EP1.13-02 Negative Predictive Value of EBUS-TBNA for Mediastinal Staging of Non-Small Cell Lung Cancer

Mediastinal staging must precede surgery in patients with resectable non-small cell lung cancer (NSCLC). According to current guidelines, minimally invasive techniques, such as endobronchial ultrasound with transbronchial needle aspiration or/and biopsy (EBUS–TBNA/B), represent the preferred first-line approach for mediastinal staging, before performing mediastinoscopy. We herein aimed to evaluate the negative predictive value (NPV) of EBUS-TBNA in a tertiary referral center and correlate NPV with primary tumor features and other clinicopathological variables. We retrospectively studied the medical records of 20 patients with resectable NSCLC, submitted, from January 2017 to January 2018, to EBUS-TBNA in the Department of Interventional Pulmonology of the first Pulmonology Clinic of Sotiria Athens General Hospital; EBUS-TBNA had been performed prior to surgical resection of the primary tumor and surgical lymph node staging in all patients. The EBUS-TBNA results were correlated with surgical staging. Among all lumph node stations sampled, there were 46 with negative EBUS-TBNA results. Post-operatively, 6 EBUS-TBNA negative lymph nodes were re-staged as positive. The remaining 40 EBUS-TBNA negative nodes were true negative, as confirmed by surgical staging. NPV of EBUS-TBNA was 87%, and thus within the range of previously published results. Among cases with negative EBUS-TBNA results, a statistically significant correlation was observed between low NPV (false negative EBUS-TBNA) and T3 tumor size (> 7cm), pre-bronchoscopy N2 disease, presence of necrosis within the primary tumor, and microscopic vascular invasion (p-value <0.05 in all cases). Patients with T3 tumors (>7cm), N2 disease before the performance of EBUS-TBNA, and those with tumor necrosis or microscopic vascular invasion may be at greater risk for false negative EBUS-TBNA results. Given the small sample size of our study and its retrospective study design, it must be emphasized that these findings are highly preliminary and must be confirmed in larger prospective series.