Real world data is information captured as a “by-product of everyday patient care”, in the words of Elsevier’s vice-president of life sciences Timothy Hoctor. It can be collected from various sources, such as patient registries, electronic health records, insurance databases, social media and patient research networks.
Although data from randomised clinical trials has traditionally been viewed as the gold standard for drug development, over the past decade regulators have begun to view real world data as a way to better inform drug approval and label expansion decisions.
This is evidenced by the US Food and Drug Administration (FDA) making its first label expansion based purely real world evidence in April this year. The drug in question is Pfizer’s Ibrance (palbociclib), which is indicated for male breast cancer, and the evidence used as the basis of the approval included electronic health records and postmarketing records of its use in male patients.
Elsevier, which collaborated with Pfizer to support and coordinate part of its clinical development programme for Ibrance, chief medical officer Dr Ian Chuang explains: “The small prevalence of breast cancer in males can result in challenges with extracting data from clinical trials, especially when strict criteria can exclude parts of the pool.
“Real-world data offers meaningful insight into the extended populations, e.g. how men with breast cancer have been treated, and how patients reacted to Ibrance since it was first licensed for postmenopausal women with advanced breast cancer in 2015.
“The amount of data gathered in such a short time to support the new licensed use of Ibrance exceeds what a clinical trial could have gathered in the same amount of time.”
Adding value to clinical trial data
Although Ibrance’s approval demonstrates the usefulness of real world data in a post-marketing setting, this type of data can be also add value throughout the entire clinical development progress.
Data-focused analytics company Exploristic CEO Aiden Flynn argues that there is a particular value in using real world data at the design stage. It allows companies to determine the “feasibility” of conducting a trial for a certain disease area and patient population. This is particularly important for rare diseases where “there is a lot of uncertainty and it is very difficult to find patients [to recruit into studies].”
Real world data consultancy Precision Xtract vice-president Kathy Lang echoes this by stating real world data “databases can be used to explore the impact of trial inclusion and exclusion criteria; identify areas where patients who meet clinical trial protocol criteria reside so that they can be cross mapped to trial sites.”
Medidata’s SHYFT Analytics head of research into real world evidence Aaron Galaznik argues this means that “good products can succeed faster” and “products that are not going to be successful can be found out faster,” thereby improving the efficiency of the drug development; research has shown that approximately nine in ten drugs fail and the cost of bringing a drug to market can reach $2bn.
Flynn further describes the impact of real world data to determine “longer term outcomes” of treatments, such as survival, hospitalisations and interactions with doctors, which are not possible to measure in randomised clinical trials. This can inform how a “drug company might price a new treatment.”
Integrating real world data into clinical trials
Increasingly “real world data and clinical data are merging,” for example, Flynn says, “we are seeing more smarter, integrated trials where you have one phase which is a standard, randomised controlled study looking at short term outcomes, but the study continues and becomes much more of a real world study where you are following the same patients over a longer period of time.”
Another example is the emergence of pragmatic studies, which are like “a real world trial… [and] run like a clinical trial where standards that you apply in terms of data collection is the same a clinical trial, but the population you are looking is a much more heterogeneous one.”
Practical knowledge about the therapy can feed back into the “drug development cycle”, because, according to Flynn, it allows “companies to evaluate where your drug might have an impact in other disease areas, because once the drug is in the real world, it is used in a much broader population of patients who suffer from lots of other conditions.”
Therefore, real world data can “better tackle [a] broader variety and heterogeneity of populations than you can recruit in a single clinical trial,” according to Galaznik, which more accurately “reflects the true divert of people with the same condition,” in the words of Hoctor.
Challenges to using real world data
However, Galaznik notes “the very thing that makes it strong – its breadth and diversity, and its reflection of real practices – also makes it challenging to work with.”
Flynn explains because this data is not collected for research purposes, there is no standardisation of physician entry of data or the types of computer systems used, which “introduces bias into the data.”
Galaznik states: “The disparate nature of regulatory bodies is a logistical and structural impediment to achieving that data harmonisation and cross dataset comparison.” Medidata global compliance and strategy principal Fiona Maini adds: “Inconsistent use of terminology is often an issue in data formats.”
Another challenge with using data from outside a clinical setting are “issues around privacy and confidentiality.” Flynn explains: “The guardians of the data are concerned about sharing, because if there is some kind of data, they are the one that will get blamed.”
Regulatory attitudes to real world data
Globally, regulators have shown their acceptance of real world data by publishing and updating guidelines on the topic. Maini explains: “The International Council for Harmonisation brings together regulatory authorities, pharmaceutical industry and various trade associations.
“They are going through a whole modernization of good clinical practices…which will include an annex on trial complexities, trial design and data sources.”
As well as continuing to update its guidance and choosing particular cases, such as haemophilia, where real world data is especially useful, Maini talks about how the European Medicines Agency and the Heads of Medicines Agencies have set up a joint task force to prepare a strategy on big data, which looks at data from electronic health records and patient registries, alongside genomics, clinical trials, spontaneous adverse drug reaction reports, social media and wearable devices.
FDA’s technological solution to challenges
The FDA has taken their commitment to real world data a step further than simply publishing and implementing guidelines liked to the 2016 21st Century Cures Act. The agency has moved to directly tackle the issue of standardisation of data by creating the open-source MyStudies app.
Galaznik explains that technology is a useful solution to challenges with real world data because it “allows you to take data from disparate sources and provide a measure of uniformity and standardisation” to make it “research ready.., and understand the various caveats and nuances” needed to use it.
The aim of the MyStudies app, which was launched in November 2018, is to facilitate the input of real world data by patients into one platform. This data can be linked to their electronic health records, as well as helping with enrolment and assessment of patients into clinical trials.
Former FDA commissioner Scott Gottlieb wrote in a statement: “This digital platform enables developers to adapt our technology to advance new ways to access and use data collected directly from patients—with the necessary controls in place to ensure patient privacy.
“Our hope is that the collection of more real world data directly from patients, using a secure app, will lead to more efficient product development and assist with safety monitoring.”