Lonza Biologics: Making Drugs Safer and More Cost Effective
Jesús Zurdo, Head of Innovation in the Biopharmaceutical Development at Lonza Biologics explains Lonza’s approach towards safer and less risk-involved methods to develop and produce drugs.
Pharma IQ: Could you perhaps tell us a bit about your role as Head of Innovation at Lonza Biologics, and what you would identify as the most important challenge right now in the pharmaceutical industry?
Jesus Zurdo: Lonza is a contracting factory primarily. We support many companies in both the pharma and the biotechnology industries. My role is to help develop and define new approaches to developing drugs in a safer, less risky way; helping our customers reduce their exposure to risk and making our processes more cost effective primarily, but also more effective in helping them be successful in their processes.
I think the main challenge for the pharmaceutical industry and all the associated companies that are supporting them is the high cost of developing new drugs. There was a piece published by Forbes showing how the costs are dramatically higher than people anticipate. This is primarily motivated by the extremely high failure rate. In some cases more than 90% of drugs entering the clinic fail. So it is a very risky business, and it puts a lot of pressure on how the costs and the risks are managed.
Clearly the blockbuster model is having issues in terms of sustainability. And because the costs are too high, it makes it difficult for the new players and very innovative companies to enter the game. There is also an increasing pressure from payers and governments as a result of the current crisis, and this is something that is going to become more of an issue in the future. I am hearing more and more from customers and from companies that how they approach the payers is becoming one of the most important things in defining whether or not they would go to the end of the development of a drug. And this is showing that it is not just the safety and even the biological activity, it is also value for money and affordable for the user
Pharma IQ: So how do you view this? What implications does this challenge have for your company and your role?
Jesus Zurdo: What we aim to offer is value for money, and of course we always strive to reduce the costs of the services. But I think, more importantly, as a service provider we share problems and difficulties as well as success. So we are looking at ways of reducing risks, and ways of increasing the efficiency of product life cycle development.
Pharma IQ: Could you perhaps give a concrete example for that? Where was this simultaneous approach successful?
Jesus Zurdo: Well, it is not a single approach. We have a number of different strategies in terms of the structure of the product development. And this is something that we alone cannot do. Particularly in biologics it is a very complicated, very cumbersome, very risky process development, and over the last few years we have been pushing towards simplifying the process, making it more robust and making it quicker.
So one clear path is to streamline the access to the clinic. It is important for everybody really, particularly for small players because at the moment, getting into the clinic and back from Phase 1 could take up to four years in some cases. And by reducing that, smaller players would be able to manage that with their current investment cycle which goes in terms of three years typically. But also for big companies, it is very important because one year less in development is one year more that you can exploit your drug. And it could be worth many billions. So that is first for us, but also over the last, I would say, four or five years, we have been quite innovative in this area.
We have been looking at ways of de-risking the drug, as such, particularly in biologics, and looking at ways of predicting and establishing stability earlier on, looking at safety, particularly immunogenicity. So the focus is to manage the risk in a more rational way. You don’t find a problem and then think of a solution, but more looking at things that can go wrong in the process by analyzing your product – your molecule. So, that is what I call implementing a true QbD – it is not just modifying and understanding your process, it is also understanding how the chemical composition or the secrets of your product, or the way it is made or designed, has a big impact in how you can manufacture, how you can develop your product. If you can engineer the right properties in the product from the beginning, you would have a much easier life in development.
Pharma IQ: Leading over to the next question concerning early risk mitigation, could you maybe describe for someone who is not familiar with the term what the most promising approach is in this field and why?
Jesus Zurdo: I think there is a single magic bullet that would solve most present issues. I think it is a collection of different approaches, and a more collegial way of approaching development. I think clearly, and not only in our company but globally and as an industry, there is a significant push towards in-silico methodologist approaches and the idea is to use the power of computers to predict what is going to be the behavior of your product during the process development during the manufacturing, but also later on in the clinic. And in some cases there are very interesting examples.
In terms of biopharmaceuticals, I think one thing that is very successful is to understand what the elements that define the behavior of your product are. For example, you have a protein that has some tendency to aggregate. If you know that you can identify where the aggregation tendencies are located in the molecule, then you can engineer it from the beginning to make it less likely to aggregate. Immunogenicity is a clear example where there has been a lot of work in recent years, very successful work.
If you can identify particular motives that are likely to generate an immunogenetic response, and I want to qualify this, it is more likely to trigger, but it is not going to be a final solution. But the idea is that you identify elements that potentially can drive the molecule to elicit an immune response, then you can try to remove those elements and make a safer drug. So the idea is that the earlier you consider all these elements, the earlier you tackle these problems, the easier, the cheaper and the more effective it is. We have found many times in process development but also with customers going to the clinic that they encounter significant issues, and this could mean stopping the program altogether, and this could be a disaster for small companies – small biotechs.
