Lung cancer staging methods

I was giving a talk last week and several questions about staging of non–small cell lung cancer came up and so I wanted to address some of these in case there are others with questions. Stage-1 and -2 lung cancer are fairly straightforward and comprise local disease. These may be T1 (<3 cm) or T2 (>3 cm) tumors but do not involve any mediastinal structures and do not cause lobar collapse. A stage-2 lung cancer involves intra-pulmonary, hilar lymph nodes.

Stage-3 lung cancer is complicated and heterogeneous. The tumor can be T1, T2 or T3 and any N is included. All told, there are 13 combinations of Tumor and Nodes that result in a stage-3, locally advanced NSCLC. Even further, treatment decisions may vary depending on whether patients have “multi-station” nodal involvement, meaning more than one N2 node is involved with the cancer. Accurate staging is critical to accurate treatment of patients with NSCLC.

One often confusing feature of lung cancer staging is the nodal stations. I can simplify this greatly by highlighting that the nodes are labeled stations 1 through 14. All double-digit stations are regional (N1) lymph nodes and all single-digit stations are central (N2) nodes. So if the pathology report indicates that a level 7 node (subcarinal) is positive, this is an N2 lymph node. A level 14 (hilar) lymph node is a regional node — N1.

A picture and table of the nodal stations can be found here.

Accurately staging NSCLC requires several steps to exclude the possibility of metastatic disease. This is critical since the treatment for locally advanced disease (chemoradiotherapy) is much more toxic than the treatment for advanced disease and because patients who would benefit from surgery must be determined.

I stage patients in this fashion: CT chest through the adrenals and liver first. After an initial CT scan, if metastatic disease is apparent, a PET scan is usually unnecessary since my treatment decisions will not be affected by the results of a PET. If there is no obvious metastatic disease by CT, then a PET scan and brain imaging (MRI preferable over CT) are performed. If still no evidence of distant disease, the mediastinum should be pathologically staged with mediastinoscopy or endoscopic bronchoscopic ultrasound. Patients with N2 disease are generally candidates for chemoradiation rather than resection so these patients need to be identified accurately with pathology. I would like to highlight that a PET scan is generally not adequate as the only test to stage the mediastinum, because of its relatively poor positive predictive value — in other words its false-positive rate is too high to rely on it solely. PET scan sensitivity and specificity are in the range of 80% and 70% respectively.

Hope that helps.