SLIDES & TRANSCRIPTS
Tuesday, June 19

PATHOLOGY/CORRELATIVE STUDIES SECTION - PATHOLOGY OF LESIONS DETECTED BY SPIRAL CT: THE CORNELL EXPERIENCE

Madeline Vazquez, MD

In order to give an overview, because we have been talking vaguely about numbers, and I wanted to get a perspective for myself about the numbers, I reviewed 1999 series of fine needle aspirations of the lung. And there were 319 total lung aspirates performed in 1999; 37 were detected on screening CT. I would like to focus on this category of lesions. Interestingly, and offering a wonderful opportunity for me in this field, is that 19 of those, more than 50%, 51.4%, were from the ELCAP series, the Early Lung Cancer Action Program at Cornell. The CT detected lesions comprised then 11.6% of the lung biopsies performed in 1999.So I'm talking only of this small series of lesions, because Dr. Katz so thoroughly reviewed all of the problems that we see in differentiating primary lesions from metastatic lesions, and the wealth of information we can obtain by immunohistochemistry and other modalities we apply to surgical pathology.

TOP

Interestingly also, like the M.D. Anderson experience, our non-diagnostic rate is higher in these CT detected lesions, because they are smaller. In this series, 13.5%. I've looked at the years 1997, 1998, 1999, 2000, and up to the present time, 2001, and we have a non-diagnostic rate of less than 5% in each of those years. So when looking at the screening study cases only, the percent of non-diagnostic cases is higher. However, I do include in the non-diagnostic series those cases which show only benign cells, macrophages only for example, that we consider definitively non-diagnostic. Two of these were less than 1 centimeter, and in six months time increased in size and one proved to be bronchoalveolar carcinoma. The other was benign. I also include any lesion that we call atypical or suspicious for malignancy, because I think we have offered you nothing. We were concerned about the lesion. It was going to be excised. Our question was how should it be managed? And we tell you there are some atypical cells. Although when I say suspicious for malignancy, I am very, very confident that we are dealing with malignancy. There are very rare cells that perhaps there is some discordance with another pathologist, and therefore I include this as non-diagnostic. There was one that was less than 1 centimeter in size, and it proved to be a bronchoalveolar carcinoma. So of the five that were non-diagnostic, three of those were less than 1 centimeter, so the smaller and more indolent lesions obviously are more difficult to aspirate. Also, there is one here that in the present time, I would call definitive for hamartoma. It was a lesion that showed just fibromyxoid stroma. This is definitively called hamartoma presently. At that time we weren't confident, and we just favored that diagnosis. I also include that in this non-diagnostic series. Others would exclude this, and obviously make this number smaller, but I want to be realistic. In other words, we did not help 13.5% of those patients.

TOP

Now benign FNAs that were diagnosed in the CT detected lung lesions comprised 24.3% in 1999. In other words, 9 of the 37 screened lesions. There were 4 total benign lesions that presented as ground glass opacities. Two of those were less than 1 centimeter. There was one that was positive with AFB organisms on the fine needle aspiration material, and it was also less than 1 centimeter. Interestingly, we had two benign neoplasms, hamartomas that were less than 1 centimeter, were diagnosed definitively. The patients have been fine. There has been no enlargement, and we can be very confident in this diagnosis, and it comprises a large proportion of those lesions that are found incidentally on CT. And we loved this case, because it was a tiny, little ectopic liver nodule of the basal portion thought to be within the lung. It was not. It was larger than 1 centimeter, however, but it was definitely diagnosed by fine needle. And there was one reactive lymph node that you will also see the surgical pathology to that case, which was also less than 1 centimeter, and diagnosed definitively by fine needle aspiration.

