Hello and welcome. I am Dr. Marilyn Bui. I serve as the President of Florida Society of Pathologists. Today, you are at our very first FSP's Precision Medicine Academy lecture series. We're focusing on cancer diagnostics and precision medicine. Next slide, please. This is a continual education program. It is all free for registration and the CLAIM CME for FSP members who attend this virtual meeting in real time. So Dr. Ching is our faculty today and I also serve as a planner and a faculty today. And this is an industry sponsored platform. However, this grant is non-restricted educational grant. Next slide, please. So Florida Society of Pathologists Precision Medicine Academy is an educational enrichment, provide a platform for continuous learning of molecular pathology and biomarker testing. It consists of six CME webinars planned between November 2024 and May 2025. And FSP will also launch a new learning management system. And our curriculum will encompass topics such as molecular diagnostics, genomic medicine, bioinformatics, digital pathology, AI, and personalized treatment strategies. And our goal is to connect the state of art diagnostics to the cutting edge therapy and facilitate multidisciplinary communication and collaboration. Let's move back to the previous slide. So next, yeah, I'd like to take the opportunity to thank our friends from the AstraZeneca and the Dianqi Senko. Next slide. Next slide. So today, let me introduce our faculty, Dr. Jin, who is the Medical Director of Molecular Diagnostic Laboratory at Moffitt Cancer Center in Tampa, Florida. Next slide, please. Our learning objectives, including, one, define the role of precision medicine in enhancing the diagnostic accuracy and management of cancer. Two, identify key molecular and genomic biomarkers used in cancer diagnostics. Three, integrate precision medicine principles into routine pathology practice to guide targeted therapy decisions. Next slide. So we're going to have some polling questions. First, which of the following is or are parts of cancer precision medicine? Please choose one best answer. A, molecular diagnosis. B, molecular risk stratification. C, molecular marker for targeted therapy. D, molecular markers for immunotherapy. E, answer C and D only. F, all of the above. Second, which of the following molecular assays is more sensitive in detecting gene mutation? Choose the best answer. A, Sanger sequencing. B, Pyro sequencing. C, next-generation sequencing. So D is not very clear here. It's covered. So, Dr. Kim, what is answer D? If you scroll down on the polling window, you should be able to see all the questions as well as the answer choices. It's a digital PCR. And the three, which of the following is or are components of next-generation sequencing? A, DNA and RNA extraction. B, extraction. B, library preparation. C, sequencing. D, sequencing data interpretation. E, answer B, C, D only. F, all of the above. So this is for CME purpose, we're going to do some polling questions to see before the lecture, what's your knowledge base? And then we'll we'll collect that information. So please enter your vote. And it's a privilege to vote because I can't even do it because hosts and the panelists cannot vote. So we'll give people some time. We'll give everyone about 10 more seconds, and then we will show the results. All right, we will now show the results. So for question number one, which of the following is our parts of cancer precision medicine? All of the above. For two, which of the following molecular assays more sensitive in detecting gene mutations? So NGS, three, which of the following is our next generation sequencing? All of the above. All right, so keep that in mind and let's welcome Dr. Jin to give us a timely lecture. Next slide, please. Okay, well, first of all, thank Dr. Bui's leadership, so we have this opportunity. And also thank Joe did the tremendous amount of work behind the scenes to make this possible. And of course, the thanks for everybody coming because that's what matters. So for precision medicine that I'm going to talk about the two aspects of it. But one is to show an example, you know, where this molecular genetic test plays in the precision medicine. And then we will get into details of how the test was done. This is the example I borrowed from the neuropathologist. And you can see that when you have a diffuse astrocyte proliferation, relatively high grade, you have a list of differentials in the morphology and immunos. But here, let's focus on the molecular test, what you can do on the molecular test, and how would that impact your diagnosis and goes beyond diagnosis. So one of the molecular tests that you would do is do the chromosome test, whether you have 1p19q deletion. If you have that, your differential diagnosis start leaning towards oligodendroglioma. If you don't have these two findings, the guidelines suggest that you refrain from making that diagnosis. So apart from the morphology is still important, but apart from those, you can see the molecular test result. It does swing your diagnosis more before a few years ago. And then the other molecular test that you will do is IDH test, IDH1, IDH2. If you have the negative results, then your differential will lean towards glioblastoma, IDH wild type. And if you have IDH mutant detected, your diagnosis lean towards astrocytoma. This is not saying the morphology is not important. The morphology is still fundamentally important for your diagnosis. But this example, you can see that when you have the mutation