Abstract 3582: Lung cancer screening in breast cancer survivors: a comparison of the USPSTF guidelines and risk-based criteria

Abstract Introduction: Improved survival rates among breast cancer survivors highlight the need to address long-term health risks, including second primary cancers. Second primary lung cancer (SPLC) is the second most common second primary cancer among breast cancer survivors and is associated with high mortality. However, the optimal approach for identifying high-risk breast cancer survivors to be surveilled for SPLC remains uncertain. This study seeks to fill this gap by comparing the US Preventive Services Task Force (USPSTF) criteria for lung cancer screening vs. a risk-based criteria leveraging a SPLC risk prediction model tailored for breast cancer survivors. Methods: We analyzed 5, 563 patients diagnosed with initial primary breast cancer (IPBC) in 1993-2014 from the PLCO Cancer Screening Trial. SPLC was defined as a primary lung cancer occurring at least six months after the IPBC. We utilized cause-specific Cox regression to build a prediction model for 5-year SPLC risk in breast cancer survivors, informed by existing literature and the Bayesian information criterion for model parsimony. Predictive accuracy was assessed through discrimination using the area under the curve (AUC) and calibration via plots derived from 200 bootstrapped samples. We compared the risk-stratification ability of the USPSTF and risk-based criteria for SPLC screening using the Aalen-Johanssen estimator. Results: In a cohort of women with breast cancer (mean age at diagnosis: 71 years; stages: in situ [20.7%], localized [58.8%], regional/distant [18.5%]), 91 individuals (1.6%) developed SPLC over 108, 783 person-years of follow-up. The SPLC risk model exhibited strong discrimination (AUC: 0.86 [95% CI, 0.79-0.92]) and good calibration in internal validation. Key model features included risk-decreasing factors (hysterectomy, menopausal hormone therapy, BMI, ibuprofen use) and risk-increasing factors (prior cancer, liver comorbidities, smoking-related factors). The stratification by a predicted risk threshold of ≥2% (90th percentile) showed that the observed incidence of SPLC was higher in high vs. low-risk groups (8.7% vs. 0.8%), surpassing the stratification performance of the USPSTF criteria (6.6% vs. 0.7%) when comparing eligible and ineligible subgroups. Conclusions: We developed an SPLC risk model tailored for breast cancer survivors. Although the USPSTF criteria showed reasonable risk stratification in SPLC risk, the risk-based approach provided superior performance. This method provides a promising strategy for identifying high-risk breast cancer survivors for SPLC who may benefit from low dose computed tomography screening. Given the potential value of integrating breast cancer screening with effective lung cancer screening strategies, breast cancer survivors could serve as a model to advance early detection of lung cancer during routine surveillance. Citation Format: Baitong Wang, Marissa Michael, Justin Lee, Yiwey Shieh, Lauren Groner, Leah M. Backus, Allison W. Kurian, Rulla M. Tamimi, Summer S. Han, Eunji Choi. Lung cancer screening in breast cancer survivors: a comparison of the USPSTF guidelines and risk-based criteria [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3582.