Real-World Evidence Is Reshaping Drug Approvals: What Sponsors Need to Know

For decades, the randomised controlled trial was the unchallenged gold standard for drug approval. Regulators tolerated little deviation. Then came the 21st Century Cures Act, a wave of digital health infrastructure, and a pandemic that forced agencies to make consequential decisions on incomplete controlled data. The result is a regulatory environment that has shifted faster than most sponsor organisations have adjusted to. A 2025 survey of 142 regulatory affairs executives at mid-to-large biopharma companies found that 61% rated their internal RWE capabilities as "insufficient" for the evidentiary demands regulators are now placing on label expansion submissions.

The Regulatory Shift Is Structural, Not Temporary

What is happening now is not a pandemic-era concession that will quietly recede. Both the FDA and EMA have codified their positions on real-world evidence in formal guidance documents, and those positions represent durable policy commitments. The FDA's Real-World Evidence Programme, now in its eighth year, has produced 13 published guidance documents covering electronic health records, claims data, patient registries, and pragmatic clinical trials. The EMA's revised framework, released in October 2025, introduced a three-tier acceptability model that explicitly links the required level of data provenance documentation to the regulatory decision type: post-market surveillance, label expansion, or initial approval.

The practical implication for sponsors is significant. Regulatory reviewers are no longer treating RWE as a supplementary curiosity sitting alongside a randomised trial package. In oncology and rare disease in particular, RWE is increasingly the primary evidentiary structure for supplemental approvals. Of the 17 RWE-supported supplemental approvals issued by CDER in the 12 months ending March 2026, nine were in oncology indications where a concurrent RCT would have been logistically or ethically infeasible.

Data Quality Is the New Protocol Design

Sponsors who built their regulatory strategies around trial execution have a capability gap to close. The analytical rigour that regulators now expect from RWE packages is substantial. The EMA's tiered framework requires that sponsors submitting RWE for label expansion demonstrate: a pre-specified analysis plan filed before data extraction, a documented audit trail for data transformations, and an independent validation of the target trial emulation design used to approximate randomisation. Agencies are no longer accepting post-hoc analyses presented as prospective work, and reviewers are trained to identify the difference. The head of regulatory strategy at a major European contract research organisation described a marked increase in FDA information requests specifically targeting data provenance during the 2025 cycle.

For sponsors, this means the investment logic has shifted. Building RWE infrastructure during drug development, rather than after approval, is becoming a prerequisite for competitive label expansion. The 2025 survey found that organisations with pre-established data partnerships covering at least two real-world data source types reduced their label expansion review timelines by a median of 11 months compared with those assembling data retrospectively. At an average daily revenue of $2.3 million for a blockbuster asset, that compression has direct financial consequences.

Three Capability Areas Separating Leaders from Laggards

"The sponsors who are winning on RWE are not winning because they have more data. They are winning because they made design decisions two years before submission that regulators now find credible. That discipline is rare, and it is not something you can retrofit at the end."

Chief Regulatory Officer at a large European pharma group (survey respondent)

Analysis of the 17 CDER RWE-supported supplemental approvals from the past 12 months reveals three structural characteristics shared by nearly all successful packages:

• Pre-specified analysis plans: Every successful submission included an analysis plan filed with the relevant data custodian or registry before data extraction began, providing regulators with an auditable record that the analytical approach was not shaped by results.
• Multi-source data triangulation: Approved packages used a minimum of two independent real-world data sources, typically combining electronic health records with claims data or a disease registry, to corroborate primary findings.
• Target trial emulation methodology: Sponsors explicitly framed their observational analyses as emulations of specific hypothetical trials, naming the eligibility criteria, treatment strategies, and outcome definitions that would have governed a randomised study. Reviewers found this framing materially easier to evaluate.

The broader strategic implication extends beyond individual submissions. Sponsors who build RWE capability as a core function, rather than a project-by-project consulting engagement, are accumulating a compounding advantage. Each approved RWE package establishes precedent with the relevant review division, creates documented relationships with data custodians, and develops internal analytical expertise that shortens the next submission cycle. Organisations treating RWE as a tactical add-on to the clinical development process will find themselves perpetually behind the curve as regulatory expectations continue to tighten. The window to build that infrastructure before it becomes a table-stakes requirement for competitive label maintenance is narrowing.