2026 IASLC CT Screening and Early Detection Symposium-WS02.09 - Talk 5: Future Directions in Risk Stratification and Research

Pdf Summary

The presentation reviews future directions in lung cancer risk stratification and screening research, with a focus on improving how high-risk individuals are identified. It emphasizes that lung cancer epidemiology and screening challenges differ across regions and populations, so a “one-size-fits-all” approach is not appropriate.<br /><br />A major theme is the unmet need for better risk prediction among individuals who have never smoked. The speaker notes that current models may fail to identify enough never-smokers at meaningful risk, highlighting the importance of developing tools that work well beyond traditional smoking-based criteria. Another key issue is that risk models should predict not only lung cancer incidence, but also lung cancer-specific mortality, while minimizing overdiagnosis and unnecessary screening.<br /><br />The talk also stresses the need for population- and region-specific models, since calibration and discrimination can vary across settings. To improve prediction, emerging biomarkers such as blood-based markers and multi-protein panels are being studied. The presentation also mentions the INTEGRAL risk model and related research examining whether biomarkers can improve identification of people at risk for lung cancer and lung cancer death.<br /><br />Artificial intelligence is presented as another promising tool. AI-based approaches, including radiomics and models such as Sybil, are being evaluated in Korean cohorts and among both smokers and never-smokers. Finally, the presentation highlights the importance of real-world implementation research to determine whether these models can be used effectively in clinical practice.<br /><br />Overall, the summary message is that expanding lung cancer screening safely and effectively requires better risk models tailored to specific populations, and future progress will likely come from combining clinical factors, biomarkers, AI, and imaging data.

Asset Subtitle

Yeon Wook Kim

Meta Tag

Concept Lung Cancer Risk Stratification

Concept Chest Screening

Concept Risk Prediction Model

Concept Smoking History

Concept Never-Smoker

Speaker Yeon Wook Kim

Keywords

lung cancer

risk stratification

screening research

never-smokers

risk prediction

lung cancer mortality

biomarkers

INTEGRAL model

artificial intelligence

radiomics

Lung Cancer Risk Stratification

Chest Screening

Risk Prediction Model

Smoking History

Never-Smoker

Yeon Wook Kim