Lundbeck's Jorrit Hornberg on the Latest Predictive Toxicology Developments

Jorrit Hornberg, Head of Section of In-vitro Studies in the Department of Exploratory Toxicology at Lundbeck, speaks to Andrea Charles from Pharma IQ, about the latest devlopments in predictive toxicology, ahead of the Predictive Toxicology 2013 Summit.  

Pharma IQ: In your opinion what is the biggest development in predictive toxicology?

J Hornberg: There surely have been several important developments. From a strategic point of view for instance we’ve seen that industry has implemented predictive toxicology into the drug discovery process. I think this is mainly in response to the large number of projects that they recall during clinical development for safety of toxicology reasons. So in addition to risk assessment we’re now also very much focused on early hazard identification and risk avoidance, and then also from a technological point of view we’ve seen that several new methods have been adopted by the field – some perhaps making more impact than others. Think, for instance, about high throughput screening approaches on safety end points including high content screening, the omics technologies – such as toxicogenomics or proteomics – which we now use for biomarker identification and also to generate a mechanistic understanding of a toxic effect that we observe and I think those technological advances have already certainly had some impact on the field.

Pharma IQ: What is the biggest challenge with integrating predictive toxicology into current strategies?

J Hornberg: I think what we’ve realised at Lundbeck – and others as well of course – is that just moving some toxicology studies to an earlier phase in the R&D value chain in a so-called kill-early strategy, that is perhaps not enough. I mean, of course it will prevent a drug candidate that has a certain safety risk from entering the development pipeline, but it doesn’t automatically secure that you actually will identify drug candidates without that safety risk and that is ultimately what you need to do in order to bring new drugs to the market.  So in order to produce those so-called lower risk candidates toxicology has to be fully integrated as a discipline into the drug discovery process and that means that we have to be there where the discussions take place on novel drug targets, where compounds are identified and selected and also optimised and to make that work – toxicologists – they need to be part of core project teams in early discovery already and also the endpoints that they bring, sort of, the safety endpoints, they need to be accepted as important parameters to optimise on.  So on the one hand you need to speak the drug discovery language well, but of course at the same time you also need to have a thorough understanding of non-clinical safety as a discipline, including regulatory aspects, and also the clinical research world which speaks a completely different language – development focused language – and these things together I think are not trivial and certainly have been a challenge that took us a few years to overcome.

Pharma IQ: What are your top tips for cost effective integration of predictive toxicology? 

J Hornberg: I think it’s very important to make sure that toxicology is accepted as an important discipline by the drug discovery organisation, so on the one hand you need to secure high level management buy-in on your strategy, but also in the daily work you need to be viewed as helping a project to avoid risk and also to make sure that it will survive in the long run and not just as the toxicologist  that comes in to kill a fantastic idea or a nice compound.  So not just make things fail early, but also help find out what might not fail and I think that’s the main thing to realise and you can make that cost-effective by, for instance, being focused on what is need-to-know at what stage and what’s not need-to-know.  So you don’t have to frontload everything; you just need to make sure you have the right aspects at the right time and then use tiered screening approaches.  So in that way you can ensure that you catch the things early that have high impact and that occur frequently and then optimise on those things – and then in the same time also accept that you know some things you will just not catch until you do for instance your GLP toxicology.  And I think also basically in general if you resource predictive toxicology well enough then it will avoid late stage failures due to safety concerns and thereby implicitly I would say it’s cost-effective.

Pharma IQ: Do you think that in-vitro and in-silico approaches will ever completely replace in-vivo methods? 

J Hornberg: In-vitro approaches already have replaced a lot of in-vivo methods and also by using in-vitro methods we make sure we don’t need to use in-vivo methods as often maybe as we used to, but still we also know that some adverse events only occur in a systemic setting, so only in the whole organism, because that requires the interplay between different parts of the body.  If you think of blood pressure, for instance, you wouldn't be able to screen for that in-vitro. I think you simply cannot test everything in-vitro – but when you can you should of course – and then with respect to in-silico methods we should perhaps have high hopes for those to make more impact on the future than they maybe do today when it comes to replacing in-vivo real experiments, but I think to fully replace them it would require that basically all our biological knowledge – so about health and disease and from molecules to organisms to populations – is integrated in-silico and that is a big task for the systems biology field and it’s going to take time – if we get there at all.  So I think, no, not in the foreseeable future at least they won't replace in-vivo methods completely, but to an extent.

Pharma IQ: Looking to the future where do you see predictive toxicology developing over the next ten years?

J Hornberg: I think ten years is a long time so I would hope by then that we would see the results of current efforts, so by that I mean that drug development projects should less often fail for safety reasons than they do today and that would make drug development more effective and ultimately would enable us to bring more novel drugs to the patients.  Maybe more specifically I think an interesting development that is on the horizon is personalised medicine or patient segmentation – not just from a pharmacological or efficacy point of view, but also from a toxicology perspective – so basically identifying those patients that will or will not be susceptible to a certain adverse drug reaction and some genetic markers have already been identified – for instance in the immune system or in metabolic liver enzymes – which then can explain why some individuals are sensitive for a certain adverse drug reaction and others are not, so we can predict who to treat and who not to treat.  We might be able to avoid certain toxicities and I think that will be great because then treatments which  would perhaps otherwise  have failed due to adverse events can actually be brought to the market. In development we are seeing more and more of this collaboration and also knowledge sharing between pharma companies, especially in the field of toxicology. I think for instance, in consortia and also with academic institutions there is a lot of collaboration, and I think the field of toxicology is already benefiting from that. If  you think about technology evaluations, biomarker identification and things like that, and in the next ten years I think it will surely benefit from that a lot more.

Pharma IQ: I understand that you’re going to be speaking at the Predictive Toxicology 2013 Summit taking place the 18th to the 20th of February 2013, for anyone who’s interested in attending, what do you plan to cover in your presentation?

J Hornberg: I will speak about our exploratory toxicology strategy at Lundbeck, so how we here at Lundbeck have integrated toxicology into the drug discovery process. I will discuss the way of working a little bit, what to do, what to screen for in various phases of the drug discovery process, why screen for which endpoints and how, and probably also give some practical examples.

Have Your Say
Rate this feature and give us your feedback in the comments section below


Please note that we do all we can to ensure accuracy within the translation to word of audio interviews but that errors may still understandably occur in some cases. If you believe that a serious inaccuracy has been made within the text, please email