Writers: Konstantin Tumanov, Ignacio Louzan
You probably remember how until recently, medical data management was a manual chore, with each patient's details recorded on paper. But the digital revolution has transformed this process. However, now medical data is not just digitized; it is also centralized in the cloud—something resembling an immense, invisible filing cabinet.
The magic unfolds when AI intervenes in this vast pool of centralized data. With lightning speed, AI identifies patterns and connections that would take humans much longer to uncover. As Mandar Paralkar, Global VP and Head of Life Sciences, SAP, told us, this makes AI 'increasingly used in drug discovery, quality control, and regulatory submissions, enhancing outcomes by managing both structured and unstructured data'. According to McKinsey & Company, AI application boosts the efficiency of initial manual assessments of drug targets by over 30%. Furthermore, AI-enhanced compound screening in silico has not only tripled the performance of chemical compound activity models but also slashed the time needed to pinpoint new leads by more than fourfold—which means no less than drugs being discovered at four times the pace. Insmed, an innovator in the field of rare disease therapies, is one of the examples of a biotech starting to leverage these tools.
As CEO Will Lewis told us: 'The company's collaboration with Google Cloud aims to integrate AI in drug discovery, development, commercialization, and internal operations to enhance efficiency and innovation.' In his view, 'AI's potential to accelerate understanding and optimize processes without replacing the human element signifies a pivotal moment in the industry.' Similarly, the CEO of Labcorp, one of the largest CROs in the world, told us that AI plays a transformative role in both customer-facing and operational aspects of the company's work. 'For example, we employ neural networking to enhance patient check-in processes at our service centers, improving accuracy and efficiency in patient identification and insurance verification. This not only speeds up the process but also allows our staff to focus more on patient care,' Adam Schechter highlights. However, AI’s power comes with a corresponding responsibility; Dr. Diem Nguyen, CEO of global health company SIGA, warned us that 'The dual nature of AI and biotechnology present significant benefits and risks. While these technologies have led to many advancements in virus sequencing and treatment development, they can also pose a threat if used malevolently.'
AI & drug discovery
A biotech specialized in the treatment of cancer, Genmab, has begun to utilize AI across several of their operations, as revealed by their creation of an AI incubator to foster innovation, generating over 140 concepts in various areas. While currently focused on simpler tasks, Chris Cozic, Executive Vice president & chief People Officer, told us that the future application of AI in their operations is poised for grander challenges. Specifically, Genmab believes that AI will play a 'significant role in predictive analysis and protein design,' which involves using AI to firstly understand and secondly manipulate protein structures. Proteins play a central role in drugs as they often serve as targets that the drugs bind to, modifying biological processes to treat diseases. Consequently, this means that with AI we can predict how proteins interact with other molecules—which enables researchers to virtually design proteins with specific therapeutic properties. This capability has the potential to be a game-changing advance compared to previous drug discovery methods, which relied heavily on trial and error and extensive laboratory experiments, making the process slower and less efficient.
AstraZeneca's CEO, Pascal Soriot, further illustrated the rapid application of AI into their drug discovery efforts: 'Our AI-enabled platforms are using generative models to identify potential drug molecules twice as fast as traditional processes and help us prioritize those that will be most effective. We are also using generative AI and machine learning in antibody discovery, cutting the time to identify target antibody leads from three months to just three days.' Service providers, like PerkinElmer, have also been quick to integrate AI in their offerings. 'Our OneSource services offering exemplifies how AI can optimize operations within biopharma R&D labs. By relieving scientists from the burdens of instrument complexity and maintenance, we enable them to focus on their primary objective: drug development. AI aids in predicting lab utilization rates, maintenance requirements, energy consumption, and other operational aspects, thereby enhancing efficiency and productivity,' Dirk Bontridder, PerkinElmer's CEO said.
The solution to equity in clinical trials?
To address the limitations inherent in the traditional management of clinical trials, decentralized trials and AI promise an opportunity for transformation, as exhibited by our conversations with innovators in the field. Seema Verma, EVP and GM at Oracle Health and Life Sciences, told us that the current state of trial management is 'remarkably manual,' characterized by 'extensive paper documentation and labor-intensive data entry processes.' This approach heightens the risk of errors and hinders efficiency of trials, with a considerable amount of time spent on data handling. Julie Ross, president and CEO at Advanced Clinical, asserts that 'the next significant challenges in clinical trials revolve around leveraging AI and business process automation. As we move towards decentralized trials, the focus is on making trials more accessible to patients in diverse locations.' Decentralization in clinical trials not only enhances the efficiency and scalability of data management but also fosters greater diversity among participants, as it breaks down geographical and logistical barriers, enabling a more representative cross-section of the population to contribute to and benefit from research. This is not a minor point. Looking at the diversity statistics of clinical trials: 'in all the medications approved by the FDA in 2020, the clinical trial participants were: 75 percent White, 11 percent Hispanic, 8 percent Black African American, 6 percent Asian. Only 30 percent of the participants were over 65 years of age and more than half of the patients were located in the US' (Teva Pharma). Consequently, if clinical trials with integrated AI could reduce the number of participants needed in placebo arms (referring to a group of participants who receive a placebo instead of the active treatment) and utilize data-driven approaches to streamline trials, these percentages could become increasingly balanced.
