What led you to specialise in computational tissue engineering?<\/h3>\n
Initially, I was working in mechanical engineering because there was no biomedical engineering course available at the time; it was subsequently introduced as a Masters programme in Belgium after I graduated. I always wanted to work in biomedical applications so, during my PhD in biomechanics, I started focusing on computational modelling, which is something that I picked up from mechanical engineering. At first, I worked on combining computational modelling and fracture healing, and then after my postdoc, it expanded towards tissue engineering. The research group that I was with had a strong interest in bone and cartilage, so I was able to combine that knowledge with my numerical expertise.<\/p>\n
What are the benefits of the computational modelling approach in the fields of tissue engineering and regenerative medicine when developing new therapies and materials?<\/h3>\n
Computational modelling gives you an additional tool that you can use to increase your understanding, improve your preparation, and enhance your insights in the results. Modelling is not linked to any particular application, or any particular phase in terms of the R&D cycle, so you can use it for basic research during the planning of preclinical, and clinical phases up until after post-market authorisation.<\/p>\n
My team and I focus mostly on the early phases, the R&D of the development of new constructs and new materials. We have people who are focusing more on bioinformatics and data-driven applications that use single-cell RNA sequencing to create a skeletal cell atlas, for instance, and that will help create a blueprint of the biology that we want to recreate in tissue engineering.<\/p>\n
We then have people that do systems biology modelling of gene and protein networks to understand how we should treat our cells to make them exhibit the desired behaviour.<\/p>\n
We can use models to design new biomaterials; we have an application where we collaborated with dentists and people that are working on 3D printing of calcium phosphate-based biomaterials. We built a mathematical model of 3D cell growth inside porous scaffolds, starting from the observation that cells prefer to grow in corners. Then we used that model to optimise the shape of a scaffold that would optimally stimulate growth. Large animal studies have been concluded with the scaffold that we optimised, and show the new design outperforms the controls.<\/p>\n
We also use modelling to better control bioreactors, for instance, in our in-house developed perfusion bioreactor setting. While experimentally we are limited to sensor read-outs from the inlet or outlet to monitor what goes on inside the bioreactor in a homogenised manner, with the model, we can actually visualise what happens on the inside, and the microenvironment that our cells are experiencing. We can really quantify that and use it in quality control.<\/p>\n
The last application domain is the simulation of in vivo<\/em> processes such as bone regeneration and cartilage degradation. We start by gathering mechanistic information or hypotheses from our experimental collaborators and subsequently translate this into mathematics. If we need to design a living implant for a knee joint for example, we can simulate what happens in the knee joint and know what the mechanical and chemical environment would be in which our implant is put. By doing so, we can optimise the constructs that we are designing.<\/p>\n Computational modelling is an important tool in the realisation of the three Rs – refine, replace, and reduce – but I would not say it is going to completely replace in vitro<\/em> or in vivo<\/em> tests. That is not our focus. Our focus is to increase our insights and reduce the trial-and-error approach that is currently still taking place.<\/p>\n Any! Tissue engineering and regenerative medicine is an application area that, compared to other types of medical therapies such as drugs and medical devices, is trailing behind due to its complexity. Medical devices are the frontrunners in the use of computer modelling and simulation. A couple of years ago, a large electronics company adapted their pacemaker so that the lead wire between battery and pacemaker would not overheat when the patient was undergoing an MRI scan; up until now, most pacemakers are unsuitable for MRIs for this very reason. They took the updated design to the FDA with a very limited amount of in vivo<\/em> animal evidence, everything else was \u2013 well validated \u2013 in silico<\/em> evidence. They simulated the device and virtual patients with two million simulated MRI sequences \u2013 a lot more than what you could do in a normal clinical trial \u2013 and their pacemaker adaptation was approved.<\/p>\n There are a lot of clinical decision support systems that are already using modelling; of course, it is always the clinician who has the final say, but it is providing additional input. With insulin pumps, for instance, you can predict when insulin should be injected based on the glucose levels. Many other medical devices have been designed using models or are using models to enhance their functionality.<\/p>\n For drugs, we have the pharmacodynamics-pharmacokinetics models that look at dosing mostly. There is a lot of push now to also accept other types of in silico technologies. There is a little more work to be done because it is harder to simulate the biology of pharmaceutical intervention than it is to simulate the physics of medical devices. Anything composed of biological tissues, or concerns the behaviour of biological processes, is a bit trickier to simulate. As a community, we are talking to different stakeholders, policymakers, and regulators to understand how we can show that the digital evidence generated by computer modelling and simulation is credible.