Non-invasive nomogram to predict EGFR/TP53 co-mutation for early-stage lung adenocarcinomas manifesting as ground-glass nodules

Background Early-stage lung adenocarcinoma often appears as ground glass nodules (GGN) on CT scans. Due to their heterogeneity, GGNs exhibit diverse molecular profiles. The co-mutation of EGFR/TP53 correlates closely with disease progression, treatment results, and prognosis. However, predictive models for this co-mutation remain underexplored. Thus, this study focused on developing a non-invasive prediction model for EGFR/TP53 co-mutation in early-stage lung adenocarcinoma presenting as GGNs. Methods A retrospective cohort study was carried out on patients diagnosed with early-stage lung adenocarcinoma manifesting as GGNs at West China Hospital between 2010 and 2020. Patients were randomly allocated into training and validation datasets at a ratio of 2:1. LASSO and multivariable logistic regression were utilized to construct the model. A nomogram was subsequently generated, and its predictive accuracy was assessed by calibration, C-index, and decision curve analysis. Results A total of 1827 patients were initially screened, of whom 473 were enrolled in the ultimate analysis. Among them, 292 (61.7%) had EGFR mutations, 66 (14.0%) TP53, and 55 (11.6%) co-mutations. Six potential predictors were finally selected for the nomogram: expectoration, cancer history, nodule diameter, lobulation sign, vascular convergence sign, and TNM stage. The area under the curve (AUC) for the nomogram predicting EGFR/TP53 co-mutation was 0.867 (95% CI: 0.785–0.948) in the training cohort and 0.850 (95% CI: 0.745–0.954) in the validation cohort. Furthermore, calibration and decision curve analysis confirmed its good discrimination ability and clinical utility. Conclusions A novel nomogram model incorporating six easily accessible, non-invasive features was developed and validated for predicting EGFR/TP53 co-mutation in early-stage lung adenocarcinoma with GGNs. The model demonstrated satisfactory discriminative performance and holds promise for clinical application.