The yield of endoscopic ultrasound in lung cancer staging: does lymph node size matter?

INTRODUCTION: Transoesophageal endoscopic ultrasound with fine-needle aspiration (EUS-FNA) is a minimally invasive procedure to demonstrate unresectability in lung cancer patients with enlarged malignant mediastinal lymph nodes (MLN). We compared the performance of EUS-FNA to show malignant invasion in enlarged versus small MLN.

METHODS: A single center analysis was performed in lung cancer patients with a suspicion for malignant MLN invasion based on the available imaging. In these patients, EUS-FNA was presumed to impact the diagnostic course since patients underwent surgical-pathologic verification only when EUS-FNA did not demonstrate MLN invasion.

RESULTS: We evaluated 100 lung cancer patients in whom MLN invasion was presumed based on the available imaging. In 75 patients (75%), there was at least one enlarged MLN, whereas in 25 patients (25%), only small MLN were found. The sensitivity and negative predictive value to detect malignancy in enlarged MLN was 96% (95% confidence interval [CI], 87-99) and 67% (95% CI, 29-92), respectively. The sensitivity and negative predictive value of EUS-FNA in small MLN was 93% (95% CI, 66-99) and 92% (95% CI, 61-99), respectively. EUS-FNA prevented a surgical (mediastinal) intervention in 88 and 52% of the patients with enlarged or small MLN, respectively (p < 0.001).

CONCLUSIONS: As the sensitivity to detect malignant MLN invasion is comparably high for both enlarged and small but suspected MLN, clinicians should consider EUS-FNA even in case computed tomography-scan shows no enlarged MLN. The impact of EUS-FNA to avoid surgical mediastinal interventions is greater when enlarged MLN are present. The moderate negative predictive value of EUS-FNA makes surgical-pathologic verification still compulsory, regardless of the size of the MLN.