Latest / Two PhDs on a Pod / Why Early Detection Matters: The Science Behind Viome's New Pancreatic Cancer Test
Transcript
- TPOP Ep 18 Draft Edit: Hey Momo, look at you. Rock star. my God. You take Halloween pretty seriously, huh? This is cool. A little bit seriously, not too seriously. this is pretty serious. Anyway, I I have to say because I'm traveling, I was not able to get my rock star costume, but I have a picture of myself as a rock star. And you are a rock star s every day, so boom. There you go. No, no, I this this picture that you see of mine was actually me cramming for a big show that I had yesterday. So this was from from a couple of days back, day before yesterday, I was cramming for that. But the big show I had was because of a business meeting and the travel that I'm here for. So unfortunately I couldn't get those same costumes with me and be on the podcast recording with the costume, but you are, so let's enjoy your costume. Thank you. Thank you. All right. Yeah. I got I got nice tats, tat sleeves, and yeah, it looks pretty cool. I'm excited to go to the office and hang out with everyone after this episode. Beautiful. so what are we doing today, Guru? Yeah, you know what? I I wanted to start from the whole Halloween mindset. You know, and everyone talks about spookiness and and you know, jump scares and what have you. And there is a kind of jump scare in medicine. Which is very, very worrisome. And it all happens because of chronic disease. For example, a heart attack or a stroke, right? So these are jump scares in my mind that actually is lurking in the background, in your body, in your biology for many years. And it's kind of like, you know, creeping. And there's a a little bit of a dread of knowing or not knowing whether something is wrong. And even if you kinda think something is wrong, you can't really find it and then suddenly it jumps at you. Yeah. That's right. So that's so that's the topic for us, except we're gonna make it very, very concrete with that particular disease. Yeah. So I think very few things make this creeping dread quite like the pancreatic cancer. And that's today's topic. We're gonna talk about pancreatic cancer and talk about our newest cancer detection test, which we actually went to market with this week. So pretty exciting. And let me just give the audience a bit of a flavor of how dangerous this disease is. So almost half a million people will die of pancreatic cancer this year. And that's different from like a heart attack or a stroke where you just disappear all at once. You'll get the news, and then you have a few months to live typically. And so that's a pretty horrible way to go, in my opinion. Yeah. And the death rates Are very high overall. So even for the earliest detection, 62% of people still die. And that's really, really unfortunate. But it gets worse. For people who get detected, whose cancer is detected in stages one and two, 38% are able to survive. But for stage three, only 9% of people survive. And for stage four, only 5% of people survive. So basically, for people who are diagnosed with stages three and four. It's essentially a death sentence. And even quite i i that that's even true for the earlier stages, but much better outcomes. So pancancreatic cancer rarely causes any noticeable symptoms in in the early stages. And this is really where we need a molecular test that is completely asymptomatic. If you're at high risk or you just want to get tested, you now have an opportunity to check whether you have this storm brewing in your body. And so today. as we will see from your data, existing diagnostic tests struggle to detect pancreatic cancer early. And that's why we've worked so hard to bring this test to the market. And we're going to continue working hard to make it cheaper and cheaper so that more and more people can test more frequently. Cheaper and better, right, Mama? We can actually make the sensitivity even higher. So just to give a sense to the audience, the current I would call it standard of care, because it's FDA approved and a lot of people know about it is The standard is CA 199 tests. And those are used mostly for management of the disease once it has been detected, but not for early detection, because it's not good for early detection at all. And we'll show you the numbers. We'll see the numbers soon, but you know, the detection rate is, you know, less than 50%, right? Sensitivity. So we need to focus on developing good tests for the early stages, which is what we're going to talk about, our test. is the current best in the market for early stage detection, which is stages one and two. And I believe that we need to be able to detect early stages for all cancers, right? Now I mean pancreatic cancer is a is a particularly deadly one, but all cancers need to go in the direction of early detection, which unfortunately, the current tests in the market, like the big ones, the multi-cancer detection tests, early detection tests, even though they're called early detection, are not really early detection because they are very s low sensitivity for stage one and stage two. They're typically much higher sensitivity for stage three and stage four. So I think we wanna bring a whole array of new tests for early detection and consider this, the the the current stake in the ground that we are putting from VIO, remembering also that we had a early detection test for cancer detect oral and throat, which is from a saliva sample, also very good sensitivity. Of 90% at specificity of 95% for oral and throat cancer. that is based on a saliva sample. But now we're going to talk about a blood sample based test for pancreatic pancreatic cancer tests. So that is the topic of the of the episode. And we are hoping to actually, there's already in the market, we are hoping for a lot of people to