Welcome to the Florida Society of Pathologists Precision Medicine Academy's webinar. Today's topic is Application of Molecular and Genomic Biomarker Testing in Hepatopathology. I am Dr. Marilyn Bui. I served as the immediate past president of the FSP, and I direct the Precision Medicine Academy program. Next slide, please. So everybody will obtain one CME if you are physicians. So this is the information of the CME, and you will receive an email, and you can answer some questions, then receive the CME, or you can obtain the link from our website. Next slide, please. The website I'm talking about is this FSP's Learning Center. So this one archives the previous programs, and if you missed certain programs, you can still come in and attend it, learn by yourself, and obtain CME. So please check out this learning program. Next slide. So today, we have two esteemed faculty. The first one is Dr. Rhys Cardero, Clinical Associate Professor, Director of Cytopathology Session, Division of Anatomic Pathology, and also the Program Director of Cytopathology and the Molecular Genetics Pathology Fellowship. He's also the Medical Director and the CLIA Delegate of Cytology at Jackson Health System and the University of Miami Health System. And we also have Dr. Knepper, who is PharmD Associate Member of Department of Pathology at Moffitt Cancer Center. He is attending of the Precision Medicine Clinical Service. So welcome both. Next slide. So now we're going to run some polling questions. You have two questions in front of you. We have three questions total. You can scroll them down for all the questions. So please take a few minutes to review these questions and try to answer it. If you get it correctly, that's brilliant. But if you don't, please don't worry, because after this lecture, you will get the correct answers. So those are single best answers. We'll give everyone five more seconds and then we will close the poll. And here are the results. So you see people are divided. So the correct answer for the first question is answer B. So 46% of the people got it. The correct answer for the second question is also answer A. So 69% of the people got this one. The last one, everybody is familiar with it. So 88% of the people got it. So great job. Thank you. So let's move to the lectures. Thank you very much for the kind introduction. Good afternoon, everyone. And as a brief outline, I'm happy to be here and grateful for the opportunity to participate in this webinar. And we'll try to explain some of the methodology that we use for clinical testing to try to identify molecular and genomic alterations in various hematologic malignancies, evaluate the interpretation and reporting of these molecular biomarkers, and then how to apply these genomic testing results to support personalized treatment and informed prognosis. As an introduction, I would like to highlight or start by, there's a little bit of a lag here, but I think it's important for us to look back, especially at this decade between, three decades between the 1985 to 2015, that really, as you can see here, highlights an increased amount of discoveries in terms of different molecular alterations in hematopathology. And it was, or it's just a little bit, it's taking a little bit of time for the slides to change, but through all these different studies, we now have, we can group the alterations into gene rearrangements. And unlike solid tumors in hematopathology, especially for, let's say, lymphocytes, these rearrangements can happen physiologically for antigen recognition, but also can happen pathologically, as we'll see some examples of different translocations and rearrangements that lead to malignancy. We also have two different mechanisms, different types of mutations in genes that can activate signaling and proliferation pathways, as well as changes in the copy of chromosomes or genes, specific genes. Also, these changes may lead to the overexpression or silencing of the RNA of different genes. And more recently, through epigenetic mechanisms, most commonly DNA methylation, but also microRNAs or long non-coding RNAs, we've been able to identify different underpinnings of cancer. Initially, these alterations have been classified into two different classes, then three classes based on the type of alteration and the impact that the presence of these genomic alterations had in genes involved in proliferation and survival for class one or how they disrupted differentiation of certain cell types and also epigenetic dysregulation. With the advent of next-generation sequencing, our understanding has significantly increased, coupled with multi-institutional efforts like the Cancer Genome Atlas Network. Now we have at least nine different classes where we can group the different mutations or alterations. And I'm sure that as whole genome sequencing is now in the horizon, this list is going to keep expanding. Brief overview of the different platforms. I would like to start with a general or a more low-power resolution, if you will, starting with karyotype that allows us to identify the structural variance and copy number changes in a resolution of around two to five megabases. It requires live cells because these cells are stimulated to undergo metaphase and are rested there so that through different mechanisms, most commonly the G-binding, we can get a representation of an individual chromosomes. And this was really what started the whole field. As we can see here, Dr. Janet Rowley, she described initially the presence of the translocation of chromosome 8 and chromosome 21 involving genes rungs 1 and rungs T1 back in the 1970s. And it is interesting to see that she made this observation and she believed that it was a recurrent abnormality in leukemia, submitted her results to the New England Journal of Medicine but then received a letter of rejection just saying that in their judgment these findings were not important. And so she ended up publishing these results in a different journal, which also included some preliminary evidence that the