go into the operational phase of the method, you know exactly what you've got to do. But if you rush this process, you are going to pick up the headache when you get into the operational phase. So really good method development is absolutely essential to ensure smooth method validation and operational use of your method.

And when the optimal method technically is developed you will have to think about the paperwork, the validation of the method and the qualification of the equipment. This is in the draft of USB 1220 in the Phase I, but as Bob mentioned, under a different name. Today, we are talking about method validation. Many people confuse this with instrument qualification. What is the difference? Let me quickly explain. You qualify instruments and equipment, but you validate processes, systems and methods such as computerized systems or analytical methods. Qualification is something you do on single instruments. You check if an instrument design is suitable for the task. If it works according to specification, if it still does that after some time, and you check that with certified standards. Qualification of instruments controlled by computers is part of the system validation. In design qualification or DQ, you define what the instrument should be able to do. In a computerized system validation process, the design qualification is called user requirement specification or URS, the most important document in the process. Initial installation qualification, or IQ, as well as operational qualification, OQ, often is offered as a service by the supplier. They just check if the instrument works according to the suppliers' specifications. This may take 1 or 2 days with all the documentation. Performance qualification or PQ is testing versus the labs, URS by the users. This is usually not done with the customer's product but with his setup, including staff. This is typically not performed by the supplier. As a test sample at a balance, for instance, you use certified weights. Or for pH, you use buffer standards. Or for UV/VIS spectrophotometers, you may use certified standards in sealed cuvettes. Our CertiRef and LinSet tools have such standard materials built in cuvettes in a small sample changer to perform repeated qualification automatically. Other than qualification, a validation is about a process. This includes more than just the instrument. You validate the process and as the instrument is just a component of the process, the instrument qualification is a part of the validation process and validation includes a lot of paperwork. Now let us talk about a general strategy of method development and validation. We are looking at GMP processes and therefore, data integrity should be in the core of our considerations. Bob, how can method development fit into data integrity?

The issue with how method validation fits into data integrity can be found in this data integrity model that I have developed. Firstly, data integrity is an integral part of any quality management system. And what you've got here is 3 levels and a foundation layer. And these levels and the foundation are meant to represent an analogy of building a house because if you build a house without a foundation, the house falls down. So you've got to have a foundation of the right culture and ethos to do the work, and that starts with 2 words that are not actually found in close proximity to one another, and that's management and leadership. This essentially is developing the culture from the top down. It is not the Chief Executives or the Managing Director, or the President or whoever, sending out an e-mail saying, from now on, we're going to assure all our data, our integrity, or ensure our integrity of our data and data quality. It is working hard to change people's attitudes and ensure things are in place. It's encouragement, it's setting objectives, but it is also having policies and procedures in place that allow people to understand what is data integrity and also to manage or to understand what is expected of them. Then if we look at the quality control levels, the first thing above that is Level 1, the right analytical instrument and the right computer system to do the job. So here, we're talking about qualification of the instrument and validation of the software. When you've got that in place, then we come to Level 2, which is what we're talking about in this seminar, which is the appropriate validation of verification under actual conditions of use. And actual conditions of use is very important. You can't just take a method from a pharmacopeia and just execute it as it's written. It doesn't work that way. You either have to use the pharmacopeia as a starting point and then verify that it works, or you develop a method and validate it de novo to show that it works over the range that you define that's appropriate for the method. When we come to apply that method using qualified instrumentation, validated software with appropriately trained and qualified staff, we come to Level 3. And if there are problems, the analyst needs to raise those problems and talk to management and say, "I've got a problem here" and investigate. There is no point in hiding things, so management have a role here to allow people to speak up. Unlike an ex generic pharmaceutical company, Able Laboratories, where your job depended on you passing every batch of material. And that's the start of the problems that we have found in the industry over the past 15-or-so years due to falsification of data. On the left-hand side of the model, you can see the quality oversight organized by quality assurance to check that everything is working correctly within the quality management system.

And where can I find this data integrity model reflected in the regulations? Which regulations must be considered and how should they be interpreted?