adopt this and start making a difference to your patient base. Absolutely. So let's get into it. We're gonna go into all the details. I'm Momo, a biochemist. And I'm Guru, a data science and AI expert. We're two PhDs on a pod. Or two rock stars. Two rock stars on a pod. Yes. Amazing. Amazing. Rock stars on a pod. Sounds good, man. We can hopefully do this more often than just once a year on Halloween. we can be rock stars any day. Any day, any day. So Momo, what's on your radar? Ooh, let's get into it. I have some slides to show. Yeah. And as always, we're gonna we're gonna start with food, cause man, we live to eat, so this is a very important topic. And I hope more people develop a very positive relationship with food and love cooking and eating and sharing and you know, cooking together and make it a social event like we do in my house. So I I see a lot of people talk about bread and I see quite a bit of negativity. and the negativity I see is that, ooh, bread is carbs, it's bad for you. And the other one is, well, if I'm gonna make bread at home, it's hard. And so here I'm gonna solve both problems at the same time and and add some more benefits. So, first of all, carbs are a huge food category and there are horrible carbs and there are very, very healthy carbs. And we're gonna make bread with very healthy carbs. So this is gonna be a healthy bread. Second, it's going to be super easy. This is a no-need recipe. There is no needing. There is no getting your hands dirty. You nothing. So very easy. And third, you can make this a social event and you could experiment in the kitchen. You can play a little chemist. So I'm not going to go over the recipe here in details, but you can pause the audio, pause the video and look at it. But also we're going to post the recipe in the show notes. But basically, this is super simple. You're going to mix flour with yeast, salt, and water, and you're gonna just mix it with a spatula so you don't to touch it with your hands. you're gonna then cover it and leave it there for overnight. And it can be twelve hours overnight or twenty four hours overnight, doesn't matter. when you're when it's done rising, you're gonna basically plop it onto a surface and just fold it a couple of times, transfer it to a Dutch oven. And that's really important. Dutch oven. Cast iron Dutch oven is by far the best and these things are now not expensive. and so you're Put it in the Dutch oven, let it rise for another couple hours, and then you're gonna bake it in the oven. It's literally that simple. And this bread is super nutritious because it has whole wheat, of course. But you can make it even more nutritious by adding nuts and seeds and other things. And I wrote that in at the bottom there. So you can just experiment and you can do this with your family, and you can make pita bread or loaves. You can make all kinds of shapes. If you have small kids, they can make all kinds of shapes. So yeah, it's a fun platform. And That leads me to the next slide, which is if you see the point here on this slide where it says add 20 grams of salt, people may say, whoa, whoa, whoa, this loaf of bread has 20 grams of salt. I thought salt was bad for me. All right, let's let's go into that a little bit. So s when people say salt, they mean sodium chloride, that's the table salt. And too much table salt that's not balanced with potassium chloride is bad for you. That's not, that's definitely not disputable. However, What I want to show people is that eating moderate amounts of table salt mixed with potassium chloride, which we would have naturally obtained from eating plants and root vegetables and all that stuff, now it's no longer that common in the food supply, and people don't eat so many vegetables. You can basically play a little trick. And this very, very solid study from a very solid journal told t tells you exactly what to do. So basically I write the recipe on the right side. You're just gonna go somewhere, wherever you'd like. I buy my potassium chloride on Amazon, just order it and bring it ship it to my house. You whenever I make I make a a a quart jar mixture of sodium chloride, just stable salt, with potassium chloride. And it's three parts sodium chloride, one part potassium chloride, and you mix that well. And now that jar is my source of salt for the whole household. So we fill our little Dishes with that that we use for cooking, and we fill our salt shakers from that jar and we use it for a month or two and then we refill it. That's simple. If your local co-op has potassium chloride in a bin, you should get it from there. And you can see in this study that this kind of a addition of just one quarter of total salt from potassium chloride made a really, really large difference in clinical outcomes related to cardiovascular disease. That are all likely related to blood pressure because there are many other studies showing that potassium chloride reduces blood pressure. And I can just tell you from my own experience since we started doing this, I've had the lowest blood pressure of my life. it just it really made a big difference fairly quickly. So it's it's it's it's a pretty awesome trick, very low cost, next to free, and it's easy to implement. So there you have it. Momo. That is very, very insightful. Great idea to replace a small part, actually one, quarter of your total amount of salt with potassium chloride instead of sodium chloride. Can you just remind me and the and the audience about the role of sodium and potassium in our biology, just so that we keep in mind what exactly is going on when all this, blood pressure variation happens because of consuming one level. So pata I'm not an expert on this topic, but here are some hints, and people can of course dive deeper with Chat GPT and other LLMs. But potassium and sodium are