Hello there, my name is Dr. Padram Gourami and I'm going to speak to you today about the impact of next generation sequencing on the diagnosis of melanocytic neoplasms. I'm going to specifically focus on spitz neoplasms. Some disclosures, I've been a consultant to Kessel Biosciences, that won't be relevant to this talk, okay? So again, we're going to be focusing on spitz neoplasms and we have to start from this study, very critical study, back from around 2014 where they were able to sequence successfully a series of spitz neoplasms and in this paper, which was published in Nature Communications, they were able to identify some chimeric fusion proteins as being the main initiating event of most spitz neoplasms and this is very important because identifying the drivers of spitz as being these chimeric fusion genes really opened the door and for the first time allowed for the development of a genomic definition of what a spitz neoplasm is, okay? So how do these chimeric fusions work? So for example, in the case of a ROS1 fusion spitz neoplasm, the ROS1 gene has the kinase domain at the 3 prime end of the gene and there's a regulatory unit at the 5 prime end and really the rearrangement of this gene or the translocation event in this case could be with a number of different partners. The partner is not so critical. What's really most important is that the regulatory domain that normally sits at that 5 prime end is lost and it's the loss of that regulatory unit which results in the constitutive activation of the kinase domain. So now with this translocation, the kinase domain of ROS is connected to this gene PWWP2A and because the regulatory unit is gone, the kinase domain is constitutively activated causing the development of this spitz neoplasm here. Now in this original study, they had 140 lesions and in just over 50%, they were able to identify a chimeric fusion gene involving one of these 5 genes that I've listed here. ROS1, ALK, NTRK, BRAP, or RET. Now importantly, they identified these chimeric fusions in spitz things they called spitz nevi, spitz tumors, and spitz melanoma. So it doesn't tell you if the lesion is necessarily malignant or not. It's really telling you the initiating event and that the lesion does in fact belong to the spitz family. And so lesions such as BRAF activating point mutations, which is very different from what I'm describing here, which in this case, in the spitz neoplasms, we're describing a BRAF fusion event, so a genomic rearrangement, is very different than the activating point mutations. And in fact, the WHO now recommends lesions with activating point mutations in BRAF or NRAS should actually be excluded from the category of spitz. So again, the really significant piece of this study is that now this allowed for the opportunity for a combined morphologic molecular definition of what a spitz is, and in the current WHO, we recommend that you have the appropriate morphology with a spitz associated genomic event, which would be one of these chimeric fusions, or the other alternative would be an activating point mutation in HRAS. And when you have either one of those kind of genomic events with the right morphology, then you can classify the lesion as really belonging to the spitz family. So now that we had this new kind of paradigm where we're able to have this genomic definition of what a spitz is, we thought this would be the optimal opportunity for us to really study behavior of spitz neoplasms in a new way that has never been done before. Because all studies before this that have really looked at the behavior of spitz neoplasms really did not look at lesions that were kind of genomically defined. So they're basically classified as spitz purely on morphology. And we know that morphology has limitations, right? So we don't know if, in fact, all those lesions were truly carrying spitz-associated genomic fusions or an HRAS mutation, or if some of them could have actually been BRAF or NRAS activating point mutations or other non-spitz-associated genomic changes. So what was really novel about this opportunity was we could study the behavior of spitz in lesions that were specifically genomically classified as spitz. So the way we started this study, we started with morphology, though. We identified lesions between 2000 and 2018 that had been classified as a spitz, either a spitz tumor or spitz melanoma. And we also included a third category, which was melanomas with spitzoid features. And I'll kind of go into more detail about that. And then, of course, we required for inclusion that we needed enough tissue to sequence these lesions, both for DNA and for RNA. And then we wanted follow-up in as many cases as possible. So this would be a characteristic lesion of one of the cases we would include as a spitz tumor. And for simplicity's sake, we actually combined spitz nevi and spitz tumor cases into one category and just called them spitz tumors. But this would be an example of a spitz tumor. We have nice epidermal hyperplasia on the surface. It's kind of this exophytic, pedunculated lesion. You can tell there is architectural maturation because the nests are large and expanse all up top, but much smaller at the bottom. If we go to higher magnification there, we can see that the cells have vesicular nuclei and kind of a somewhat abundant eosinophilic to slightly amphiphilic cytoplasm. So very characteristic spitzoid morphology there, right? So this would be an example of one of our cases in the spitz tumor group. This was an example of what we classified