That's an interesting question. Here, we've got the United States or the Food and Drug Administration, 21 CFR, 211 current good manufacturing practice regulations. So if we start at the bottom of the slide, we've got 211.63 looking at equipment. So essentially, we've got appropriate design, adequate size, suitably located for intended use. So this defines what you've got to say about the instrument. Sitting above that is under the laboratory control section of the CFR is 211.160(b), which basically says, whatever you do in the lab, you want scientifically sound and appropriate specifications and test procedures. Those test procedures will include, not just the analytical procedures and methods, but also how you're going to qualify your analytical instruments. Then we have sitting above that 211.160(b)(4) talking about calibration of instruments. Because the regulations were first published in 1978 and haven't been updated substantially since then, essentially, we're dealing with calibration as the only method of control. Whereas since then, we've had introduced the term "qualification". So most of the FDA citations will be against calibration. But essentially, you need to have a protocol for qualification of the instrument with limits of accuracy and precision. And what it does say after that is if you haven't met those limits of accuracy and precision the instrument must not be used. The calibration of the instrument or the qualification of the instrument, as we call it these days has to be recorded in 211.194(d), which contains the records and any periodic calibration of that instrument. And then sitting above that at Level 2 in the data integrity model, we have 211 194(a)(2), talking about analytical methods being verified under actual conditions of use. So this is demonstrating that the method is suitable in your laboratory.

Okay. That was what the regulations say but there must also be guidance documents. What are the guidance documents that help us understand the regulations?

Essentially, you have help from both official and industry guidance publications. The first are pharmacopeias, and we'll use the United States pharmacopeia, or USP as one instance. We have general Chapter 1058, which looks at integrated approach to both instrument qualification and computerized system validation. There is also the GAV5 on computerized system and also a good practice guide on validation of laboratory computerized systems, second edition that was published in 2012. Sitting underneath that, we have the draft general Chapter 1220 on analytical procedure life cycle. This would move us away from precision and accuracy as measures of parameters to measure in an analytical procedure, to measurement uncertainty. Welcome to the 21st Century to the pharmaceutical industry, and welcome to the Rest of the World because measurement uncertainty is also used in ISO 17025 laboratories. Then we also have sitting underneath that general Chapter 1225, which is the interpretation of ICH Q2(R1) and also a general chapter on verification of analytical procedures, which are 1226, which is essentially how to check and see how you've interpreted any pharmacopeial procedure for your laboratory. The one thing that is important to understand is that pharmacopeial methods are deemed validated. The only problem is that the instruction set you get needs to be interpreted. And it's not just reading the individual general chapter or the monograph, you also have to read instrument general chapters and the general notices. So Chris Burgess and I have published an article in Pharmaceutical Technology online in October 2020 that you can find more information on that.

And what about global harmonization? In the introduction, you mentioned the ICH, the International Council for Harmonization of Technical Requirements for Pharmaceuticals for human use?

Yes. We have -- currently ICH Q2(R1) that's been published and is probably about 15 years old now. The problem is that it is focused very much on chromatography and does not consider method development at all. It is currently being revised to become ICH Q2(R2). And hopefully, will expand on the validation to include other methods of analysis such as spectroscopy. The idea would be to go wider, but that is still undecided as we haven't seen any drafts to date. ICH Q14 is being written in parallel. That is looking at analytical procedure development, which very much overlaps with part of the draft ICH 1220. Now I think as 1220 is further along the line than Q14, there may well be elements taken from 1220 and incorporated into ICH Q14. But this is still very much works in progress. So you can find out when the draft are published for industry comment. But these are still probably 2 to 3 years away from finalization and final publication.

Okay. Now what are the milestones for validating methods? What effort can be estimated? How should it be resourced? And how long should each step take?

