DR. WESTRA: Well, unlike more central bronchogenic squamous cell cancers, which progress through a very well defined sequence of metaplasia, dysplasia, carcinoma in situ, very little has actually been known about the early progression of glandular neoplasms of the lung, and in particular, that precursor lesion from which these fully malignant peripheral lung adenocarcinomas arise, have very successfully alluded detection.

Today I'm going to propose that the precursor lesion has indeed been identified, and moreover, this lesion can both be appreciated at the light microscopic level, as well as more fully characterized at the molecular genetic level.

Now in the interest of time, I'm going to pick up an historic pursuit of that elusive lesion in 1988. And this of course was the year that Dr. Roberta Miller first made some very interesting analogies between adenocarcinomas of the colon - which at that time it was very well known that it progressed through a progressive series of morphologic stages - and then adenocarcinoma of the lung.

Now what Dr. Miller recognized is that up to 10% of those lungs resected for fully malignant lung adenocarcinomas also harbored additional small foci, what we now refer to as atypical adenomatous hyperplasia, or AAH for short. Grossly, those are seen as these very small, subtle nodules, generally ranging between 1 up to 7 millimeters.

Histologically, these tend to have very irregular borders. Lesional cells tend to be cuboidal to low columnar, demonstrating varying degrees of nuclear atypia. And then line intact, albeit somewhat thickened septa walls.

Now Dr. Miller reasoned that because of the strong association with fully malignant lung adenocarcinomas, and because of this characteristic cytologic atypia, that these lesions represented foci of dysplasia, possessing some potential to progress to fully malignant adenocarcinomas.

Now initially, I've got to tell you that many were unwilling to accept this proposal, mostly because these lesions morphologically could be so easily dismissed as some other process. For example, many believed that these simply represented reactive pneumocyte hyperplasia. And reactive pneumocyte hyperplasia can be seen in response to a variety of types of lung injury.

And here, the cytologic atypia may even exceed that seen in certain well differentiated adenocarcinomas.

Others believed that AAH really just represented a creeping, a local extension from an adjacent adenocarcinoma. And here for example, is a non-mucinous bronchoalveolar carcinoma. And in this particular tumor, as you look at the periphery, the tumor cells become less columnar, and their cytologic atypia seems to just fade away. So at the periphery of these lesions, they become morphologically indistinguishable from atypical adenomatous hyperplasia.

And then finally, a third possibility, which was very difficult to totally dismiss on morphologic grounds was a possibility that these may in fact represent small intraparenchymal metastasis.

So as you can see, at the morphologic level, it becomes very difficult to discern the true nature of atypical alveolar hyperplasia. And moreover, these lesions, given their very small size, are very recalcitrant to evaluation at the genetic level. But more recently, with the refinements we have seen in PCR amplification and tissue microdissection techniques, it has really opened the door now to the exploration of these lesions at a fundamental genetic level.

So now we certainly have the technology, but we needed some rational strategy for evaluating these lesions at the genetic level, specifically addressing their neoplastic nature. And one of the best targets we could think of looking at was the K-ras oncogene, for a number of reasons.

For one, mutations of the K-ras oncogene are fairly common lung adenocarcinomas, where they are present in about 30-40% of these tumors.

Secondly, they are fairly specific for glandular neoplasms of the lung. They are much less frequently encountered in squamous cell cancers.

These mutations are fairly easy to identify, even in very small tissue samples. These mutations involve very simple base substitutions at very consistent sites.

And then finally, we had at least indirect evidence to suggest that activating mutations of the K-ras oncogene probably occur very early on during the development and progression of these glandular neoplasms.

So we evaluated 41 AAHs. These were obtained from 28 patients who had undergone lung resection for primary lung cancers, the vast majority of these being adenocarcinomas. And then we found that 40% of these almost actually harbored mutations of the K-ras oncogene. So we would argue that the presence of a clonal proliferation of cells carrying a point mutation to a known tumor oncogene provides very compelling evidence pointing to the neoplastic nature of these lesions.

But even if it points to the neoplastic nature of these lesions, it still doesn't necessarily confirm them as being very early independent glandular neoplasms. And keep in mind what we have already said. Some people believed that these represented very advanced forms of tumor spread, either by creeping extension, or by metastatic implantation.

But if this were in fact the case, then we would expect that these lesions would consistently harbor a same pattern of base alternations seen in their asynchronous, fully malignant paired lung cancers. But that, in most cases, was not what was happening.

And here is just an example from a single patient, with actually two separate adenocarcinomas, and then multiple atypical adenomatous hyperplasias, demonstrating the pattern of base alternations in codon the K-ras oncogene. Again, just emphasizing the divergent nature of these base alterations, suggesting not only are these neoplasms, but they are early independent neoplasms, not metastatic implants.

Now that we were fairly confident that we had identified one of the earliest morphologic stages of glandular neoplasia, we now could use that as a model, a target for exploring the genetic progression of these lesions, now beginning to separate out early from late genetic events. And one of the mysteries here was the timing of P53 alterations. Inactivation of the P53 tumor suppressor gene occurs in about half of the malignant lung adenocarcinomas. But it really wasn't known when the timing of this alteration occurred.

So we evaluated about 40 AAHs for P53 alterations, initially using an immunohistochemical approach. And here is one of those lesions, again, demonstrating a strong nuclear immunoreactivity.

And here is actually the sequencing gel in that AAH, demonstrating this point mutation in exon 7.

However, these P53 alterations were only noted in about 10% of these AAHs. So based on this finding, we were willing to accept the fact that P53 alterations probably are not a common event in early glandular neoplasia.

Now about the same time, a group in Japan was actually subclassifying these lesions on the basis of certain specific morphologic features. They used the term low grade for those lesions demonstrating very low cell density, small cell size, minimal pleomorphism. High grade lesions were those with increased cell density, now to the point of cell crowding, larger cell size, more appreciable pleomorphism, and now some thickening of these septal walls.

An AAH-like carcinoma was a term they used for a lesion in morphologic transition. A transition between high grade AAH, and a small well-differentiated adenocarcinoma. This lesion is characterized by further increase in cell density, even to the point of nuclear stratification, more significant pleomorphism, and now fuller thickening and fibrosis of these septal walls.

Keep in mind that even though this lesion does demonstrate more significant cytologic and architectural atypia, we are still dealing with lesions less than 5 millimeters generally, without invasion, or distortion of lung parenchyma.

Now we have found that the subclassification of AAH based on these morphologic criteria actually had a profound impact on P53 alteration. P53 staining was noted in none of those low grade lesions; 10% of the high grade lesions. But in 50% of these lesions in transition of the AAH-like carcinoma. Which seemed to suggest to us that even though P53 alterations may not be common in early glandular neoplasia, it nonetheless may represent a very critical step in triggering benign to malignant epithelial growth.

Just by way of summary then is this current model for the progression of glandular neoplasia. Most patients clinically present way at this end of the spectrum, as biologically and clinically advanced tumors, with little hope for cure.

Under the microscope now, we believe we can recognize the earliest morphologic stages of glandular neoplasia as is embodied by atypical alveolar hyperplasia. But let's face it, these still remain incidental, microscopic findings, usually noted in lungs resected for fully malignant carcinomas. At least in the past, these lesions have been virtually invisible to more traditional screening mechanisms. But now we are beginning to understand some of the underlying genetic alterations driving this progression forward. We talked about activating mutations of K-ras, we talked about inactivating mutations of the P53 tumor suppressor genes. But much more is know about other important molecular genetic events, again driving this process forward.
Thank you.

TOP