as one of our spitz melanoma cases. And this lesion is ulcerated, has very expanse all nesting, poor maturation as their nests are still very expanse all at the bottom. At higher power, we can see that ulceration more clearly. We can see the expanse all nesting. If we look at this large nest near the base of the lesion, in addition to the expanse all group, we see clustered mitotic activity. Here are two very easily identifiable mitoses in this nest. And in general, mitoses were easily identifiable in this lesion. And so this would be an example of what we classified as spitz melanoma. Again, this initial classification is based purely on morphology. Perhaps in some cases, fish and some IHC were utilized to help distinguish whether we thought it was spitz melanoma versus spitz tumor. And then the third category was melanoma with spitzoid features, which is basically essentially a conventional melanoma. But maybe there are some morphologic features, such as a little bit of epidermal hyperplasia, perhaps some clefting. Maybe there's a mixture of epithelial and spindle cytology that makes the lesion look somewhat spitzoid, enough so that perhaps a referring pathologist considered the possibility of a spitz tumor. But in the end, we thought the morphology didn't quite fit for a spitz. So for example, in this case, even though there's a little bit of epidermal hyperplasia and clefting, the cell morphology wasn't quite right at the cytologic level. Rather than abundant eosinophilic cytoplasm, the cells had more of a vacillated to pigmented type of cytoplasm. And not all the cells really had the typical vesicular nuclear appearance of spitz. And so we thought this lesion really was just a conventional melanoma with some spitzoid features. So we had 80 cases in the spitz tumor category, 24 in the spitz melanoma category, and 22 of the melanomas with spitzoid features. Now what this table is showing is that we sequenced these cases, both DNA and RNA-seq, so we could identify. And the RNA-seq was whole transcriptome, so we could identify any potential fusions. And out of the 80 spitz tumor cases, we found a genomic rearrangement in 73%. That's actually substantially higher than the initial study describing fusions in these cases. And so overall, it seems our morphologic classification was not too bad. Most of the cases that we classified as being in the spitz family or spitz tumor did, in fact, have spitz-associated genomic alterations, with ALK being the most common, then MAP3K8, then NTRK, ROS, BRAF, NTRK3, RET, MET, et cetera. Now we did have a few cases, as you'll see at the bottom here, a total of three that we classified as spitz tumors that actually had activating point mutations in BRAF. And so really, if we wanted to be very strict on our classification and follow the latest recommendations of the WHO, those cases should not have been classified as spitz tumors. But that just kind of speaks to the imperfections and limitations of pure morphologic classification. But overall, I think we did a pretty good job. Because as I said, the vast majority, 73%, did have some type of genomic fusion or genomic rearrangement involving a spitz-associated gene. Now you'll notice that MAP3K8 was quite common in our cohort. That was kind of newly described in this study and wasn't mentioned in the original study identifying fusions. And maybe that perhaps could be that some of the MAP3K8 cases are not necessarily a fusion, but actually just a truncation. So the kinase domain is just activated because there's a truncation effect where the regulatory unit is removed from the gene. And that results in this protein with constitutive activation of the kinase domain. So that's our spitz tumor category. Now let's skip to the melanoma with spizoid features category. And you'll notice we actually did quite well there too. 82% had activating point mutations in either BRAF, NRAS, or NF1. That's what we should expect in that category, because they're kind of just conventional melanomas. So we were quite accurate there as well. Now let's go to this category of spitz melanoma. And what you'll notice here is that about half of them did have some type of genomic rearrangement with a spitz-associated fusion gene. But there was another 30 or so percent that had activating point mutations in either BRAF or NRAS. And what that says is the way we were using the term spitz melanoma and classifying lesions in that category was actually quite heterogeneous. And it was really a mixed category of lesions, where we're including some lesions that do truly belong to the spitz family, and then some lesions that did not, these ones with the activating point mutations in BRAF or NRAS. And I've looked at data from other institutions which has been similar. So I think the way that we and many other places have been using the term spitz melanoma, until the time of this study at least, was not in a very specific and fine-tuned manner, and really was kind of almost like a hodgepodge of different types of cases which had spitzoid features, but we thought reached a point of malignancy. And I think that really needs refinement. And I think we're going to, like studies such as this, are kind of lay the groundwork for allowing that, OK? So let's go to the next slide here. So again, this is a heat map. It just kind of demonstrates that same thing, that basically in the spitz tumor category, we had a lot of fusion events, which is the purple. And the genes involved are listed on the side. In the melanomas with spitzoid features, we had a lot