Philadelphia chromosome, which had been described a decade earlier, was not a deletion as it initially was thought of, but it was actually a translocation that we now all can recognize. And it is through this identification of the molecular underpinnings that are driving cancer that have helped improve the overall survival in patients, in particular like in this graph that we see here where the overall survival for chronic myeloid leukemia was around 11% 10-year overall survival and how that has increased to 84%, probably much closer to 90% in current years. Another method is FISH, fluorescent in situ hybridization, that also allows us to identify the structural variants and copy number variants at much higher resolution. We can also use in situ hybridization technique to identify different viruses. For example, EBP. It can be performed on whole cells, either on the smears or in FFP tissue. And what it does is we introduce fluorescently labeled DNA probes that hybridize to complementary chromosomal regions, and that allows us to visualize on a cell-by-cell basis specific genetic abnormalities. So we need to know what we're looking for, be it chromosomal translocations, deletions, amplifications, or other numerical changes. There are different types of approaches for FISH, but most commonly we use two different types of probes, fusion probes, that when two different areas of the genome come closer together, the probes emit a different fluorescent signal like the one shown here in yellow that then we can recognize and call as a positive translocation. Similar mechanism, but in an opposite way, we can have probes that are initially bound together, and when a translocation or a rearrangement happens, then they're taken apart, and we see two different signals. So it's important when one is interpreting the results to understand what are the fusions that are being used. Another tool that some of the laboratories use are the arrays, comparative genomic hybridization, or SNP arrays. These allow us to visualize copy number variants at a much higher resolution. It is usually performed on non-FFP samples. However, there are some technologies that allow for FFPE or testing FFPE, and probably the adoption, especially currently, is not as widely adopted or used in many clinical molecular labs, and one of the reasons is that it's a lengthy process. One needs to take the patient's DNA and then compare that to a reference DNA. Both DNAs are labeled with different probes that then are hybridized to this microarray containing thousands of probes, like as we can see here in the figure, each one of these dots is a specific type of probe, and then the instrument calculates fluorescent ratios between the patient and the reference DNA, and that allows to identify gains, losses at a much higher resolution than, let's say, karyotype. Depending on the array being used, additional pieces of information like loss of heterozygosity or uniparental disomy can also be derived from these assays. However, they cannot detect balanced chromosomal rearrangements like translocations or inversions, so they provide higher resolution but are still a little bit limited. Now, PCR, polymerase chain reaction, which is the bread and butter of a molecular laboratory, depending on how it is designed, one can identify structural variants, copy number variants, insertions, deletions, or single nucleotide variants to a single base per resolution. We can also use this methodology to detect viruses, bacteria, and other organisms, and it really relies on the exponential replication of specific DNA sequences. So it can only amplify known targets, and you want to have high-quality nucleic acids as the substrate, given that degraded DNA or RNA may lead to amplification failure. I don't want to get into all these details, but just to highlight that there are numerous different types of PCRs, but they all rely on the common principle of taking a strand of DNA, denature it with higher temperatures so that by introducing specific primers and nucleotides on the polymerase, if the alteration is present, this can anneal to the templates and then start expanding. And so from a single molecule after one cycle, you end up with two molecules, and after another cycle, four, and so on and so forth. And finally, sequencing. Initially, with Sanger sequencing, viral sequencing being low-throughput targeted approaches, and more recently, next-generation sequencing that allows for high-throughput, massively parallel detection of multiple targets, multiple patients at the same time. And depending on how the assay is designed, one can detect a number of genomic alterations. And this really allows comprehensive genomic analysis. It's also important to have a good substrate of high quality and quantity of nucleic acids for this to work. And this is just to put it in perspective. On the left side, what comprehensive genomic profiling could look like where on the most outer circle, we have the chromosomes and then the transcriptome, the copy number changes, different mutations, even methylation, or structural variants where if we think of single-gene assays, we're only going to get one single piece of information. Now, these molecular alterations become very relevant in hematopathology given that currently we have two different classification criteria, the World Health Organization, that as you can see here on the left, includes mutations, rearrangements, copy number changes to define specific entities, and the same is true for lymphoid malignancies. Similarly, the ICC classification also incorporates molecular findings to define specific entities, and even though they have significant overlap, as we'll see with Dr. Todd's presentation, there are some minor differences in terms of introduction, new terminology, BLAST thresholds, and overall, they both integrate molecular genetics into the different entities and really recognize also germline predisposition for certain entities. So with this, I'll pass it on to Dr. Knapper to continue the