inside and outside cells in different ratios, and they maintain basically the ionic strength, the pro proper ionic strength in blood and inside cells, and they also enable proper nerve signal conduction. And so, ionic environment, that that's what they establish. And so, It's really the ratio of potassium to sodium. If you have too much sodium, but you have the right ratio, you can just pee it out. But if you have too much sodium and too little potassium, that's a problem. And so as long as they're balanced, your body will figure out how to get rid of excess if you're eating too much, but you shouldn't be eating too much, of course. but yeah, so it's That's a important thing. That's a great that's a great start. Thank you. That I the the the ionic balance, you know, among among the different elements of your of your biology like the the tissues inside and outside the t cells, extracellular you know tissue that you have and so on, right? So everywhere you need to have a balance of sodium and potassium ions. Yeah. So okay, that's that's a good start. So what else do you have on your radar? I just have one picture because also I think check that going out going out that is going super beautiful. going out and exploring places, meeting new people, and so on, I think is a super important part of life. And and this is this is like one of my favorite places to travel. This is the the Jungfra region of Switzerland, and this town is called Vengen. It hosts a very, very famous downhill ski race that you know I used to watch as a child every year, you know, it was a big anticipation. But now we actually stay there and the views are just spectacular. And the other thing I want to mention, I may have mentioned this before, is that I now travel exclusively with my iPhone. And I used to be a super, super like geeky photographer. I used to have like a two professional DSLR bodies and like six lenses. And I used to be a semi-pro photographer. And I I basically abandoned all of that complexity, weight, just Today's iPhone, and I I have the 17 version, the 18 version is a little a little better, but it's just a little better. But the Pro phones enable you to take raw images and then process them to taste in Photoshop or other s programs. And they they're just fantastic photography tools and you just carry it in your pocket. it's it's it's it's such an exciting time we live in. So I wanted to share that. Cool, cool, cool. I think I'm gonna have a few mountain pictures to share next time we have an episode. But until then so what's on your radar, Guru? What's that? What's on your radar? My radar. Yeah, I actually want to do a little bit of a throwback to the lentil discussion we had last time. Anytime. Lentils are always a good topic. Listen, I you know, I I counted the number of times I ate lentil in the last two weeks since our last episode. Turns out I've eaten lentils five times in the last two weeks. And that just tells you, how important it is in in my cuisine, but also many other things that we're doing. so I want to just show you a little bit of a picture of one of my latest meals, which actually had two elements that were amazing to me. So of course, you have the lentils in the bottom half. I also added a few black beans, which are super amazing for for my style of lentil making which is actually in a category called sambar India and I I can actually share the recipe for making sambar which is a very very standard dish in South India and I I have add all all kinds of different things including beans to it and then at the top this is known as a palya in South India. Palya basically just means different types of vegetables. You can also add beans to it if you want and you spice it up in in a more or less dry fashion. No, it's not like a gravy-oriented thing. It's more or less dry. But th there's also the internal moisture of the vegetables themselves that actually make them much softer. It's very f fast to make. And I make this all the time with beets and with some other root vegetables that are very similar to beets. Like I I use Daikon, for example. Daikon I really love It's a root vegetable, one of my favorites. It also has a little bit of a crisp a feel to it. Yeah. And it's it has a little you know, it's not just it, it's it has a little bit of spiciness into to my palate and it adds a great texture to to this palya, particular palya that I make. And one more thing, which is covered here, is that I put quinoa at the bottom. Quinoa is one of my favorite grains or or or basic, you know. underlying, let's call it, background for most meals because it not only has great carbs and fiber, but it also has a significant amount of protein in it. So it's it's just by itself, it's a great thing. I I think we should have a whole episode on quinoa. And all the great things that come come about because of quinoa. Anyway, I just want to share this because I thought this was one of the best lentil based meals I had in the last two weeks out of the five that I had. So one of the Michelin Star restaurant beauty beautiful meal. Thank you, man. Thank you. That's amazing. Love it. It's yes, it's definitely one of my favorites. And the other thing I want to share is you and I showed how we are dressed as rock stars for this Halloween, but I want to show you a couple of pictures from some of our colleagues, our other scientists in bio, who've also dressed up beautifully for for Halloween. So the man here on the left is Dr. Eric Patridge, who's a biochemist. And so he has all these potions, like Harry Potter, and and he has this whole emporium on on his left. You can see all the different little potions and he makes things and he's a biohacker. He actually does a lot of biochemistry on on himself and has a lot of great stories. That's one. On the right hand side is actually our our