SLIDES & TRANSCRIPTS
Tuesday, June 19

SURGERY SECTION - SENTINEL LYMPH NODE BIOPSY IN LUNG CANCER AND ITS POTENTIAL APPLICATION TO CONSIDERATIONS OF RESECTION OF SCREENED NSCLC

Michael Liptay, MD

But be that as it may, sentinel lymph nodes really have taken off in the treatment of melanoma and breast cancer. And the primary use of that technique in these two tumors is really to limit potentially morbidity and non-therapeutic formal nodal dissection if the sentinel node itself is negative. I'm putting lung in the category of the present data on colon cancer in that with colon cancer, the nodes are removed routinely with the mesentery. Likewise, in lung cancer surgery, at least in my opinion, the morbidity is really of the thoracotomy and the parenchymal resection, and not from the lymph node dissection. So I don't think we are going to spare morbidity really to all that great an extent by limiting the mediastinum node dissections. But what we will be able to do is to help our pathology colleagues, and direct them, with their more sensitive techniques, such as immunostaining, RT-PCR, and the like, to the more likely nodes that will harbor occult micrometastases or other prognostic information.

TOP

Just quickly, we don't have to belabor the fact that lung cancer is the number one cause of cancer-related mortality, and lymph node involvement is really the strongest predictor of survival in localized tumors, the ones that surgeons routinely deal with. Nonetheless, up to 40% of patients reported by our pathologists as node negative by H&E staining do relapse within two years of resection. And it's our hypothesis at least that some of this can be accounted for by undetected occult metastases in the lymph nodes at the time of resection.

TOP

Some of the questions that I had as I was preparing this talk, and I was really left to do some soul searching, do lymph node micrometastases in lung cancer really matter? Is a node dissection therapeutic, prognostic, or neither? Skip metastases, which I will just discuss briefly, are they relevant? Do they behave more like N1 disease or N2 disease? And of course we have had a lot of discussion about whether lobectomy is required for these small tumors or not.

TOP

Well, regarding micrometastases, some work out of Dr. Paslik's group in Germany looked back at 117 patients that they resected, and they went retrospectively and repeated immunostaining for a cytokeratin marker on patients that were all called node negative. They found 25 patients that were actually positive for micrometastatic disease, and they plotted the survival after following them all for five years. They basically found that these patients with micrometastatic disease detected retrospectively pretty much had a survival curve superimposable for stage II disease. Likewise, the patients that were truly node negative had a far better survival.

TOP

This is a fairly familiar map, certainly to most of my surgical colleagues, but for those of you who aren't necessarily too familiar with the anatomy of lung cancer, would that it was so simple that a peripheral lung tumor decided to spread to the closest node, and make its way up the chain in an orderly fashion. Then certainly sentinel node technique would have far less appeal, and I wouldn't be up here talking to you today. But I will let the secret out that this is unfortunately not usually the pattern, and I will try to show you some data on that.

TOP

. Skip metastases refer to patients that have mediastinal lymph node involvement, or positive N2 nodes, with negative lobar or hilar nodes. The incidence is reported up to 20-30%, in the patients that do ultimately have positive N2 nodes, and in this case are classified as stage IIIA.
Skip metastases are really not relevant to the sentinel node paradigm, because if the sentinel node is in the mediastinum, that's the first site of drainage, whether it is N1 or N2 by our classification. That's really not germane.

TOP

Dr. Yoshino from Japan did look at 110 patients with stage IIIA positive N2 disease. And he found that there were 33 of these patients that had the skip pattern, and no positive nodes in the lobe itself or in the hilum. These patients had about a threefold better survival, closer to the survival of stage II than those with the garden variety of N1 and N2 positive pattern.

TOP

So certainly skip metastases do seem to have a different prognosis. This is some more data from Dr. Takizawa and his group in Japan, just describing small tumors less than 2 centimeters. He performed 157 lobectomies, and he found 17% of those patients had lymph node metastases in these very small T1 tumors.
Interesting to me, over two-thirds of those patients grossly at the time of surgery, when the surgeon was doing his node dissection, the nodes were not grossly apparent to him. Also, almost half of these nodes or 40% of the patients that had lymph node metastases had metastases to only a single node, and this perked my interest regarding the sentinel node technology. Likewise, they found nodal metastases, segmental lobar, interlobar, mediastinal nodes, all over the place.

