For decades, the randomized controlled trial (RCT) has reigned supreme as the "gold standard" of medical research. Built on the bedrock of meticulous randomization, rigorous bias reduction, and the pursuit of causal proof, RCTs have delivered the breakthrough therapies that define modern medicine. Yet, as the field of oncology becomes increasingly granular—driven by genomic insights and personalized medicine—a persistent tension has emerged: the rigor that makes clinical trials so powerful often creates a "narrow aperture" that excludes the vast majority of the patient population.
Today, only about 5% of adult patients with cancer are enrolled in clinical trials. This statistic, highlighted by Jeff Allen, president and chief executive officer of Friends of Cancer Research, underscores a critical gap in our collective medical knowledge. When the vast majority of patients are treated outside of the highly controlled environments of trials, clinicians are frequently forced to make critical decisions—regarding treatment sequencing, drug combinations, and dosage—based on data that may not reflect the diverse, messy, and complex realities of the patients sitting in their exam rooms.
The Paradigm Shift: Defining Real-World Data
The medical community is currently in the midst of a data-driven transformation. Real-world data (RWD)—information derived from routine clinical care, electronic health records (EHRs), pharmacy claims, genomics, patient-reported outcomes (PROs), and even wearable technology—is no longer a secondary consideration. It has become a primary tool for understanding disease progression and treatment effectiveness.
Historically, RWD was limited to structured claims data—essentially a digital paper trail of billing codes. Today, it has evolved into a rich, longitudinal tapestry of the patient experience. The integration of artificial intelligence (AI) and large language models (LLMs) has been a watershed moment, allowing researchers to curate and analyze unstructured data—such as physician notes and pathology reports—that were previously trapped in silos.
A Chronology of Evidence-Based Evolution
The transition from viewing RWD as "supplementary" to "pivotal" evidence did not happen overnight. It was the result of a deliberate, decade-long push toward regulatory modernization.
- Pre-2016: RWD was largely used for retrospective observational studies or post-market safety surveillance (pharmacovigilance).
- December 2016: The passage of the 21st Century Cures Act served as a foundational catalyst. The legislation mandated that the FDA evaluate the use of Real-World Evidence (RWE) to support regulatory decisions for new indications of previously approved drugs.
- December 2018: The FDA released its formal RWE framework, providing the industry with the necessary guardrails to ensure that RWD could be used with the same confidence as clinical trial data.
- 2020–Present: The integration of AI-driven curation and the rise of "external control arms" have allowed RWD to play a role in drug development lifecycles, enabling researchers to identify biomarkers and test drug targets before even launching a clinical trial.
Donna Rivera, executive vice president of clinical evidence modernization for Canal Row Advisors and former leader of the Oncology Real-World Evidence Program at the FDA’s Oncology Center of Excellence, notes that this evolution has fundamentally changed how we prove effectiveness. "After the 21st Century Cures Act, the evolution towards evidence for effectiveness has shown that real-world data can serve as pivotal evidence, especially in oncology and rare diseases," Rivera explains.
Supporting Data: Addressing the Representativeness Gap
The most compelling argument for the expanded use of RWD lies in its ability to bridge the gap between "idealized" trial populations and the "diverse" patient population.
Consider the common debate over eligibility criteria in oncology trials. Dr. Ankit Kansagra, a hematologist-oncologist and associate vice president at the Tampa General Hospital (TGH) Cancer Institute, points to the example of hemoglobin levels. Trial designers often set strict cut-offs for hemoglobin, but these limits can exclude a significant segment of patients who might otherwise benefit from a therapy.
"If you lower the hemoglobin threshold from eight to seven-and-a-half grams per deciliter, what does it really do? Does it lead to poor survival in the real world?" asks Dr. Kansagra. "Real-world data has the power to help make those decisions in day-to-day life for clinicians as they design clinical trials."
By utilizing RWD, researchers can simulate the outcomes of different patient cohorts, allowing for more inclusive trial designs that better reflect the clinical reality. However, this relies on a crucial caveat: the representativeness of the data. If a dataset is derived exclusively from high-resource academic medical centers, it will inevitably miss the nuances of community-based oncology practice. Ensuring that RWD captures a diverse spectrum of facilities and patient demographics is essential to its validity.
Official Perspectives: Fitness for Use and Regulatory Rigor
As the reliance on RWD grows, so does the scrutiny regarding its quality. Regulatory bodies, led by the FDA, emphasize that RWD is not a universal panacea. Before asking if the data is "fit for use," researchers must first ask if the clinical context is appropriate.
"Is the clinical context appropriate? Does it align with real-world evidence generation, or is a clinical trial perhaps more appropriate if feasible and ethical?" says Rivera. This "context-first" approach ensures that RWD is used where it adds the most value—such as in rare diseases where patient populations are too small for large-scale trials, or in oncology, where rapid sequencing changes are common.
Once the appropriateness of the study objective is confirmed, the evaluation shifts to two pillars: relevance and reliability.
- Relevance: Does the data contain the necessary variables (exposures, outcomes, covariates) to answer the specific research question?
- Reliability: Is the data accurate, complete, and trustworthy?
A significant hurdle in reliability is the "missingness" of data. Clinical documentation is designed for patient care, not for research. Consequently, information such as biomarker testing status is often inconsistently recorded. To combat this, the industry is increasingly moving toward "data linking"—the process of stitching together EHRs, genomic data, and pharmacy claims to create a comprehensive patient journey.
Implications for the Future: Precision Care
The path forward for RWD is paved with both technological promise and structural challenges. While AI and data-linking are improving the quality of evidence, there remains a persistent lack of standardization. Definitions of "lines of therapy" vary from hospital to hospital, and the uptake of recommended biomarker testing—which should be near 100%—remains stuck between 70% and 80% in many analyses.
However, the tide is turning. Collaborative international initiatives are working to harmonize data guidelines, and new technologies like "ambient scribing"—which uses AI to document clinical encounters in real-time—promise to improve the completeness and structure of RWD at the source.
The ultimate goal is the achievement of precision care: a system where real-time, point-of-care recommendations are generated for an individual’s specific condition, based on the collective experience of millions of patients who came before them.
As Dr. Ming He, senior medical director at Verana Health, observes, the integration of these disparate data streams is not just a research necessity—it is a moral imperative. By bridging the gap between clinical care and clinical research, the medical community is moving toward a "living" model of medicine, where every patient experience informs the next treatment decision.
While the randomized controlled trial will remain a foundational tool, the future of oncology lies in the thoughtful, transparent, and rigorous integration of real-world data. We are moving toward a new, inclusive gold standard—one that prioritizes the unique individual in the context of the broader, real-world population. In doing so, we ensure that the lessons learned from cancer treatment are not limited to the few, but are available to benefit all.
About the Author