Depends on the complexity. Firstly, you need a good method development report that defines all of the work that has been done with the analytical control strategy. What are the critical things that you need to control. Coming into that, you would define what you're going to be doing in a method validation plan. It will define what you're going to be doing and the various experiments. That may be repeatability, intermediate precision, limits of detection. One of the biggest issues is, don't just copy all of the parameters in method -- in ICH Q2(R1). If you're looking at assay, you don't need to do limit of quantification or limit of detection. So use your brain to say which are the critical parameters in relation to what I'm trying to achieve with this method. The next thing is that you do the work properly, take your time and certainly, this is where method verification or development comes for pharmacopeial methods. The validation plan you need to make certain you've got your objectives and please define the acceptance criteria before you start the work, and those acceptance criteria have to be scientifically sound. It is important to understand that if you've done method development correctly, the validation is merely confirmation of your good work in the development phase. If you skimped on the development, you will have problems in the validation or when you pass it to quality control to do the routine work. Typically, the method validation plan will -- you'd be over a template and it should be pretty quick to develop for your particular method, literally a day or 2 would be my approach. Because if you've got the experiments, you just need to change and add in new ones or take out the old ones. So that should be pretty quick. In fact, the review process typically may take longer than actually writing the thing, depending on the availability of staff within the lab and QA to sign this off.

Okay. How do you specify exactly what you need the instrument or system to do? How do you define fit for purpose?

Fit for purpose is essentially document does the method in your validation plan meet your analytical requirements. Have you demonstrated that the method has been verified under actual conditions of use as required by the 21 CFR 211.194(a)(2) because if you haven't, you've got a problem.

I guess you will somehow have to fulfill what's in the analytical target profile. Can you explain what this means?

An analytical target profile is a simple statement of what you want the analytical method to do. It essentially is independent of the analytical technique you're going to use, and it will mention items such as the analyte or analytes you want to measure. Whether the method is going to be qualitative or quantitative. If it's quantitative what's the range over which you want the analyte or analytes to be measured? What's the matrix, be it a solution, a tablet, a cream, an anointment, or even, I hate to say this, but a suppository. And the point is that from this ATP, you then start to design the analytical method. And you start to say, "I'm going to measure the analytes here. With this ATP, I'm going to need a balance or I'm going to use chromatography. I'm going to use UV. It's a melting point, something like that. So it's a starter and you start to develop the actual procedure from the analytical target profile.

Okay. Do you have any example where this went wrong?

Oh, yes. The one thing that is very good is that the Food and Drug Administration published not only warning letters, but also key 483 citations from various companies on their website that's publicly available. We can see here with 1 company where laboratory controls should -- do not include the establishment of scientifically sound procedures. Essentially, there is -- there are numerous examples. Here's one where the lab has been cited for using a wrong standard when they're doing Karl Fischer water content determination. The standard used was far away from where the lab was measuring. This is wrong. Ideally, you need to do top and bottom of the range that you expect to demonstrate that the instrument is performing as intended on the day of analysis. This gives you confidence in the data that you are generating from the samples. If you don't do that, you are open to question the data and the integrity of those results.

In September 2020, there was published a new draft of general Chapter 1220. What are the major changes? What impact does that have on existing methods?

What we now have is a life cycle approach to analytical procedures. So it's analytical procedure life cycle management, APLM. And what this starts with, which is missing from ICH Q2(R1) is method development. You have, in essence, an analytical target profile, which is your user requirement specification for the method. From that, you determine the instrumental approach and what you are going to do to develop that method. It may be something totally de novo and/or it may be adaption of an existing method for a homolog or something equivalent. But you need to spend time developing what we would call the design space. You're going to see, as things develop more and more with new methods, they're going to need to define a design space. From that, you will go into the method validation, which should confirm the good quality method development that you've done. And then from there go into the operational use of the method. Any changes that are made in the operational use of the method that are within the design space can be made without revalidation. If you have something that goes outside of that, then you go through change control and revalidation of the change. If you're changing the way the method operates, perhaps incorporation of new analytes or different operating parameters, you may need to come back and redevelop the whole method and start again. That's where perhaps you will change the analytical target profile and start through the process again. Also, what you may have is transfer of the method, either from the development group to quality control or from one site to another. That transfer should be documented. It should be tested to ensure consistent results with the originating laboratory. But what ever you do, remember the requirement to verify under actual conditions of use. That is the critical thing.

Okay. Assume you have validated a method and now your instrument becomes obsolete or you want to change the supplier. Will you have to revalidate with a new instrument?

