Focus and Alignment from Bench to Bedside: Dr. Sean Khozin Discusses the Importance of Rapid RWE throughout Healthcare.
Dr. Sean Khozin is an experienced oncologist who has deployed his experience in clinical, business, and regulatory settings. In this interview, he shares insights on the transformative power of real-world evidence in medicine. Delving into gaps between traditional clinical trials and real world patient outcomes, the conversation explores how RWE offers a more comprehensive picture - especially in oncology. Amidst challenges in drug development, from rising costs to regulatory complexities, RWE offers hope for change. Dive in for intriguing examples, to discover the promising work of Atropos Health, and to envision the future of personalized medicine.
1 - Please share a bit about your background.
I’m a board-certified oncologist and, because of an early fascination with mathematics and programming, my career has been as much about white coats and stethoscopes as it has been about algorithms and big data. In addition to advising companies at the intersection of biology, technology, and AI, I’m a Senior Partner at a new venture firm called Braven where we invest in innovations in highly regulated industries. Previously, I was the CEO of CancerLinQ and the global head of data science innovation and data strategy at J&J. I also had a tenure in federal government as the associate director of the oncology center of excellence and the founding executive director of Information Exchange and Data Transformation (INFORMED), FDA’s first data science incubator I was fortunate to establish under special federal authorities from the Department of Health and Human Services. I also have basic science and clinical research experience at the US National Cancer Institute, having started my clinical research work focusing on the development of molecular profiling strategies in thoracic malignancies, at a time when “driver mutations” were just starting to work their way into our vernacular.
2 - How would you describe the role of Real-World Evidence (RWE) in medicine?
Real-world evidence plays a pivotal role in healthcare and biomedical research today because it captures the experience and characteristics of patients outside of clinical trials, where the majority of patients are under active monitoring and treatment. In fact, the results of clinical trials are nothing more than an attempt to estimate the experience and outcomes of patients in the real-world. Unfortunately, this estimation can sometimes be very difficult because we enroll patients that can often be very different from real-world patients in important ways. For example, in my field of oncology, only about 5% of adult patients with cancer are represented in clinical trials and they tend to be younger, healthier, and from more advantaged socioeconomic backgrounds. This can lead to what we call external validity deficits in clinical trials, which would make it hard to extrapolate the results to tailor personalized treatment decisions at the point of care. Therefore, real-world evidence can help address such external validity deficits by providing actionable insights on patients who are excluded from traditional clinical trials.
3 - What are some of the biggest challenges facing the pharmaceutical industry in its quest to develop new medications for patients in need?
In the past hundred years, we’ve made giant leaps forward in developing innovative new medicines and interventions. Vaccines, antibiotics, and more recently precision therapeutics such as targeted therapies in oncology have significantly improved health outcomes on a global scale. However, the discovery of new medicines today is an increasingly complex undertaking that requires a much deeper understanding of underlying disease mechanisms. Furthermore, the development of promising new therapies has become very expensive due to the rising costs of conducting clinical trials in the backdrop of a highly fragmented and inefficient healthcare system. In addition, the downward pressure on drug pricing we see today around the world poses significant challenges since the modern tools of drug development are quite costly and developing therapies more efficiently calls for fundamental changes to existing workflows and new collaboration schemas, a daunting task in an industry serving a system that has traditionally had a zero-sum game approach to advancing its priorities.
4 - What opportunities do you see for evidence throughout the drug development process, from early research to post-market surveillance?
If we give the right therapy to the right patient at the right dose and time, mathematically speaking 100% of patients should respond. Needless to say, this is rarely the case, especially in real-world patient populations where outcomes tend to be inferior to those observed in clinical trials due to important differences between patients in the real-world and those with access to clinical research. Characterization of the differential responses to treatment using real-word data is an opportunity to discover new mechanisms of disease and patient groups amenable to new treatment modalities. In addition, using real-word data to inform new clinical trial designs, such as those utilizing external controls, can bring new efficiencies to advancing clinical development programs. Lastly, real-world evidence can give us unique insights about physician prescribing patterns and patients experiences, data points that can significantly help advance market access opportunities for bringing the right drug to the right patient at the right time in their journey.
If we give the right therapy to the right patient at the right dose and time, mathematically speaking 100% of patients should respond. Needless to say, this is rarely the case, especially in real-world patient populations where outcomes tend to be inferior to those observed in clinical trials due to important differences between patients in the real-world and those with access to clinical research.
- Sean Khozin MD, MPH | Life Science Advisor to Atropos Health
5 - Can you elaborate on how you see real-world evidence (RWE) transforming the pharmaceutical industry?
I mentioned specific examples such as the use of external controls previously. If I were to summarize the most significant driver of impact on the biopharmaceutical industry as a whole, it’s about increasing focus on the experience and outcomes of patients at the point of routine point of care vs the rather artificial constructs of traditional clinical trials where not only patient populations but also they treatment protocols and standards are very different than in the real-world. Bringing clinical research and drug development closer to the point of care will undoubtedly enable more seamless discovery, deployment, and commercialization of precision therapies.