This new format of decentralized clinical trials is being implemented by Medidata, a Dassault Systèmes brand and leading provider of clinical trial solutions to the life sciences industry. CEO, Anthony Costello, told us that 'patients now participate actively from their locations, contributing data through mobile phones and wearables, significantly changing the landscape of data collection and increasing the authenticity and applicability of clinical trial data in real-world settings.' Additionally, Costello explained that 'recently we have focused on harnessing AI to leverage data from the 33,000 trials we have conducted, which aids in optimizing trial designs and operational efficiency.' By leveraging data from trials, AI tools can identify the most suitable research sites, refine trial protocols, and extend the benefits of medical research across different demographics.
Verily, Alphabet’s health technology branch, is also dedicated to making clinical research easier for its biopharma clients. Verily Viewpoint, part of the company’s suite for enhancing clinical research, includes services aimed at optimizing clinical trials through technology. ‘This encompasses patient recruitment, consent processes, and clinical trial management. Viewpoint, also includes Workbench, which connects data providers and researchers on a collaborative health care data platform for analysis and insights and, exemplifies Verily's innovative approach to clinical research, enabling more efficient, data-driven studies,' Verily’s Chairman and CEO Stephen Gillett told us.
The bones of innovation
As we witness the exponential advances made through software, hardware advancement seen in robotics, glucose monitoring and other medical devices craft the bones of the life sciences. Globus Medical, originally focused on spine health, has expanded its influence into orthopedics and trauma care. With the advent of robotic technology, surgical procedures have undergone a transformation, 'revolutionizing the field by providing surgeons with high precision tools for tasks like pedicle screw placement, allowing them to focus more on patient care rather than the technical aspects of surgery.' Dan Scavilla, CEO of Globus Medical, told us. Similarly, Zimmer Biomet is making strides in the integration of robotics to the field of extremities care. The company's ROSA Shoulder system is the world's first robotic-assisted surgery system for shoulder replacement, and has recently received FDA clearance. Their hardware devices are not only advancing per se, but also by connecting to the cloud and leveraging the power of interconnectivity. As president & CEO, Ivan Tornos, told us, 'Virtually all products in our portfolio either collect data or feed data to the rest of the interconnected ecosystem of care. With partnerships with Apple and Microsoft.' If one sector can leverage on the potencies of others through interconnectivity, we will be looking at exponential progress and benefits.
A company that was founded with the very idea of applying AI to medical devices, and as early as 2006, is iRhythm Technologies. Its objective is to provide a more effective alternative to the traditional Holter monitor, which is a wired portable electrocardiogram (ECG) that records the electrical activity of the heart for up to 48 hours. IRhythm's device, called Zio, is a patch ECG monitoring device designed to be worn continuously for up to 14 days. 'We know that many cardiac arrhythmias are missed with the Holter monitor due to the limited time in which they are worn by patients. Our device outperforms the conventional Holter monitor in effectiveness and convenience, backed by substantial clinical data—helping physicians reach a definitive diagnosis and decrease time to treatment,' iRhythm's president & CEO, Quentin Blackford, told us. The data Blackford refers to is analyzed by iRhythm's FDA-cleared AI algorithm, compared against a database of various arrhythmia readings, validated by trained cardiac technicians and presented to physicians as an actionable report.
Another illustration of the integration of hardware and software comes from one of the leaders in thrombectomy devices, Penumbra. The company has developed tools which, with computer assistance, can remove blood clots much more seamlessly. As Penumbra's Co-Founder & CEO, Adam Elsesser, explained, the traditional, manual approach often involved significant blood loss as well as the risks coming from the intervention of a human hand. 'Lightning Flash and Lightning Bolt 7 represent significant innovations in thrombectomy by improving the efficiency and safety of blood clot removal. These computer-assisted vacuum thrombectomy technologies are built on the concept of mechanical clot removal, which is becoming preferable to drug-based methods due to the latter's risk of bleeding and the necessity for careful monitoring. Our approach, focusing on aspiration, aims to maximize safety, speed, and simplicity and avoid damage to the delicate structure of the blood vessels,' Elsesser said. Both Lightning Flash and Lightning Bolt 7 employ sophisticated algorithms to recognize the difference between blood clot and blood flow, making lives of both surgeons and patients easier.
