From Passive Sampling to Precision Intervention: A Standardized Radiomics-Driven Workflow for Molecularly Adequate Lung Biopsy.

purposeTo develop and temporally validate a predictive framework for molecular adequacy in computed tomography (CT)-guided transthoracic needle biopsy (TTNB) of lung lesions, integrating procedural variables with radiomic features derived from pre-procedural CT imaging. MATERIALS AND

methodsThis observational derivation-validation study included 670 CT-guided TTNB procedures performed in 633 patients. A retrospective twin-center cohort (522 procedures) was used for model development and internal validation, while a prospective single-center cohort (148 procedures) served as an independent temporal validation set. Molecular adequacy for next-generation sequencing and PD-L1 testing was the primary endpoint. A multivariable model based on procedural predictors was developed and evaluated, followed by an exploratory radiomics sub-study integrating quantitative imaging features. Model performance was assessed using area under the curve (AUC), calibration metrics, and transportability analyses. Procedural complications were classified according to the modified CIRSE classification system.

resultsOverall molecular adequacy for next-generation sequencing was 83.9%. The prospective validation cohort, managed under a standardized procedural framework, demonstrated a significantly higher adequacy rate compared with the retrospective cohort (91.2% vs. 82.6%). The procedural model achieved an AUC of 0.86 in the derivation cohort and 0.84 in the validation cohort. Integration of radiomics further improved discrimination (AUC 0.88), particularly in subsolid lesions. Complication rates were comparable between cohorts, with no increase in major complications.

conclusionMolecular adequacy in CT-guided lung biopsy can be meaningfully improved through procedural standardization and further enhanced by integrating radiomics. The proposed framework provides a foundation for future prospective evaluation of decision-support strategies in interventional radiology.