No, but it requires verification under actual conditions of use. So if for argument's sake, you've got a new instrument coming out or you're moving from one supplier to another, then what you want to be able to show is that the results on one instrument are equivalent to the other. Not exact, but equivalent because obviously, you'll have instrumental error. So you will do a confirmation that the new instrument produces equivalent results to the old instruments. So you'll do a comparative test. You'll have the same samples, run them on both instruments over the various ranges that you're going to be working under and check to see that they are within 95% confidence limits or whatever, they are equivalent. You will have a plan for this, along with acceptance criteria that are predefined before you execute. How you compare the results will be defined and the acceptance criteria will be defined. That's the critical thing.

Okay. And what do you think how much work is involved? How long will this take?

Obviously, it depends on the complexity. But again, writing the comparison, if you've got templates should be done pretty quickly and approved again within a couple of days. The work anywhere from a couple of days to 5 working days, depending on the complexity. And then the reporting should be done pretty quickly after that. You should look at this as trying to check, plan, perform and report as quickly as possible, otherwise, you can get swamped with routine work. You need to be able to do this properly. When you look into the draft USB 1220, there are some differences in terminology. The team that were writing the 1220 was to equate this to process validation guidance some 8 to 10 years ago from the FDA. So we have the requirement. Our specification is the analytical target profile. And then there are 3 stages in the life cycle. And essentially, instead of calling it method development, we call it, procedure design and development. So this is where you try and find out the critical factors, develop the instrumental process, the sample preparation, any other material preparation, buffers, et cetera, and you work out what are the critical control parameters and assess the robustness, develop the design space for the method. For more complex methods, you may invoke or use experimental design software. But essentially, the aim is to develop a robust methodology. From there, you go into the -- what is known as procedure performance qualification, PPQ. This is a method validation under any other name. Once you finish that, it will move into the production environment for procedure performance verification. And the key thing here is you are monitoring how the method works on a day to day basis. There are various feedback loops of continual improvement, and that's important. Changes, as I mentioned just now, changes within the design space can be made without revalidation. Changes outside need to be considered from the perspective of change control and just how far back in the process you go. But sometimes you may want to make a change that takes you right the way back to Stage 1 or Part 1 where you have to revisit the analytical target profile and then go through the whole process again.

So can you please summarize in a nutshell what the changes are?

The whole thing is, for the first time, you have the whole life cycle of an analytical procedure in one document. Up until now, ICH Q2(R1) has not included method development or any formal way of defining what the analytical procedure has to do. And this is the major change both with 1220 and also what is, I think, going to be proposed for ICH Q14. And perhaps one of the things with ICH is that you should actually merge Q14 and Q2(R1) into a single new process around analytical procedure life cycle management or APLM. So what we have with 1220 is an integrated approach to development, validation and operation of analytical procedures, that is far better than we've had in the past.

Now to one of my favorite questions always, what are the typical pitfalls or the common mistakes? How do you avoid death by compliance?

I think one of the first things of death by compliance is following ICH Q2(R1) blindly. Essentially, if you're measuring an analyte between 80% and 120% of target, why do you need to worry about LOQ and LOD. I think it's totally a waste of time. What you find is, ah, it's in ICH Q2(R1). It is, but why are you doing it when you're going to be measuring something possibly 2 or 3 orders of magnitude higher, it's a waste of space. The other thing that is very common is that companies will take the pharmacopeial limits for instruments and actually make them tighter, "look at what we are doing". They're in the pharmacopeia, for a particular reason, don't change them and don't make them tighter. You are going to have problems with failure when you come to qualify your instrument and you will have to explain it. And you may have failures when you come to run the system normally in operational use. So the key thing here is, don't be smart. If the pharmacopeia says, this is a particular limit, use it, don't change it.

I have once read that the qualification is more a journey than an event. Can you please comment on that?

Yes. I think it's important that when someone qualifies or calibrates an instrument at the start when they do an operational qualification, it's the start of a journey, it's not the end of an event. And what you've got to be able to demonstrate throughout the time that use the instrument is showing that the system is still fit for use, can be related back to the requirements in the laboratories, user requirement specification and it still meets those requirements throughout the life cycle. When you come to an annual preventative maintenance and requalification, that you can check the requalification is the same as previous requalifications, and the instrument hasn't changed its operation or changed its measurement capability, this is very important. It gives us long-term confidence in the work that you do.

Okay. Interpretations of regulations evolve, and so does technology. How to keep a method current with regulations, requirements or interpretations, but also keep it technically state-of-the-art?

