How real-world evidence supports oncology drug development

Executive summary

• Clinical evidence expectations are rising across regulators, HTA bodies, and payers
• Oncology and rare disease submissions now depend on multimodal, geographically diverse real-world data
• Europe is central, but many evidence packages must also support decisions in APAC and the Americas
• High-stakes programs need early evidence design, defensible comparators, and traceable data provenance
• The strongest teams identify and resolve evidence gaps before submission and reimbursement review begin

Rising evidence expectations in oncology drug development

Real-world evidence oncology programs face rising expectations as regulators, health technology assessment bodies, and payers publish clearer standards for evidence quality, transparency, and methodological rigor.

In Europe, the European Medicines Agency created DARWIN EU to generate real-world evidence from healthcare databases across the European Union. This initiative signals that access to data alone is no longer enough.¹

Under the EU HTA Regulation, sponsors operate in a more coordinated joint clinical assessment environment. This approach increases scrutiny of comparators, assumptions, and evidence transferability across markets.²

NICE makes the same shift explicit in reimbursement settings. Its real-world evidence framework asks whether data provenance, relevance, quality, transparency, analytical methods, and uncertainty support the decision at hand—not simply whether the data exists. The FDA framework takes a similar approach by emphasizing fit-for-purpose data and evidence generation for regulatory decisions.⁴

In other words, organizations must move beyond supportive data and produce decision-grade evidence.

Why oncology and rare disease face the most evidence pressure

Oncology and rare disease are not the only therapeutic areas facing these challenges. However, they provide some of the clearest examples because evidence constraints are often more stringent. For cancer therapies, sponsors may need to support decisions in biomarker-defined populations, single-arm settings, or rapidly evolving treatment landscapes. In these situations, randomized comparisons can be difficult to conduct or may become outdated quickly. In rare diseases, small patient populations, fragmented care pathways, and clinical heterogeneity often make evidence packages harder to build and defend.

Common evidence challenges

The stakes are high. A weak comparator can undermine a submission. Missing endpoint logic can damage credibility. Unexplained country variations can reduce the value of a treatment pathway analysis. Reviewers and payers do not reject these studies because they lack ambition. Instead, they test whether the evidence package can withstand scrutiny.

As the FDA notes, regulatory-grade real-world evidence requires fit-for-purpose data and rigorous methods. In oncology, this requirement extends beyond exposure and survival data alone. Guidance on external control arms shows that comparator selection, baseline imbalance, endpoint definition, and timing bias can materially affect study conclusions.^4,5

In rare disease research, experts emphasize early planning, cohort aggregation, and careful management of uncertainty when sample sizes remain small.

How multimodal data support oncology evidence generation

When teams talk about submission-ready real-world evidence, they typically mean a dataset that can answer multiple questions under regulatory, HTA, and payer scrutiny. This level of evidence usually requires more than one data type.

Why data integration matters

Clinical records may show treatment patterns, while pathology confirms diagnosis. Genomic data can identify the relevant patient subgroup. Imaging may support response assessment. Clinician notes can help identify progression, relapse, or line of therapy with the precision needed to defend the analysis.

This is one reason oncology and rare disease submissions now depend on multimodal data. Researchers often find critical variables across multiple databases and record systems. To answer the evidence question, they must link those data sources in a credible and transparent way.

The need for geographically diverse data is rising for a different reason. Standards of care, testing patterns, coding practices, and access pathways vary across countries and regions. A comparator that appears valid in one market may face challenges in another. Likewise, pathway analyses that ignore geography can overlook differences that matter to regulators, HTA bodies, and payers.

However, not every evidence package needs a global design from the start. Teams should determine early whether the decision requires multicountry relevance. They should also confirm that the underlying data architecture can support that goal without introducing bias.

That does not mean every evidence package needs to be global from the start. It means teams should know early on whether the decision they are supporting will require multicountry relevance, and whether the data architecture can support that without creating new bias.

Europe’s role in global oncology evidence planning

Europe plays a central role because it combines policy change, fragmented data access, and cross-border complexity in a single operating environment. The European Health Data Space (EHDS) provides the clearest signal of future policy direction. It establishes a common framework for the use and exchange of electronic health data across the EU.⁶ However, today’s environment remains highly fragmented. Country-specific governance models, varying data availability, and different interpretations of privacy and access continue to create challenges.

Sponsors that can build an evidence strategy across Europe’s fragmented landscape are often better prepared for evidence planning in other regions. Many organizations also need evidence packages that support regulatory, HTA, and market access discussions across APAC and the Americas. As a result, Europe has become more than an important market. It increasingly acts as a design constraint that influences evidence planning from the outset.

Operating discipline as a strategic advantage in oncology drug development

Building a defensible evidence package

Operating discipline only matters when teams connect it to concrete decisions. In practice, that means building the evidence package early and rigorously enough to prevent key assumptions from collapsing during submission or reimbursement review. Teams should define the study question, confirm the required variables are available, test comparator logic, and document data provenance in a way reviewers can easily follow.