discussion. Great. Thank you so much, and thank you, everyone, for joining us today. It's great to be with you. So for this next section of our webinar, we're going to focus on the application of molecular testing, predominantly next-generation sequencing, as was just introduced by Dr. Ariz Cordero, and in terms of application into clinical care within specific hematologic malignancies. So molecular testing, next-generation sequencing, has a multitude of roles in the diagnostic as well as the treatment paradigm and kind of all-encompassing for clinical care, particularly in heme malignancies. The emergence of next-generation sequencing, as it was introduced, allows us to interrogate hundreds or maybe thousands of cancer-related genes that have, and this has been transformative, to the field of malignant hematology, where the molecular profile and the determination of the specific alterations present in a specific malignancy will give us insight and allow us to make better clinical decision-making when it comes to therapy selection on the basis of specific molecular alterations that could provide an indication for a specific target of therapy. If any of you had joined any of the previous lectures with a serious focus on solid tumors, you've probably heard a lot about that, where the field of solid tumor molecular care, precision oncology, has really become increasingly reliant on the detection of these specific alterations in specific tumors providing indication for targeted therapy. And this is true, but there's multiple other uses for next-generation sequencing, and I think the field of malignant hematology has really led the way in that. Some examples, like we have laid out here, include using the determination of the genomic profile for risk stratification. We'll go into this a little bit in our disease-specific examples, but some examples include the ELN risk criteria for AML, different prognostic scores that have been developed in the case of MDS, as well as MPNs. Also, beyond helping to aid in diagnosis, multiple hematologic malignancies now have specific subtypes that are defined molecularly, so the presence of specific molecular alterations in combination with the other diagnostic criteria better defines or explicitly defines a specific molecular subgroup of the disease, and then potentially, if not now, maybe in the future, can lead to differential treatment strategies based on these different molecular classifications. Next-generation sequencing, in terms of both the identification of these founder clones and then more advanced and deeper sequencing, allow for measurable or minimal residual disease, MRD testing, and the various roles this can play for treatment monitoring and response monitoring allow us to provide the best care for patients as well. These next-generation sequencing tests and panels, although are predominantly intended to detect somatic mutations, meaning acquired mutations that are present in the tumor only or the malignancy only, these tests are capable of detecting and often do detect and report alterations that can be germline as well, so it's important to be able to recognize when you have potentially detected a germline alteration, particularly those that are associated with potentially hereditary cancer syndromes or other hereditary diseases, and for clinicians' perspective, to know what to do about them. So when it comes to the clinical considerations that a clinician who has a patient in clinic with them think about, kind of think about some kind of the core questions that clinicians will think about. So number one, do the results that were reported back to me on a next-generation sequencing report make sense based on the diagnosis? And here, clinicians will lean very heavily on their pathology colleagues to make sure that the diagnosis fits, that there's something that's kind of unexpected detected and kind of helping to reconcile that. But beyond that, other questions that are typically thought of are, are there genetic alterations that are detected here that provide additional prognostic information, those risk categories that we talked about? and then are there indications for particular therapies? Because increasingly the detection of specific alterations is tied to FDA approved therapies on the basis of detection of that mutation. Many of these, most are tied to a specific indication, but we're increasingly seeing, particularly in solid tumors, maybe with a little bit of a hint of that coming in acute leukemias, most of them are tied to a specific indication, but in solid tumors, some are kind of tumor agnostic, solid tumor agnostic therapies. But not only the detection of specific alterations can give us indication for therapy, they can also give us contraindications for therapy, telling us if you detect this mutation, you should not use this therapy, or the detection of this mutation, acquired resistance mutations, indicate that resistance is occurring or will be occurring in the near future, and giving us that type of information. And then finally, being able to recognize and differentiate between germline and somatic alterations, particularly those related to disease risk. So getting into our specific disease areas here, we're starting with MDS, where we see where genomics plays important roles throughout care for patients, including diagnosis, prognosis, treatment, as well as monitoring. In our top left panel, we see the genomic landscape of MDS reported from large cohorts of patients, and we notice a really long tail, lots of mutations are recurrently, lots of genes are recurrently mutated in MDS. Over 50 recurrently mutated genes, none of which actually occur in more than a third of patients. So there's not any particular alteration that you're gonna see very regularly. It's kind of a mixture of alterations that have kind of various signaling pathways and various mechanisms. Important genomic abnormalities stand out there when it comes to the disease classification systems from WHO and ICC, where some of which are detectable by