Abstract
Background
CT-guided percutaneous lung biopsy (CT-PLB) remains a cornerstone for the diagnosis of pulmonary nodules (PNs), although several alternative techniques, including robotic-assisted bronchoscopy, are increasingly adopted. This study aimed to evaluate CT-PLB in a real-world cohort, focusing on predictors of diagnostic yield and complications.Methods
We retrospectively evaluated patients who underwent CT-PLB between 2020 and 2024. The diagnostic yield was defined according to the recently published ATS/ACCP criteria. Clinical, radiological, and periprocedural variables were assessed as predictors of diagnostic outcome and complication [pneumothorax (PNX) and pulmonary hemorrhage].Results
Six hundred fifty-seven patients were included (mean age 69 ± 11 years; 54% male). The diagnostic yield was 84% with higher yield observed for nodule size (diameter ≥ 10 mm), density (solid), and location (peripheral). Increasing subcutaneous fat thickness was independently associated with a modest but significant reduction in diagnostic yield. PNX was radiologically present in 44.4% of cases; however, only 5.9% required hospitalization, and 4.7% required chest-tube placement. Greater parenchymal depth, trans-fissural needle passages, pleural tags and supine and left-lateral decubitus positions were associated with increased PNX extent, while intraparenchymal lesion and greater chest-wall thickness showed a protective effect. Radiological pulmonary hemorrhage >2 cm occurred in 38% of patients, whereas only 1.5% experienced major bleeding requiring intervention. Greater parenchymal depth and non-solid density were associated with increased hemorrhage severity.Conclusions
In a contemporary diagnostic landscape, CT-PLNB still demonstrates high diagnostic yield and an acceptable safety profile when strict diagnostic criteria are applied. These findings support its continued role as a primary diagnostic approach for PNs.