Design and Creation of a Racially Diverse Lung Cancer Registry with Detailed Genomic and Environmental Annotation

Area of Science:

• Oncology
• Genomics
• Epidemiology

Background:

• Increasing proportion of lung cancers in never-smokers harbor EGFR mutations.
• Limited understanding of risk factors and prognostic indicators for EGFR-mutant lung cancers.
• Scarcity of integrated genomic, clinical, and environmental datasets hinders research.

Purpose of the Study:

• To create a comprehensive database for studying EGFR-mutant lung cancers.
• To facilitate research on risk factors, treatment, and outcomes in lung cancer patients.
• To leverage natural language processing for enhanced data extraction.

Main Methods:

• Developed the Meyer Cancer Center Molecularly Enhanced Lung Cancer Database (MCC-MELD).
• Linked lung cancer cases to institutional cancer registry, EHR, and genomic testing results.
• Utilized NLP for extracting unstructured genomic data and smoking history; linked geocoded addresses to area-level data.

Main Results:

• MCC-MELD includes 9,573 lung cancer patients (1988-2024) with diverse demographics.
• Identified 1,092 (11.4%) EGFR-mutant lung cancers; NLP identified 397 additional cases.
• NLP demonstrated high accuracy for EGFR status (97%) and smoking history (90%-98%); 16% of cases were never smokers.

Conclusions:

• MCC-MELD is an NLP-enhanced database integrating clinical, genomic, and environmental data for lung cancer patients in an urban setting.
• This resource supports studies on lung cancer risk factors, treatment, and outcomes stratified by EGFR mutation status.