detected, not detected, it does swing your differential diagnosis towards certain directions. And if you have the astrocytoma diagnosis made, the next thing you do, you grade it. Grade two, grade three, grade four. Now, here there's another gene you need to pay attention to, that's the CDKN2A and 2B, because these two genes are one next to the other. When you have the mutation on this one, it will push the grade higher to the grade four. If you don't have it, it may not have that push for your grading. Now, if your diagnosis is glioblastoma, then you have other molecular tests to do, and one of them is EGFR mutations. If you detect this mutation, it gives you a potential target for treatment. And I say potential, you know, in the non-small cell carcinoma, we have EGFR mutations, which are well-known for certain EGFR inhibitors for the treatment. Now, for glioblastoma, the EGFR mutations are different from those that you saw in the non-small cell carcinoma, but the research is going on right now to identify the treatment for those unique EGFR mutations. So it potentially has the impact on that one. So from this example, we know that the molecular tests play more and more roles in diagnosis, risk stratification, and treatment options. So this just to show example of it, every one of you work in the frontline of pathology that you can find some similar examples in your subspecialty that this molecular test play more and more important role in it. Now, for precision medicine, there's many aspect of it, but one of the aspect is the molecular test, molecular genetic test, that it gives you answers for different genes, whether you have mutations or any other changes. And we have different technologies to do that. And one of the most important one in recent years is next-generation sequencing. And I would like take this opportunity to do a introduction about this technology, what does that means, what it can do, and what you need to know so you can use this technology. Next-generation sequencing, of course, it's sequencing. To understand that one, we'll start from our traditional sequencing. Now, traditional sequencing is you sequence one target at one time. That's the traditional sequencing. Next are some details of it. I hope you can get that one because that's important for the rest of our presentation. This is double-stranded DNA. The red part is your target sequence that you want to interrogate. And to do that, here is the important technique details. You need to PCR first. You have primers on one strand of DNA. You have primers on another strand of DNA. This is actually the real PCR was designed. You have pair of PCR primers, but it's on different strands. And only that, that you can really do your PCR. You put the polymerase, you put the nucleotide in it, you can synthesize a copy of it. And this copy will serve as a template. Then it will serve as a template and allows you to make more copies of that one. So the PCR will go in like this. That's what we commonly talk about PCR cycle because it's on two different strands. They use each other's product as template. So this PCR can go on, can make many copies of it. And this is a fundamentally important for the molecular test actually. Without this step, you only have one molecules as your target. No instrument is sensitive enough to perceive the signals that you generated. Normally you have to do the PCR, you make many copies of the same molecules. So when you do sequence on those, you are sequencing many same molecules, all the signals generated will be the same and will generate together. The signal actually will be strong enough for your cameras or sensors to detect it during the molecular test. Now for sequence, you have your sequence primers again, and that will locate it close to your target. And you can use the polymerase nucleotide again to synthesize copies. The difference between regular PCR and the sequence is this time around, you use certain method which allows you to see the sequence itself, like each nucleotide is added, what nucleotide is added, and therefore you can detect mutations. You can do it on the other direction, do the same thing. So this is called forward sequencing and reverse sequence to get your results out of it. So this is a simple way to describe traditional sequencing. You interrogate one target and you get the result for that one target. Now in the human genomes, we have a billion space pairs in the genomes. If you use this method to sequence one person's genome, it will take decades. That's a human genome project. It takes decades, billions of dollars to do that one. Now the next generation sequencing revolutionized this process, and it can do the sequence much faster. The key part is it can sequence many target at the same time. And how does it do this? Let's say this is a double-strand DNA. Let's say this represent the whole genomes. The next generation sequencing, basically somehow you make this into small pieces. You make the whole genome into small pieces like that. And you sequence each of these pieces at the same time. The sequence part is the same as you saw before, but you sequence it at the same time and you get a lot of information for it. Now, when you get this much information, the humans, it's not able to interpret it because too many for our brain to handle. So you would need to have a computer to help. You have a reference genome sequence here. The computer will help you to organize all these sequence results and line them up with the reference genome. Okay, and you can see