OR WAIT 15 SECS
IntroductionAfter decades of promise, we have begun to realize the fruit of “-omics” technology. Recent advances in proteomics, genomics and metabolomics have enabled us to understand the molecular basis of disease at both the diagnostic and treatment levels. Equally important, a growing suite of biomarkers now provides predictive value for diagnosis, disease progression and cure/remission.
After decades of promise, we have begun to realize the fruit of “-omics” technology. Recent advances in proteomics, genomics and metabolomics have enabled us to understand the molecular basis of disease at both the diagnostic and treatment levels. Equally important, a growing suite of biomarkers now provides predictive value for diagnosis, disease progression and cure/remission.
Today’s techniques are increasingly targeting new sources of information from patients, recognizing the uniqueness of individual subjects and producing massive quantities of data. At this point, the clinical trials toolbox has grown so extensively that the research community is no longer limited by the tools themselves - we are limited instead by our ability to combine, deploy and manage these complex tools in ways that enable us to drive innovation.
Indeed, the drive to innovate has prompted researchers and regulators to explore novel and more complicated ways to investigate promising new products, yielding trial designs that are faster, more flexible and more targeted. In many respects, the future of clinical research lies in the successful completion of these complex clinical trials, many of which require the simultaneous development of novel combinations, including drug/biologic, drug/diagnostic, drug/device, etc.
In keeping with our commitment to understand and simplify complex research needs, three of Theorem’s leading thinkers have each selected and examined a prominent area of clinical research that is undergoing rapid change as a result of technological advancement. As you will read, each of the chosen areas will make a profound impact over the next several years.
Data Integration and Visualization
Today’s complex clinical trials yield data sets that are almost unimaginably large and complex. As a result, the process of transforming trial data into usable knowledge simply cannot be accomplished using traditional presentation methods such as conventional charts and graphs. Instead, research organizations are beginning to intelligently apply data integration and visual analytics tools in ways that enable researchers to both explore and interact with aggregate data. What results is a new level of discovery with in-depth insight that would never have been achieved through a traditional review of the raw data.
Wielding these advanced tools and rendering data in this expanded capacity are challenges requiring the deployment of well-established visual analytics theories, sophisticated software and a talented pool of programmers. Yet despite the challenges, when properly applied, these data integration and visual analytics tools bring real-life potential to emerging concepts that would otherwise be relegated to “wish for” status:
To demonstrate some of the specific ways in which data integration and visualization are changing the clinical trials landscape, let’s take a moment to unpack one: risk-based monitoring.
Risk is an unavoidable component of living and working in an active society, but risk can be reduced by identifying and interceding at points with the greatest likelihood of failure. For example, crossing a busy city street can be risky, but by implementing appropriate safety measures - traffic lights, pedestrian lights, walkways - the risks are mitigated. The probability of being hit by a car is reduced when procedures are properly followed.
Likewise, the challenge in clinical trials also lies in associating a risk factor to the most likely failure points, establishing procedures to reduce risk at those points and monitoring closely. This process is known as risk-based monitoring (RBM). RBM is the study of identifying and mitigating the points where a trial and its subjects are at greatest risk for harm through data negligence, lack of education or incorrect conclusions.
Enabled by advanced data integration and visualization techniques, RBM improves real-time (not just retrospective) visibility of a continuous flow of data by automating centralized data collection and connecting the personnel analyzing it, helping them make informed and impartial decisions.
RBM takes its underpinnings largely from signal detection theory, which describes the ability to discern between real information-bearing patterns (signal) and random patterns that distract from the information (noise). By employing concepts of signal detection, we can create monitoring protocols better designed to detect anomalies effectively and quickly.
The intent is to bring intense focus onto the most likely failure points, looking at the data for what is NOT being revealed during study conduct. RBM employs several important mechanisms, including:
Risk can be reduced by monitoring sites for subject visit information and verifying the transcription process of source data being dutifully entered into an electronic data capture (EDC) system. However, is that where one should look for risk? Can the opportunity for uncovering data entry error or EDC entry field interpretation truly drive the reduction of risk for study results, or is it a review of the source data itself?
Studies have shown - and an experienced monitor will confirm - that the risk lies at the source of the information.