next-generation sequencing, like SF3B1 mutations and TP53 mutations that could be defining MDF subtypes. As far as risk stratification goes, the IPSSM model is a relatively recently developed prognostic model that takes into account clinical factors, laboratory path of values, but also genomics, where the study has utilized and identified what they described as 16 main effect genes and then 15 residual genes. So all of this can be incorporated into a calculator, the clinical factors plus the molecular factors and using that to define patients in the different risk groups. And we can see from this Kaplan-Meier curve here, correlated very strongly with survival. And then finally, the relatively new approval kind of started in AML, but now it's FDA approved in MDS as well, where the detection of susceptible IDH mutations in patients with myelodysplastic, with MDS, provides an indication for treatment with an IDH1 inhibitor for patients with relapsed or refractory disease. So you can see here in this case where molecular testing is a necessary component across the treatment paradigm for patients with MDS. So the other kind of question that has come up and then you might like to see is determining whether or not a alteration detected on an NGS report is somatic or germline, the kind of a colloquial way that physicians might talk about this, is this mutation real? And one resource that I like to use in this space actually just comes from the NCCN guidelines has a nice table, which is sort of replicated here, which gives us the common or likely typical somatic mutations across a variety of myeloid malignancy associated genes. And you can see here certain alterations like where ASXL1, typically non-sensor frameshift mutations are gonna be somatic, whereas we may see a missense mutation or is almost in all likelihood gonna be not a somatic mutation, a germline polymorphism, potentially a sequencing artifact, et cetera. But this table I think is really handy. Just wanted to kind of highlight a couple of things here. The DTA mutations, TET2, DNMT3, ASXL1, common mutations that we often see in SHIP for patients maybe without a diagnosed hematologic malignancy. We can see really any of these mutations detected in the blood in a patient without meeting the diagnostic criteria would be potentially indication of SHIP at DTA mutations in particular. And then highlighting DDX41 as a gene that is also inherited germline mutations of the gene are associated with hereditary myeloid neoplasms. Continuing on to AML, fairly similar to MDS in the sense that NGS is really used across the paradigm, starting with the diagnostic classification system where the molecular alterations are really key there. So kind of starting at step one, meeting criteria for AML, but then with disease-defining abnormalities, as we can see in that green panel on the upper right, certain disease-defining abnormalities, many of which are detectable by next-generation sequencing as your molecular platform, and then sort of moving from left to right across the hierarchy. Again, kind of different disease classifications of AML dependent on a detection of specific molecular abnormalities, either by mutation, by NGS, or potentially by cytogenetics. Similar, but perhaps even more formally than an MDS, the classification system in AML, the primary classification system for risk classification is the ELN risk classification system, which again, relies very heavily on molecular sequencing, detection of specific alterations. And here, depending on your risk classification, will influence the therapy selection, including in the frontline induction therapy setting. So important to have that, and really necessary to have that risk stratification system and the information that you need to make the classification upfront. And then finally, in the bottom right, we see genome-guided targeted therapies that are FDA approved in patients with AML on the basis of specific molecular alterations. Beyond just the single IDH1 mutation and targeted therapy approval in MDS, we have multiple drugs for IDH1 mutations. We have IDH2 mutations. We're providing an indication for treatment in the relapsed refractory setting. FLT3 can also be incorporated into care in terms of in the induction therapy setting, or potentially the relapsed refractory setting, depending on the specific clinical scenario. And the newest approval, targeted therapy approval, that I alluded to, a bit more of a multi-cancer approval for the detection of AMLL rearrangements, or KMT2A rearrangements in patients with acute leukemia, including AML, providing an indication for the treatment with repumenib. So moving to B, ALL, we're looking at the molecular testing. Again, incredibly important. We have two major molecular subgroups defined by the presence or absence of the T922, translation 922, the Philadelphia chromosome, BCR-ABL. Within the Philadelphia chromosome positive group, of course, the presence of this alteration provides the indication and standard of care for the use of tyrosine kinase inhibitors up front. Subsequently, again, for patients on tyrosine kinase inhibitors, the next generation sequencing or molecular testing is going to be used for resistance monitoring and then deciding what subsequent tyrosine kinase inhibitors you use based on what resistance mutations are detected and their various sensitivities. So incorporating NGS into care for resistance monitoring and treatment selection. And then finally, we're increasingly seeing in ALL, as well as other diseases, the use of MRD testing as a primary endpoint on major clinical trials. So the use of deeper sequencing to detect low levels of mutations or low levels of disease in order to be a primary outcome to assess the, kind of compare the benefit between multiple therapies in a clinical trial setting. When we move to Philadelphia negative BALL, it actually gets more complicated with a really growing number of molecular subgroups that are defined by the presence of specific genomic abnormalities. The newest update