each one of them, we line up to certain segments of it. And you can see the gap between those, and you have some other segment fill up those gaps. Eventually the computer will stitch all this information together and give you a result, just a whole human genome sequence. It's like that. Wherever you have a mutations, it will also detect that one. So this is how the next generation sequencing works. So now we can sequence a whole human genome in couple of days and it's $1,000, you can do it. And when you talk about the next generation sequencing, the NGS sequencing, and you often heard a few words. One is, what's the depth of it? We talk about the deep sequencing. And they ask you, how deep is that NGS sequencing? So what's the depth of it? The other thing you heard about is how many reads you got. You got this mutation, how many reads you got? What does that mean? And let's look at this one. Say, in terms of depth of it, say this segment on the left side of the panel, you can see that there's several reads line up for that one. And on the screen, the lower left corner is the 8X. That means on that part, the depth is 8X. 8X means you have eight reads. So you see the depths and the reads are the same thing. So you have eight reads. Now moving a little bit towards the right side, when you have an overlapped reads there, you will have more reads than eight. You have 15 now. So you have more depths on this part. You have 15 reads. The more reads you have, the more reliable your data is. Basically, you sampled more, okay? So that's the depths and the reads. And that is the important part for your data. If you have many reads, you find the mutation is more reliable. If you have very few reads, then people may question that mutation is real or it's some artifact or it's a noise background. So that is a depths and reads people use very often in the next generation sequencing, okay? And this is, when we say fragment the DNA, and you have different way to fragment it. And one of the way we call the amplicon assay, you just use a PCR to fragment it. How'd you do that? So you basically, you have your PCR primers. This is oversimplifies as you know, because we talk about the primers has to be on the opposite strand. But just for the simplicity, you have PCR line to the certain part of the genomes. You amplify it. The PCR product, actually it's the equivalent of the fragment of your target, the DNA, and you can sequence it in this way. So this is one of the ways to do it. And this is the real, the data that you got. And on the screen, you see the, each bar is a reads from the sequence. It's a homogenous looking, and because you didn't see the mutations. Now among this homogenous reads, you will see some blue bars, red bars, and those indicates a mutation to found in next generation sequencing. Okay, let's zoom in. We can see better. So in the middle of the screen, you see the AAA on the background of a homogenous reads. So those AAA indicates a mutation found. And if you look at the bottom of this screen, there's a reference sequence there. The corresponding to A, you have a T. So you know the wild type is T, and now the mutant, it's A. Okay, and if you look at the left panel of the screen, then you can see that on the top part, this has a DNA. So the top part of this data, it shows the DNA. The lower part is for the RNA. So in this next generation sequencing, they interrogated both DNA and RNA. And you can see this A mutation, and not only on the DNA, it also transcribed in the RNA. So that's, you can see, depends on how you design your NGS. Sometimes you have DNA only, sometimes you have DNA and RNA. And so that gives you a different information for it. And the homogenous is actually, there is a sequence in the background. They make it homogenous, so you can see the mutation easily. If you want to see all the sequence, you can click a certain buttons, it will show you like this one. It show all the sequence, but you can see the rest of the sequence. It's the same as the reference. Okay, so that's just computer make it convenient for you. And there's another type of mutation you can see on this one. This is on TP53. On the one reads, you can see T, G, so two point mutations. On the left side of it, you can see a deletion. So you miss one nucleotide, that's a deletion site for it. And three mutations happens on one read. So this is a complex mutation. The molecular lab can give you the final result, what all these changes means. Usually like indel, this type of thing, and give you the significance of it. And also you can have insertions on the data. So this is the insertions. If you look at the left side of this panel, among those sequence, you can see a I, the blue I, in every reads of it. That I is abbreviation of insertion. If you click on that I, they can show you what the insertion is. Some are longer, some are shorter. So basically this is how the NGS data looks like. Okay. Now, there's other things that you need to know about NGS. So this is to say, we have a sample, we have different DNAs in it. And you have a sample, say this one. This is all tumors. Every string cells are the tumors. And in the middle of this screen, that the X shows you the mutant. You can see every DNA has a mutant, which means 100%. So here, we introduce a concept called the mutation allele frequency. Mutation allele frequency means how many mutant you find in all the target you have interrogated. In this case, it's 100%. Okay. So 100% tumor, 100% of mutation. A new frequency, 100%. The next one, you can see in