Hello, I'm Kirk Frey. Chief of the Division on Nuclear Medicine and Molecular Imaging at the University of Michigan Hospitals, and the lecture we're about begin is the second in a series of four on the use of FDG-PET in the imaging non-small cell lung cancer. This lecture will consider the imaging and interpretation of lymph nodes in the thorax in patients with non-small cell lung cancer. These are my disclosures. I don't believe any of these will have a significant impact on the materials we discus. We begin the lecture with a question to keep in mind, as we go through the material. The question to address is the least reliable characteristic of metastatic lymph node involvement is a, the size of the node. B, the anatomic location of the node. C, the metabolic intensity in the node. Or d, the shape of the node. So hopefully, we'll address this question and you'll be able to select the best response by the end our presentation. The learning objectives for this lecture are these. Participants will achieve understanding of thoracic lymph node anatomy relative to sites of primary malignancy in the lung. Participants will obtain understanding of diagnostic performance of FDG-PET in nodal staging and of the relationship of nodal size, and metabolic intensity in FDG-PET interpretation. So, let us begin. At first, we need to understand the schema and atomic classification of the thoracic lymph nodes. At this schema, we'll be presented I am sure in other lectures and modules focus on anatomic imaging of the thorax and perhaps also on surgical approaches. Lymph nodes in the thorax relevant to lung cancer are divided primarily into two groups. The first of those, which are mediastinal. Numbered one through nine here on the right panel of these set of images. Double-digit lymph nodes are those associated with pulmonary visceral locations and include those that are hilar, interlobar and lobar in location. In the TNM staging system for thoracic malignancies, the nodal stage or N stage is defined as indicated here. Patients are classified as N0 when they have no evidence of metastatic disease in regional, that is the thoracic lymph node. N1 stage is when metastases are present in ipsilateral hilar nodes only. N2 stage is when metastases occur to ipsilateral mediastinal and, or subcarinal lymph nodes in the mediastinum. Patients are classified as nodal stage N3 when metastasis are seen in contralateral mediastinal stations or in hilar nodes or to scalene, or supracalvicular nodes. And in this case, the higher nodes involve are contralateral to the primary lesions. Knowledge of the most likely pattern of nodal metastatic disease in primary lung cancer is an important background in the interpretation in imaging and staging. This comes from a very interesting paper published in the International Journal of Radiation Oncology in approximately 2006. Summarizing in this table, the probability of lymph node involvement has surgical staging in patients with lung tumors that are primarily located In these lung lobes. So right upper lobe, right middle lobe, right lower lobe in these columns. Left upper, left lower lobe and then overall, left lung or right lung in these last two columns. These rows indicate the nodal station according to our numbering system and those which have both left and right sided laterality have two entries, one for the nodal station on the left and one on the right and the numbers in each of these cells represent the percentage probability of having metastatic disease in that lymph node. You can see this is a very powerful up priori structure regarding the likelihood of metastatic disease in any one thoracic lymph node station, which is perhaps best exemplified by looking here at the local long node status. So that is the 10L and 10R stations being the hilar nodes and 11L, 11R the lobar nodes within the thorax. And you can see, for instance, here that in right sided primary tumors, the likelihood of having a left hilar lymph nodes be involved by the tumor is zero. The same is essentially true for the contralateral possibility of right higher involvement by a left of lung cancer. This kind of schema is important to keep in mind when addressing the image results and the likelihood of metastatic disease seen in thoracic lymph nodes with FDG-PET. Now, what is the overall performance of FDG-PET scanning in staging in thoracic lymph nodes in patients with non-small cell lung cancer? The answer is that FDG-PET has a reasonable specificity, but is poorly sensitive compared to surgical pathology. This paper from the European Journal of Cardiothoracic Surgery in 2009 is typical of most patients addressing this in the published literature. And here, we studied 51 lung cancer patients. FDG-PET positive was defined as a lymph node with an SUV of 3 or greater. And with that criterion, the sensitivity for identifying lymph nodes as positive that were later surgically proven was only 40%. Whereas the specificity was much more acceptable at 85% and subsequent papers including this one that additionally summarized five or six prior studies are all fundamentally quite similar in that the sensitivity of identifying involved lymph nodes with PET imaging is unacceptably low to stand alone as the fundamentally, and predominant approach to rely on for nodal staging. I'll show