TOP

We found sentinel node mapping in lung cancer has been useful, in order to direct pathologic examination, to alter stage, and to detect micrometastatic disease. We certainly are able to detect a pattern where N2 nodes might be the first site of drainage. And one potential which we haven't exploited yet is if it can be shown to be a reliable test, it has the potential to either limit or expand a nodal dissection, depending on the status of the sentinel node.

TOP

I just want to go over some data from our series, the last time we reviewed things back in the end of March. I'm reporting on 100 consecutive patients. We took all comers. Again, we aren't at the stage the people are with breast cancer. If they have hilar adenopathy on CT scan, even if they received chemo and/or radiation therapy up front, we are still trying to see if the technique works. We have 95 patients that ultimately had non-small cell lung cancer with the sentinel node procedure.

TOP

I'm just going to go over a couple of photos of how we perform the technique. We have actually made some modifications. I initially started with 2 millicuries of technetium, and then my fingers started to turn red, and other kinds of radiation necrosis, so I decided to decrease the dose to 0.25 millicuries, and really what it did is it cut down on the background of radiation of the tumor. We injected in a four quadrant pattern, and then we proceed with our routine dissection and resection. One thing we do pay attention to is to leave the peribronchial tissues for last, because that's really where the majority of the lymphatics reside.

TOP

After a period of at least 10 minutes - we have narrowed it down to that - we are able to do intrathoracic sentinel node reading with a hand-held Geiger counter.

TOP

And one of the things that we have also done is we also dissect the lymph nodes off of the lobar specimen, and take ex vivo readings, just to make sure that our results aren't being colored by shine through effect from the tumor. Because the tumor certainly remains the hottest radioactivity in the chest. Once we separate the tumor off of the nodes, we get a true reading.

TOP

Our results are as follows. We had successful migration of the radioisotope in 78 out of the 91 patients. In 69 out of these 78 patients, we accurately identified a sentinel node. And what that means conversely is that 9 patients had a negative histologic sentinel node, with positive disease elsewhere. That is why they were not thought to be accurate sentinel node identifications. The sentinel node itself was positive in 21 out of the 78 patients. It was the only node positive in 9 out of the 21 or 42%. And we detected the sentinel node as being positive only by micrometastatic disease when we used immunohistochemistry for cytokeratin, or serial sections at about 10 micron slices for 33% of our sentinel nodes.

TOP

And detecting the micrometastases with immunohistochemistry or serial sectioning, we upstaged four stage IA patients to IIA, two stage IB to IIB, and one stage IA patient to a IIIA.

TOP

We also found 20.5% of our patients had skip metastases sentinel node pattern with mediastinal nodes as their first site of nodal drainage. Six patients had level 7 nodes or subcarinal, 5 had paratracheal, and 5 had AP window nodes.

TOP

This is the figure that I think is the most important slide I'm going to show. I went through our series, and we have 17 right upper lobe tumors where we did a right upper lobectomy and mediastinal node dissection, 17 of them with a T1 lesion of less than 3 centimeters in size, the size varying from 3 millimeters to 3 centimeters. The sentinel node location is as follows: 3 in the paratracheal region, 7 in the tracheal bronchial angle, 2 in the segmental region, and five in the interlobar region.

TOP

So it is certainly something that we have convinced ourselves at least that it's feasible and safe to do. It doesn't appear to prolong the operative resection. It seems to be relatively - or at least relatively, we say highly - accurate. And I think it is going to allow our pathologists to provide more precise nodal staging in our patients.

TOP

The technical issues that we have to deal with as surgeons is that a VATS technique is in development. We have to have some help from industry in developing special probes. I personally have some trouble figuring out how I'm going to be able to do an adequate nodal sampling, much less a sentinel node procedure with a wedge resection. The background radioactivity from the nearby tumor is something that we are still grappling with, because the accurate interpretation of our radioactivity often requires ex vivo readings. And I think at this point until we get a multicenter validation, we still require mediastinal node dissection to validate our conclusions.

TOP

So we are in the validation phase now. We have got a multicenter setting hopefully set up. We are going to keep working on modifying the technique for VATS, and hopefully come with at the end of all of this, a standardization of our procedure.

TOP

Then I would say we hopefully move on to the application phase, where we can begin to answer the question whether a mediastinal node dissection was necessary at all times. We can hopefully maintain accurate lymph node evaluation, while also limiting resection for these small tumors. We are also going to help our pathologic colleagues in ultra staging these lymph nodes, and getting a true reading of the actual stage, and not be including stage IA with tumors that are actually more advanced. Hopefully, we may also get some information and influence a new staging system. Thanks for your attention.

TOP