Assessment of Surgical Mediastinal sTaging in cN1 Lung canceR

Few reports in the literature evaluated the final pathological stage distribution of patients with resectable and operable non-small cell lung cancer (NSCLC) with clinical stage cN1. These retrospective series demonstrated that patients with computed tomography (CT) based cN1 often had clinically occult mediastinal lymph node metastases (N2/3 disease). Hishida et al. reported that 30% of 143 patients with cN1 were diagnosed N2/3 by mediastinoscopy3. Watanabe et al. reported that 37% of 168 patients with cN1 were diagnosed N2/3 by mediastinoscopy 4. Adding FDG-positron emission tomography (PET) to CT might enable the detection of N2/3 disease among these cN1 patients, but negative PET findings do not necessarily exclude N2/3 disease. Kim et al reported that 19,2 % of 99 patients with cN1, in whom cN2 was ruled out by PET-CT scan, were found to have pathologic N2 disease at pulmonary resection with mediastinal lymph node dissection.5 In conclusion, 20-30% of patients with cN1 nodes on imaging, and normal sized FDG-negative mediastinal lymph nodes on CT and PET have malignant involvement in their mediastinal nodes.

The ACCP guidelines state that invasive preoperative mediastinal staging should be performed in these cN1 patients 6. The updated ESTS guidelines recommend mediastinal staging by echo-endoscopic or mediastinoscopy.1 Non-randomized trials suggested the potential of linear endosonography for mediastinal staging 7-9. However, the patients with cN1 disease form only a minority in these studies. A recently performed prospective ASTER 2 trial (N=100) showed a sensitivity of echo-endoscopic for mediastinal staging of 38% (ITT analysis), while the prevalence of mediastinal nodal disease was 24% (unpublished data Aster 2) 2. The conclusion made by ASTER 2 is that a negative endosonography must be followed by a VAM. However, the investigators consider such double approach not cost-effective in a setting with N2 prevalence <30%. Therefore, it seems reasonable to perform a VAM instead of an endosonography in cN1 patients, which is one of the proposed strategies in the recent ESTS guidelines.1 However, there is no prospective study to date that assessed the sensitivity, NPV and accuracy of VAM in a well-defined group of cN1 patients.

Several publications have demonstrated a lobe-specific mediastinal nodal drainage for upper versus lower lobe NSCLC. Shapiro et al conclude that in early lung cancer, including cN1 disease, lobe-specific mediastinal dissection is warranted 10. However, in this study the only patient with a positive subcarinal node, upper lobe tumour, and negative superior mediastinal nodes had positive N1 nodes. To the investigators knowledge there is no study focussing on mediastinal nodal dissemination patterns in cN1 patients.