6 - Why, in your view, is it crucial for pharmaceutical companies to embrace the use of RWE in their processes and decision-making?
I don’t see any other options but to embrace the use of RWE across the full spectrum of the decision-making process since the performance of therapies in the real-world will undoubtedly be of increasingly concern to policy makers, regulators, physicians, and patients in a world where analyzing and interpreting such data is becoming more ubiquitous and streamlined.
7 - What are the key pain points that pharmaceutical companies typically encounter when it comes to accessing and utilizing real-world data and evidence?
There are several. First and foremost, it’s ensuing that the data is of high quality and contains enough variables that allow for in-depth assessment of the longitudinal experience of patients vs cross sections of insights in inadequately characterized populations. Availability of such data remains a significant challenge due to the fragmented nature of the healthcare system where interoperability issues and misaligned incentives complicates the prospects of following the longitudinal experience of patients in the real-world. Lack of expertise in managing the intricacies of real-world data can also be a challenge for organizations with inflexible cultural norms and workflows. Lastly, lack of regulatory clarity and harmonization can lead to complexities in developing appropriate use cases that can advance R&D goals in predictable ways.
8 - Can you share some specific examples of how the incorporation of RWE has led to more effective treatments or better patient outcomes?
Visible examples include the use of RWE for expanding drug indications, such as pablociclib in male breast cancer. Less visible examples include characterization of target populations to inform enrichment strategies in clinical trial designs where RWE has been very beneficial in our ability to focus on patients who are most likely to benefit from treatment.
9 - How does the use of RWE translate into benefits for patients, especially those who are currently underserved (i.e. women, elderly, with rare diseases, rural, socioeconomically diverse, racial and ethnic minorities) and/or have complex conditions like cancer?
The external validity deficits in the results of traditional clinical trials I mentioned earlier means that the body of clinical evidence available to physicians can be inadequate for tailoring treatment decisions to the individual needs of patients. RWE can therefore provide important insights for optimizing the treatment of patients who are under-represented in traditional clinical trials. Without real-world evidence, anecdotal evidence and clinical judgment, albeit very important, would remain the primary driver of many treatment decisions.
10 - Can you speak more about the potential impact of RWE in the field of oncology, and how it might revolutionize treatment and patient care?
The potential of RWE in oncology is significant and we’ve only scratched the surface. This is due to our progress in recent years in precise characterization of patients using, for example next-generation sequencing. Oncology benefits from data generated from a range of validated diagnostics and assays, including liquid biopsies, that are available on the market today. Our ability to objectively characterize patients in the real-world using these tools is very unique to oncology given diagnostic criteria in many other disciplines, such as neuropsychiatry, remains a subjective effort based on clinical observations and examinations that can be blunt tools and highly dependent on the physicians’ level of training and experience.
11 - How do regulatory agencies like the FDA interact with RWE, and what role do they play in validating and integrating this data into healthcare decision-making?
The FDA plays a critical role since many of the most impactful use cases for RWE involve supporting regulatory decisions. Following the enactment of the 21st Century Cures Act in 2016, the FDA has taken a proactive approach to bringing clarity to the use of real-world data. Since then, the FDA has developed a framework and several guidances. But ultimately, it’s the use cases they see and the consensus that is established in the community that drives a substantial part of FDA’s position on RWE. So while regulatory clarity is crucial, it’s incumbent on industry to be bold and venturesome in pushing the boundaries of using RWE to advance discovery and development.
12 - From your perspective, what challenges need to be addressed to facilitate the wider acceptance and use of RWE in regulatory decisions?
Data quality standards and mechanisms of creating more data liquidity are critical to address. For the most part, real-world data continues to reside in siloes with antiquated standards and technologies that haven’t fully embraced modern tools of information exchange. Today we can land airplanes in zero visibility using sensors, predict that structure of proteins based on their amino acid sequences using AI, and interact with nearly the entire corpus of human wisdom captured in written text using large language models but our electronic health records remain frustratingly inefficient and fax machines, telephony, unstructured emails, static PDF files, and pagers continue to be the dominant forms of communication and health information exchange.
Today we can land airplanes in zero visibility using sensors, predict that structure of proteins based on their amino acid sequences using AI, and interact with nearly the entire corpus of human wisdom captured in written text using large language models but our electronic health records remain frustratingly inefficient.
- Sean Khozin MD, MPH | Life Science Advisor to Atropos Health
13 - What strategies or technologies can be implemented to enhance the connectivity and integration of RWE across different platforms and stakeholders in the healthcare ecosystem?
