Related Experiment Video
Updated: Jun 13, 2026
Construction of An Orthotopic Xenograft Model of Non-Small Cell Lung Cancer Mimicking Disease Progression and Predicting Drug Activities
Published on: May 10, 2024
Real-World Emulation of Landmark Lung Cancer Trials: A Registry-Based Reconstruction of KN189, KN407, IMpower133, and
Irina Surovtsova1, Wilfried E E Eberhardt2, Dorothea E Meschke2
1Clinical State Registry Baden-Württemberg GmbH, Baden-Württemberg Cancer Registry (BWCR), 70191 Stuttgart, Germany.
Abstract:
Background: Checkpoint inhibitors (CPIs) improve the survival in advanced NSCLC and ES-SCLC, but reproducibility of randomized trial results in routine practice remains unclear. Target trial emulation enables approximation of randomized trial designs using observational data while reducing common biases. Methods: Using the Baden-Württemberg Cancer Registry, we emulated four pivotal trials: KEYNOTE-189 (eKN189), KEYNOTE-407 (eKN407), IMpower133 (eIMP133), and PACIFIC (ePACIFIC). The eligibility criteria, treatment strategies, and outcomes were aligned with the original trials as closely as possible within registry constraints. Time-zero was defined as the treatment initiation for systemic therapy cohorts (eKN189, eKN407, eIMP133) and as completion of chemoradiotherapy for ePACIFIC. The treatment groups were balanced using propensity-score-based inverse probability weighting. The outcomes included the overall survival (OS) and the objective response rate (ORR). Results: A total of 4015 patients were included (eKN189: 1762; eKN407: 467; eIMP133: 1190; ePACIFIC: 595). After weighting, the baseline covariates were well balanced. ICI-based regimens were associated with an improved survival: eKN189: HR-0.59 (95% CI: 0.52-0.66), 5-year OS-22.7% vs. 4.7%; eKN407: HR-0.69 (95% CI: 0.54-0.89), 5-year OS-11.8% vs. 6.6%; eIMP133: HR-0.7 (95% CI: 0.62-0.79), 5-year OS-12.8% vs. 5.3%; ePACIFIC: HR-0.51 (95% CI: 0.39-0.66), 5-year OS-47.3% vs. 23.4%. The OS curves demonstrated long-term plateaus across CPI-containing arms. The ORRs were consistently higher with CPI-containing regimens (eKN189: 41.6% vs. 29.6%; eKN407: 44.8% vs. 31.5%; eIMP133: 51.7% vs. 43.1%; ePACIFIC: 56.6% vs. 47.1%). Conclusions: This registry-based target trial emulation reproduced the survival patterns observed in four landmark lung cancer RCTs within trial-like real-world populations, supporting the external validity of pivotal immunotherapy trials while reinforcing the role of high-quality registries as a powerful complement to randomized trials in oncology.
Insights
Checkpoint inhibitors (CPIs) improve survival in advanced lung cancer, and this study confirmed their efficacy in real-world data. Registry-based emulation of clinical trials supports the validity of immunotherapy findings in routine practice.
Area of Science:
- Oncology
- Clinical Trials
- Real-World Evidence
Background:
- Checkpoint inhibitors (CPIs) have improved survival in advanced non-small cell lung cancer (NSCLC) and extensive-stage small cell lung cancer (ES-SCLC).
- Reproducibility of randomized controlled trial (RCT) results in routine clinical practice remains uncertain.
- Target trial emulation using real-world data offers a method to approximate RCT designs and mitigate biases.
Purpose of the Study:
- To emulate four pivotal randomized trials of CPIs in advanced NSCLC and ES-SCLC using real-world data from a cancer registry.
- To assess the consistency of CPI efficacy observed in RCTs within a real-world patient population.
- To evaluate the external validity of landmark immunotherapy trials in lung cancer.
Main Methods:
- Emulation of four key trials (KEYNOTE-189, KEYNOTE-407, IMpower133, PACIFIC) using the Baden-Württemberg Cancer Registry.
- Alignment of eligibility criteria, treatment strategies, and outcomes with original RCTs within registry constraints.
- Propensity-score-based inverse probability weighting was used to balance baseline covariates between treatment groups.
Main Results:
- A total of 4015 patients were included in the emulated trials.
- CPI-based regimens demonstrated significantly improved overall survival (OS) across all emulated trials compared to control groups.
- Objective response rates (ORRs) were consistently higher in CPI-containing arms, with long-term survival plateaus observed.
Conclusions:
- Registry-based target trial emulation successfully reproduced survival patterns from landmark lung cancer RCTs.
- These findings support the external validity of pivotal immunotherapy trials in lung cancer.
- High-quality cancer registries serve as a valuable complement to RCTs in oncology research.
