Evidence-Based Screening Criteria for Second Primary Lung Cancer Using a Risk Prediction Model for Lung Cancer Survivors

• Activity Number: 463 - Recent Innovations in Method Developments for Risk-Based Screening and Health Policy Modeling for Cancer
• Type: Invited
• Date/Time: Wednesday, August 10, 2022 : 2:00 PM to 3:50 PM
• Sponsor: Section on Risk Analysis
• Author(s): Summer S Han* and Eunji Choi and Sophia Luo and Heather Wakelee
• Companies: Stanford University and Stanford University School of Medicine and Stanford University and Stanford University
• Keywords: risk prediciton; lung cancer; competing risk regression; cancer screening; validation

Abstract

With the widespread adoption of CT screening, lung cancer (LC) survivors are rapidly increasing in number. Although mounting evidence suggests LC survivors have a high risk of second primary lung cancer (SPLC), there are no consensus screening guidelines for SPLC. In this study, we develop and validate a statistical model for predicting a 10-year SPLC risk among LC survivors, utilizing large population-based data that include smoking history, initial tumor characteristics, medical history, and treatment data from MEC and NLST. We apply a set of competing risk regression and regularization methods for feature selection using time-to-event data in the presence of competing risk of death. Using real-world patient outcome data from Mayo Clinic, we evaluate the performance of risk-based screening for SPLC versus the existing clinical guidelines by NCCN. We consider a hypothetical semi-annual screening program implemented over five years after initial diagnosis and estimate overall sensitivity, specificity, and NNS to detect one SPLC case. We show that an evidence-based screening strategy for SPLC using a prediction model provides increased efficiency versus the existing guidelines.