Sorry, you need to enable JavaScript to visit this website.
AI in Pharma

Understanding cardiovascular diseases like never before

patient segmentation

Some years ago, a close friend had a stroke. He was in his early 60s and I saw firsthand how debilitating it was to both his professional and private life. Unfortunately, I am just one of many who has seen someone dear to them affected by cardiovascular disease. It is indiscriminate in who it targets, and the World Health Organization reported close to 18 million people died of cardiovascular diseases in 2016, 85 percent due to heart attack and stroke. 

With my pharmacist hat on, I found myself thinking about how we could treat cardiovascular diseases in the future: What if we could identify patients at risk of a stroke before it even happened? What if we could find the most effective treatment for each heart failure patient? With the ambition of making these “what ifs” a reality, and I joined Bayer 15 years ago to find medicines to treat and prevent cardiovascular disease. 

 

Over the years I have witnessed a change in the industry landscape. This includes societal megatrends such as population aging, bringing with it a higher prevalence of cardiovascular disease which places greater pressure on healthcare systems. Especially with chronic disease, it is becoming more and more difficult to gain the insights we need with classical research methods. 

 

Clinical trials in cardiovascular disease are enormous, the cost implication is equally vast, and the probability of success does not get any higher. My role at Bayer began to take a new turn, and together with my team, I began to explore artificial intelligence (AI) and how it could facilitate and improve the R&D process. It is clear to me that as an industry, we are behind the curve with such technologies – we need to truly harness machine learning and advanced computing infrastructures, as they can provide solutions which transform how we find new therapies for patients. 

 

 

Putting patients and their diseases under the AI microscope

 

One example is finding the right treatment, for the right patient. We currently divide patients into groups based on their characteristics and how they respond to certain treatments – it is a practice long-used in research and development to optimize the development of medicines. However, applying artificial intelligence and machine learning to this process of patient segmentation provides an even deeper characterization of patients. 

 

We can derive novel insights from the vast amount of data generated during the delivery of every day healthcare and drill down to truly understand diseases that affect patients. We could predict which patient is predisposed to cardiovascular diseases and how and why some treatments could be more beneficial to treating them than others. This is a huge opportunity to move towards personalized treatment and the prevention of cardiovascular diseases. 

 

This represents a huge step in how we treat and prevent cardiovascular diseases. AI empowers us to make better informed decisions. It opens new avenues outside of current R&D processes, which no longer serve the growing demands of cardiovascular patients, and transports us to a new R&D realm with in-depth disease characterization that supports research and more accurate patient segmentation.

 

With this level of granularity, we could solve the issues that currently render drug development for cardiovascular diseases so challenging, namely, better design of clinical trials. This ultimately decreases R&D risk and reduces costs with smaller and shorter trials. For patients, this holds the potential for faster and more efficient development of targeted treatments that could change their lives.

 

With artificial intelligence, I can see the “what ifs” morphing into real possibilities. And in turn, I am comforted by the vision of a future where fewer patients have to suffer the way my friend did.


 

Author
Andrea Derix
Andrea Derix
R&D Global Program Management, Thrombosis Portfolio, Bayer Pharmaceuticals