Clinical Researcher—June 2024 (Volume 38, Issue 3)

DATA & DEVELOPMENTS

Stephan Ohnmacht, PhD

 

 

 

The potential of personalized medicine presents an opportunity for life sciences to leverage big data to target therapies to specific patients better. With artificial intelligence (AI) and machine learning technologies continuing to develop, research and development teams can finally bring this vision of personalized medicine to life, provided that the data they are using are clean, standardized, interpretable, and secure.

Understanding a treatment’s potential for a specific patient requires biopharmaceutical companies to bring together data from multiple disparate sources. Some sources, like patient demographics, electronic medical records, and quality-of-life scores, will be consistent across all therapeutic areas. However, most of these data sources—including genetic information, imaging, and activity data from wearable devices—will be unique to each individual. With personalized treatments, clinical effectiveness and safety profiles will vary from patient to patient, so relevant stakeholders must be able to trust the data to make medical and business decisions confidently.

A new approach to quality, ownership, and interoperability will ensure the data are useful, even when working with millions of datapoints. Biopharmaceutical companies are also rethinking existing ways of working to get to first-time-right submissions. With access to a clean data foundation, they can identify which functional areas are most important in accelerating time to market so that patients get the innovative treatments they need sooner.

Getting Treatments to Patients Faster

In the past, data collection initiatives have been broad in scope and ambition, ranging from sequencing, imaging, and electronic health record data to text-based data, such as interactions with health authorities and conference abstracts. These initiatives strived to achieve data completeness, but the scale of datapoints collected made it challenging to develop useable insights.

Now, the go-to-market and approval requirements for personalized treatments are highly complex. Biopharmaceutical companies are looking to use their study data appropriately earlier in the process, shifting focus from data collection to governance and ownership. This shift in control and oversight will change the companies’ relationships and contracts with third parties. In turn, connected technology is becoming critical, as it allows relevant stakeholders to maintain constant visibility into the data rather than waiting for data to be shared in specific meta-data formats or final text-based documents.

With clean data, biopharmaceutical companies can more easily pinpoint the inefficiencies most detrimental to the clinical development phase—a vital step in lessening time to market so that personalized medicines remain commercially viable. Analyzing data on the cycle times between two critical clinical milestones could determine where inefficiencies and operational challenges typically arise, whether during protocol design, site selection, or elsewhere. With a single source of accurate data, companies can gain a competitive advantage by improving decision-making on patent filings, patient recruitment, or efficiency gains in outsourcing, procurement, or portfolio rationalization.

While analytical and data science capabilities have improved, limitations remain. Raw data are not standardized, and limited reference models exist. However, if common pain points around cleaning, ownership, and standards can be resolved, data will be more plentiful and accessible. Establishing a data model with stringent user access controls is essential to address privacy and cyber-security concerns.

Clean Data are Useful Data

When prioritizing data initiatives, clarity of purpose is key. Suppose the primary goal is significantly reducing the time from “first patient, first visit” to database lock. In that case, selecting a group of experts before data are collected and cleaned to decide the approach and exact use cases is the best approach. Data scientists, subject matter experts, and even external experts (e.g., healthcare professionals, key opinion leaders) could all take part in these decisions and test hypotheses for this key clinical development milestone.

For many organizations, the problem is not assembling the right talent and technology, but rather effective governance. Solving this may require collaboration between functions that typically don’t interact, such as research and information technology, and certainly requires commitment from leadership. Testing different working models can help narrow the options to one that best suits the company’s culture.

Once roles and responsibilities are established, align people, processes, and technology to the broader corporate goals, problem statements, and hypotheses. Ensuring resource flexibility is key. For example, an urgent drug safety issue might have immediate clinical and downstream commercial repercussions for a company, meaning that the right experts must be available to tackle it. Having clean data from a single source is critical so that all the stakeholders, from statisticians to chemists, can work efficiently to get back on track.

As the use cases for big data are defined and executed, collaboration between teams within organizations will improve as they all work toward the goal of effective, personalized medicine. The result will be higher-quality documentation, reduced cycle times, and more right-first-time submissions. The growing need for direct data application programming interfaces with regulatory and health authorities, contract research organizations, and other third parties could lead to more cooperation and faster regulatory decisions. That’s good news for biopharmaceutical organizations and patients who stand to benefit from the treatments.

The Wide Reach of Smarter Data Use

Developing personalized medicines challenges even the most efficient research and development functions due to the associated costs and risks. Smarter data use will help organizations better manage the complex drug development journey and identify which issue to solve first.

Personalized medicine isn’t the only area that stands to benefit from big data that are clean, standardized, and interoperable. Other possibilities include finding novel biological targets or net new patient populations. Eventually, a centralized approach to data management could support the long-held ambition of connecting real-world data—such as patient data, electronic medical records, and digital therapeutics—to clinical development, improving the patient experience. Fundamentally, these advances with big data move us toward the shared goal of providing life-enhancing medicines to patients who need them.

Stephan Ohnmacht, PhD, is Vice President and Head of European R&D Business Consulting for Veeva.