Real‐World Evidence in Oncology
- Zhifei Zeng
- 1 day ago
- 5 min read
Introduction
In the fields of medical research and drug development, randomized controlled trials (RCTs) have long been regarded as the most robust method for evaluating drug efficacy and safety. However, RCTs often have limitations: they are typically conducted under strictly controlled conditions with relatively homogeneous study populations, which do not fully align with the complex and diverse patient scenarios encountered in the real world. Therefore, relying solely on RCT results may not fully reflect a drug's performance in routine clinical practice.
Against this backdrop, Real-World Evidence (RWE) has gradually taken center stage. RWE originates from Real-World Data (RWD), such as electronic health records, insurance claims databases, patient registries, and information collected by wearable devices. By systematically analyzing these data, researchers and policymakers can gain deeper insights into the efficacy and safety of drugs across different populations and clinical settings, thereby providing more comprehensive references for clinical decision-making and regulatory approvals.
The value of RWE lies not only in supplementing the limitations of RCTs but also in advancing more patient-centered medical practices. It assists physicians in selecting the most appropriate treatment options and accelerates patients' access to novel therapies.
Real-world evidence helps describe the effectiveness of oncology treatments
In oncology research, RCTs provide crucial information on drug efficacy and safety, but their strict inclusion criteria often exclude large patient populations commonly encountered in real-world clinical practice. For instance, many RCTs exclude patients with poor performance status, central nervous system metastases, cardiovascular comorbidities, or HIV infection. Consequently, RCT results may not fully reflect treatment outcomes in complex clinical settings. In other words, the “efficacy” demonstrated in RCTs is often ‘diluted’ in real-world practice, leading to a significant reduction in a drug's “effectiveness” in the real world.
RWE holds significant value in this context. By analyzing data from electronic health records, cancer registries, and insurance claims, RWE can help describe treatment outcomes for patient groups that are underrepresented or entirely excluded in RCTs. For instance, while sorafenib significantly extended survival in advanced hepatocellular carcinoma patients during clinical trials, subsequent RWE analyses revealed that patients receiving treatment in real-world settings experienced far less benefit than trial results indicated. Similarly, RWE studies of docetaxel plus prednisone for prostate cancer demonstrated shorter survival and higher adverse event rates. These findings suggest significant potential for bias when translating RCT results into clinical practice.
However, RWE studies also have methodological limitations. The absence of a randomized design makes comparisons between treatment effects susceptible to selection bias. While statistical methods like propensity scoring and multivariate regression can partially correct for this, they cannot fully replace randomized controlled trials. Therefore, RWE is better suited to supplement rather than replace RCTs, offering unique advantages in evaluating tolerability, long-term safety, different treatment sequences, and patient benefits across regions. For instance, in HER2-positive breast cancer, pivotal trials did not cover pertuzumab as first-line therapy, whereas RWE helps supplement data on different treatment sequences to guide clinical decision-making.
Furthermore, RWE can reveal variations in treatment outcomes and economics across countries or regions, providing policymakers with insights to balance drug accessibility and cost-effectiveness. In summary, RWE not only tests the external generalizability of RCTs but also delivers more actionable evidence for patients and clinicians in real-world settings when evaluating treatment efficacy.
Real‐World Evidence for Rare Molecular Subpopulations
In the traditional new drug development pathway, anticancer drugs typically require demonstrating efficacy in Phase II clinical trials before undergoing randomized Phase III trials comparing them with standard treatments to obtain marketing approval. However, as tumor molecular characteristics continue to be revealed, many driver gene mutations exist only in a very small number of patients. The feasibility of conducting large-scale RCTs in these rare molecular subgroups is severely limited: researchers must screen vast numbers of patients to find a small number of eligible cases, which is not only extremely costly but also often lacks sufficient motivation from pharmaceutical companies.
