Pharma

Medidata wants to help control your single-arm clinical trials

New York City-based Medidata is building a case for the efficacy of its “synthetic control arm” service for clinical trials, which takes matching historical controls to a whole new level.

Paper made silhouettes with one of them of orange color to stand out from the rest

At the start of June, more than 30,000 oncology experts will flood into Chicago for the 2017 American Society of Clinical Oncology (ASCO) Annual Meeting.

The focus, as always, will be on data from the most exciting cancer drugs and tests. Expect updates on CAR-T immunotherapies, checkpoint inhibitors, companion diagnostics and the like.

Amid the noise, New York City-based Medidata will be presenting new evidence to support the use of so-called “synthetic control arms” or SCAs, which address a major limitation in clinical trial designs.

According to a company statement, “Medidata is the first technology company ever selected by ASCO to present its scientific innovation to advance clinical research and further drug development in oncology and beyond.”

In other words, it’s not often that they give the mic to someone discussing a tool for clinical trials.

Founded in 1999, Medidata is a software as a service (SaaS) company that offers a diverse range of products and services to both trial sponsors and CROs. Its team will present two studies at ASCO this year. One concerns an algorithm designed to generate new genomic information; the other discusses a new approach to the use of historical controls.

Randomized controlled trials (RCTs) have long been the gold standard for proving therapeutic safety and efficacy. From the beginning, patients are randomly assigned to two or more groups. One cohort is dosed with a placebo or the standard of care, and the other receives the therapy. When the data read out, the trial sponsors and regulatory experts can directly compare how the drug performed against a placebo or the standard of care.

Unfortunately for trial sponsors and the FDA, RCTs aren’t always an option based on the patient population, the nature of the disease, or the resources of the clinical trial sponsor.

Plan B is to look at patients taking the investigational drug and compare their outcomes against historical controls i.e. how patients with the same disease have typically fared in the past.

One of the major concerns with this approach is that trial sponsors could (intentionally or unintentionally) bias their results by selecting historical controls that make the effects of their drug more pronounced.

Medidata combats this flaw in several ways. The first is by taking ownership of the control arm construction. As a third-party group, its team is theoretically less likely to bias the selection of patient controls.

The second and arguably more powerful differentiator is Medidata’s process for mining data on historical controls. The company has access to an archive containing data from over 3,000 clinical trials. With a large pool to draw on, the patient matching is more efficient and potentially more relevant.

At ASCO, Medidata will present findings from a recently published journal article that analyzes the creation of an SCA for a Phase 1/2 single-arm trial in acute myeloid leukemia (AML).

For the best possible controls, the study authors matched a number of variables, such as recency (all trials were completed in the last five years), eligibility criteria, and baseline covariates. Data cleaning and standardization were then used to ensure consistency across the data fields.

David Lee, Medidata’s chief data officer, noted via email that the advanced approach has a number of applications.

“SCAs utilize advanced matching algorithms that enable better estimation of treatment effect from a single-arm trial, and therefore provide better prediction of subsequent later-phase RCTs.  Thus, adoption of SCAs will decrease clinical development failure rates, estimated to be close to 90% from Phase 1 to regulatory approval. In the future, SCAs will also enhance understanding of disease sub-groups, aid in the selection of new endpoints, and reduce the number of patients in control arms with futile standard-of-care or placebo treatments.”

SCAs aren’t going to unseat RCTs as the gold standard for clinical trials anytime soon. However, the study authors believe the approach has the potential to inform adaptive clinical trials as sponsors and the FDA troubleshoot difficult and rare diseases, short timelines, and more.

Photo: FotografiaBasica, Getty Images