the middle of the slides, the brown color is the tumor and the rest of it is stroma. And on the left side of the line picture, you see the X, only a few X there. Yes, thank you. That you can see three mutations among all the interrogated ones. So now this time, it's not 100%. It's only certain percent of mutant in it. That a new frequency is important for you to understand it because you know how many mutant among the wild type. Sometimes it's proportional to the tumor cell percentage. Sometimes it's not. The next one. Joe, I guess I lost control. Okay. So the next case you see, it's all tumor cells, 100%. But if you look at the mutation, it's only one mutation is found on one reads. That means among all these tumor cells, only one tumor cells has a mutation and the rest of it is not. So this is called the tumor mutation heterogeneity. And this is a commonly found phenomenon when we have NGS technology. You find that this new frequency is not always corresponding to the tumor cells you have. And this is important to know. And when you issue the report sometimes, this is important for the clinicians. You don't just say it's a mutation found. You tell them what's the ideal frequency. Make up an example. Let's say this is the BRAF mutation and you've got only 1% mutant. It's better to tell them. Says, you know, among all the cells, you have only 1% had this mutation. If you use BRAF inhibitor, just so you know that not, may not be every cells can sensitive to that. If they treat the patient, the response is not well, then you can tell them and say, tells you it's just small part of it has that mutation. So this is a important basic concept for ideal frequency and your tumor cellularity from the sample you submitted for molecular test. Okay. We talk about the DNA sequence. We also talk about RNA sequencing. And for the RNA sequencing, similar technology, but for RNA sequencing, usually it's for fusion detection. And there's several good reason for that one. We don't get into details of that, but we look at the technology of the RNA sequencing. You need to fragment it like every NGS. Somehow you fragment it. For RNA, you transcribe it into cDNA because RNA is not stable. cDNA is more stable to work with in the lab. And you also add adapters to either side of it. Actually, all the NGS will do that. We just use this example to show you use adapters. These are the technique details. Basically, you know, for NGS sequence, you need to do that. And the lab will taking care of it. When you have that one, you can do the amplification like every molecular test. You make more copies of it. Okay. Now, here is the part you see that a small circles with a bar, that is called a probe. And this is when you use this specific probe, which was designed for your target. It will pull out your target PCR product. And for the rest of it, it won't. And in this diagram, on the left side of your initial RNA, fragment RNA, you have ones. It has nothing to do with the fusion. At the first stage, it will be amplified. It will be tagged. Adapter will be added. But during this purification steps, only target will be pulled out. And this is how you enrich your target. And all these target will be amplified again so that you increase your sensitivity of the test. And you can start a sequence. And this is example of the sequence. As you can see, the left side arrow is a primer. And then it is synthesized. The first segment of the product is a brown color. So it's from one gene. But all of a sudden, it become blue. And that is from another gene. So when you have this kind of sequence result, you know that you sequenced two genes, which usually is not in the same place. So you know this is a fusion. So this reads, support your fusion. And the same token, if you sequence on the other end, reverse sequence, you got the same thing. It's just another gene shows up first, and then the next gene shows up. These are the reads tells you you have a fusion. So this is called a split reads. So this is a supporting reads for the fusions. When you have the results, the fusions, people tells you we have how many supporting reads for this one. And this is where the reads comes from. Then you understand that the more reads, the more reliable of your results for it. And there's another similar situation when sequence data comes out. You know, the forward sequence give you one gene result. So at this point, you don't know whether there's a fusion or not, because it just shows one genes. But your reverse ones get the results of another genes. And that also tells you, oh, you have a fusions now. That's what you are interrogating. So this is also supporting reports for it, okay? So in the fusion, you always have a gene one, gene two, what results you got from. So this is another way to look at the RNA sequence of it. This is the real data. On the right side, you see a homogenous, that is a sequence on one genes. It looks like no mutations. But when you come past the middle line, all of a sudden you see all those nucleotides shows up, which means it's not expected on that gene, which means it come from a different genes, which means it is a fusion. So this is how fusion looks like, the data looks like on the NGS output. So this is the essential concept about the NGS. Apart from those, there's more details about NGS. I'm going to use this example to show you what can happens. This is MDS patient and they do the NGS test. You can see many mutations is found and each mutation have a certain allele frequencies. And what's in the report