Opening the aperture :
As originally conceived, personalized medicine referred to the tailoring of medical treatment to the individual characteristics of each patient,1ultimately leading to a shift in the clinical treatment paradigm from a trial-and-error approach to “the right drug, for the right patient, at the right time.” Today, a combination of public investment, biotechnology development, and digitization of health profiles has evolved personalization beyond therapy selection and into the realm of drug discovery, how care is planned for and delivered, and increasingly, to how we as consumers engage with companies seeking to improve health (exhibit). Driving this transformation are advances in diagnostics, digital devices, and imaging, alongside an arsenal of analytics tools working across a multitude of institutions and stakeholders. Encompassing this entire ecosystem, medicine will be driven by three key components: 1) Data collection through diagnostics and behavioral devices that capture us in various states of health and disease; 2) Individualized solutions through advanced analytic engines and personalized therapies; and 3) Business models necessary to sustain value and incentivize continued growth.
In this article, the introduction to a broader compendium, Precision medicine: Opening the aperture, we discuss the recent advances in each of these three areas, the challenges the industry faces going into the next five years, and the implications for key stakeholders. The compendium also includes articles on the following:
Previously, we thought about precision medicine as data from targeted genomic panels informing therapy selection. Today, with the explosion of data collection at the population level with multiple data points, it’s common to say that data has become the “oil” for our time. The sheer scale of data proliferation is breathtaking. According to a 2017 white paper from Stanford University School of Medicine, 153 exabytes2 of healthcare data were produced in 2013, and an estimated 2,314 exabytes will be produced by 2020, a 48 percent growth rate annually.3 This growth rate is so extreme that we can say that 99 percent of the world’s data has been created in the past 18 months—a staggering statistic. Analyst reports estimate the market size for big data in healthcare at between $53 billion and $69 billion by 2025, with CAGR of up to 27 percent.4
Beyond generating such vast quantities of data, health systems are getting better at integrating data sets to more easily aggregate them and gain better access. In the United States, the Department of Health and Human Services (HHS) has invested $35 billion in healthcare IT5 ; this has rapidly advanced medical data storage through the proliferation of electronic health records (EHRs). Such integration and aggregation of data is allowing us to close the loop to fully understand patients from their symptoms, to treatment, to outcomes. Roche is an interesting example of starting to “own the patient” from end to end. With the organization’s recent acquisition of Flatiron Health ($1.9 billion) and the remaining stake in Foundation Medicine ($2.4 billion), it now has access to genomic data from thousands of oncology patients.6 ,7 Combining this data with innovative, targeted therapies, such as those coming from their acquisition of Ignyta ($1.7 billion), could provide Roche with a continuous data loop from identifying a patient, confirming genomic signature, treatment selection, and on to monitoring outcomes.8
While genomic data in oncology is still a critical part of healthcare today, we have “widened the aperture” to understand all of the ways in which we can personalize healthcare: continued growth of genomics beyond oncology, additional modalities to understand our molecular phenotype, and collection of behavioral data through devices. In the not-too-distant future, we envision that every patient will have his or her own data ecosystem, a closed loop of continuous learning based on ubiquitous data, enabling each patient to benefit from insights generated by the collective experience of the entire medical community. Challenges of siloed data collection, interoperability, and policies for data sharing have slowed the realization of this vision, but these are slowly being overcome. We discuss the implications for pharma, providers and diagnostic players in terms of how they can compete in this digital, data-driven world.
In the original concept of precision medicine, insight generation centered around univariate analysis: that is, understand what mutation leads to what disease through retrospective research, and use that algorithm to prospectively identify mutations in a new population. This has become increasingly powerful in oncology, through integration of genomics, EHRs, and advanced analytics. However, across therapeutic areas, we are seeing an explosion in the availability of data over multiple dimensions, which in turn leads us to a much broader set of questions to solve.
The advent of new technologies and mobile medical apps has allowed us to actively track a patient’s physiology in real time. Whereas previously we collected descriptive statistics of discrete populations, we can now take this multidimensional data and create predictive algorithms, which use the collective learnings to predict outcomes for an individual. This approach implies a cyclical, dynamic feedback loop whereby processes and underlying capabilities are constantly modified based on the inputs from patients. To continue to push the potential of precision medicine, healthcare stakeholders are actively trying to build capabilities along three dimensions: data acquisition, data analysis, and analytics-based decision-making. As the number of data inputs increases and the level of analysis becomes more and more sophisticated, we are seeing both start-ups and established technology players with core competencies in advanced analytics also trying to enter the healthcare space.
One recent example is Tempus, which describes itself as a “technology company that has built the world’s largest library of clinical and molecular data and an operating system to make that data accessible and useful, starting with cancer.”9Recently valued at $2 billion, the company has established data partnerships with large cancer centers across the United States, including Vanderbilt-Ingram Cancer Center and ASCO. It provides a proprietary platform to ingest unstructured data (clinical notes, pathology images) and structured data (next-generation sequencing) to deliver actionable, personalized insights. More established players in this space include IBM and Google. Google’s DeepMind recently published impressive results analyzing 3D optical images, outperforming experts in making referral recommendations for a range of retinal diseases, while IBM’s Watson has continued to improve its ability to tailor treatment options to a patient’s genomic profile. While still a work-in-progress, both the level of commitment and investment by major technology companies to advance AI in medicine is a harbinger of things to come.10 11
Would you like to learn more about our Pharmaceuticals & Medical Products Practice?
