Zhangheng Huang ¹ Yuexin Tong ² Lujian Zhu ³ Binbin Yang ¹ Kai Chen ² Peilin Dai ^4*

• ¹ First Affiliated Hospital of Wenzhou Medical University, Wenzhou, Zhejiang Province, China
• ² The Second Affiliated Hospital of Wenzhou Medical University, Wenzhou, Zhejiang Province, China
• ³ Jinhua Central Hospital, Jinhua, Zhejiang Province, China
• ⁴ Department of Radiotherapy, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, China

Background: Lung adenocarcinoma patients are often found to have developed bone metastases at the time of initial diagnosis. With the continuous development of technology, we have successfully entered the era of immunotherapy. This study aimed to determine the efficacy of immunotherapy in lung adenocarcinoma patients with bone metastases (LABM) through a multicenter retrospective analysis and to develop a novel tool to identify the population that could benefit most from immunotherapy. Methods: To assess the impact of immunotherapy on LABM in terms of overall survival, we used analytical tools such as Kaplan-Meier analysis, Log-ranch test, and propensity score matching (PSM) method. A predictive model for constructing overall survival was constructed using Cox regression modeling. Based on this, we developed a risk classification system depicting Kaplan-Meier curves for subgroup analysis to determine the optimal beneficiary population for immunotherapy in different risk subgroups. Results: A total of 20073 eligible patients were enrolled in this study, of whom 8010 did not receive immunotherapy, while 12063 patients received immunotherapy. After 1:1 PSM, 15848 patients were successfully coordinated, yielding a balanced cohort. Kaplan-Meier survival curves showed significantly enhanced overall survival (P < 0.001) in patients who received immunotherapy compared to those who did not. The results of Cox regression analyses showed that age, race, sex, primary site, immunotherapy, surgery, chemotherapy, brain metastasis, liver metastasis, lung metastasis, and marital status were independent prognostic factors. The area under the curve for all three cohorts was close to 0.7, indicating that the model was well-discriminating. The calibration curves further proved that the model had a high predictive accuracy. Decision curve analysis demonstrated that the model could achieve a high net clinical benefit. The risk classification system developed based on the model successfully screened the best beneficiary population for immunotherapy. Conclusion: This study provides convincing evidence that immunotherapy provides a significant survival advantage for LABM. Secondly, the clinical tools constructed in this study can help clinicians identify the optimal population to benefit from immunotherapy in LABM, thus enabling precise treatment and avoiding the waste of medical resources and over-treatment of patients.