usually is a pathogenic. So this is one test patient has. Patient was treated to follow up the test and they get the more results out of it. So this is another test. Some mutations still there, some are not detected. Another follow up, some is there, some is not detected. Now you got this report, you need to come up with interpretation that what does this mean? You know, some detected or some are missing. There could be many factors will impact those results. I just borrowed this example to show you those. This is the most problematic ones. They happen to have all kinds of issues, but it's good ones for us to show you what could happen to it. And for example, you have ASXL1 mutation, that's a pathogenic mutation. Next two times you don't see it. And the question is, well, I don't see it. What does that mean? It disappeared or what? And this is the time you call the lab. The lab pull the data, they can see that mutation. You see on the left side, you see many TTTT and that's the mutation shows. And on the right side of TTT, you see a different scattered mutations around. Those are the noise background. So you understand that the NGS sequence does have a lot of noise background. So much so when they update the new version of this test, they give up this segment. They said, because too many noise background, the result is not reliable. And therefore actually this test on the first column of results, it's done one version of NGS and the next two is the newer version of it. Newer version because of the noise. So they give up that part of the sequence. It's not covered anymore. And that's why you don't detect it because it's not covered. So you know that for this mutation, first time you detect it, the other time we actually don't know if it's there or not because it's just not covered. Okay, what about the other genes here? Let's pick up one, say SIBO. The SIBO, let's look at the SIBO is, you have two SIBOs on the table, the second SIBOs in the last column, you don't see the mutation there. And the question is why? And in the lower panel, you look at the SIBO, there's a one, you have a blue frame, frame it, a little arrow point down, point out a number that is 0.04. And that one is a new frequency, 0.104 means 4%, which means in this day, 2017, you detect the 4% of that one. Why not show up? Because each panel, they have to cut off. This panel cut off is a 5%. So 5% is below the cutoff, so it's a cutoff. So call the lab so you know that it's not totally gone, but it's a low frequency still there. So this is another reasons you don't see it. And now let's look at the other one called the EZH2. EZH2 on the first test, you found the one mutation called the Q55. In the last test, you find the Q47. In middle, you don't see anything. What does that mean? Is that different mutations or what is this all about? And the lab can goes into detail to look at this one. So if you look at the EZH2 on the lower panel on the right side, the top part of it, it says Q47. The lower part of it says Q55. That means that if you use different transcript, it will come up with different numbers. Actually, it's the same mutation. The different transcript is people do the research. They have to work on the transcript. Some people work on the one transcript. Some people work on different transcript, come up with different numbers. So that 55 and 47 actually is the same. So you know the first test and the last test, you found the same mutations. Now, how about in the middle? You don't find anything. Lab can goes into the details to look up that one. When they look it up, they look at a coverage. So in the lower panel, you see it's all EZH2, but in the middle of two rows, EZH2, exon three, no coverage. So this is a NGS assay. It's not a perfect test. You test the many genes at one time, but that doesn't mean you cover every inch of it every time. And this is the time that exon three was not covered. There happened to be that mutation in this not covered region. So in the middle test in 2016, nothing detected. Doesn't mean that mutation goes away. It just means that day, that run, this mutation is not covered on that one. All this detail information, the lab tells you. And when you have something that doesn't make sense, call the lab, that's very important. Okay, now for NGS, there's a different scales of NGS. We can see you have a whole genome sequencing. You have whole exon sequencing. You have transcriptome sequencing. You have, let's skip all these details. You have targeted gene panel sequencing. So different types of it. You might want to say, why don't you just do the whole genome sequencing? Nice and simple, it covers everything. Well, it's nice, but it's not as simple as that because the whole genome sequencing, you take more real estate in your sequencing part of it. It's more expensive. The turnaround time will be longer. So it costs you more. Besides, when you cover the wide range, the depths of sequence has to be shallow. And when you have the less depths, then your sensitivity goes lower. You may have low allele frequency mutations, which are important, but in a whole genome sequence, you're going to miss it. So you cannot do every case for whole genome sequencing. So every sequencing method has these applications. In our own college, most often used is a targeted gene panel sequencing. So that's most common test that you might order for now. And they come from different sizes. Some have like 54 genes, 26 genes, and that covers certain genes that you needed for your subspecialty and diagnosis. You