High-stakes programs increase the importance of strong study design because the consequences of weak design are greater. Examples include single-arm oncology submissions, rare disease programs with limited patient cohorts, label expansions in small biomarker populations, and evidence packages linked to reimbursement negotiations. In these situations, uncertainty can directly affect access timelines and pricing. Small design flaws can quickly become strategic problems. Teams should identify risks early and build mitigation plans into the evidence strategy.

NICE is explicit that teams should evaluate data source suitability and study design against the decision problem from the start. Regulatory and observational research literature also emphasize this point, with repeated focus on fit-for-purpose data, transparent methods, and clear study planning.^3,4

Simply stated, practical discipline means testing whether an evidence package is truly defensible before it comes under high-stakes regulatory, reimbursement, or market access scrutiny.

Why the strongest real-world evidence strategies start early

Addressing evidence gaps before review

The strongest evidence strategies start early because teams rarely solve the most difficult problems at the end of a study. Instead, they address them while there is still time to refine the protocol, test cohort logic, confirm endpoint traceability, and identify data gaps. Once submission timelines tighten, opportunities to correct weak assumptions shrink quickly.

This is consistent with formal guidance. NICE encourages early assessment of whether a source and design are suitable for the intended use.³ The FDA framework similarly emphasizes evaluating relevance and reliability against the specific research question. In rare disease research, early protocol refinement and thoughtful cohort aggregation can reduce uncertainty before it becomes a major review challenge.

For leaders, the implication is clear. Access to real-world data is no longer the primary challenge. The real question is whether an organization can transform that data into decision-grade evidence that withstands scrutiny from regulators, HTA bodies, and payers. Achieving that goal requires earlier planning, alignment between the evidence plan and decision context, and the methodological discipline to defend comparators, endpoints, provenance, and geographic relevance. In short, evidence readiness has become a strategic advantage.

Practical guidance for pharma leaders

The oncology evidence readiness model

Organizations can pressure-test an oncology evidence strategy by assessing readiness across six dimensions:

1. Decision clarity: Define the regulatory, HTA, reimbursement, or market access decision the evidence package must support
2. Multimodal data availability: Confirm that critical clinical, genomic, pathology, imaging, outcomes, and treatment variables can be accessed and linked
3. Comparator defensibility: Test whether external controls, treatment pathways, and standards of care are credible for the intended decision context
4. Endpoint traceability: Validate that progression, response, survival, line of therapy, and other decision-critical endpoints can be identified consistently
5. Geographic relevance: Assess whether country-level differences in care pathways, testing, coding, and access could affect interpretation
6. Provenance and governance: Document where the data came from, how it was transformed, and how privacy, security, and data custodian control are maintained

When these dimensions are addressed early, evidence teams can identify gaps before they become submission, reimbursement, or market access risks.

Questions leaders should ask

For a pharma leader, there are five questions that really matter:

1. What decision is this package meant to support? Approval, HTA review, reimbursement, and portfolio decisions do not require the same evidence design.
2. Can the team trace the critical variables? Diagnosis, biomarker status, progression, line of therapy, and mortality often sit in different places and need different validation approaches.
3. Will the comparator remain credible across markets? External controls and treatment pathways can break down quickly if local standards of care are ignored.
4. Can the team explain data provenance clearly? Reviewers increasingly expect to understand where the data came from, how it was transformed, and why it is fit for purpose.
5. Where is the uncertainty? Small cohorts, underrepresented populations, and cross-country variations should be surfaced early, not buried until the end stages of the study.

The bigger shift is that evidence expectations are no longer defined by access to real-world data alone, but by the ability to produce evidence that is credible, transferable, and fit for increasingly coordinated decision-making. In oncology and rare disease, where small cohorts, external controls, biomarker-defined populations, and fast-changing standards of care are common, that shift turns evidence design into a strategic capability rather than a downstream analytics exercise.

Europe is now a design constraint, not simply an important market. Joint clinical assessments under the EU HTA Regulation and the implementation of EHDS send a clear signal: evidence plans must anticipate cross-country scrutiny earlier in the process.^2,6

The real competitive advantage, then, is not data volume. It is evidence readiness: the ability to assemble multimodal, geographically relevant, transparently sourced data into a package that can support regulatory, reimbursement, and access decisions across markets.

References

1. European Medicines Agency. Data Analysis and Real World Interrogation Network (DARWIN EU). European Medicines Agency.
2. European Commission. Joint Clinical Assessments. Public Health: Health Technology Assessment.
3. National Institute for Health and Care Excellence. NICE Real‑World Evidence Framework. NICE, June 23, 2022.
4. U.S. Food and Drug Administration. Real‑World Evidence. FDA, updated June 3, 2026; U.S. Food and Drug Administration. Framework for FDA’s Real‑World Evidence Program. FDA, December 2018.
5. U.S. Food and Drug Administration. Considerations for the Design and Conduct of Externally Controlled Trials for Drug and Biological Products: Draft Guidance for Industry. FDA, February 2023.
6. European Union. Regulation (EU) 2025/327 of the European Parliament and of the Council of February 11, 2025, on the European Health Data Space. Official Journal of the European Union, March 5, 2025; European Commission. European Health Data Space Regulation. Public Health.