other colleague Tundi, who Momo and I both work with regularly on various projects. But The real nice thing here is this is our office in in New York City. So Eric is sitting sa standing in one corner of our office. Tunji is standing in another corner of our office where we have a a little counter, kitchen counter where we make a a nice portion every week. We make this potion from different fruits and vegetables and we all share it on a Wednesday. We do that every every week. On this particular Wednesday, Tunji is serving. a blood bar, right? So it's freshly squeezed blood bar, you know, as as a vampire. And he has different blood types if you want, right? There's a there's the A and the B and the O plus and all that stuff. So that's his scary Halloween class. Super cool. Super cool. The final one, I have Angel, another one of our colleagues that we work with regularly, who is treating a doctor. He's treating a doctor. And this doctor is totally down and out. and Angel's asking them, Tell me, what's eating you? Right. You know, it's it's a great, great picture. I just loved it, the whole setting in another corner of our office. So I thought I'd share some of our favorite colleagues and you know, there's many others that we'll bring in off and on. We bring them on as guests, but I just want to show some of our colleagues on the show as well. Love it, love it, love it. Awesome. Yeah, so anyway, with that, I think we should go back to the main main body of our episode today, Momo. Which you know, I have to start sort of from the big picture standpoint. The pancreatic cancer, right? Pancreat cancer, yes, it's a sudden disease, but there are many risk factors that can actually increase the odds or the probability that somebody gets pancreatic cancer. And there are several risk factors that I've listed here on the page. The first most important thing is just age. It's true for all cancers, but in particular for pancreatic cancer, once somebody crosses the age of fifty, the risk of pancreatic cancer prevalence, actually the risk of cancer because of its prevalence goes up by two and a half times compared to what it is before fifty. So that's a big shift in the in the in the risk because of the prevalence. So I just want to say that I just want to say that b below fifty doesn't mean no risk. There are unfortunately people being diagnosed and dying of pancreatic cancer before before the age of fifty. Absolutely. I mean, I you know, we you and I talked about a case that was, you know, somebody with you know 36 years old, right? Who who had pancreatic cancers in one of our studies and so on. But yes, there's you know, there there is a you know a a risk before fifty, but it increases dramatically after fifty. So once you cross the the threshold of 50 years, make it even more important in your own health journey to track what is happening in your body and do all the early cancer detection tests that you possibly can, especially if you have other risk factors, like for example, in this case, nuance diabetes. There's just so many people who get diagnosed with diabetes every year, type two diabetes, especially. Here we've shown the incidence of type two diabetes of twenty-five to thirty million people a year. Every year. Globally. Globally. Globally. So this is prevalent in many places. I'm you know, I come from India. I know that the prevalence and this incidence of type 2 diabetes is very high. In India, it's actually the diabetes capital of the world. So there's a lot of people who need to worry about this. And the three year incidence for nuanced diabetics is 0.62%. Which is actually much higher than people with no diabetes. And also for those people, sixty-one point eight percent of cancers happen within the first twelve months. So that's a big risk factor. That's a really dangerous period basically of your life. Basically, after type two diagnosed di type two di diabetes diagnosis, th those next three years are really dangerous because six out of a thousand people will get diagnosed with pancreatic cancer and the vast majority of them will die. Yeah. Yeah it's a very, very sad and concerning story. Yeah. Also remember that diabetes itself takes a long time to develop. So if you have early markers of diabetes, you can reverse that. You but even before you get to prediabetes or even after you get to pre diabetes, you can reverse that. So you never actually get into the diabetic situation where you're increasing the probability of all these other downstream problems that you may end up having. So do your health checks for let's say biome functional pathways and all the other things that we have in our biome portfolio and other people's portfolios that can help you prevent and reverse diabetes because diabetes is a risk factor for many other, even more deadly diseases, right? Yeah. I can s I can just mention something here. in in in our healthcare provider with our healthcare providers, we struggle to order the fasting insulin. That's not a normal test. And that should really be a completely standard test, especially for people over fifty. but it doesn't really matter what age you are. You should insist with your doctor to order the fasting insulin because that's gonna be one of the earliest signs on the journey to developing type two diabetes. And then when you catch it early, you have an opportunity to reverse it, like Guru said. So really important point. Yeah, yeah. And and and you know, I I I always wonder, you know, if with respect to smoking, another risk factor. I always wonder, what gets people into smoking. there's many, many things that get people into smoking, but there's a lot of ways of getting out of smoking these days, especially. So smoking being such a high risk factor for many different types of diseases, pancreatic cancer is definitely one of