now some examples of lymph nodes. This is a patient with a right lung lesion, which you can see on the whole body representation in the MIP images. But if you look carefully at the rest of the thorax, you see that there are bilateral prominent hilar lymph nodes and even some in the middle mediastinum. All of which have a very similar overall metabolic intensity and on the transaxial images, you can see that the pulmonary nodule in question here is FDG would be considered a core positive primary lesion. In this patient, nodal staging revealed that the symmetrical appearing hilar and mediastinal lymph nodes were involved by histoplasmosis. A very common fungal infection particularly in the Midwest in the Ohio River Valley. Characteristics of this scan that allow one to suspect, if not conclude that these lymph nodes are not related to this patient's cancer relate to the bilateral symmetry both of nodal involvement and of metabolic intensity. This is quite typical of what we see in hilar nodes that are not involved by a metastatic disease from a localized lung cancer. And here's another example where we see some lung bronchial abnormalities, but even more strikingly widespread involvement in an irregular fashion of a bilateral hilar middle mediastinal and perhaps even supraclavicular lymph nodes. And I would point out as well that there is a retroperitoneal collection of lymph nodes here in the abdomen, as well. This is a patient who has biopsy proven sarcoidosis. So, a none malignant granulomatous disorder relatively less common than infectious inflammatory changes in the thorax. But nevertheless, an important mimic, if you will of potential lymph node involvement in patients with lung cancer. And then finally, another mimic, which is easier to identify when you look at the transaxial and coronal images in greater detail, but you can see that there are apparently focal areas of increased FDG activity present in both supraclavicular fossa regions as well as in several paraspinal regions in the thorax. When one looks at these in the transaxial imaging, you can see that these areas that metabolic intensity are co-registered with areas of very low density in the matching CT scan. This represents metabolically activated brown fat. It is a thermogenic response to cold exposure of the patient prior to conducting the PET imaging, then this can be avoided by instructing patients to avoid cold exposure. Keep the rooms where patients are a place for uptake of the radio tracer at a relatively warm temperature and using appropriate warmed blankets, and other features to keep the patient from becoming chilled during the FDG uptake period. So, let's return to the question for this lecture. Again, the question to address is what it the least reliable characteristic of metastatic lymph node involvement and that's with FDG-PET? The possible answers are a, the size of the lymph node. B, the anatomic location of the lymph node. C, the metabolic intensity in the lymph node. And d, the shape of the lymph node and here's our recommended best answer at the question. So first, the size of the node does not necessarily dictate whether it's involves by metastatic disease or not. The larger the node, the more likely the metabolic activity is to reflect the average pathology within that node, but metastatic disease through a normal or even subnormal size lymph node can be detected with PET. The anatomic location of the node is very important and we've shown that foreknowledge of the anatomic pattern of lymph nodes spread can be important in accurate interpretation of the FDG-PET scan. Answer d, the shape of the lymph node is also not a reliable indicator of this involvement by metastatic disease. So the best answer c, the metabolic intensity in the node and we have seen lymph nodes that are metabolically very intense that reflect benign disease. And conversely, it is possible that a normal metabolic lymph node may be infiltrated, but not overwhelmed by cancer. Hence, the use of surgical pathology and staging for the most accurate nodal metastatic staging in lung cancer. So, the take home points from this lecture we hope are that PET has insufficient sensitivity to replace mediastinoscopy and the use of microscopic pathology to complete nodal staging in patients with primary non-small cell lung cancer. The effective lymph node size does not have a great impact on our ability to image this with PET and large lymph nodes may be reactive in nature or may alternatively be involved by metastatic disease. The effect of the primary tumor metabolic lymph rate, we didn't explore in great detail, but the general structure that we employ is to assume that lymph node metastatic deposits will have a similar metabolic signature or appearance to that of the primary tumor and this worth keeping in mind. So a very intense primary tumor with symmetrical and low level bilateral mediastinal, and hilar lymph nodes is not likely to represent widespread nodal metastatic disease. And again, probably the most important aspect one needs in advance knowledge of metastatic patterns of disease in order to optimally interpret thoracic PET imaging in this context. So, thank you for your attention to the topics in this lecture. The second in a series of four on the use of FDG-PET imaging in non-small cell lung cancer.
