• Non-smoking high-risk lung cancer population (50-80)

• Family history of lung cancer

• Air pollution, secondhand smoke exposure

• History of TB or COPD

• Post-surgery early-stage (Stage I) lung cancer

Led by Prof. Gee-Chen Chang, the Lung Cancer Diagnosis and Treatment Research Center at Chung Shan Medical University Hospital launched the FORMOSA Trial, a prospective multicenter study to develop a lung cancer risk prediction model for never-smokers. By integrating epidemiological data, air pollution exposure, biomarkers, and imaging characteristics, the study identifies high-risk individuals for precision screening to improve early detection and reduce lung cancer mortality. The model is being prospectively validated through collaboration with eight major medical centers across Taiwan. Our team was also the first in Taiwan to contribute lung cancer data to the National Health Data Alliance platform.

To date, the FORMOSA Trial has enrolled 2,439 participants and developed the FORMOSA Lung Cancer Risk Prediction Model for Taiwan’s never-smoking population, achieving a sensitivity of 90–96% and a negative predictive value of 98–99%.

The Center also participates in the Taiwan Cancer Moonshot Program and collaborates with the U.S. Cancer Moonshot initiative. Using over 14,000 proteogenomic datasets, the team developed a four-marker ELISA panel for early recurrence prediction, supporting personalized treatment and improved patient outcomes.

• Prospective enrollment of over 2,400 never-smokers in one of the world's most comprehensive lung cancer screening cohorts.

• Taiwan's only prospective LDCT screening study integrating environmental exposures, multi-omics biomarkers, and AI radiomics.

• Taiwan's first air pollution–integrated lung cancer risk prediction model for never-smokers.

• LDCT detection rate of 2.54%, exceeding international benchmarks from NLST (0.65%) and NELSON (0.9%).

• Taiwan's only prospective study on air pollution and lung cancer risk.

• Novel ELISA-based blood test panel for early detection of lung cancer recurrence.

• FORMOSA enrollment: 2,436 (target 3,127)

• LDCT detection rate: 2.5% (NLST 0.65%, NELSON 0.9%)

• 2023 CSMUH lung cancer total cases: 871

• Lung Cancer Moonshot: 14,000 proteogenomic entries

• 89 patient plasma samples, 35 candidate proteins validated

• 2024 TFDA LDTs certification achieved

• AI-assisted image interpretation improved diagnostic sensitivity by approximately 39.3%

• Development of a precision screening strategy through a multidimensional risk-scoring system

• Particularly high screening yield among never-smoking women, individuals aged ≥60 years, and those with high air pollution exposure

• Overall lung cancer detection rate of 2.50% (61/2,439) in the prospective cohort

• LDCT radiation dose meets international standards

• AI reduces missed diagnoses

• High sensitivity avoids overtreatment

• Multi-center IRB strict ethics review

• High protein marker accuracy

• International QA laboratory standards

• Provides early screening for high-risk groups

• Reduces unnecessary surgery and treatment

• Improves post-op recurrence monitoring

• Enhances early diagnosis and survival

• Transparent evidence-based pathway

• Cross-institutional collaboration enhances service

• Cell 2020 (IF 64.5): East Asian non-smoking proteogenomics

• JTO 2023 (IF 20.1, Editor's Choice): Family-history LDCT

• Lancet Respir Med 2024 (TALENT trial) lead author

• 2023 IASLC cited our air pollution research

• TW-US Cancer Moonshot 2.0 collaboration

• Invited STC-D panelist

• Provides evidence for national health policy

• Drives AI imaging and precision medicine implementation

• National Core Facility integrates 7 research units

• Builds commercializable early recurrence prediction platform

• Includes non-smokers in international lung cancer screening guidelines