Actually, this is probably one of the most difficult parts to consider because if it's a registered method within your marketing authorization of product license for use of the method. It is sometimes very difficult to change that method and trying to keep it current. For drugs that have been on the market for some considerable time, you may be using very old methods. And one of the problems is that if you want to change it to a different analytical technique, you have to register it with the individual regulators that you have marketing authorization for and that costs money. So many companies are rather loath of methods that are currently on the market and enrolled, are rather loath to change. And this is probably a big issue with the inflexibility of licensing of pharmaceutical products. For newer methods that are still being developed, current keeps things up to date, simply on the basis that before you put it into a marketing authorization, providing you change control it, you can make those changes without major impact on your regulatory dossier. But this is one of the more archaic parts of pharmaceutical regulation that you have to go through regulatory approval to make changes with existing methods. And prevents people from actually being current with new methodologies and new analytical techniques.

So you say it's difficult to keep your methods current, current with latest technology and maybe requirements. What are the options to make changes in your methods?

If you have a registered method, you can make changes. And within the European Union, there are 3 types of change that you can make to a method. These are Type Ia, Type Ib and Type II. So essentially, a Type Ia variation is a minor change with no impact on product quality or the specification. And if you determine this is a Type Ia, you just make the change. If there is a change that again is a minor change, it may have an impact on product quality or the specification, you would have to prepare a justification, submit it to the regulator stating that it's a Type Ib variation. And if you haven't heard back within 30 days, you can make that change. More bigger impacts require a more formal approach and this is covered under a Type II variation where it impacts the specification, for instance or a major change on the method. Here, you need a proper change request to the regulator, justifying why you're making it. You may also have to provide evidence to show that what is the impact of the change and that is reviewed by the authority, and you get formal allowance to make that change if the regulator is happy that you've done the job properly and they're happy with the justification, they will allow you to make the change. A type II is almost a reregistration. You're making a major change to the method. So here, you might be changing the analytical technique.

You mentioned it is recommended to leave enough design space to be able to make changes without affecting the marketing authorization. Can you please explain that?

I think the key thing you've got to think about here is that most companies will actually write the analytical procedure that they register very much like a pharmacopeial method. Very generic. That allows them the flexibility to change. If, for example, you specified a specific make of model of instrument to do a particular measurement, if the supplier discontinued that measurement, you've got a problem because your method is wedded to that instrument. On the other hand, if you have more flexibility, you can keep current. And also as part of keeping current without having to change a method, you have the ability to implement technical controls within software to ensure data integrity, that wouldn't affect a marketing authorization.

What are the measures if instrument firmware is updated? Repeat, redo the instrument qualification, repeat the validation?

I think the key thing here is to look at what release notes have been given with the instrument. You certainly need to requalify the instrument, but the firmware update will typically be done as part of the preventative maintenance, and then you will get a follow-on of the OQ to show that it meets the requirements in the lab user requirement specification. What I would also do is from the release notes, do you need to run any reference standards, samples or whatever, just to see what the differences are between the old firmware and the new. It's very difficult to give concrete rules here because you never know what the impact is of the firmware update. But be careful. And certainly, the minimum you need to do is an instrument requalification typically after the preventative maintenance visit. And from there, you would need to determine do I need to do any retest. I wouldn't repeat the validation, but that will be the normal course of events. But we need to understand what has changed and how that affects the way the instrument works, and that's where the release notes are so critical. Well, Christoph, you've been asking me all the questions, perhaps it's my turn to ask you something. What can a supplier do to help support their customers around method validation and operational support?

This feels like an infomercial. Thank you for the question, Bob. Actually, we do have a lot of material services and support to offer. Like many companies our service offers IQ and OQ, but I should better start with the support in the selection phase. Our sales consultancies are trained to find suitable instruments for your needs. Paper-wise, for your documentation, we can offer a form for your user requirement specification, or short URS, for a computerized system. That document template can save you quite some time. Measurement uncertainty, determination for titration applications is a service that guides you through some measurements and creates a comprehensive report for your documents. We call this Newpack. I mentioned that. GWP or good weighing practice, along with a number of similar packages for all kinds of instruments are consultancy packages for best practice and minimizing operational risk. They give tips and hints for better use of the instrument with the aim of getting better results and extending the lifetime of the instrument, et cetera. For our rather new UV/VIS spectrophotometers, we offer accessories to automatically perform a number of pharmacopeial tests with certified standards. The standards are in sealed cuvettes and located in an auto sampler. They run in an automated procedure and make a test report for you. Our service for this instrument family includes a new tool for measuring the photometric accuracy as recently added to the performance test requirements by the U.S. and the European pharmacopeias. To summarize, there is a lot we can do. Yes, this brings us almost to the end of this webinar. Bob, what are the conclusions from your end?