can be more coverage, hundreds of genes, depends on your needs. So you can order one. So when you order a gene panel, it's better you know what's covered, what's not covered. And there's pros and cons. Basically, the more they cover it, the less steps they will have. There's always trade off on that one. Also recently, there's a concept called the comprehensive NGS test. What does that mean? I look up literature. I didn't find a good definition for it. However, I can give you some concept, the ballpark of it. Comprehensive NGS testing. Usually you're talking about the panel will cover hundreds of genes. And then it will cover fusions, which means that DNA, RNA, both will be sequenced because the RNA usually is designed for the fusions. And besides that, it will also cover tumor mutation burden. Tumor mutation burden in simple way to say it is how many mutations you found among all the nucleotide you have sequenced. If you find more than 10 mutants among millions base pair you have sequenced, they call the mutation, mutation, tumor mutation burden high. So if it's tumor mutation burden high, which means the chances you have a new antigen is higher. And therefore the immunotherapy may work better. So that's the idea of tumor mutation burden test. You also can have MSI high, MSI stable. So MSI, everybody knows we used to use for the Lynch syndrome diagnosis. And now it also used for the immunotherapy. When MSI high, the chances to response to immunotherapy is better, that one. And there's another test, it's called the HRD test. It's a homologous recombinant deficiency test. If you find those, the patient may response to the PARP inhibitor test. And if you have about four of this, like hundreds of genes, the fusions, tumor mutation burden, MSI status, you generally people will call, this is a comprehensive NGS test. And better still, if you have HRD add to it, it's comprehensive test. So it's a language people use. It's just an idea, tells you how is that. Now for the NGS test, certainly this is a key technology, revolutionized technology, which contribute to the precision medicine. And that's why we spend this much time talk about it, because in the following lectures, you will encounter this, and hopefully this will give you some foundations to start understand the rest of the lectures and a paved way for you to be a master of precision medicine. Now that doesn't mean the other molecular test is not used at all. Actually, in reality, we are still using other molecular tests. And I just want to quickly go through those. So you get a little bit ideas of that one. Let's say sequencing, a single sequencing. I have to skip the details because you just gone through this one. For the single sequencing, the way you see the sequence result is through the dideoxynucleotide. You don't know the technique details, that's perfectly okay. You run the gel, and this is a capillary gel, and you will see the result. The result will be depicted as an electrophorogram, like a pix. So it just shows you what it looks like. This is an actual electrophorogram, and I would like to draw your attention to the middle of the panel, and maybe not, because I cannot point it to the peak you want to see, which is okay. Let's focus on the two lines of text below. The Sango sequencing, basically it can sequence hundreds of base pairs. So that's one of the features of it. The other feature, unfortunately, it's not very sensitive. So it's not sensitive, which means in your sample you need to have a certain amount of tumor cells for the mutation to be detected, and many labs will say you need 40% of tumor. So as a pathologist, when you submit it to your sample, you make sure you have that much tumor cells in the sample. If lower than that, you got a negative result. False negative will be always in your mind. You say, oh, this may be the false positive. There is a possibility there, because the sensitivity of this test. Okay. Now, another sequence called the pyro sequencing. It's similar, but it's quite unique. Again, we can skip all these technical details. How is it done? This is a pyrogram. You can see it reflects the mutations. Labs will do this for you. They will read all these and give you a report of the mutation. Now, what we focus on is the three lines of text at the bottom of this slide. It is more sensitive than single sequencing. The sensitivity goes down to 5%, which means you need 10% of tumor cells to be reliable for the test. Now, why 5% sensitivity? I need 10% of tumor cells. It's because the tumor cell mutation, normally we talk about the somatic mutation, which means two DNA we have, only one has that mutation. Therefore, you need the 10% tumor cells to satisfy that sensitivity. When we say 5%, that's generally speaking. Some mutations, it can go below that one, like BRAF mutations on the pyro sequencing. If you have 2%, it can be detected. They can see the peak. They can tell you, say, I saw the low peak. It could be positive. It gives you some hints on that one. Also, this assay needs a very small amount of DNA. How small? We usually use a method to measure the DNA. Sometimes the measurement comes out as zero. Even a sample like this one, you may still have a chance to do the pyro sequencing and get your results. I don't mean it's a false negative or false positive. It's a real result because the technique is so good and there's internal controls in it. When you try the pyro sequencing, it doesn't matter what the measurement of your DNA is. It's still worth to try it. That's the advantage of