TOP

Now I'll divide the CT detected lesions that were diagnosed as malignant by FNA into those that were nodules, and also that presented as localized opacities. Of the nodules, there were 14. Interestingly, of these 14, only 5 were less than 1 centimeter. I was very rigid about these numbers, because we keep alluding to the incidence, and really there is very little data on that. But two of these were less than 1 centimeter, and proved to be adenocarcinomas. Interestingly, one was a squamous cell carcinoma, and it was peripheral, and it does happen. There was one that couldn't be classified further, it was a poorly differentiated carcinoma, classified as non-small cell carcinoma. It was 1.2 centimeters, so does not fall into this category of sub-centimeter lesions. And interestingly again, carcinoids were frequently found in this category of lesions. Two out of five, less than 1 centimeter, definitely diagnosed by fine needle aspiration on-site, with no discordance in the final interpretation, with confirmatory immunohistochemistry. The carcinoid lesions both were positive for chromogranin, and for cytokeratin, and for TTF-1, where there is a debate as to whether that is positive in endocrine lesions. Interestingly, one was a small cell carcinoma, but it was 1.3 centimeters in size, and it was a peripheral lesion, so it does occur. So 14 of the CT detected patients who went for screening had malignancies diagnosed by FNA detected as nodules. Five of these were less than 1 centimeter in size.

TOP

But what I find even more interesting are those that present for screening, have a new opacity, nodules are localized opacity, one where we see the markings of the lung, hamartoma beyond the nodule, their density. And 14 of those also similar, same number presented as opacities. And here the lesions, the diagnostic entities are completely different. And fascinatingly, they range from a typical bronchoalveolar proliferation that could even be reactive, to one that is neoplastic, to atypical adenomatous hyperplasia, to bronchoalveolar carcinoma, to adenocarcinoma with prominent bronchoalveolar features, we say favoring bronchoalveolar , because we know we need to look at the entire neoplasm in order to see how much invasion is present. But interestingly all of these presented as ground glass opacities, or localized opacities, partly solid, non-solid nodules. Also interestingly, 50% of those were less than 1 centimeter in size. These two proved to be adenocarcinomas, one purely bronchoalveolar, one with a little microscopic foci of invasion in the Gucci type 2 lesion. They classify as atypical bronchoalveolar proliferation when there are just a few clusters of cells, when the lesion is smaller than 5-8 millimeters by CT, according to Dr. Yankelevitz's measurements, and when the nuclear morphology is minimal, and I'll show you that one does not feel there would be concordance among cytologists in rendering a malignant diagnosis. I feel we are extremely accurate in this. If we had mucinous bronchoalveolar carcinoma, we have no doubt in making the diagnosis in that those cells are also minimally atypical, but the background of mucin should not be there. As we know, in the presence of mucin, we have an adenocarcinoma. We diagnose it as mucinous adenocarcinoma, and when those nuclear features are like those that we will describe, we favor it at the bronchoalveolar subtype. Why is this necessary? Because we are talking about, theoretically, sub-centimeter and this larger lesion was not much larger than 2 centimeters, most of them ranging from 1-2, with an average of about 1.4 centimeters. So we are talking about very close to 1 centimeter lesions. We want to identify those lesions that may be managed differently by the surgeon, or that maybe need to be interpreted differently clinically. The rest of the lung must be looked at very carefully. Multiple slices of the lobe need to be looked at to see if there are other lesions. We need to address the entire environment of the lung a little differently. So that is why we try to classify these lesions in this manner, rather than rare atypical cells. That's very different than the atypical I described before.

TOP

. So in summary, the CT detected malignancies diagnosed by FNA that were less than 1 centimeter in 1999, this comprised 32.4% of all lesions that were detected on screening. More fascinating, 7 were bronchoalveolar carcinomas. Two were invasive adenocarcinomas, but with prominent bronchoalveolar features. Intriguingly, there was a squamous cell carcinoma, and of course the carcinoid tumors.

TOP

This is a ground glass opacity that was aspirated, I'll say localized opacity. It is irregular. There was cystic change.

TOP

And I emphasize once again in I think a much more dynamic photograph, the importance of immediate on-site interpretation. Usually, Dr. Yankelevitz is sitting here, but he has recently cut his hair. And so since I don't want anyone to confuse him with another player, we have included one of our resident.