of activating point mutations, which is the green. And in the spitz melanomas, we had a mixture of both, so kind of heterogeneous group, right? Now, we had follow-up. In these cases that were studied, we had follow-up on 101 cases. And seven of these had some type of recurrence. Two resulted in death, four with distant metastasis, and three just were local recurrences, OK? Now, let's look at the cases that actually had these recurrences, OK? These are the seven cases, 1, 2, 3, 4, 5, 6, 7, listed here. And what you'll notice is, independent of the morphologic diagnosis that we rendered, whether we called it spitz tumor, spitz melanoma, or melanoma with spitzoid features, a common underlying factor is that almost all of them had activating point mutations in BRAF or NRAS with the exception of one case, which had a MAP3K fusion. Now, this case, which we morphologically diagnosed as a spitz tumor, was on the hand of an eight-year-old kid. And the lesion was re-excised. And then right outside of the excision scar, he popped up with another lesion with very identical morphology. And that was then also re-excised. And it's been like about three or four years since then. And the kid has had no subsequent recurrences. I kind of suspect that that case was more an example of a admanent presentation of spitz rather than a true recurrence. All the cases that had any distant metastasis, and the two cases that died, again, were BRAF or NRAS. And TERT promoter mutation was also very common, as you see in this column. All the cases had activating events in TERT promoter, except this one MAP3K8 case, which, again, only had that local recurrence, which, again, I think was just an admanent presentation. Now, let's go on to look at the survival curves, or relapse-free survival curves. This is the relapse-free survival curve of the three categories, spitz tumor in red, spitz melanoma in green, and melanoma with spitoid features in blue. And there's no significant p-value in describing these different Kaplan-Meier curves. But what if we take advantage of our new ability to have genomic definitions of these lesions? And in doing so, we eliminate the cases that had the BRAF activating point mutations from the category of spitz tumor. And we're very strict about that, and only include the fusion cases. Once we do that, what we see is that the Kaplan-Meier curve for the spitz tumor and the spitz melanoma cases kind of start to align. And they become very distinct from the melanoma with spitoid features group. And that makes sense with the table that I just showed you, which showed that really pretty much most of the actual true recurrences, distant METs, deaths, et cetera, were actually in the melanoma with spitoid features group or the BRAF-NRAS mutated cases. And what we find is now there is a statistically significant difference. The p-value becomes significant in comparing the Kaplan-Meier curves of the spitz tumor or spitz melanoma with the melanoma with spitoid features group. And the spitz tumor and spitz melanoma group actually are very similar, but they're distinct from the melanoma with spitoid features group. So what this says to me that most of the cases that are true genomic spitz, meaning they have either the right morphology with a chimeric fusion gene that's associated with spitz, such as ROS, RET, NTRK, MAP3K8, et cetera, or an HRAS activating point mutation, whether we call them spitz tumor or spitz melanoma, the vast majority of those cases actually have an excellent prognosis. And the cases that actually do poorly are more the ones that we call melanoma with spitoid features or that can mimic spitz. Now, the next case I'm going to cover was nice and wedge-shaped and symmetric spitz. As you see here on higher power, that has epithelioid and spindle morphology. The low power really couldn't ask for any better spitzoid cytology. But yet, this case was not a true spitz because it had that BRAF V600E point mutation. This was, in fact, a nevoid melanoma mimicking a spitz. And so now that we have these genomic tools and these genomic definitions, I think this gives us our best opportunity now to really separate out these spitz mimickers and true spitz lesions. And the point is also not to say that all cases with genomic fusions or true spitz genomic defining events are benign. That's not the point because there's clearly examples in the literature of malignant spitz neoplasms or cases of MAP3K8 or BRAF fusions that are truly malignant and have terp promoter mutations. The point is more that no matter what setting you're practicing in, whether it's a small community private practice or a big consultation practice, et cetera, the cases that we need to be most weary of and are most likely going to cause problems for all of us are really not those truly malignant spitz because they're actually quite rare. They happen, but they're quite rare. The cases that are common and that are going to cause a lot of problems are the spitz mimickers. Those BRAF and NRAS mutated cases, such as the one I just showed you, that mimics a spitz. So those are the cases we have to be most cognizant about. And those are the cases that our newer genomic tools will give us the opportunity to separate out from true spitz. And I'm going to end the data plasm. This data is published in this paper in the JID. And with that, I'm going to end my. presentation, and I thank you all for your attention.

next-generation sequencing

melanocytic neoplasms

spitz neoplasms

chimeric fusion proteins

genomic rearrangements

prognoses