The digital revolution in this medical space is here to stay. There are a couple of things that are very critical for us to move at pace. One is data. Internally within AstraZeneca or externally within hospitals and health systems, how is data captured, and is there connectivity in the data space?
Once you can connect data and follow the patients longitudinally through their journey, you can make a better-informed decision through machine learning or AI or all of that. So, I believe that we need to get the fundamentals right first, which is about data. And once we have data, I think machine learning and artificial intelligence can be applied to better understand and predict which patient may be responding to a treatment, which patient may not be responding, and how we get early markers in that space.
The other thing that I think is critically important, specifically in the lung space, is that patients get diagnosed earlier. When you look at US screening rates naturally, it’s less than 5%. So how can we use data, artificial intelligence, and other things to get patients screened earlier and detected earlier, so they can benefit from those treatments?
See our whole interview with Arun Krishna from ASCO 2023 here.
With the technological advances that are taking place in the world in general, it’s appropriate that they should be applied to medicine as well. And some of the ways these are being explored currently in medicine are typically in the realm of aiding either pathologic review or radiologic review of either the pathology specimen or the imaging.
They can be used to help ensure an objective read or at least provide a preliminary read, and then humans can come in and apply their expertise to make sure this preliminary read is working. In other instances, it may be being explored to help with diagnostic considerations or differential diagnoses. It’s conceivable that, in five to ten years, this sort of technology may be used regularly or routinely within clinical practice in those capacities.
See our whole interview with Gregory Lubiniecki from ASCO 2023 here.
I think this is the next step for everyone in drug development to move on because you can progress and move so much faster. As proof of that, Eisai is opening a new chapter in our storybook called deep human biology learning. Essentially, what we are doing now is emphasising having a deep understanding of the science and then translating that understanding into new ways to discover and develop drugs using AI, cutting-edge platforms, and technologies that will make more efficient what we do every day.
See our whole interview with Corina Dutcus from ASCO 2023 here.
I’m really excited about artificial intelligence coming into the research space. Oncology is very complex, and as we advance in genomics, we want to better understand how genetics impact outcomes and help us tailor which therapy will be best for the patient to treat and hopefully cure cancer. The data is vast. It’s very complicated, and that’s why AI will be extremely useful to help us to analyse and build the next level of patient-focused therapies.
See our whole interview with Edmond Chan from ASCO 2023 here.
If you think about targeted therapies, the companion diagnostic piece is a huge issue. If you were to develop an algorithm, for instance, that could screen standard pathology slides to predict whether or not that patient is a carrier – that can help you pre-screen and identify patients much more quickly without having to do the test in every single patient – and that’s already a huge advancement.
See our whole interview with Martin Vogel from ASCO 2023 here.
From being able to detect a certain adverse event or disease, we can see AI tools capable of predicting before something happens. And as that window of prediction continues to increase – I’ll give you an example: CRS. Cytokine release syndrome is one of the critical negative effects of some of the very powerful drugs, CAR-T. These drugs are very helpful to the patient. At the same time, the risk of CRS is so high that many patients are either stuck in the hospital or just simply drop out. We already see now that we can predict with reasonable accuracy the risk of a grade three or higher CRS.
What this means is the physician can take action of saying, “Look, this patient’s risk is relatively so low, I’m not going to burden this patient by making them stay in the hospital. Let them go home, and as long as they’re within this two-to-three-hour range, I can bring the patient back if they do end up getting the grade 3 CRS”.
Imagine this technology becomes even more predictive, and you can pinpoint which patients are likely to get it even before they’re infused with the drug. And all the rest of the patients now have the freedom to be in a hotel nearby. So, I think as the predictive abilities improve, our ability to personalise treatment and personalise approaches is going to continue to grow.
Takeda has invested a lot in digital because customers want digital content, and they want to interact digitally. We try to partner with companies that utilise digital means to communicate scientifically. We’ve also done a lot with AI in predicting things like responses to certain medications, and that’s all in an effort to get the right drug to the right patient at the right time, and I do see this continuing to play a role in pharmaceuticals and in medicine. The hope is you can better predict how a patient will respond to a certain drug, thus limiting unnecessary toxicity and hopefully for better outcomes.
See our whole interview with Jennifer Elliott from ASCO 2023 here
I would say from a technology point of view, you’ve got AI, you’ve got digital, etc., but what that is resulting in is faster development in areas we already know are very promising. And I’ll give you an example. If I had to list the three hottest areas in oncology right now that are benefitting from some of this technology, they are: the progress with antibody-drug conjugates, immuno-oncology, and targeted protein degradation. I think that’s great for patients. I think you’re seeing those advances move very quickly, and at Astellas, those are all areas we’re participating in and are making significant investments in.
From an artificial intelligence perspective, within our research and development group, we’re actually leveraging that now to see which patient types are going to be more receptive to this therapy and which patient groups we should look at and enrol in clinical trials, which is actually going to help us with guidelines so we can get these therapies to the most appropriate patients moving forward. And then, from a digital technology standpoint, we know there’s a huge unmet need right now.
Everyone here knows and understands about cell therapy, but a lot of these patients are still working and operating in the community, and the big challenge we have, and how we’re leveraging digital technology, is: how do we get to these patients? How do we get to the community physicians to educate them about cell therapy so they can get referred early enough, potentially receive these therapies, and get a cure?
See our whole interview with Warner Biddle from ASCO 2023 here
So many ways. I don’t know where to even start. It starts with prevention and screening, right? There’s a lot of behaviour change that is associated with that space, whether it’s tobacco cessation or whether it is managing insomnia or stress. There are a lot of multifactorial conditions associated with cancer that we treat, and digital has a huge role to play in managing those conditions, but more importantly, to drive behaviour change.
We know behaviour change is also very hard. And this is where you also have to reduce not just the incidence, but also the mortality. And so, digital plays a role in early detection. It plays a role in precision medicine. You think about digital radiology, digital pathology, digital biomarkers, and next-gen sequencing; These are all very much digital enablers for treatment. So, it plays a role in prevention and screening. It plays a role in early detection. It plays a role in accurate diagnosis, treatment, symptom management, and survivorship. You tell me where it doesn’t play a role.
If you think about data interoperability, you think about infrastructure building. For example, a patient that has to go through multiple phlebotomists, radiologists, care navigators, care centres, infusion clinics, clinicians, diagnostic labs, and pharmacists.
There are so many intervening conveners and individuals that are collecting this data. What are we going to do with all this information that is out there? How can we streamline that? There’s so much efficiency that AI and, essentially, digital tools can bring that can streamline care to reduce the point solutions and to reduce the gaps that still exist in the whole supply chain of healthcare.