in the ICC classification system added nine new molecular entities that are classified on the basis of specific molecular alterations, be it chromosomal rearrangements or molecular recurrent sequence mutations that are detected by NGS in addition to five provisional entities. In the figure there, you can see the Kaplan-Meier curve showing a really great differentiation in overall survival on the basis of that molecular classification system indicating the importance as far as prognostic goes there and potentially leading to the development of differential therapeutic strategies as well. And even within the common subgroup, the Philadelphia light subgroup, we're seeing further subclassifications. The Philadelphia light subgroup defined by a gene expression profile similar to the Philadelphia positive, but without the BCR-ABL translocation as three different subgroups defined by ABL1 class rearranged groups. So it's gonna be ABL1 partner gene, but not with BCR, other partners. JAK-STAT activation, including the most common mutational rearrangement that we'll see in Philadelphia, like the IGHTRLF2 rearrangement, and then others as well. And then finally, like I highlighted on the previous slide, that presence of MLL or KMT2A rearrangement provides an indication for specific targeted therapy. For patients with CLL, we have here again, a list of common mutations in the gene here, like MDS, where we saw less than, no particular mutation was present in more than a third of patients. Here in this study of about a thousand patients with CLL, no particular mutation was present in even more than 20% of patients. But the most common mutations in CLL, we see here, SF3B1, ATM, NOTCH1, TB53, BRT3, and so on. So what's the role of these mutations? It's kind of changed throughout time as therapy has changed. So as we know from cytogenetic markers, the influence on prognostic between, you know, inflation 17P, for example, as kind of the most typical unfavorable risk, but now adding on and thinking about the molecular alterations as well, where in the pre-targeted therapy in the chemo immunotherapy era, several of those recurrent mutations predicted poorer outcomes and poor response to chemo immunotherapy, and then shunted patients to receive treatment with targeted therapy. As we can see in that Kaplan-Meier curve to the right, where we have in blue, patients who received chemo immunotherapy and green patients who received targeted therapy. And the two different colors or the different hash marks tell us that TB53 mutated would be the ones who have the poorer survival than those that are TB53 wild type. So in general, TB53 mutated, so in general TB53 mutations really still stands out as the poor prognostic higher risk marker in patients with CLL, and standard of care treatment for any patient is gonna be targeted therapy. And when we think targeted therapy, oftentimes we think potential acquired resistance, and that's certainly the case here in CLL. And you can see the table here, particularly with the two classifications of BTK inhibitors, the covalent BTK inhibitors, the ones that have been on use for longer, Brutinib, Acalabrutinib, Xanabrutinib, we see characteristic acquired resistance mutations, including the BTKC41 mutation, PLC gamma, two activating alterations, and then with BCL2 characteristic missense mutations. And we'll get into that a little bit more in the cases later. These next two sections will be a little bit faster, focusing on select PCL lymphomas, where in diffuse large PCL lymphoma, for example, not seeing next generation sequencing too much in terms of applications into routine clinical care currently. A lot of the effort in DLBCL is helping to define and establish different molecular subgroups. There have been multiple research groups and clinical groups out there that have developed classification systems or are working to develop classification systems based on the presence of characteristic alterations, which again, related to overall survival, but nothing yet that is leading to differential treatment strategies or consensus finding. Follicular lymphoma, on the other hand, is a little bit different. Again, kind of a different characteristic translocation, but here the presence of specific alterations and activating mutations ECH2 gene will provide an indication for targeted therapy with HZAC inhibitor, tazometastat, where patients who have the ECH2 mutation and follicular lymphoma are able to receive treatment with tazometastat earlier on in care. And then finally, MPMs or myeloproliferative neoplasms, where we have characteristic mutations that are kind of defined as disease driver mutations that are included in the major diagnostic criteria of the disease, where we see JAK2 mutations typically and V617F, but also exon 12 in a minority of cases, but if you kind of include them together, the vast majority of PV is gonna have JAK2 mutations, but then also MIPL mutations and CalR mutations in the broader landscape of myeloproliferative neoplasms are important components of the diagnostic criteria. And then beyond the disease driver mutations, we have a so-called clonal driver mutations, which have important roles in the risk stratification systems, defining of high molecular risk or others that are associated with adverse prognosis or risk of transformation. And again, an increasingly number of risk stratification systems that are being developed based on these alterations. So this was just a bit of a summary and how clinicians are thinking about the use of these next generation sequencing results and why it's so important to get this testing done and delivered to clinicians. And so with that, I'll hand it back to Dr. Ruiz Cordero. Thank you. Yeah, so to be able to generate high quality results that can be used by pathologists, clinicians to make diagnosis and establish therapy, it's important to know the types of samples that we're dealing with and that we would be processing in the molecular lab. So for lymphoid malignancies, frequently we're gonna get either