Clinical Development in a Post-pandemic world | Raj Modi | DSC Europe 2022

Yeah, thank you. Thanks so much for inviting me. I'm really delighted to be with you here this morning. As Bob says, my name is Raj Modi. I work within Oracle's Health Science organization. My personal background is 15 years in management consultancy, and my presentation will be all about the clinical R&D process. What I'd like to do is share with you examples of how digital platforms and data science is helping research organizations such as biopharma, CROs, other research organizations bring medicines to patients faster. So just to first of all introduce Oracle: Oracle is a cloud technology company. We build technology for clinical trials and for safety, and we also build cloud capabilities in analytics, in data science, in high performance computing, data storage, and many other areas as well. So to summarize, our expertise in this industry is really around data and digital for clinical R&D.

Our focus in this industry is on two areas. The first is that we've built the next generation of platform to simplify clinical trial systems and reduce cycle times, and that platform can handle the scale and the velocity of the data that we foresee coming over the coming years. And then the second area of our focus is on our merger with Cerner. Cerner is an electronic health record system, has a very strong footprint around the world, and Cerner is now part of Oracle. And our vision is to integrate electronic health records into our clinical platform, link EHRs to wearables and sensors, and enable EHRs to improve greater diversity into clinical trials.

A little bit about clinical trials as we enter this post-COVID-19 world: we're seeing that clinical trials are evolving incredibly quickly. They need to be faster, they need to support new models like decentralized clinical trials and other models, they need to handle much more data, and they need to accommodate new treatment models like, for example, cell and gene therapy. And we believe that research organizations will need to achieve a new level of agility and responsiveness to be able to handle these new pressures.

In terms of where we see the opportunity, on this diagram I've illustrated a few use cases of innovation across the drug discovery life cycle, from scientific research to pre-clinical development to clinical development, and then supporting post-launch with engagement with patients and clinicians, and more broadly with healthcare management. And what I'd like to do is just show you a few examples that bring these use cases to life.

So the first example I'd like to start with is regarding scientific research. Oracle has partnered with the University of Southern California and the Ellison Institute to support cancer research. A recent breakthrough has been to train predictive algorithms for breast cancer diagnosis. Many of you will be aware that if you really want to get value from deep learning in a field like this, it requires vast amounts of well-annotated data sets. In fact, we often need millions of annotated images to learn statistically significant relationships. However, partnering with the Institute here, we've gotten around this challenge by first pre-configuring the deep learning network to recognize or fingerprint individual tumors, and this can then leverage large unannotated data sets which are widely available. And this has supported groundbreaking research into providing accurate predictions of clinical subtypes of breast cancer.

The next example, moving from scientific research to pre-clinical development. This example talks about addressing the challenge of the high failure rate in developing new medications, which is around 90%. Typically, when we develop a drug, drug compounds are tested on animals before they go into human trials. But often, the animal testing methods don't always replicate how the drugs actually affect human organs in real life, and they can have unexpected toxic effects on human tissues. So the Wake Forest School of Medicine are creating real organoid models of the human body on chips. So, for example, they're growing real cells of lungs, of the heart, of the kidneys, of the liver on plastic substrates, and they're testing the active drug compound to measure the toxicology and get early insights into the safety and efficacy of the candidate drugs. And this doesn't require any animal testing. And as they do these tests, they're generating huge amounts of data: they're generating assay data as well as other scientific analysis data. And that shows how and why certain compounds metabolize differently in certain parts of the body. So Oracle's role in this is to provide the data and the data science platform, the expertise to understand the data points, and to start analyzing and understanding these experiments.

So moving on from pre-clinical to clinical development, so this is running trials in humans. This is a case study about recruitment and engagement of patients at speed and scale. The U.S. government, as part of the COVID-19 program, wanted to recruit a million people into COVID-19 vaccine trials as part of Operation Warp Speed. This was around the March 2020 period. We collaborated with the NIH, the CDC, and the FDA to build a recruitment pre-screening and waiting room platform, and the platform supported the recruitment of 600,000 patients in the first six weeks. In fact, it's the largest and fastest clinical trial recruitment drive in history. And we were able to build and implement that platform in 30 days. We're now using the same approach for other clinical trials, and we're very proud to be supporting clinical trials in HIV vaccines around the world. 40 million people have HIV, and we're using the same approach to recruit 7,000 subjects into HIV vaccine trials. Now, what's been remarkable here is that we've been able to quadruple the number of patients, or number of people volunteering for these clinical trials for HIV vaccines versus previous HIV vaccine trials. Now, this is really important because recruitment and retention of patients is a really big challenge. In fact, independent data shows that for every 100 people that usually show eligibility and interest into a trial, of those 100, only seven people make it to the end of the trial. And enrollment can cost as much as 40% of the overall cost of a clinical trial.

