Establishing a predictive model for tumor mutation burden status based on 18F-FDG PET/CT and clinical features of non-small cell lung cancer patients.

Background: Tumor mutation burden (TMB) has emerged as a promising biomarker for immune checkpoint inhibitors (ICI) response, but its detection through whole exome sequencing (WES) is costly and invasive. This study aims to establish a predictive model for TMB using baseline metabolic parameters (MPs) of ¹⁸F-fluorodeoxyglucose (FDG) uptake on positron emission tomography/computed tomography (PET/CT) and clinical features in non-small cell lung cancer (NSCLC) patients, potentially offering a non-invasive and cost-effective method to predict TMB status.

Methods: A total of 223 NSCLC patients with baseline ¹⁸F-FDG PET/CT scans and TMB detection results were retrospectively enrolled from January 2019 to September 2023, and were divided into two groups: TMB-high (≥4 mutations/Mb, 96 patients) and TMB-low (<4 mutations/Mb, 127 patients). Twelve clinical features and five PET parameters were assessed. Univariate analysis was conducted in all patients to reveal the preliminary associations between variables and TMB status. All patients were randomly divided into a training set (n=135) and a validation set (n=88). Feature selection was performed using lasso regression and logistic regression analyses. A predictive model and nomogram were established with the features selected above. Decision curve analysis (DCA) was performed to assess the clinical utility of the developed model.

Results: Two clinical features and two PET parameters were identified through lasso regression and logistic regression analysis including pathology type, cancer antigen 125 (CA125) level, maximum standardized uptake value (SUV_max), and metabolic tumor volume (MTV). The predictive model exhibited an area under the curve (AUC) of 0.822 [95% confidence interval (CI), 0.751-0.894], and internal validation yielded an AUC of 0.822 (95% CI, 0.731-0.912). The model was well-calibrated. The developed nomogram, incorporating the four selected variables, showed promising potential in evaluating TMB status in NSCLC patients.

Conclusions: In this study, a predictive model combining ¹⁸F-FDG PET/CT and clinical features of NSCLC patients effectively distinguished between TMB-high and TMB-low status. The nomogram generated from this model holds significant promise for predicting TMB status, offering valuable insights for clinical decision-making.