core biopsies or excisions that are fixed in formalin embedded in paraffin and then sliced. For bone marrow biopsies, we frequently get the aspirate in EDTA or different media. We usually get a core biopsy that depending on the method used for decalcification that may damage the nucleic acids, making it unusable for molecular testing. But we could also potentially use the cells on the smears or similar to the lymph nodes, a portion of the aspirate can be placed in formalin and processed as a clot section. And we can also get samples from peripheral blood or patients that have acute leukemia with circulating blasts or as Dr. Knepper mentioned, more and more frequently liquid biopsy as a measure to track response to therapy. So keeping this in mind, we need to make sure that our practices align with the different sample types so that we can guarantee the quality of our processes. Similar to what we do in other areas of the laboratory, we have a pre-analytical phase where not only do we need to focus on proper laboring and transport of the specimens, but also having workflows for tissues that are fixed in formalin or that are fresh. We need to validate the extraction of nucleic acids to make sure that we have adequate amounts, but also the quality is optimal for downstream molecular testing. And just as an example, I like to highlight here one of the methods that's used in the lab. So for example, for this particular sample on the left side, we have, after extraction, we run the DNA, like in this case, through an instrument called the tape station. There are a number of those that use capillary electrophoresis to quantify the size of the fragments based on what we see here in column A of the DNA sample. And so we have a ladder that then we can use to compare our sample. In this particular case, in column B1, we see a bit of a smear. We don't see a distinct line. We just see a darker area that is reflected here in this graph, where we see that we have a number that the DNA is fragmented significantly, and we have fragments that are around 200 bases to 3,000, maybe 4,000, and a little bit higher. So even though we have a good concentration of nucleic acids, these are highly fragmented. And the instrument calculates this number, the DIN number, which stands for DNA Integrity Number. It goes from zero to 10, 10 being the highest quality of DNA, and zero heavily degraded. And in this particular case, it was 2.7. So it's an indication of how fragmented DNA fragments are. And for certain applications, this is suitable and would yield results, but for others, it may not be the quality that is required. In comparison, we have this other sample where probably cells were better fixed and extracted quickly. It also provided a good amount of nucleic acids. And you can see here that the DIN number is much closer to 10. So, this is really high molecular weight DNA that is reflected here in this higher peak that's even beyond the 48,000 ladder, top mark of the ladder. And we can see here in the column C that we do observe a distinct band, you know, beyond the upper limit of the ladder. So, with this quality of sample, there are a number of applications that we can use it for. And then that's the first portion. With that, we move to the second part, which is the analytical phase. For this, we want to make sure that the assay that we're using in the laboratory has been adequately validated. We need to, as part of the validation, confirm the analytical sensitivity, the specificity, the accuracy of the results that we are generating. We need to establish a limit of detection in terms of what is, for example, the allelic frequency at which we detect all of the alterations. Sometimes we detect some below the lower limit of detection, and the problem there is that there may be some artifacts at similar allelic frequencies. And so, depending on what the finding is, it may be difficult to determine if it's real or if it's an artifact. So, this exercise is important to be able to report high-quality data. And all clinical runs, unlike sometimes research, on the clinical setting, we need to run positive control, negative controls, no template controls to make sure that we don't have inhibitors or degraded samples or especially cross-contamination between different samples that we're testing. A few words on analytical sensitivity that are worth mentioning. Like we see here in panel A, we have 10 cells, four of which are larger. These are tumor cells. And so, we have 40% tumor fraction. But it's important to keep in mind that usually cells have two alleles, maternal, paternal. And in total, we have 20 alleles, four of which have a mutation highlighted here in red, and the rest are wild-type. And so, this becomes important when we are trying to establish our analytical sensitivity of the assay to make sure that we're able to detect, like in this case, 20% of mutant alleles. And this is, in part, the explanation as to why entities like we see here on the right, classic Hodgkin lymphoma, where the tumor cells are scattered and few and majority of the cells are benign reactive cells. You can imagine that after extracting all the DNA from the cells, majority of the DNA is going to come from the benign cells, and that will dilute the tumor-derived DNA. And so, we may end up with false negative results. And this graph sort of like highlights that. Ideally, we want to test highly cellular samples with high tumor fraction because that really increases our chances of success. Sometimes we have samples that are not as cellular, but they have high tumor fraction, and there is some risk of failure in these specimens. And when we have samples that are highly cellular, but the tumor fraction is low, then we run the risk of false negative results. And then, of course, samples with low tumor fraction, low cellularity, they pose a high risk of failure. It's also important to understand how the assay, the next-generation sequencing assay, is designed, or for that matter, any assay. For example, here in this graph, I'm