<\/p>\n When presenting the assessment of credibility, we have to ensure the implementation is correct (verification) and that the model represents reality (validation). We must also quantify the uncertainty \u2013 e.g. in the value of model parameters \u2013 and how that affects the behaviour of the overall model. There is now a recognised standard for medical devices from the American Society of Mechanical Engineering<\/a> (ASME, V&V40), which comprises an official document on the degree of verification and validation needed for the credibility assessment of your digital evidence.<\/p>\n Compared to ten years ago, the field has matured tremendously. With all of the key stakeholders aligned, we are making a big push now for the transition towards industry and the clinic.<\/p>\n When it comes to personalised medicine, it is relatively simple in the sense that you can personalise the design of an implant, starting with medical images and then ensuring your medical device has the right shape. You could go one step further with patient stratification if you know the behaviour of certain tissues or cells and how, for instance, they change with age or sex. The real pinnacle of personalisation would be to have measurable biomarkers that are also parameters in the model and that model would be based entirely on a patient\u2019s own data. We are most likely to start with a generic avatar of a human that gradually gets fed information (the patient\u2019s height, weight, age etc.) and then for specific medical questions, let\u2019s say for the heart, you can attain medical images of the heart and then make a personalised geometry, or have an ECG readout that you could also use.<\/p>\n There is a lot of work being done on omics technologies where we could identify biomarkers that can be correlated to specific processes. If, for example, you have a biomarker for inflammation that is higher in your system and you want to simulate the progress of your osteoarthritis, you could include that biomarker in the model and make some changes based on that biomarker, then you will have a much more accurate prediction of the progress of the osteoarthritis.<\/p>\n Another very active field is wearables and how gathered information could be used to essentially feed your digital twin. The challenge is amalgamating the models, data, infrastructure, and high-performance computing and making this available to the end user.<\/p>\n Tissue engineering applications were the first advanced therapeutic medicinal products (ATMP) on the market, but also the first to leave the market, unfortunately. The manufacturing challenges are significant, which makes it very costly. There needs to be much more of a process engineering mindset, not just thinking about putting living implants together, but also how to do that on a larger scale for the patient following a quality-by-design approach. Quality analytics become very important, and models are frequently used, for instance, in the biopharma industry as part of that quality pipeline.<\/p>\n The manufacturing aspect is being developed yet, despite the increased attention it is being given, it still has to gain more traction. For researchers focusing on the basic biology, it is really hard to think ahead to the manufacturing stage.<\/p>\n Applications of the computational tools that we use in tissue engineering that are set for the clinic include the aforementioned 3D printed scaffold. We have finished the large animal experiments, and once all the data becomes available, we will see whether it is ready for human trials. In this case, it is not the model that is destined for use in the clinic, but the model was an essential step in creating that new design and the product that goes towards the clinic. Computer modelling and simulation, in a lot of cases, is used early on and followed by the classical chain of in vivo<\/em> and preclinical evidence that the regulator is used to seeing. There are some success stories in tissue engineering, where a model has been important in arriving at a certain mechanistic conclusion or for designing a product. The commercially available bioreactors have all had a quantification of their environment in terms of shear stresses, oxygen distribution, and fluid flow by computer models, so it is not because you do not see them in the end product that they have not been there.<\/p>\n When it comes to computer modelling and simulation, there is also a maturation period that a field needs to go through. The acknowledgement and validation of in silico<\/em> technologies in tissue engineering was very different when I started in the field 15 years ago. But now there is an active group of people who seek out teams like ours that are doing modelling, to try and understand what it is about, or because they are already convinced that there is an added value in it. Tissue engineering is a strongly interdisciplinary field and can never be done by any one discipline in isolation. It needs to be an interactive endeavour in order for it to be successful, and computer modelling and simulation is adding an additional tool to this mix to further increase the chances of success.<\/p>\n Professor Liesbet Geris<\/strong>Which areas of healthcare could benefit from this technology?<\/h3>\n
![\"\"](\"https:\/\/www.innovationnewsnetwork.com\/wp-content\/uploads\/2022\/09\/Image-2-1-300x169.jpg\")
\u00a9 iStock\/ChooChin<\/figcaption><\/figure>\nHow could computational models benefit the field of personalised medicine?<\/h3>\n
How do tissue engineering and computational tissue engineering fit within current medical practices? What are the key challenges in terms of translating research and data into the clinical arena?
\n<\/strong><\/h3>\n
\nResearch Professor in Biomechanics and Computational Tissue Engineering<\/strong>
\nUniversity of Li\u00e8ge and KU Leuven in Belgium<\/strong>
\nwww.biomech.ulg.ac.be\/team\/liesbet-geris\/<\/strong><\/a>
\nhttps:\/\/www.linkedin.com\/in\/liesbet-geris-7b086563\/<\/strong><\/a>
\nhttps:\/\/twitter.com\/liesbetgeris?lang=en<\/strong><\/a><\/p>\n