the big ones because of the, you know, the the morbidity rate and the mortality rate of it. 74% higher for current smokers, whereas it's 20% for former smokers. So the better it is that you stop smoking as soon as possible. There's many programs available for doing that. So please take advantage of those things. Also, there are other risk factors, including other benign pancreatic diseases like chronic pancreatitis or pancreatic lesions that are benign and so forth, also known as IPMN. Those are factors that you want to take more seriously if you if you get diagnosed with those. Chronic pancreatitis, you can get a lot of symptoms. that that makes it very clear. Benign pancreatic lesions are a little bit harder. But if for some reason you've already saw seen pancreatic lesions, take it very seriously and get whatever you can do to detect pancreatic cancer early. And then there are other standard risk factors like obesity, family history. Family history of pancreatic cancer is very important. Easy to survey that in your family. I've actually went back to my grandparents and got a good history of all the health statuses of everybody in my in my family that that passed away. so I know what kinds of things I should be aware of and so forth. So do that if you can. And then all these germline mutations for these specific genes that that are listed over here, Momo, you may want to talk about some of these genes, but some some of these are very well known, the BRCA one, BRACCA two genes and so forth. But just just take us through what whatever you know people should keep in mind for these specific genes, Momo. I mean, these are not routinely done. And so there's really i if you don't have a family history of of any cancer, you're very unlikely to test for these. but there are some some of these tests are available and you can go pay for them. and Yeah, it's just unfortunately not a really common thing. Yeah. Okay. All right. Well, it is what it is. In some cases, people get it. And if if you notice or if you if you do get your your DNA sequence and your mutations analyzed, these are the germline mutations that you should be you know aware of and be proactive about about testing for cancer. So so so Momo, let's go to the next step of what can we do to screen or or or test for pancreatic cancer through simple molecular tests. So give us a a summary of the biome assay and the test that we have we have built. And and I I know before before I start I just want to mention something for our audience. You and your team have done a stellar job in taking an academic test and turning that into a commercial test within a matter of three or four months. And and that's an amazing feat and it required a lot of things that I'm not sure we'll be able to cover, but I just want everybody to know that this is not normally done. In most cases, people kind of do some simple pilot tests and they put, you know, they write a paper in there and they put everything on the shelf and move on to the next project and next paper and so forth. It never gets commercialized, never gets in the hands of clinicians and so forth. But here in Vio, we believe that we need to take all these scientific advances, turn them into commercially viable projects. Products that we can put in the hands of individuals, whether they are consumers or doctors, patients, we want to put it in the hands of people who can use it and take advantage of it and improve their own health. So with that, Momo, tell us how you did it. Yeah, so this is a very interesting story from many angles. And so yeah, I I read papers all the time, of course. And in February of 2025, this paper came out, which is on the next slide, from Oregon Health Sciences University or OHSU. And I was look I was wowed by the ability to detect early stages of pancreatic cancer because of its importance as we discussed. And so I when I when I look at these kind of diagnostic or detect cancer detection papers, the first thing I focus on is did they run a robust independent validation? Because that seems to be the most common deficiency where people do machine learning and they present something on a very small number of cases, but it's a cross validation. And we discussed that in your AI episode. That how important that is. But these guys did a very, very nice, thorough, and robust independent validation. And the performance of the test was so good. That's one aspect. And the second aspect, I'm an analytical biochemist and I understood how this works. And I thought, okay, we can actually commercialize this. So I literally just emailed those professors and said, Hey, would you guys believe be interested in discussing licensing this? And an hour later they replied with yes. And so, like the next day we had a call. And we discussed all the things. These two gentlemen are right here. We're now good buds. So yeah. So I had a chat with them back in February of twenty twenty five. And we were excited to license this from them. They were excited about Vyome. And then you know, th then I went to the leadership team at VIOM and we decided to pursue this as a commercial project. And then our CEO, Naveen worked with OHSU to get a license. And that took a while, involving lawyers and and negotiations. It took about like a year and four months or so to get a worldwide exclusive license to the entire platform technology. And that's really key that we did not license a pancreatic cancer test. We licensed the technology. And that goes back to what you said that we at VIO want to have a true multi-cancer early detection test. And this one will have dramatic advantages over others, not just in terms of accuracy, but in terms of throughput and cost and and accessibility to much larger demographics. Because let's be honest, who can afford a a thousand dollar gallery test? That's