showing a next-generation sequencing assay that only has amplicons that are highlighted here in orange that span most of the axons in TP53, but do not cover axon 1 or axon 11. And so, we can see here on the top in blue bars showing the coverage of these different amplicons and axons. But if a mutation were to occur in axon 1 or axon 11, we would not be able to pick it up just because the design of the assay does not allow for the detection of such mutation. And then, in the post-analytical phase, it's important to review the QC metrics from the sequencing runs. There are specific cut-offs that need to be met to be able to report high-quality results. And with the use of different classification systems and guidelines, we can then classify these findings as being pathogenic, likely pathogenic. Frequently, we find variants of uncertain significance over those that are common in the population that we classify as benign or likely benign. And it is also a frequent practice to cross-reference with different databases to make sure that we make an adequate interpretation. And then, after all of that, we need to integrate the information into a structural report. And so, there are different levels of evidence proposed by these guidelines. For example, KD has levels of evidence for therapeutic, diagnostic, and prognostic markers. The FDA has a three-tier system. And finally, as a last step, once we've issued the report, we need to continue the quality assurance practice by participating in proficiency testing, interlaboratory testing. The laboratories most adhere to CLIA, CAP, and probably in the near future to some FDA regulations. And these guidelines establish what the performance needs to be, and similar to anatomic pathology and cytopathology, for how long the files and the data needs to be stored. And it's a requirement also to have continuous quality improvement and measures to mitigate errors. And so, then the guidelines also indicate the way in which the reports need to be generated and the type of information that must be included in the reports. You know, of course, we're going to have the patient name and the identifiers. We need to highlight the sample type, the peripheral blood, tissue, bone marrow, collection date, any relevant clinical history. There needs to be a portion with a comment integrating or interpreting the results, and then a list of those results, including the gene, the type of mutation, the classification of that mutation, and other pieces of information like allelic frequency or coverage. And finally, it's also important to include the test description that includes the sensitivity, specificity of the assay, the methodology that is used, sample preparation, the sequencing platform, bioinformatics pipeline, and specifically test limitations like the one that I showed for that particular case of TP53. And so, it's also important that after we do all of that and we have a molecular, you know, NGS report or other report for a molecular test, that we integrate and generate a unified report that includes the clinical, morphological, and immunophenotypic information, and that it's not only an additional report that's floating in the patient's chart. Thank you. So, for this next section, I'm going to move through it relatively quickly to leave us time for questions, but this is going to be essentially a couple of clinical cases that incorporate next-generation sequencing into clinical care and some of the questions that might arise through that. These tables here are just relatively up-to-date, as of maybe a couple weeks ago, of the FDA-approved precision oncology therapies that are tied to specific FDA approvals or specific molecular alterations and specific malignancies, those highlighted in red, or heme malignancies, those highlighted in green, or those tumor agnostic approvals, like I talked about a little bit earlier, but you can see here a growing number of those agents. So, for our first case, we have a patient who presented to his primary care physician with complaints of progressive fatigue, dyspnea observation, ultimately, for sake of time, worked up and was diagnosed with AML. So, NGS was done up front, and the results are listed here. I won't go through all this for us, but we can notice this NPM1 mutation, a FLT3 missense mutation, and a CEBPA alteration as well. So, that's going to carry us through the next section. So, kind of that first question, again, is, do these results make sense? Does it provide us with diagnostic information? Of course, here, with the presence of kind of looking at that algorithm that I showed us earlier, kind of starting at the beginning, is there an AML-defining recurrent genetic abnormality? And here, the case of the NPM1 mutation, our answer to that would be yes. So, obviously, this makes sense based on the diagnosis. In fact, it gives us a specific classification. Next, moving on to the prognostic information. So, we see with the NGS results in the upper right, we have an NPM1 mutation, and we have a FLT3 mutation. So, that actually is going to influence, is this a favorable risk? If it's not a FLT3 ITD, or if it is a FLT3 ITD, it might indicate intermediate risk. We also see a CBBPA mutation. The question would be, is this a BZIP mutation, which is the new update of the classification system changed us from bioallelic CBPA to BZIP CBPA? And then, finally, are there myelodysplasia-related gene mutations that provide adverse risk? So, there's two different types of FLT3 mutations. For the sake of time, we'll kind of skip through this, but we have the FLT3 ITD and the FLT3 TKD, which are going to be different. The CBPA alteration, like I highlighted, we can see the change from 2017 ELN to the 2022 ELN. Notice that it requires BZIP in-frame mutations. We see the BZIP domain listed here from amino acids 272 to 358. So, our CBPA mutation here happening at second codon. So, at the very, very beginning, not a BZIP mutation. So, we kind of pulled that back, right? So, that FLT3 mutation that we have is a FLT3 TKD