RBM allows earlier action to be taken for operational study conduct and provides awareness of the potential for in-study subject risk, enabling corrective action to be taken sooner. The opportunity for analysis of near real-time data analysis becomes a value-based objective.
As a result of focusing on areas of risk, areas of non-risk receive lower-level scrutiny in the study monitoring process, bringing efficiencies and cost reductions. Source document verification to the EDC entry system becomes less important as predictive analytics drive determination of risk.
Implementing RBM can be challenging, but many research sponsors have forged a path to success by following these steps:
With proper oversight, RBM offers numerous benefits. Electronic health records data can be mapped into EDC systems, thereby automatically populating integrated systems. RBM also offers the benefit of integrating new data streams into familiar reports, whereby the process of visualizing and understanding the information is much simpler. What’s more, when RBM-created data is presented via visual analytics, decision makers have the ability to see data relationships in new and intuitive ways. Finally, RBM can usually be integrated into existing technologies; there is rarely a need to start from scratch.
Yet for all its promise, RBM is saddled with several misconceptions and challenges. The largest misconception is that RBM is a cost reduction process. While cost reduction is an expected secondary result of proper implementation of RBM, the primary benefit is quality. In addition, RBM is based on big data, and big data takes big brainpower; deploying the right technologies and techniques requires serious expertise and sustained effort. Finally, RBM presents a multitude of HIPAA challenges that can’t be overlooked.
In the end, the promise of RBM will only be realized when it is appropriately customized. Indeed, all of the advancements supported by augmented data integration and visual analytics tools require effective customization. Since a one-size-fits-all approach isn’t feasible, sponsors may benefit by partnering with an experienced CRO as they work to design a robust data integration and visualization strategy.
Although the term sounds straightforward, Personalized Medicine is a complex and multifaceted construct focused on the successful application of advanced technology to the synthesis of (and the blurring of lines between) data, technology, drugs, devices and diagnostics. In the near term, the most promising aspect of Personalized Medicine is the use of currently available technology to administer today’s therapeutics to patients in highly individualized ways. In the long term, Personalized Medicine holds the promise of delivering custom-designed therapeutic agents manufactured to match an individual's specific genomic profile.
Consider how far we’ve come in the prevention and treatment of America’s No. 1 killer, cardiovascular disease (CVD). Elevated arterial blood pressure (BP) has long been associated with increased risk for heart failure, myocardial infarction, stroke, kidney disease, etc. Yet, while there are many individual factors that can lead to elevated BP, the treatment for hypertension has historically consisted of little more than a trial-and-error choice between one or more diuretics, beta blockers, calcium channel blockers, angiotensin-converting enzyme inhibitors or angiotensin receptor blockers (ARB).
Fortunately, new research is providing a clearer focus. Scientists are looking beyond a simple assessment of the degree to which agents lower BP; they are measuring their ability to change specific CVD biomarkers. For instance, the recently initiated ATTEMPT-CVD study is the first large-scale study to focus on the efficacy of ARB therapy to the changes of biomarkers.With the results from this and other groundbreaking research, clinicians may soon be able to select more appropriate individualized therapy based on each patient’s urinary albumin creatinine rates, plasma brain natriuretic peptide, serum high sensitivity c-reactive protein, urinary 8-hydroxy-deoxy-guanosine, serum adiponectin or high molecular weight adiponectin.
But the movement toward Personalized Medicine extends far beyond CVD. Today, many diseases have specific diagnostic biomarkers that allow pinpoint treatment. Specific biomarkers yield better diagnostic criteria, more specific treatments, better outcomes and potentially fewer side effects, all of which show promise in shifting the balance in the risk-benefit continuum.
Which areas of Personalized Medicine research show the most promise? We believe the most exciting potential - and the most challenging complexity - lie in individualized data and combination trials.
In many cases, clinical research is shifting away from its reliance on patient population averages and is
instead seeking to identify the best treatment for each individual based on his or her unique characteristics. Key person-specific biomarkers are now being identified for many diseases, and disease treatment/management strategies have emerged that are focused on specific disease variants.