TOP

And at that time, we immediately assess the cells using the Diff-Quick stain. This Diff-Quick stain, we render a diagnosis of atypical bronchoalveolar proliferation on-site, and leave it at that. Why? We want to save ourselves from marker studies. We want to save them in thin prep for -- all the biomarkers can be done, many on the methanol basis cytotech preparation. We want to be able to look very closely at the nuclear morphology on Papanicolaou stains. We want to look at them in cell blocks. We want to save the material. We've got one tiny lesion. We may get this pass, that's it, one slide. And what I see is a population of cells that is very monomorphic in appearance. There appears to be intranuclear inclusions on this air dried smear. The red blood cells here make a nice measurement comparison. They are 7 microns in size. Sure, they are actually pretty small, but they are very uniform and monotonous. This, to me, is very suggestive of bronchoalveolar carcinoma. Of course we cannot make that diagnosis on cytology. We need to see if there is invasion, but we do know that it's an atypical bronchoalveolar proliferation.

TOP

And on pap stain, what we do see more clearly is the intranuclear inclusions, which become very prominent. There is very little nuclear abnormalities, and this is very different from what we see in other adenocarcinomas, and you will see that. But in this case, I think the proliferation is so extensive, there were multiple sheets of this. And the lesion was only 6 millimeters in size. Now you can imagine on histology in order procure this number of cells and clusters, clearly there has to be a tremendous proliferation of cells. I was confident to call this adenocarcinoma. The cytologic features we discussed, bland, smooth membranes, intranuclear inclusions, minimal nuclear abnormalities, favor bronchoalveolar carcinoma, and it was.

TOP

This is a histologic section, which I leave to Dr. Flieder to discuss at length. He actually has an image of that resected specimen.

TOP

This is a larger lesion, not one of our screening cases, and also not one less than 1 centimeter, because here I think we could say that it's a little bit larger than that. But it is one that Dr. Yankelevitz showed. And it shows a localized opacity, which proves to be BAC.

TOP

But in this case it's a little bit more interesting. It seems that the cells are feathering out, as we see in colon cancer. That emphasizes a cytoplasm, which is not visualized very well here on this colored background. But we see that the cells do have a very clear columnar cytoplasm that we will see on different slides a little bit better. Another prominent feature of these cells, although they are very smoothly marginated, they are similar in size, there is not a lot of nuclear abnormality, but there are very prominent nuclear grooves. I find this a very important finding, and I emphasize it significantly on many of my specimens, especially if there is an inflammatory component. Because we tend to dismiss atypical abnormalities in cells when accompanied by inflammation as being reactive changes. But when you see these grooves, in my experience, I think that this is very commonly a mucinous bronchoalveolar carcinoma. Now that prior picture is not the typical picture for mucinous bronchoalveolar carcinoma, which tends to be quite white, very, very dense, because of the mucinous material.

TOP

But in this case, there was very little mucin in alveolar spaces. And here we see the cells that mimic those that we had seen on the fine needle aspirate. Columnar cells that haven't even lost their polarity, lining the thickened alveolar as single cells in a lipidic growth pattern, but very little mucin in the alveolar spaces. When there is a lot of it, they tend to be solid nodules, not ground glass opacities. So I find that an interesting case.

TOP

Here just this thin, wispy mucinous material, very, very thin, not sufficient to make it a dense nodule. On the high power of the histology, showing those exact features we saw in cytology, cells with grooves, and that feathering look. There was very minimal nuclear abnormalities. In fact, look at this lymphocyte just adjacent to the cell. You see that it's only two times the length in diameter, very pinpoint, but conspicuous nucleoli, irregular chromatin pattern, but the nuclear grooves I find very, very helpful in making the diagnosis.

TOP

One more picture; I'll show it. This is an irregular complex opacity. We have to remember that even when the lesions are big, they may represent this category of lesion. Even if it's big, if you feel it represents this category of lesions, it's very important to point that out cytologically. This was an adenocarcinoma in the scar. The cells was scanty. I called this atypical, suggestive of adenocarcinoma. I felt it to be in the scar, but I'm pointing out is that there may be a problem on frozen section and on resection. And we want to make the surgeon and the pathologist aware of this. And I will add that this case, when resected, was very controversial amongst the pathologists in the international group that we see at the Weill Cornell Center, with a final important diagnosis of adenocarcinoma in the scar. I'm going to leave the forum to my colleague, Dr. Flieder.

TOP