this technology. Now, downside is it can only sequence dozens of nucleotides long. It sequences short distance. The test has to be specifically designed for your test. The primers are located very close to your target. That's a pyro sequencing. There's other things. This is a real-time PCR assay. Basically, your primers are specifically designed for certain mutations. You amplify the target and continuously monitor your PCR product. Continuously modify it so you see a curve on that one. It's more reliable on this test. It's more sensitive. This assay is often used for the MRD test. For example, PCR-able test uses real-time PCR for this one. Also, you have the digital PCR. Digital PCR is a relatively new technology. What it does is you look at the left side. You have a test tube. You have your reagents in it. You have your samples in it. You put some lipids in it and shake it. It will generate a lot of bubbles in it. If you look at the right side, it shows you many bubbles in it. Because there are many bubbles, you get the right concentration of your sample, which results in each bubble has one target sequence in it. When you do the PCR, basically, each bubble is one reaction. This bubble, you may have one positive mutant in it, like the bubble schematically on the upper side. Then the lower side, you have a negative one in it. These two bubbles have two different results in it. One is positive. One is negative. When you read this result, you read it just like the flow cytometry. You can run it through an instrument like flow cytometry. A single file of bubbles goes through, just like each cell goes through. The results are depicted like the flow cytometry. It tells you how many mutants you found. There are other mutations like fragment analysis for gene rearrangement. This has been a very classic test that we're still doing. This is a polyclonal one. This is a single-clonal one. You detect it. Those are the things. Then there's a methylation assay for MGMT. This is not common, but MGMT is common for the neural part of it. The bottleneck is the bisulfate treatment. You have your samples. You bisulfate treat it. Non-methylated cytosine will transfer into the uracil. Then you sequence it. You find what's supposed to be seen now is the U or equivalent of it is T. You know that methylation doesn't happen. Otherwise, methylation happens. This is a methylation assay. It's still used for it. These are the other technologies apart from NGS. It has been used and continues to be used in the precision medicine service. It depends on what your needs are. You can choose the test that you have. This is pretty much the basic ones that I would like to go through quickly so you get an idea of the technology landscape of precision medicine. I'll be more than happy to answer your questions. Wow, this is excellent, Dr. Chin. You covered molecular pathology very comprehensive. At the same time, you make it very easy for us to understand. It's like a reader's digest version. And at the beginning, you also showed exactly how the molecular biomarkers help the neuropathologists classify gliomas, neurocytomas. Then you covered some common techniques. We're encouraging people to type your questions in the Q&A. We see there is one coming from Dr. Andrew Bukowski. Excellent presentation. Thank you, Dr. Chin. What role would AI play in the interpretation of the sequencing data in the near future? Okay. The AI will play a very important role. What we can see most easily is the NGS data interpretation. Currently, the pathologists who sign out the NGS case, they visit many databases to get answers, like OncoKB, ClinVar, and other knowledge databases to come up with that one. AI technology should be able to visit this much faster than a human can do and come up with an interpretation for hundreds of those mutations they found. We can see that one. Currently, people haven't used that one because the application part is not mature yet. We just came back from the AMP meeting last week. We talked about the AI application. In short, there is a potential. I'm sure in the future, it can, I hate to say, use the word to replace the pathologist, but it can do a whole lot of work for the pathologist to come up with interpretation to save pathologists a lot of time on that part. Of course, there could be many other parts, but the most obvious thing is interpretation of NGS data. Very good. The next question is, Dr. Qin, what are the current limitations of NGS in cancer diagnostics? What is the limitation? I guess the answer can come from two aspects. Look at the positive side. We have no limitation on that one because the technology is so powerful, but on the other side, we do have all kinds of limitations on that. Let's say, for example, we sequence RNA and detect the fusion. When I do the presentation, it looks like easy breezy. You just run it and you can see that actually, as long as you know one part of the genes, you don't have to know the other part of the genes because they will sequence through into the other genes and use the database to find out what is the other genes. But there are limitations. If you have a segment of a sequence with a high GC ratio and the polymerase doesn't work well on those things, it will make all kinds of errors, generate a lot of noise. That noise is a killer for the sensitivity part. We do have all kinds of technology to overcome or minimize those problems to improve it. On one side, it's a limitation of that sensitivity. On the other side, we can use molecular barcode to label the