Development of a predictive model to predict postoperative bone metastasis in pathological I-II non-small cell lung cancer.

Abstract

Background: Bone is a common metastatic site in postoperative metastasis, but related risk factors for early-stage non-small cell lung cancer (NSCLC) remain insufficiently investigated. Thus, the study aimed to identify risk factors for postoperative bone metastasis in early-stage NSCLC and construct a nomogram to identify high-risk individuals.

Methods: Between January 2015 and January 2021, we included patients with resected stage I-II NSCLC at the Department of Thoracic Surgery, West China Hospital. Univariable and multivariable Cox regression analyses were used to identify related risk factors. Additionally, we developed a visual nomogram to forecast the likelihood of bone metastasis. Evaluation of the model involved metrics such as the area under the curve (AUC), C-index, and calibration curves. To ensure reliability, internal validation was performed through bootstrap resampling.

Results: Our analyses included 2,106 eligible patients, with 54 (2.56%) developing bone metastasis. Multivariable Cox analyses showed that tumor nodules with solid component, higher pT stage, higher pN stage, and histologic subtypes especially solid/micropapillary predominant types were considered as independent risk factors of bone metastasis. In the training set, the developed model demonstrated AUCs of 0.807, 0.769, and 0.761 for 1-, 3-, and 5-year follow-ups, respectively. The C-index, derived from 1,000 bootstrap resampling, showed values of 0.820, 0.793, and 0.777 for 1-, 3-, and 5-year follow-ups. The calibration curve showed that the model was well calibrated.

Conclusions: The predictive model is proven to be valuable in estimating the probability of bone metastasis in early-stage NSCLC following surgery. Leveraging four easy-to-acquire clinical parameters, this model effectively identifies high-risk patients and enables individualized surveillance strategies for better patient care.