OBJECTIVE: Mediastinal staging is of paramount importance prior to surgery for non-small-cell lung cancer (NSCLC) to identify patients with N2-disease. Mediastinoscopy remains the gold standard, and sampling from at least three lymph node stations is generally recommended. It is unknown whether biopsy volume has any influence on the result of conventional cervical mediastinoscopy. In this study, we investigated the influence of biopsy volume and the number of lymph node stations biopsied during mediastinoscopy on the probability of demonstrating N2-disease in patients with NSCLC. METHODS: We identified 678 consecutive patients who underwent mediastinoscopy for staging of NSCLC during an 8-year period (1999-2007), but 111 patients were later excluded from analysis because of misclassification or of missing data. All patient charts and pathology reports of the remaining 567 patients were reviewed retrospectively. Demographics and the number of lymph node stations biopsied were recorded, and the volume of biopsies from each lymph node station was calculated. RESULTS: Multivariate logistic regression analysis demonstrated that larger biopsy volume was significantly associated with increased probability of demonstrating N2-disease (p<0.001). However, sampling from several lymph node stations was significantly associated with a decreased probability of demonstrating N2-disease (p=0.015) and volume was significantly larger per station when fewer stations were sampled (p<0.001). CONCLUSIONS: Biopsy volume from lymph nodes during mediastinoscopy was significantly associated with the probability of demonstrating N2-disease; however, contrary to common belief, sampling from several lymph node stations was not associated with an increased probability of detecting N2-disease. Although purely speculative, these findings may be explained by a perioperative clinical decision by the surgeon: large volumes are secured from macroscopically large and suspicious lymph nodes if detected. Consequently, further dissection and possible complications were avoided.
European Journal of Cardio-thoracic Surgery, 2010, Vol 37, Issue 1, p. 26-9