Introduction
Pulmonary masses are a diagnostic challenge in the field of endoscopic ultrasound (EUS) tissue acquisition (TA), due to both an unclear diagnostic accuracy of other available approaches and the need of enough tissue for molecular profiling. Traditionally, percutaneous CT-scan guided and trans-bronchial tissue acquisition are still the most common techniques. However, centrally or peri-oesophageal lung masses require a different approach, and moreover the development of third generation needles for fine-needle biopsy (EUS-FNB) permitted to acquire through EUS-FNB more tissue from lung masses compared to the transbronchial approach. Our study evaluate the feasibility, diagnostic yield, accuracy and safety of trans-oesophageal EUS-FNB of pulmonary lesions.
Materials and methods
Consecutive patients undergoing EUS-FNB of parenchymal lung masses suspected for malignancy were enrolled between December 2020 and August 2023 in a prospective registry at our tertiary research and referral institute. All of the EUS procedures were performed by experienced endosonographers (defined as at least 50 mediastinal EUS-FNB). Outcomes of interest were feasibility, diagnostic accuracy, diagnostic yield, diagnostic sensibility, diagnostic specificity, specimen adequacy and safety.
Results
We included 42 patients with lung masses undergoing EUS-FNB (43 total procedures). The mean age was 70.0±10.3, and male were 73.8%. About ninety percent of patients had a positive history of smoking, with 55.8% of them being active smoker at diagnosis. Procedures were performed mainly under deep sedation (95.2%) and some of them (50%) using double channel laryngeal mask. The mean lesion size was 51.1±21.6 mm, and patients had mostly a single lesion (n=35, 83.3%). Most of the patients had an advanced stage at diagnosis (stage IIIA, 32.4%; stage IIIB, 16.2%, stage IV, 40.5%), and the most common lung cancer was not-small cell lung carcinoma (NSCLC, 45.2%). The histological diagnosis was adenocarcinoma in 10 patients (23.8%), squamous carcinoma in 1 patient (2.4%), mesothelioma in 2 patients (4.8%), neuroendocrine tumours in 2 patients (4.8%) and metastasis in 1 patients (2.4%). Median follow up of the whole cohort was 49 days. Diagnostic yield was 90.7% (CI 95%, 78.4% to 96.3%), and the diagnostic adequacy rate was 93.02% (CI 95%, 81.4% to 97.6%). Moreover, diagnostic accuracy was 97.4% (CI 95%, 86.8% to 99.55%). Furthermore, diagnostic sensitivity was 97.3% (CI 95%, 86.2% to 99.5%), while diagnostic specificity was 100% (CI 95%, 34.24% to 100%). AEs were reported in two procedure (4.7%).
Conclusions
Trans-oesophageal EUS−FNB is a feasible and safe diagnostic method of tissue sampling for lung masses. Our findings on EUS-TA confirm the high sensitivity, specificity, diagnostic yield, accuracy and specimen adequacy of EUS-FNB for lung masses.

Characteristics of patients (n=42) undergoing EUS-FNB of lung masses

Outcomes of EUS-FNB of lung masses

Gabriele Rancatore

Appearances