We need to reimagine the fundamental underpinnings of health information exchange. While interoperability is important, it’s not a panacea and I think we need to develop thoughtful approaches to capture structured data at the point of care without causing undue burden on physicians and other healthcare professionals. Having said that, implementation of standardized data models and terminologies is important and can enhance the uniformity and clarity of data across various platforms. It would also enable more efficient data sharing among stakeholders, as well as better assimilation and comparisons of RWE. In addition, leveraging advanced analytics and AI can add significant value. Machine learning algorithms, for instance, can be used to analyze vast amounts of RWE and provide insights that can not only advance research but also clinical decision making. To facilitate data capture at the point of care, we should take advantage of tools that allow passive data collection, such as voice recognition, mobile health technologies, and wearable devices. This will help reduce data entry burden on healthcare professionals, while also capturing more representative and timely patient data. We should also consider implementing patient engagement platforms and tools that can help to empower patients to share and manage their own health data. This will help ensure that the patient's voice and experience are properly represented in RWE.
Lastly, forging strategic partnerships across the healthcare ecosystem is paramount. This could involve collaborations between healthcare providers, researchers, technology vendors, regulators, and patients. Such partnerships could help drive innovation, best practice sharing, and coordinated efforts towards improving RWE connectivity and integration.
14 - Why did you decide to join the Clinical Advisory Board for Atropos Health, and how does this role align with your wider goals and vision in the healthcare sector?
I strongly believe that the guidelines of tomorrow will be algorithms delivered as decision support tools to clinicians at the point of care, allowing them to tailor their treatment decisions based on the unique attributes of their patients. In essence, what Atropos health is working on is a fundamental building block of precision medicine, the ability to administer the right therapy to the right patient at the right time and dose at the level of the individual, the n of 1. As a physician, that’s an exciting proposition.
I strongly believe that the guidelines of tomorrow will be algorithms delivered as decision support tools to clinicians at the point of care, allowing them to tailor their treatment decisions based on the unique attributes of their patients. In essence, what Atropos health is working on is a fundamental building block of precision medicine, the ability to administer the right therapy to the right patient at the right time and dose at the level of the individual, the n of 1. As a physician, that’s an exciting proposition.
- Sean Khozin MD, MPH | Life Science Advisor to Atropos Health
15 - How does Atropos Health address the challenges and opportunities associated with RWE in the life sciences industry?
Atropos Health provides near real-time insights based on the experience of individual patients in the real-world. This can ensure that new therapies can potentially reach their safety-efficacy targets in a shorter time frame and in smaller studies that are optimized for clinical benefit by selecting the right patients using features extracted from RWD. In today’s world, being able to develop innovative new medicines that address high unmet medical needs of patients expeditiously and efficiently requires turning our focus to the point of care, where the majority of patients are under active monitoring and treatment.
16 - Given your expertise in AI, how do you foresee the integration of AI with RWE, and what potential breakthroughs could this bring in areas such as personalized medicine and oncology?
AI is instrumental to maximizing the utility of real-world data. At the most foundational level, AI is helping us parse through the vast volumes of real-world data from disparate sources. Identifying patterns in these complex datasets using machine learning methods is helping us uncover new insights about treatment outcomes, including predicting the likelihood of patients responding to therapies outside of the tightly controlled environment of clinical trials. This is starting to contribute significantly to developing personalized approaches to treating patients. AI, when applied to RWD, is also helping us bridge the gap between routine care and clinical research, helping us for example, identify what patient populations to target, what endpoints to measure, and in some cases even predict study outcomes using simulation. Moreover, AI is helping us parse through the RWD to identify potentially eligible patients for clinical trials, speeding up the traditionally slow process of patient recruitment. In the area of post-market surveillance, the analytical power of machine learning methods are creating new opportunities for monitoring drug safety in the real-world as large and diverse patient populations are exposed to approved therapies outside of traditional clinical trials.
17 - What are the ethical considerations or privacy concerns associated with the collection and analysis of real-world data for pharmaceutical purposes? How can these be addressed to ensure patient privacy and data security?
The collection and analysis of real-world data for pharmaceutical purposes, while promising, is fraught with ethical and privacy concerns that need to be carefully managed. Protecting the privacy of patients is of critical importance as is implementation of robust data security measures for safeguarding the data and preventing breaches that could lead to privacy invasions and potential harm to patients. Cybersecurity is a constantly moving target and it’s critical for organizations working with RWD to keep abreast of the latest advances in the field. Privacy-preserving technologies, such as differential privacy, federated learning, or homomorphic encryption, can offer additional layers of protection while still allowing for valuable data analysis. Equity is another major concern as there is always a risk that the data collected could mirror and amplify existing biases in healthcare, furthering inequities in health outcomes. Like maintaining cybersecurity, mitigating bias in the data requires proactive measures to ensure that the data is not only representative of diverse patient populations, but also regional variations in practice patterns and access to care.
To discuss a life science collaboration, contact info@atroposhealth.com