that's not something that most people can afford on a yearly basis. And and I also want to mention that what we didn't mention before is that while screening yearly for most cancers is sufficient because of the growth rate of pancreatic cancer, Screening once a year is typically not sufficient, meaning that in one year it can go from undetectable to deadly. And so you really need to attest more often. Anyways, so had a had a great conversation with these folks. Naveen did a great job of licensing the technology, and now we are in June basically of 2026, and we are now ready to convert this academic test into a commercial laboratory developed test. And it's taken us Yeah, like you said, three to four months, which I don't know if people understand how complex it is to take an academically developed test into a commercial CLIA test. But I would say that the vast majority of diagnostic companies that license the technology from a university fail to do that. And these companies are not like dumb or anything. These are like super smart people, super hardworking, experienced people, and they still fail to do that because it is an extraordinary challenge. And in fact When we tried to run the test the first time in our lab, it didn't work because, you know, y you don't know where you're gonna get hit. But you know, we b we had a different instrument in our lab, d literally the same model instrument, but just a different instrument. And that instrument gave us completely different results from their instrument. And so initially we were like, wow, this is completely different, but then we worked through it and we figured it out. So let's go to the next slide, Guru. And now I want to share with the audience. the the the concept here and why this why we believe this technology platform is really going to take over the diagnostic market. So first of all, let's distinguish solid tumors from liquid tumors. And so we are focused here on solid tumors, which is the majority of tumor tumors of course, including pancreatic cancer. So all solid tumors must, and I'm emphasizing words all and must secrete these enzymes called MMPs, matrix metalloproteases, in order to grow and spread. They must do that. If they don't, then they cannot grow and spread. These matrix metalloproteases, as the word says proteases, they basically digest the tissue around the tumor and enable the tumor to grow. So MMPs or matrix metalloproteases are known to be very, very important biomarkers of cancers. However, There are many difficulties with detecting just MMPs, meaning that there could there are many times MMPs floating around, but they're not active. And that doesn't mean anything. So you really want to measure activity of these MMPs in blood. And you want to do that at a very, very, very low level, which is very challenging. And so if we go to the next slide, this is sort of the core of this technology. So the Protease activity nanosensor, nanosensor that measures protease activity is the key, the foundational basically technology here. And I'm going to show you how it works and I'm going show you how easy it is. And that's just really has an incredible potential. So, in the box on top, what I'm showing is we order a peptide, literally just order it via a website. That's the blue string. It's a very special peptide, of course, it's not a normal peptide. Yeah, that special peptide has non-natural peptide bonds in all linkages except for the one. And that one enables cancer-specific protease to cleave the peptide in half. Not quite half, but close to. And then on the left side, that black ball is a magnetic nanoparticle. And on the right side, the red sun is a fluorophore. A fluorophore is simply a molecule that when you shine a laser on it with a very specific wavelength, It emits light with a very specific wavelength. a glow in the dark sticks, but you have to shine a laser on it. So we add that's the nanosensor. What's what's on the left side of that box? That's literally the nanosensor. So you put that nanosensor into a tube and you add the serum from a patient. And if that serum contains this protease, the protease will cleave the peptide in half. And so now we have separated the magnetic nanoparticle from the fluorophore. And that's the principle that we're going to use for the assay. So what you what you see below is how the assay works. We basically take 12 microliters of the nanosensor inside a tube, and it's already pre-formulated in the reagents lab. The lab technician is going to add eight microliters of serum and mix. And then after 45 minutes, we're going to apply a magnet to that tube. And so now The magnetic nanoparticles are all going to go to the side of the tube and we're going to pipette out what we call the supernatant, which is the liquid in the middle of the tube. And if there was no protease in the sample, then all the fluorescence will stay with the nanoparticles and it'll stick to the side of the tube, and there will be no fluorescence in the solution. But if there was a protease that releases the fluorescence, that now fluorophore is not tied to a magnetic particle and it'll stay in solution. We remove it from the solution and we put it in a plate and measure the fluorescence. So literally we contact blood with the nanosensor, wait forty-five minutes, we apply the magnet for two minutes, we transfer the supernatant to a plate and put it on the instrument, and the instrument takes one minute to read it. That's the assay. It's incredibly simple and robust. And this is what we developed and validated. And I can spend an hour talking about all the steps that it took to Elevate it to a level of clinical robustness because we had to add absolute measurements for everything and we had to add a positive control, a negative control, and then quality controls, these