Moreover, our understanding of complex and sinister diseases has continued to evolve with the knowledge gained at the genetic level. Consider that, at one time, breast cancer was considered to be a monolithic disease. Over time, and with the evolution of technology, different tumor types have been identified and treatment strategies altered to coincide with this new knowledge. With the discovery of the estrogen and progesterone receptors and their role in diagnosis and treatment came even further refinements in prognostic indicators and treatment strategies. Therapeutic options have continued to evolve as we learn more and more about this disease. The HER2 receptor yielded one of the first great Personalized Medicine success stories with the approval of Genentech’s Herceptin®, which significantly changed the face of treatment and survival for breast cancer patients. Today, treatment strategies for breast cancer and other malignancies are guided by multiple individual factors, including tumor histology, receptor status and even gene mapping. As a result, diseases that were once attacked with a pharmaceutical bomb are now attacked with laser precision. The ultimate in individualized data, of course, will be reached only when we have the ability to quickly and inexpensively map the complete genome of each patient, a goal that may become a reality in the not too distant future.
Interestingly, the progress toward individualization is yielding a new dynamic in the design of certain research protocols. In numerous recent trials, each subject is his or her own control. While this type of study design is not yet widespread, continued movement in this direction should be expected.
Finally, we must not close our discussion of individualized data without acknowledging the interrelationship between it and the rare disease/orphan drug domain. To a significant degree, the very recognition of many rare diseases is a function of scientists’ newly developed ability to parse larger disease categories into smaller and smaller subsets. Most rare diseases are caused by uncommon genetic mutations, so rare disease trials are targeted at an extremely focused patient population. In the end, the study of orphan diseases is Personalized Medicine.
Combination trials are particularly useful during parallel development of two or more regulated components (drug, biologic, device, diagnostic) and can support the development of two distinct categories of products. Combination products comprise two or more regulated components that are physically, chemically or otherwise combined or mixed to produce a single entity. One example is a drugeluting cardiovascular stent. Noncombination products comprise two or more regulated components thatare used jointly but are produced as separate entities; one example is the HER2/neu receptor diagnostic, which is used in concert with Herceptin®, noted earlier.
In the context of Personalized Medicine , some of the most exciting combination trials are centered on diagnostic + therapeutic combinations. According to the Tufts Center for the Study of Drug Development, the FDA currently lists more than 100 approved products with pharmacogenomics information on the label, dozens of which have labeling suggesting or requiring use of a companion diagnostic. Another exciting arm of diagnostic + therapeutic research is the integration of drug delivery systems with molecular imaging, which has enabled us to find and target diseases in ways that were unattainable even a decade ago. As the level of sophistication continues to grow in diagnostic technology, we will be able to identify, assess and treat diseases with greater precision, sensitivity and specificity.
Despite rapid growth in the arena of Personalized Medicine , important challenges still exist:
The promise of Personalized Medicine demands clinical trial strategies that are increasingly complex and have evolved with the technology and advances in science. Being able to determine the registration pathway for these individualized products is equally complex and requires knowledgeable clinical and regulatory strategy. As the science of clinical development struggles to keep up with technology, the development paths for many new products have no precedents; meeting this challenge requires not only an in-depth knowledge of clinical development and strong science, but a team of regulatory strategists who can interpret guidance where none exists and articulate a successful strategy. Just as with other products in development, the process for these challenging entities needs to be guided by high science, strong ethics and flawless execution.
Mobile technology has come a long way since the introduction of the early wearable devices. Remember your excitement when you first encountered a mechanical pedometer. At the time, it seemed truly amazing - a belt-clip device that could actually count your steps for you! Then came the ability to compare one day's step count to the next, and then the power to compute the calories those steps burned. Today, similarly sized wearable devices track sleeping patterns, heart rate, eating habits and more. Now in development are smartphone applications that can detect epileptic seizures, contact lenses that measure blood sugar and glasses that automatically and dynamically correct vision.
As seen through the lens of clinical research, mobile medical technology may be viewed as either an end, which entails the testing of the technology in clinical trials in order to achieve FDA approval of the technology itself, or as a means, which examines the many ways in which today’s mobile technology is changing the conduct of drug and medical device trials. Let’s address each of these briefly, beginning with an overview of the landscape surrounding FDA oversight and approval of mobile technology.