mutation. So, it's not an ITD. That's not a BZIP mutation of CBPA. And it's not, we have no myelodysplasia gene mutation. So, this would just be MPM1 without FLT3 ITD. From the therapeutic perspective, we have FLT3 inhibitors that are approved on the basis of detection of FLT3 mutations, whereas quisartinib is only FDA approved for FLT3 ITD, whereas myelitometastorin and giltritinib have approvals for FLT3 ITD and TKD. So, that could be a consideration on the basis of kind of the specific clinical scenario here on the basis of the FLT3 mutation. And this is kind of an overall summary of looking at the molecular to really guide us in terms of diagnosis, risk stratification, favorable risk with the MPM1 mutated, and then targeted therapy indication. We're able to use myelitometastorin as part of frontline induction therapy on the basis of that FLT3 TKD mutation. Second case, patient with CLL with DEL17P, historically started treatment with chemo immunotherapy with FCR, NGS here, NOTCH1, SF3V1, TP53. Patient experienced progression, and then started treatment on zanibrutinib, was on zanibrutinib for a while, again, experienced signs of progression, and NGS was ordered. We can compare that first NGS test to our current NGS test and see those same three mutations are there, but in addition, a new BTKC41S mutation is detected. So, interestingly about these mutations in patients with CLL treated with BTK inhibitors, these mutations actually will precede actual clinical progression by a meaningful period of time. So, just the presence of these mutations alone is not an indication that's changed treatment. It would be potentially these mutations, but really it's the clinical progression that's changing care, and then subsequent treatment should include or could include patients with a PCL2 inhibitor, and then subsequently there's this new FDA-approved, the non-covalent BTK inhibitors that have activity against the CLL with the acquired resistance mutations. So, I know I went through that quick, but again, I want to say this quickly, but I want to save time for questions. So, hopefully our objectives have been met for you all, and you have a better sense, and I'm happy to take some questions. Wow, what a comprehensive and informative presentation. Thank you so much for your knowledge and expertise. So, there's one question here that really resonated with me. Besides IDH1, what other mutations may be found in multiple types of solid tumors and the hematologic malignancies? I'm a bone soft tissue pathologist, and IDH1 for me, it's for the chondrosarcomas. I know neuropathologists also use it, so it's really interesting to see hematopathologists use it. So, can you give us some examples that's the biomarker that across the board? Sure, I can. You know, one that comes to mind is, for example, BRAF that we can see in thyroid papillary carcinoma. We can see it in melanoma, and we can also see it in hairy cell leukemia. And, you know, I kind of briefly mentioned like the ONCO-KB and the FDA classification. Sometimes, like a BRAF inhibitor is an FDA-approved drug for, in this case, let's say PTC, for example, but it may not necessarily be used or approved to the same degree in different entities. However, sometimes if there is enough biological information that it would make sense to give that to the patient, then, you know, oncologists can use these target therapies. Yeah, that's a great example. I'm also thinking of like Erdheim-Chester disease as well, where there's data supporting that. I think that's probably, you know, the best example. As we know, like IDH1 is interesting because there's FDA approval, like in cholangiocarcinoma, and atriocytoma, and oligodendroma, but then there's data from other solid tumors that it's not as effective. Potentially other ones, maybe we could say like ALK, where it's a common alteration in non-small cell lung cancer, but as well as N-plastic large, ALK-positive N-plastic large cell lymphoma. There's not a whole lot. I mean, there's FGFR1, where we have labilo-lymphoid neoplasms, and we have FGFR1, FGFR family of alterations in solid tumors, like urothelial carcinoma and others, and it's kind of emerging data there, but I think it speaks to the point that it's important to evaluate these alterations within their specific disease context, because even in solid tumors, you might have the same biomarker, and an FDA-approved drug in a certain malignancy, it might not work as well in another. Very good. So we have other questions, but they're more tactical. So let's move on to the following slides. So again, please scan this barcode, QR code, so you can go to the FSP's Learning Center to access to this recorded lecture, to access to CME link, and also to previous recorded lecture. In addition, this is the Learning Center for all FSP educational information, so even you missed a new meeting or previous other lectures, you can have access to this. Next slide, please. So this is a promotional information of the FSP future meeting. This one is nearby, it's in July, in Miami, JW Marriott, Turnberry, Miami, those are the three faculties. We have a hematopathology topic, and then we have dermatopathology topics. So what's the other? What's a GYN topic? So please register and join us there. The next slide, please. So next year, February 13 to 25, is our annual meeting, and in Disney Grand Floridians, and here are the faculty. Next slide. All right, thank you very much for attending this meeting. Our next one is effective communication and collaboration across the multidisciplinary team. Faculty from University of Miami, Dr. Williams and Dr. Walco from Moffitt Cancer Center. So please complete the evaluation in the Online Learning Center to claim CME. Thank you very much for attending. I hope you enjoy the rest of the day. Take care. Bye. Thank you.

Precision Medicine

Molecular Biomarker Testing

Hepatopathology

Next-Generation Sequencing

Genomic Alterations

Hematologic Malignancies

Prognosis

Treatment Strategies

Germline Mutations

Clinical Decision-Making