How Big Data Is Transforming Drug Development

By Gail Dutton

When the FDA approved Genentech’s Herceptin in 2010, it became the first genetically targeted therapy for breast cancer. Today that tight targeting is becoming possible for other drugs and diseases, as pharmaceutical companies turn to big data to help with drug development.

Big data is beginning to deliver on the promise of personalized medicine that matches drugs to individual patients rather than the aggregated results that have characterized medicine for centuries.

Cracking genetic codes

Pharmaceutical makers are linking gargantuan amounts of data from genomics and other research to help understand more about disease and drug interactions.

Janssen Research & Development, an arm of Johnson & Johnson , focuses its immunology research on diseases with the greatest unmet needs, such as rheumatoid arthritis. The data it captures on how diseases affect physiology allows for deeper research, resulting in more effective drugs, says Mark Curran, Janssen’s vice president of immunology systems pharmacology and biomarkers.

Diseases typically are triggered by the confluence of multiple genetic and external factors, rather than a single gene. “The technologies and scientific strategies applied in big data efforts are enabling deeper understanding of these complex disease mechanisms,” Curran says. One relatively new research area, for example, studies the microorganisms that share our body and the broader environment to better understand what causes genes to be reprogrammed.

Improving clinical trials

Biopharmaceutical companies also are discussing how to use large data sets to improve clinical trials. One of the most promising areas is determining which patients may benefit most from a drug and, therefore, should be included in a study. Big data strategies also are being developed to improve data quality in clinical trials, as well as the relationships between the pharmaceutical company and trial physicians.

“Big data improves clinical trials by managing a larger amount of patient data, processing it faster and more efficiently,” says Max Dufour, partner, strategy and innovation, at the management consultancy Harmeda LLC.

Big data analytics also make it feasible to target specific patient profiles for a clinical trial by processing many data points. A clinical trial could focus, for example, on female breast cancer patients over age 50, of Asian ancestry, with the presence of the BRCA1 gene – which is linked to breast cancer – who had failed first round chemotherapy. “This would not be possible without big data,” Dufour says.

It also lets researchers examine multiple studies across the industry.  As Michael Mentesana, partner in the pharmaceutical R&D advisory services at PricewaterhouseCoopers, points out, “In the past, drug developers looked at a study. Now they can look at virtually all studies involving that class of drug, including outcomes, adverse events, deviations from protocols and new drug applications to mine information for predictive analytics.”

Additionally, Dufour says, “Big data allows investigators to monitor trials in real-time. Automating and accelerating clinical trial analysis speeds the entire drug development process.” That reduces costs by identifying issues, such as adverse events, earlier in the process, bringing innovative drugs to market sooner.

Regulatory concerns

The caveat in all this is that data must be de-identified to comply with global patient protection and privacy regulations. “As companies analyze large samples of data from multiple sources and collaborate with other organizations, they must ensure they have the authorization to actually leverage and share that data,” Dufour says.

The FDA and regulatory bodies throughout the world are moving “extremely fast” to accommodate big data, Mentesana notes. The FDA recently announced plans to allow public access to their data set for adverse events, recalls and labeling data. “Access helps companies expand their drugs to other indications more easily, as well as become aware of trends and adverse events that may affect their own development pipeline,” Mentesana says.