Prytime Medical Devices is also bringing innovative tools to the aid of surgeons with the aim of minimizing blood loss. Its focus is severe internal bleeding that is often the result of car crashes or war-related traumas. The Texas-based company's CEO, David Spencer, shared: 'The challenge with internal bleeding, especially in the case of severe trauma, is that traditional exploratory surgery to locate and address the source of bleeding has a high mortality rate.' To address that, Prytime has developed a tool called REBOA which involves inserting a balloon catheter into the aorta through a small incision in the leg, effectively blocking blood flow to control hemorrhage and maintain vital blood supply to the heart, lungs, and brain. 'This approach significantly reduces the invasiveness of the procedure and increases the chances of survival by preventing further blood loss,' Spencer told us. The REBOA technology is notably used in Ukraine. 'The inability of the Ukrainian military to evacuate casualties quickly due to drone threats has made our technology crucial for providing soldiers with a chance to survive by 'turning down the faucet' on hemorrhage, thus buying time for medical intervention,' Spencer shared, noting the similar urgent necessity in civilian car crash cases.
Next Generation Sequencing (NGS), having revolutionized the study of the human genome, may be a familiar term by this point. But the exponential growth in sequencing data means that, even though costs have been brought down tremendously compared to a few decades ago, significant computational resources are still required for the use of genomics in diagnostics and drug discovery. We spoke with a Japanese start-up, Mitate Zepto Technica, which, through the use of semiconductor technology, claims to be able to make DNA analysis up to a hundred times cheaper. 'Traditionally, analyzing DNA sequencing data necessitated outsourcing or the use of supercomputers, a process that could take weeks and was prohibitively expensive,' Keisuke Harashima, Mitate Zepto Technica's CEO shared. Still in prototype phase, the company aims to launch its technology in early 2025. Simultaneously, proteomics, the study of proteins, has not received the same amount of attention as genomes. Proteins are coded by genes. There are 20,000 genes in the human genome, and these create over 1 million proteins that comprise the human proteome. Our understanding of the proteome is pivotal, since this could allow for a deeper understanding of biological processes, leading to insights into the mechanisms of diseases, such as Alzheimer's, whose progression can be traced via proteomic signatures. Omid Farokhzad, a pioneer in the field and a Founder of Seer, told us: 'Seer was founded to address the key challenges faced in the field of proteomics, with the aim to elevate our understanding and technological capabilities up to par with those of genomics.' He is certain that advancements in proteomics like the Proteograph are yet to reveal many previously unknown biological insights. Referring to the company's analytical device, the Seer Proteograph, Farokhzad added: 'Since the introduction of our product, the tools at the ready for scientists to analyze the proteome have significantly advanced. We have moved from being able to analyze a few hundred proteins in a handful of complex samples, such as plasma, to detecting up to 10,000 proteins in plasma and conducting studies with thousands of samples.' Seer has already announced important partnerships with companies like Thermo Fisher and SpaceX.
Enhancing diabetes management
According to the World Health Organization, approximately 422 million people suffer from diabetes, of which Americans represent 37.3 million. The prevalence of diabetes is rising steadily, disrupting the lives of those diagnosed by demanding both management and monitoring. Yet, this task has become much simpler through innovation in monitoring systems. A glucose monitor features a sensor that goes easily inserted beneath the skin, where it can measure glucose levels in the fluid around cells. The sensor then generates an electrical current based on the glucose concentration, and the monitor then translates this current into a readable glucose value. Before the development of glucose monitors, people with diabetes primarily relied on urine glucose tests to manage their condition. As Timothy T. Goodnow, president & CEO at Senseonics, explains, 'CGM (Continuos Glucose Monitoring) represents a leap forward in diabetes management, providing continuous glucose readings that offers the user a dynamic view of their glucose levels in real time. This technology moves beyond traditional finger-stick tests, offering glucose data every five minutes'. Goodnow further highlighted the future direction, stating that 'innovations will aim to further reduce the management burden on people with diabetes by eliminating external components and integrating directly with personal devices for seamless glucose monitoring.' Similarly, the president of Ascensia, Rob Schumm, emphasized the role of software integration, 'The company's cloud-based solutions and partnerships aim to provide a more cohesive and user-friendly experience, allowing for real-time data sharing with health care providers and caregivers, thus improving disease management.' Once again we stumble upon the recurring trope interconnectivity, enabling predictive analytics and real-time data sharing, and rendering diabetes management far simpler for millions.
This pace of innovation keeps on increasing; AI is speeding up drug discovery to four times its usual pace, and cloud-based systems provide an interconnectivity of elements that were previously impossible to connect. And then there is the hardware—robots that bring precision to surgeries that were once fraught with greater risks. These advancements represent big leaps—maybe too big for us to fully understand just yet—towards a future where health care is radically faster and more integrated than ever before.