are low c low QC and high QC for making sure that it all works. and so we had to create those. Yeah. Yeah. Before before we go into the the process side of it and making sure that the quality is good and the you know scalability. is is good and all of that stuff. I just want to just make sure that I completely understand the the science behind it. So I'm gonna ask very basic questions. Okay? So the MMPs, the proteasis in a let's call it a cancer environment, right? That has has gone into the bloodstream, correct? From the from the the the solid tissue correct that has now been somehow translocated into the bloodstream. Exactly. Blood is what we we extract from the individual at a certain volume, which I, I'm hoping that it's small enough that it doesn't become another big, step in the process by itself. So the the blood is what a a a patient sends over to the lab. And so tell me about the process of getting the blood. How much blood do we need? What kind of transport is required for doing that? that that stuff is important. Yeah, it is important. Thanks for asking that question. And I th and that's basically the the the assay's one of the assay's biggest weaknesses is that what the assay requires is serum. And so what it requires is venous blood draw. and fortunately we can draw any amount of blood. The smallest vacuum, a single one, it provides Plenty of blood because for example, the smallest vacuumer is about three milliliters, which provides at least a milliliter of serum. And as you can see here, we only need eight microliters. So there isn't a large blood draw, but unfortunately the cost and logistics of a blood draw, first, and then second, that serum has to be stored on dry ice immediately. So and shipped on dry ice. Basic question, basic question over there. Why could we not use the finger prick blood draw that we normally have in our biome tests already? Right. So we cannot do that because these proteases are actually produced inside blood cells and they're stored inside blood cells already active. They're just not secreted. And so the and so if you have if you lys the cells, which the blood car does, yeah, then you're gonna release the proteases that are inside the cell and those are not a biomarker of cancer. Uh-huh. Uh-huh. So so in the in the case of the normal finger prick blood draw, you would you would have lysed the cells and those would have been false markers of cancer. Exactly. So you don't want to use that. What you want to do is you want to pick you you wanna you want to collect blood from venous draw. For now. For now, Guru. But but we are literally already we already kicked off the project for collecting blood from finger prick that it's a it's a little microfluidic device that at a person's house separates serum from blood cells right there in the microfluidics device. And so we are working already on the project to bring this to at home collection by yourself and room temperature shipping. Now beautiful that's that's really the end game here that we're talking about, right? That is makes it end game easy. Makes it easy for people to collect the samples at home, not having to go to a a big process of collection and shipping and et cetera, et cetera. Well, the the the the actual process today is not really that complicated because we send a phlebotomist to your to people's home. And so we ship a collection kit to someone's home, whoever orders the test, and then a phlebotomist literally comes to their home and draws the blood. So it's not really complex, but it is expensive. A phlebotomist visit costs money, and overnight shipping with dry ice costs a lot of money. So in fact Greater than ninety-five percent of the cost of this test is coming from the logistics and not from the laboratory analysis, which for me is very frustrating, of course. And so we're gonna obviously try to work that out. yes. So great. So I have one more question besides that. So in the in the basic biology that you explained here, which is that there's there's this peptide target which gets cleaved by the proteases that comes from the sample, right? So these proteases are active for how long? what is what is the purpose of the dry eye storage and all of those things? Yeah. well, the purpose of dry eye storage is to prevent protease from dying, basically, because it is an enzyme and currently we have not clinically validated the stability. So for now we're just launching with dry ice because that's how it was originally developed. we don't know how long the enzyme is stable for. in serum and we're gonna be doing those validations soon. So we're gonna try to avoid dry ice as soon as possible. Got it, got it. Okay, perfect. So so now let's get to the process stuff, right? So you in an in an academic environment when somebody's just doing a pilot project or something for the purpose of a publication, they would probably not have all the quality controls that will make this scalable and repeatable and reproducible, all of those good things that everybody needs in commercial test that is used in a clinically meaningful way, right? So tell us what are the basic principles of quality management for an assay like this? Okay. So there are several aspects. One is reagent preparation and qualification. So reagents need to be prepared using protocols that are standardized. They need to be qualified, meaning their performance of every reagent has to be Qualified unknown samples. So samples you've already tested and you know what the expected result is. So think of you prepare a standard of cholesterol that has 186 milligrams per deciliter and you have that stored. When you prepare a new reagent for testing cholesterol, you're gonna apply that reagent to that stock of cholesterol. And if you get 191 or 182, great. If you get 250, something's wrong, right? So reagent