RWE may play a pivotal role in this context. As molecular tumor testing becomes more widespread, increasing numbers of patients with defined molecular alterations are identified and receive targeted therapies, even when these drugs were not originally designed for that tumor type. For example, HER2-targeted therapies have demonstrated confirmatory evidence in breast and gastric cancers, while also showing activity in colorectal and other tumor types. However, the presence of a molecular alteration does not guarantee efficacy. Consider the BRAF V600E mutation: Vemurafenib demonstrates significant efficacy in melanoma but shows limited benefit in colorectal cancer and other tumors unless combined with other agents. This illustrates how isolated case reports or unsystematic use can overestimate efficacy due to publication bias, potentially leading to ineffective or harmful treatments.
Therefore, the core challenge lies in ensuring patients receive potential benefits while accumulating scientific evidence in the absence of traditional RCTs. Compared to uncontrolled off-label use, a more ideal approach is prospective registry studies. For instance, a proposed “Access to Medicines Program” establishes targeted drug formularies and matches patients based on molecular profiling. While administering treatment, physicians collect efficacy and safety data, and patients consent to contribute their information to a research database. Typically overseen by academic institutions or professional societies to ensure data quality, these projects resemble “academic clinical trials” in form while integrating research with practice. They hold promise for informing future regulatory decisions on new indications.
Within this framework, all stakeholders benefit: patients gain access to previously unattainable treatments, physicians advance cutting-edge therapeutic practices, pharmaceutical companies gather drug activity data in novel settings, and payers avoid covering unvalidated therapies. This demonstrates that for rare molecular subgroups, combining RWE with prospective registry studies may serve as a vital complementary pathway for populations traditionally underserved by conventional RCTs.
Real-World Evidence for Treatment Safety
In cancer treatment, safety and tolerability are no less important than efficacy. When selecting therapies, physicians and patients need to understand not only whether a drug is effective, but also what side effects may occur, their severity, duration, and management strategies.
While clinical trials provide vital safety information, they have limitations. Due to limited patient enrollment and stringent screening criteria, trials often reveal only common adverse reactions, potentially overlooking rare or long-term toxicities. For instance, cardiac toxicity associated with immunotherapy went largely unnoticed in pivotal trials but gained attention after the drug entered clinical use.
This is where RWE holds value. By collecting medication data from real healthcare settings, we gain a more comprehensive understanding of a drug's safety profile across diverse populations, particularly regarding:
Rare side effects: These are more easily detected in large-scale real-world datasets.
Long-term toxicity: Clinical trials have limited follow-up periods, whereas real-world data can supplement information on adverse reactions after prolonged use.
Special populations: For instance, elderly patients or those with comorbidities are often excluded from trials but represent common groups in real-world settings.
Adherence and tolerability: Real-world studies also reveal whether patients maintain long-term treatment compliance, as toxicity, costs, and treatment complexity can all impact adherence.
Of course, RWE faces challenges in safety research. Medical records often lack comprehensive documentation of side effects, especially overlooking patients' subjective experiences. Therefore, specifically designed post-marketing studies or incorporating patient self-reported data can more accurately describe drug safety.
Conclusion
Overall, real-world evidence (RWE) provides valuable insights for clinical practice by revealing the actual outcomes of treatments in routine healthcare settings, particularly among patient populations that are underrepresented or excluded from randomized controlled trials (RCTs). RWE also offers perspectives on treatment efficacy and cost-effectiveness across different geographic regions or economic contexts. However, due to the absence of randomization and strict trial controls, RWE remains limited in assessing the relative efficacy of different treatments, and its data quality may be inferior to that of clinical trials.
For targeted therapies in rare molecular subgroups, systematically planned and designed prospective academic programs often yield more reliable and robust methodological evidence than fragmented, non-standardized clinical use. Although adverse event documentation in real-world settings may be less detailed than in clinical trials, RWE can reveal treatment tolerability across larger, more heterogeneous populations, providing clinicians and patients with more comprehensive safety information. In summary, RWE serves as a vital complement to RCTs, helping optimize cancer treatment decisions and advancing precision medicine.
Sources
Assessed and Endorsed by the MedReport Medical Review Board