In 2017 and 2018 we also saw approvals of two truly individualized therapies, Yescarta and Kymriah, for leukemia and lymphoma. These CAR-T therapies are a type of immunotherapy where a patient’s own immune cells are genetically modified to fight cancer cells. Other gene therapy techniques, most notably CRISPR, are in active development, and we expect more and more individual therapies to be approved in the next 5–10 years.
While we can all appreciate the importance of data and insights, it’s much less clear how to derive value from those insights, and who will pay for it. Each stakeholder grapples with this differently in the precision medicine ecosystem. Diagnostic players need to understand the market for future tests, whether the value will lie in the test itself or the insight generated, and the best commercial model to support that. In addition to providing quality, personalized care, providers are considering how best to aggregate and mine patient data, and what insights could be commercialized from that data. Pharma companies are partnering with payers and providers to be part of the data ecosystem, and are also trying to determine the best commercial model when considering smaller and smaller patient populations. There is no gold-standard business model to pave the way for how to unlock value. Some are giving away genomic sequencing in order to own the data (for example, Nebula Genomics, Tempus), while others are solving tangential problems for health systems (such as billing, tumor board management, data aggregation) in an effort to access and own the data; however, most pharma and med device players are tapping into the precision medicine ecosystem to bolster conventional business models (for instance, selling more tests or therapies through conventional channels). Given all of the changing healthcare dynamics, the time is ripe for disruption through business model innovation. In this compendium, we consider the five aspects of business model innovation (value proposition, economic model, delivery model, production model, operating model) through the lens of key stakeholders, and also explore the potential for new entrants to unlock further value in the precision medicine ecosystem.
Given the rapid pace of change in this field, we expect the fundamental ways in which we deliver healthcare to transform. Within that process are a number of unknowns as the technology evolves with different stakeholders taking multiple approaches to the importance and value of data, how its worked with and analyzed, and how that’s applied to patient care. With this in mind, there are three key questions we should be asking ourselves as all players in the industry consider and shape the future of this space.
In an information economy, the ability to gather data, generate insights, and then transform those insights into impact in the real world form the backbone of value creation. Such capabilities exist natively in the tech ecosystem because, in large part, digital companies are the ones that have taught us how to build wildly successful businesses around data. Accordingly, tech players—Apple, Amazon, Google, Facebook—possess a tremendous advantage as they enter biomedicine, being the incumbent experts in the components of the operating model that we expect to drive value in precision healthcare: data stewardship, excellence in analytics, agile product design, and superior analytics talent.
Yet, to this point, the biomedical incumbents have remained unthreatened in the delivery of healthcare and the development of therapies; while there have been interesting partnerships, acquisitions, and enabling technology development, there have been no at-scale examples of healthcare disruption by a major tech player to date. An important open question is whether one set of players will ultimately win the day, or whether the coming years will see greater collaboration and joint product development that will ultimately transform how most patients interact with the healthcare system.
The ability to collect more and more information about our health comes at a cost: molecular diagnostics can be hundreds to thousands of dollars and are not always reimbursed, while devices for collecting behavioral data are a costly personal expense (for example, Apple Watch, Fitbit). Health systems are continuing to evolve, and we could see stakeholders that would benefit from lower costs of care (providers, payers) helping to defray costs in order to allow more patients access to such tools. Additionally, the cost of targeted therapies resulting from personalization can run from tens to hundreds of thousands of dollars a year. As we discover the importance of all of these data inputs to drive clinical insights and inform new treatments, and find tremendous clinical benefit in novel personalized therapies, how broad do we expect access to be, and how quickly will it scale?
Macro factors will invariably determine the trajectory of innovation in precision healthcare: from the way that information is regulated (for example, GDPR) and how payment for medical services is rendered to how new therapies are tested and approved. These factors will have an impact at least at the regional level but, in most cases, at a national level or beyond. Traditionally the United States has seen the most funding, innovative technology and therapy development; however, China is investing heavily and has a much less stringent regulatory environment. Interestingly, many European countries have healthcare models that are most aligned with the value proposition of precision medicine, and have been quite aggressive in developing initiatives around individual patient data collection, especially for large, de-identified patient data sets (for example, UK Biobank). However, data privacy issues and regulations could dampen this momentum. Given the diversity of possible approaches, where will precision healthcare accelerate the quickest?
Over the past five years, healthcare’s collective description and understanding of what constitutes precision medicine has evolved for the simple mandate of “one patient, one drug” to a more complex data, analytics, and business model ecosystem. We look to the next five years to see how far this data revolution in precision medicine will go, and what transformative new therapies it will usher in.
Download Precision medicine: Opening the aperture, the full compendium on which this article is based (PDF–28.2MB).
Jacob Aptekar, Meredith Reichert, and Erika Stanzl are consultants in McKinsey’s Silicon Valley office, where Edd Fleming is a senior partner; Nicholas Donoghoe is an alumnus of the Silicon Valley office; and Kevin Webster is an associate partner in the San Francisco office.
We just sent you an email. Please click the link in the email to confirm your subscription!