FDA regulators recognize the potential benefits that mobile medical technology can provide in helping patients gain access to useful information in managing their own health and wellness. The FDA published guidance in 2013 intended to create a framework enabling the agency to appropriately oversee the development of such technology without undue restrictions. In short, in order for any particular technology to receive FDA oversight, it will first need to fall within the definition of a medical device, characterized generally as an instrument, apparatus, appliance, machine or contrivance intended for use in the diagnosis, cure, treatment, mitigation or prevention of disease.
The FDA decided to implement a three-tiered approach in considering mobile medical technology: 1) medical devices that the FDA will regulate, 2) medical devices for which the FDA will exert enforcement discretion and 3) applications that do not meet the definition of a medical device. For more information on the definitions, standards and processes for each of these, please see the paper, (How) Will FDA and the EU Regulate my Mobile Medical App? by Brian D. Bollwage, JD, Theorem Vice President of Global Regulatory Affairs.
Let’s turn our attention now to some of the most important ways in which mobile health technology is changing the conduct of drug and medical device trials. Today’s research landscape is being transformed by rapid advancements in several areas of mobile technology, but one of the most exciting developments is the ongoing rollout and adoption of on-body and in-body sensors. These medical devices can be used to monitor a patient’s physiologic state (e.g., respiratory status, glucose levels, electrocardiogram readings). They will also be increasingly able to provide contextual information (e.g., body position, activity level) and environmental characteristics (e.g., moisture detection, temperature changes).
These wireless medical devices can provide patients an ongoing and comprehensive view of their health over time. Clinical trials that invite patients to become active participants in the investigational process by means of mobile medical technology are inherently more patient-centric, benefiting both the patient and the research. Indeed, this simple step - engaging patients in their own care during study participation - can have a profound benefit in patient adherence and persistence. Moreover, cutting-edge, patient-centric trial designs create opportunities for patients to take immediate action in response to biometric or symptomatic changes.
While the benefits of informing patients about their own health status are undeniable, the enhanced promise of always-connected on-body and in-body sensors begins to be realized as the data collected by these medical devices is communicated automatically and continuously to investigators, site managers and clinical trial databases.
In contrast to traditional data collection, which consists of a mix of self-reported data along with periodic bioassays that give a limited snapshot of a patient’s health status, the automatic and continuous reporting of patient-specific trial data can move risk-based monitoring beyond the site to the individual, where it can improve patient safety and guarantee data integrity. Such data collection and transfer eradicate delays, eliminate transcription errors and reduce missing/incomplete data.
Finally, we need to note the patient-friendly efficiency of BYOD (bring your own device) smartphone apps. On-body and in-body technology that collects and automatically reports crucial study data using a patient’s own existing mobile phone has obvious and significant benefits compared with technology that requires an additional device. These benefits include lower cost of deployment, less training, higher adherence, easier scale-up, etc.
In our view, it is vitally important that clinical trial sponsors and CROs move aggressively to adopt mobile technology whenever and wherever possible. The benefits of this emerging field are real and expanding quickly. Moreover, the overlap between mobile technology and several topics reviewed earlier - including Personalized Medicine , data integration/visualization and complex combination trials - is significant. Each feeds on the others: Genomic advances lead to complex trials, which yield drug/mobile medical device combinations, which enable individualized therapy, which makes diagnostics real time and continuous, which creates more data, which can lead to even more genomic advances.
Looking forward, we can only anticipate that the future will bring even more rapid technological changes, and researchers will have to adopt and adapt these changes. Ongoing development and deployment of data integration/visualization, Personalized Medicine and mobile technology will continue to change the science and practice of clinical research. Our challenge, then, is to participate in and respond deftly to that change in ways that lead to ongoing success. After all, as Charles Darwin is said to have reminded us, "It is not the strongest of the species that survive, nor the most intelligent, but the one most responsive to change."
Just as Personalized Medicine holds the promise of delivering custom-designed therapeutic agents manufactured to match an individual's specific genome, we at Theorem view every engagement as an opportunity to craft a unique, custom-designed development plan built around your specific needs. Here’s to the future, when the changes we embrace today will lead to a healthier world.
John Potthoff, PhD, President and CEO
Mark Penniston, Executive Vice President and General Manager, Clinical Analytics
Marc Hoffman, MD, Chief Medical Officer
D. Lee Spurgin Jr., PhD, Senior Vice President and General Manager, Medical Device and Diagnostic Development
Related Content:Online Extras