DNA target at the first place. Later on, when those things have been PCR-generated multiple times, we know that is the one target that just amplified multiple times. Later on, we don't count that one as our frequency estimation will be more accurate. It is a balance. Yes, there are limitations on that one. In reality, when we order the test, I think currently a limitation is we need to know what we want to know. You choose the right test to order. When the result comes back, you know how to interpret those results. When something doesn't make sense, make sure you call the lab because there's more details in that one. We said there were details and good information is also in the details. Thank you. The next question is, how can pathologists effectively communicate the NGS results with oncologists and other members of the cancer care team? Yes, that comes to two aspects. First, we need to understand the results. We need to understand the limitations of that, like allele frequencies, how many reads you got. When you understand that one and you communicate it with the clinician, basically, since you do it in your report, you say this mutation was found. What's the allele frequency? Make sure you have this information there. I know a few years ago, when we asked clinicians, do you need an allele frequency? We got the two answers. Hemo oncologist says yes. The solid tumor says no. Now you ask the same question. Everybody says yes because it's so important. You give them the allele frequency and you tell them what's your tumor percentage in your sample. These two parameters will make a lot of sense for them. If they still don't understand, you can explain to them, say, hey, this sample have 1% of mutation. I have 100% of tumor, which means only a small portion of the tumor has that mutation. When you use a targeted treatment, you have been informed only a small percentage of tumor cells may respond to that. If it's other information, you have the more complex, you probably need to explain that one. Also, when we say allele frequency, we're talking about a mutant among the wild type. When you talk about the tumor percentage, that's physically how many tumor cells in it. This tumor percentage, you need to have two times because we normally have the somatic mutation in our tumor sample. After saying that, of course, there could be GNI mutations. When GNI mutations landscape is changed, GNI mutations usually it's 50% or 100%. However, there are exceptions because deletions, copy number changes, it will change that one. These are the things that if you know it, you are in a better position to communicate with a clinician. Also, willingness to talk to them, I think it's very important because after all, you know better than they do in terms of your sample and the test results. Very well said. As a pathologist, we are close to patient sample. Then we control the quality of the sample. When the result comes, we are the best person to make informative educational communication with our clinicians to direct their effective therapy. This is great. We are at the end of the hour. A couple of housekeeping messages. First, people ask, how do I claim CME? There are two ways you can do it. First, on the devices you're attending this meeting, you will see a link posted. You can just claim your CME there by filling that brief form. Or alternatively, you will receive an email with that same form feedback and then you can claim your CME. Only if you attend a live meeting like this, you will receive CME. Otherwise, this lecture will be recorded, posted on a platform, but then people are not qualified for CME. For the people who attend the meeting today, if you continuously attend the meeting live, claim CME, complete this course at the end, you're going to receive a certificate showing completion. Let's thank Dr. Qin again for a fantastic job and thank FSP for providing this educational opportunity for all of us. The last plug I'm going to put in is to talk about our annual meeting in 2025, February 14 to 16 in Orlando, Florida, Grand Floridian Hotel. So in this agenda, we will have an academic institution provide educational symposium. And this year is going to be our Mayo Clinic Jacksonville group of pathologists provide that educational event. And then we're going to have other lectures, focusing one on prostate cancer, GU cancer, and by Dr. Ming Zhou, and then followed by interesting and relatively new entity in GU by Dr. Jess Rimman. And we have the day two is going to be immunohistochemistry by Dr. Jason Hornick, and then followed by a lot of events for residence fellows, interesting cases, poster presentations, and also a CAP president will be visiting us. We'll have fellowship fair, and on Sunday we'll conclude on cervical cancer, cytology, and by Dr. Rachel Nair, and thyroid cytology, and also interesting cytology cases. So we have an excellent educational program built for you guys. So I will see you hopefully in our next webinar. So do we have the next webinar information? So that one comes first. Yeah, we don't have that right now, but we will be sending that by email. So it's in January. So I will see you at that webinar first, and then we'll see you in February in Orlando, Florida. With that said, thank you everyone, and happy Thanksgiving. Bye. Thank you. Bye.

Precision Medicine Academy

cancer diagnostics

molecular diagnostics

genomic medicine

next-generation sequencing

molecular pathology

artificial intelligence

chromosome testing

CME credits

Florida Society of Pathologists