Predictive factors for non-diagnostic results of computed tomography–guided percutaneous core needle biopsy of the lung: focus on computed tomography findings

PURPOSE: Computed tomography (CT)-guided percutaneous core needle biopsy (PCNB) of pulmonary lesions is widely used for tissue diagnosis, but non-diagnostic results remain common and can delay diagnosis, prompting repeat biopsy or prolonged follow-up. This study aimed to identify clinical, CT, and procedural predictors of non-diagnostic PCNB results and to develop a scoring system for risk stratification. METHODS: This retrospective study included consecutive CT-guided PCNBs performed for pulmonary lesions at a single tertiary center between January 2019 and March 2021, comprising 358 PCNBs in the training cohort and 126 in the validation cohort. Clinical, radiologic, and procedural variables were compared between diagnostic and non-diagnostic groups. Independent predictors were identified using multivariate logistic regression with Firth's penalized likelihood method to reduce small-sample bias, and a weighted scoring system was developed. Discriminative performance was evaluated using the area under the receiver operating characteristic curve (AUC). RESULTS: Non-diagnostic rates were 20.4% (73/358) and 22.2% (28/126) in the training and validation cohorts, respectively; 34.7% of non-diagnostic results were ultimately confirmed as malignant. Five independent predictors of non-diagnostic results were identified: lesion size ≤ 1.5 cm [odds ratio (OR), 2.434; 95% confidence interval (CI), 1.111-5.204], patchy morphology (OR, 2.423; 95% CI, 1.210-4.794), atelectasis adjacent to the target lesion (OR, 2.221; 95% CI, 1.096-4.426), absence of pleural invasion (OR, 8.418; 95% CI, 2.084-77.273), and inaccurate needle tip positioning (OR, 3.669; 95% CI, 1.513-8.907). The scoring system demonstrated modest discriminative performance in the training (AUC, 0.731; bootstrap optimism-corrected AUC, 0.692) and validation cohorts (AUC, 0.769), with a stepwise increase in non-diagnostic probability across four score-based risk groups. CONCLUSION: Lesion size ≤ 1.5 cm, patchy morphology, atelectasis adjacent to the target lesion, absence of pleural invasion, and inaccurate needle tip positioning were independent predictors of non-diagnostic results after CT-guided PCNB. The derived scoring system showed consistent performance in a temporal validation cohort. CLINICAL SIGNIFICANCE: The derived scoring system, based on readily assessable CT and procedural factors, may enable pre-procedural risk stratification, facilitate individualized biopsy planning, optimize sampling strategy, and improve diagnostic yield.