Probably the most contentious areas of lung cancer management is stage IIIA NSCLC, with N2 nodal involvement, the nodes outside of the lungs, toward the middle of the chest but on the same side as the main tumor. One of the key issues is that the staging is the same whether there's a single microscopically involved lymph node or multiple enlarged lymph nodes in a few areas of the mediastinum (mid-chest, between the lungs). But the outcomes of these groups of patients is very different, so it may be worth thinking about them a little differently. As shown here, from a retrospective review of just over 700 patients in France who had N2 mediastinal nodes involved (abstract here), outcomes were much better for the subset who had a single lymph node area involved (called a nodal station), and no clinically enlarged nodes (meaning that on a CT scan they didn't appear abnormally big, defined as more than a centimeter):

(Click to enlarge)

The curves show that the patients with a single (L1 for one level instead of L2 for more than one, in the legend above) non-enlarged (m for microscopic instead of c for clinically involved) nodal area involved have a long-term survival in the range of 35%, while the outcome for patients with visibly enlarged mediastinal nodes and/or more than one level/station involved isn't as favorable, although there are still long-term survivors. But this retrospective series is limited because it pools together people who had differing rigors of staging, some receiving chemo after and some not, and otherwise just a very heterogeneous population. That's somewhat helpful for teasing apart signals within that broad range, but it helps to look at patients treated somewhat uniformly.

We just got some of that information from thoracic surgeons at the University of Alabama, Birmingham, a very respected program led by Dr. Robert Cerfolio. They just published results from their single center experience of 148 patients who had all undergone a chest CT as well as a PET scan and had no findings suggestive of mediastinal involvement. Dr. Cerfolio does a very thorough mediastinal dissection and identified these patients who had been found to have unsuspected mediastinal N2 node involvement, nearly all of whom (93%) subsequently receiving post-operative chemotherapy, and a minority also getting radiation. He reported a 5-year survival of 35% overall, and a more favorable survival among the patients with a single node involved compared with multiple nodes (40% vs. 25%):

Our usual approach if we know someone has stage IIIA NSCLC before surgery is to recommend pre-operative chemotherapy or chemoradiation if we are thinking of pursuing later surgery. But for patients without evidence of nodal disease by careful pre-operative imaging, it's not clear that this is superior to doing surgery and then following with chemo +/- radiation afterward, now that post-operative chemo is a standard approach for higher risk resected patients.

At our institution, we routinely do mediastinoscopies looking for mediastinal node involvement before surgery even in many patients who have completely unremarkably findings in the mediastinum on CT and PET. Uncommonly but not rarely, our surgeon finds a microscopic focus of cancer in 1 or 2 N2 nodes on that mediastinoscopy just before planned surgery. While this has usually led them to stop the procedure, close up, and send them to me for induction (pre-operative, also known as neo-adjuvant) therapy, this work suggests that it's reasonable to proceed with surgery in some cases and pursue the chemo later.

The downside? Some of these patients may not get through it all, such as the 76 year-old otherwise fit man who underwent surgery for an early stage NSCLC tumor who had a single unsuspected node at medistinoscopy, went through surgery, and then saw me for chemo. I recommended and treated him with his first cycle, and even though he did well with it from as far as I could see and measure, he said he wasn't interested in any more treatment. I think he could potentially have done better if he were to have received 3-4 cycles of chemo along with surgery, and maybe he would have been able to do that if we started with chemo and then went on to surgery. But the surgery is the more important and potentially curative treatment here, so it would have been far worse to have had him develop significant problems with pre-operative chemo that kept him from getting surgery, which could cure him even with little or no further therapy.

I'll just close by saying that these concepts all depend on a meticulous per-operative staging workup and a good surgery that removes many lymph nodes. Dr. Cerfolio and the surgeons I work with typically remove a dozen or more nodes, so we can be confident that involvement of no nodes or 1-2 nodes is actually the case. We can presume this if many other nodes have been removed or examined, but if only a handful of nodes were resected, we can't know if there are only a few nodes involved because that's the true extent of disease or because that's all they looked for. Lung surgery is a setting where training, experience, and attention to detail are very critical. When a well-trained lung surgeon finds just a single N2 node involved, those patients can do much better than the general numbers for that stage would suggest.