Conclusion from my end, if we go back and look at the data integrity model, we need properly trained staff. But if we focus on what we're looking at for method validation we need, the underlying requirement is a qualified analytical instrument and a validated computer system. And then what we need to think about is not just what it says in the regulations, but also to include the pharmacopeias and ICH documents. What's coming down the line is not just focusing on method development, but USP 1220 where we look at analytical procedure life cycle management. And here, we take an all-encompassing view and the really critical part of the whole of the life cycle is get the method development right. If you don't get that right or rush it, you are going to have problems in the validation and more likely in the operation of the method. The better you design document and understand the design space, the better your method is going to run in real life. So that will be my conclusions from this talk.

So for those of you who are interested, here is a collection of some of the literature we used for this webinar. And I'd like to open the discussion for a short Q&A session.

Any questions. I opened the chat. Can we get copies of the slides later? No, you can't. But there is -- this will be repeated in, I think, June next month. So that's the date June 22, next -- not next month, next year. Okay. So I hope that answers this question. But there is a lot of literature available, and I could look what I can make you available. Here is a book, which I can't send to you, but this was published in 2019. It's still quite current. That was a book by Bob McDowall, about data integrity and data governance. And if you enter the voucher code Data20, you can get a 20% discount for this book. It has 600 pages, almost. Okay. So will there be a certificate? Yes, you can have a certificate for attendance. Send me an e-mail. Here is my e-mail address, and you will get the certificate that may take about half a week or so. But yes, you can get that. I need your name, your company and I don't know if we put the country even on it, but it should have some value, and therefore, we just need a little bit more than just a name. So are there any method validation seminars resourcing within biologics and analytical methods? Maybe there are, but not from me. I did this, of course, because we provide analytical instruments. So -- and we have solutions in the total complex of computerized system validation. So analytical instruments with software. Our solution is LabX. But for biologics, probably you would need to look on the Internet for other people who give this type of presentation. But I interpret this as a complement that you would like to have more from us. And actually, this is a good hint. We do have, as I mentioned earlier, a number of webinars, maybe not going so deep into it, but we have next year in January 19th, there will be this webinar on Lean laboratory. So to improve the efficiency of your lab by using these tools from the lean concept, and we will have more -- yes, more basic information about data integrity in here, cost-effective computerized system validation, so how to avoid death by compliance, what do you need to do? This is a webinar about so-called firmware instruments because some people give you -- like to give you the impression, it doesn't need a computer for a computerized system according to 21 CFR Part 11, and we will explain why this is quite dangerous. This is Lean Lab, next month. And instead of this flexibility of compliance, we will have a new topic which will be released. I think in May, next time, next year. This is about misconceptions, myths and lies about data integrity. I think this is going to be almost fun. Yes, as I mentioned next time, we will repeat this webinar that will be in June 2023. Maybe it's -- I mentioned it at several points already. So in May this year, USB 1220 was already released. So it came out. And also the ICH Q2(R2) is also -- the draft is public. Bob said, it's not really great. ICH Q2(R2) should be better, but USB 1220 came out quite well, was his comment. Yes. So if you have any comments, you still can send me an e-mail, down here is my e-mail address. You could ask questions. You could also ask questions to Bob McDowall. He's not with us today so the things I showed you was recorded. But -- yes, he will shortly answer any questions that you have that I cannot answer. Yes, for the certificates anyway, and I'm happy about any type of feedback. So if there are no more questions, I will close the session. Thank you for attention -- for your attention. Sorry for taking a little bit longer than anticipated. And have great holidays and good rest of the day and rest of the week. Bye-bye.