This is another example, more focused on how do we reduce the operational speed of running clinical trials. Keelan Pharmaceuticals were seeing a lot of complexity in the way that they set up their trials. They were, when they were setting up a clinical trial, they were running lots of different systems. And they were having to configure those systems. So these are systems with regard to the clinical study systems, with regard to data management systems, with regard to setting up the analytics, and systems with regard to dealing with supplies. In fact, there could be as many as 20 plus different systems and 40 plus different integrations, and this leads to tremendous complexity. In addition, setting up oncology studies has its own additional level of complexities. For example, there are systems of adaptive trials and master protocols and umbrella trials, so there's lots of different approaches which add more complexity. So we worked with Keelan and we supported them with an interoperable platform which provided them with all the capabilities to set up a clinical trial quickly and efficiently, from data collection to data submission. So they're not designing a clinical study with 20 different systems, they just design that clinical study once. And the outcome is that they're able to set up a Phase 3 oncology study in days rather than months. In fact, there was recent independent benchmarking comparing the speed at which it takes to set up oncology studies, and they compared our new agile microservices-based platform with legacy systems. And what they found was that setting up an oncology study is 58% faster for data collection, 91% faster for randomization and trial supply management, 65% faster combining data collection and randomization and trial supply management, and 94% faster overall for study deployment. And the underpinning reason for this is the extent to which we're able to automate the study setup and simplify the configuration in our platform. Another example of how automation can really transform the setup of clinical study: working with Transcellarate on their digital data flow program, we're able to take a standards database and use that to automate a study build in minutes. And that's the entire oncology library, so all the forms, all the workflows, all the visit structures, all that contextual detail typically that takes weeks to set up in an EDC system, and we're able to set them up in minutes. Moreover, once the study is created in an EDC system, it usually takes many more weeks to set up the same study in an RTSM system. And this is where our platform can do two jobs at the same time: when the study is built for data collection, we can concurrently in parallel build the same study in RTSM, and that dramatically reduces the time taken. So the point I'd like to make here is that you can reduce the time it takes to set up clinical studies by weeks through using automation.

The last example I'd like to share with you is an example regarding public health surveillance using real-time genomic sequencing across 20 international markets. We think that we might be coming to the end of this COVID-19 pandemic, but the truth is we don't know for sure. And there might be a more severe COVID-19 variant, or there might be another pathogen around the corner. In partnership with Oxford University, we've built a global system to rapidly identify and track new COVID-19 variants. And what we're doing is the platform enables the sequencing of the virus and enables the processing of it, and then informing governments, institutions, and public bodies and health systems what they need to know about that variant, that pathogen, across a number of different reference sites around the world. All of this information is captured in real time and it alerts national governments and public health researchers around the world. The genomic data that is created is very granular, it requires enormous computational power to assemble, analyze, and make sense of it. And each site here doesn't need to have its own data center, it's a very simple system to operate with visualization tools where scientists can look at their own data and compare that across the world. So it's one system that runs across 20 countries, which supports safe and secure data sharing across those markets.

So I've shared a few examples here, and let me just share a little bit more about the underlying capabilities that are enabling these use cases. So we have an extensive suite of digital capabilities that span the entire lifecycle of drug development. We have capabilities in clinical operations, capabilities in clinical data, we have capabilities in safety which include signal detection and processing for adverse events, and we have significant capabilities in data science. And then more broadly with healthcare, for example, genomic data management. Now all of these capabilities are integrated into one cloud platform. The notable word here is 'open'. Open means that these capabilities can be adopted incrementally to fit into the existing clinical IT estate. They will fit in like Lego blocks with those existing clinical systems. And what it means also is that new innovations can be developed on the fly to support new processes, because the whole platform is extensible. So to conclude, the common denominator to all these case studies is data. It's data flow and the analytics around that data. And that is what we as an organization are reimagining. Our mission is to help people see data in new ways, discover insights, and unlock endless possibilities. And we think this is really the method, the approach to supporting medicines reaching patients faster. So I'd like to really thank you for your time today this morning, and very happy to take questions.

Yeah, thank you for that question. That's a great question. We've just, as I said, we've just acquired Cerner, which is the electronic health record company. We also have an extensive suite of capabilities that span the broader health ecosystem. So I've talked about our clinical development platforms, we also have systems with providers, we have systems with claims management and payers, we have national health record systems. And what we're doing is we're now interconnecting all of these systems together. Because I think what we've seen with the pandemic is we've seen that data does not flow efficiently across the whole health ecosystem. And that is preventing care being delivered in a cost-effective way, and also facilitating new breakthrough research around therapies and treatment interventions. So our mission is really to be able to try and join up the data across the e-health ecosystem. Data is our heritage as an organization, so we think we're very well suited to be able to support that.