Victoria Pauli
PhD Student,
Technical Research and Development,
Novartis Pharma,
Basel

umerous initiatives from sponsors and regulators are currently evaluating how Big Data can accelerate drug development, maintain drug safety, and affect the regulatory landscape.

In the session Overview of Major Big Data Projects across EU, US, Japan, panelists from the EMA, FDA, and PMDA gave an overview of ongoing initiatives in Europe, Japan, and the US, attempting to find common pathways and align regulatory strategies.

Key Takeaways

• The ‘real world’ is represented significantly better by RWE (Real World Evidence) than by substantial evidence from well-controlled clinical investigations. Using RWE in regulatory decision making is especially recommended in the rare disease and oncology sectors where clinical data might be limited. However, it is vital that we gain a deep understanding of the quality of the available data before making any regulatory decisions.
• Among the expected Big Data challenges are p-hacking, which affects how data are analyzed and interpreted, and data ownership, an issue that arises when datasets are solely licensed to the pharmaceutical industry from independent partners.
• Two FDA-endorsed projects are the first steps to make the collection and utilization of Big Data a new standard in health care: the Sentinel Initiative and the FDA MyStudies App.
• A recently founded EMA Big Data taskforce is set out to characterize the Big Data landscape and investigate how it can benefit regulatory decision-making.
• Japan currently has two established Big Data projects: MIHARI (the Medical Information for Risk Assessment Initiative), which enables the systematic use of electronic healthcare records, and MID-NET (the Medical Information Database Network), which currently combines 23 hospital records.

Real World Data (RWD), such as electronic health record data and insurance claim records, vary in quality depending on the data source, and must be treated differently than, for example, clinical trial records or genomics data. However, the interest in RWD has risen significantly over the past few years, in large part because the use of RWD promises to reveal previously unnoticed patterns and create Real World Evidence (RWE) on numerous health-related topics.

It was pointed out that the inclusion of RWD in the regulatory process, especially in the case of rare disease and cancer treatments, is very much in line with the 21^st Century Cures Act, a US law enacted in 2016 to accelerate the approval process and a drug’s time-to-market. The FDA-endorsed Sentinel Initiative and the FDA MyStudies App follow up on the Cures Act to make the collection and utilization of Big Data a new standard in health care:

• The Sentinel Initiative allows rapid and secure access to information from numerous data registries supplied by a diverse group of data partners, allowing to widely monitor the safety of medical products after they have entered the market.
• The FDA MyStudies App allows researchers to gather real-time, contextual data about drug use and other health issues via patient questionnaires.

To ensure consistent data quality and to fully harvest the information that Big Data can provide, we need international data standards; political support through, for example, data sharing rewards; and proper patient and healthcare personnel (HCP) involvement.

To accommodate these needs, the EMA Big Data taskforce, launched in March 2017, will create detailed descriptions of the Big Data landscape and provide recommendations for using RWD in the regulatory process based on how such data are already used in pharmacovigilance. In Japan, the Medical Information for Risk Assessment Initiative (MIHARI) was formally launched in 2014 to enable the fast, systematic use of electronic healthcare records in the evaluation of possible safety issues and create a database of post-marketing studies. In addition, the Medical Information Database Network (MID-NET) currently combines 23 hospital databases that contain records of numerous medical examination histories; disease and prescription information; and laboratory, physiological, and radiographic inspection data – all aiming to benefit regulatory decision-making in the future.