I mean normally you would encounter difficulties in the clinic, but on many occasions, modern people want to believe that the issue has become apparent already during the manufacturing process. It is very important to get things right. You can make it at the right level, you can make it in a stable formulation, and you can make it in a composition that is compatible with the route of administration. And also we have seen often drugs coming back. They fail in the clinic maybe because of immunogenicity issues. This is typical in biologics. So these are things that can be successfully tackled early on, and I think we are helping our customers in developing this. There are many things that we are interested in developing, and the whole approach is helping them to go through development as easily as possible. There is always going to be the hurdle of biological activity and all those elements that we cannot incorporate into the design early. But our approach is helping them reduce that level of risk.
Pharma IQ: Are there cases where early risk mitigation is sometimes not possible?
Jesus Zurdo: In principle in all cases it is possible. The beauty of in-silico methodologies is that you can run them before you have done any experimental work. It is extremely useful to look at the potential issues that you would face before you embark on a cell line construction or process. These things are not cheap. It is worthwhile slowing down a bit and making sure your product has the right properties before you move forward. I have many examples in which I have seen products just being pulled because of developability issues; because of manufacturing issues, stability aggregation and also immunogenicity.
We had recent cases where products were pulled out of the clinic because of immunogenicity issues. Now, these things could be addressed at the very beginning and I think what is important is that people see that the cost of pulling something out of the clinic, even if you can go back and redesign it, of course you can do that, but it involves the loss of many years. Typically, in biology it could be 3 to 4 years at least – a cost that could sometimes be too high, particularly for small players. So again, looking at this issue at the very beginning, the use of in-silico methodologies particularly offers a very adequate and high throughput and relatively low costs, and this is required for experimental work.
Pharma IQ: What are the advantages of in-silico tools for process optimisation in comparison with in-vivo or in-vitro?
Jesus Zurdo: I think you always need the three of them. You can’t reproduce a human being on a computer. The day we can do that, it will be fantastic. But it is still a
long way. We are just scratching the surface. We are starting to understand how the properties of molecules influence their behavior in the process but also later.
So the advantage of the in-silico is that it can be done very early on, just when you are thinking about your product. Sometimes, particularly in biologics, you can handle as many sequences as you want. The throughput is virtually limitless. There is no limit to the amount of sequences or molecules that you can analyze. It does not require any specials setting or experimental work. It just requires a good model and a computer. So you have very low cost. And this makes it very attractive to seek and screen sequences and concentrate on those that are potentially low risk. Of course it has an intrinsic error, it is not perfect, but you have a large number of sequences and you can really substantially limit the risk. Now you can always apply this to engineer molecules to make them more effective and you would always need to use in-vitro systems to validate that. The challenge I suppose for in-vitro systems is to develop a predictive assays that reproduce the nuances of a biological system.
For example, in immunogenicity, there are currently ex-vivo systems being used in the industry where you can reflect population viability. So you have samples extracted from human patients, human donors, and you can have a pretty good idea how responses occur across the population. I mean, in-vivo from a process perspective, and particularly safety, always shows the true picture. But it is expensive and it has a number of significant ethical challenges. So what you would like to do is make sure you reduce the risk first in-silico, then in vitro, and then you go with the one molecule that you really believe is going to behave appropriately. It still might fail, but the likelihood of failure in the clinic is going to be reduced. And also this is good news for patients. It is good news for drug developers and potentially it could be good news for payers as well, because it should reduce the costs of developing drugs.
Pharma IQ: So the challenge for in-silico is how to maximize the effectiveness of your computer simulation in comparison with the more traditional tools of in-vivo and in-vitro.
Jesus Zurdo.: Yes, and there is still a long way to go. But I have to say that over the last few years, this field has exploded. There are much better tools out there. The main issue with computational models is that they are intrinsically linked to the nature of the data that you can feed in. If you instrumental data is not complete enough, or is not a good predictor of the final viable that you want to observe or predict, then your model is not going to be very effective.
And we have put quite a lot of effort in trying: A, to understand what the models are really telling us, and B, to be honest about it. I mean it is not going to tell us if this drug is going to be perfect or not. It is just telling us what is going to be the level of risk. And there is always going to be a level of risk. We reduce it, we just don’t eliminate it. But also it is important to understand what the nature of the data you are feeding in to build this model is. So how do they reproduce or represent the true picture of what happens in the process. For example, instability: Instability manifests in many different ways. So I do not have a single number to tell you what is going to happen in the early stages of development, what is going to happen in formulation etc. You need to have a collection of tools to explore all these avenues. And you would likely not have a number. You would have kind of a threshold, a kind of confidence level – am I more likely going to have more issues or not. And once you have that threshold, what you do is that you use that to eliminate sequences. It is not a question of comparing, it is a question of getting those that are more likely to be successful. Or when you have something that you know is problematic, having the confidence that whatever modifications you introduce would have a high likelihood of solving those issues.
Pharma IQ: Thank you very much Mr. Zurdon. That was very interesting. Thank you for this interview.
Disclaimer: All these opinions belong to Jesús Zurdo, and do not reflect necessarily Lonza’s views on the biotechnology industry, the drug development process or any potential challenges the industry might be facing. Lonza does not necessarily share or endorse Jesús Zurdo’s views on the company strategy focus or plans, both present and future.
Have Your Say
Rate this feature and give us your feedback in the comments section below