preparation, storage, stability. formulation and performance are all SOPs now and everyone has to do it. Standard operating procedures. Standard operating process. Standard operating procedures. And that means that they're prepared exactly the same way every time. And that was quite a big of an effort to translate from an academic lab that used to kind of wing it and adopt a an absolute measurement method that is not instrument dependent where We now have the ability to prepare these reagents in any lab with any instrument because it's based on the performance, not on the it's not dependent on the instrument. I have to say here that the the two academics that we worked with, Jared and Jose, did a phenomenal job. And that was one of the attractive features that attracted me to commercialize this test is because they've already done more than what normally is done. But we had to do quite a lot, so quite a few steps more. So that's on the reagent side. on the assay side, when you're actually performing it, you want to make sure that that day with those reagents in that lab, with that temperature, and with that technician, the assay performs well. And so we have multiple positive and negative controls that have all quality control acceptance criteria. So these are validated samples whose performance has already been qualified, and they when whenever they're retested, they have to fall within that narrow range. If they're if they're if they don't provide the signal within that narrow range, then the first suspicion is that the assay didn't work properly. And then we have to go back and troubleshoot. Yep. and so these controls are again reagents that have to be prepared and qualified and standardized, and their storage has to be very specific and their stability has to be validated. You can't just assume that it's stable. So all of that was a part of the validation, stability of the reagents, stability of the controls. Then we go into standard operating procedures for performing the assay. No matter who performs the assay has to be done according to the pr procedure. Think of that as a checklist for a Boeing 737. It doesn't matter if you bring a new captain. They have to follow the exact same checklist in order to fly that plane safely. They can't just say, I have 20 years of experience. I'll fly this plane the way I want to. No. Every that doesn't doesn't matter how big of experience you have. If you come to a new plane to you, You're going to go line by line, exactly the way that airplane needs to be flown. Your experience makes no difference. You have to do exactly what the aircraft requires. And here, the assay requires all the steps. And so that's another part. all people have to have a life science, all lab techs have to have a life science degree and have to be proficiency t competency tested, meaning that their their performance is blindly evaluated to make sure that they can perform that. Then we have proficiency testing, which is blinded samples that are tested over and over again. So the technicians don't know that they're doing this. There's many additional layers in CLIA. All instruments have to meet performance specifications continuously. So that's called IQOQPQ. And then all of the storage has to all the freezers have to be temperature monitored, humidity monitored. And then last but not least, in a CLIA environment, Everything has to be documented and everything has to be checked and signed off by a technical supervisor and the lab director. If there is no paper trail as to who did what, when, how, with what control lots, with what c control criteria, then it didn't happen. And so all of that records keeping takes place. Yeah. that's a pretty extensive quality management system you have in there. And as as we've discussed before. CLIA stands for Clinical Laboratory Improvement Amendments, which has been a a a legal kind of a framework in which these kinds of certifications happen for laboratories so that they can create and maintain a an acceptable standard of quality so that people can depend on the reproducibility and the robustness and everything from that lab, that that moment where you just described. So going through all of those things and creating the clinical Test is what takes usually a lot of laboratories, several quarters, maybe even years, and many of them just don't even do it. But many three to four months, which is which is amazing. Yeah, I would say go to minimum expected time in any company would be twelve months, but like minimum. But yeah, but most companies don't ever do it. And so it's a very big challenge. And you know, I just want to give you a a a little microcosm. of this. We have to have a LIMS system that tracks all the samples with barcodes and all reagents with barcodes and their expiration dates and so w an inventory in the line. It's a laboratory information management system, LIMS, right? Correct. So we have to develop that. You know, most people don't even know that that exists, but yeah, that's like a glue that holds everything together in terms of sample tracking, reagent tracking, and quality control. And then downstream, your team takes the raw data out and does all the bioinformatics and model and we'll go into that in a minute. In a minute. so so the the there are there are so many different steps to develop and every one of those we can spend quite a bit of time as to setting the requirements, looking at the performance, setting qu quality control acceptance criteria and so on. So the fact that we did that with a modest team of about ten people in four months, I mean I'm really proud of the team and how how well they've done that. Of course, I have so much experience with this and you have so much experience with this that for us it's not a big deal. But it's it yeah. I'll I'll talk about the the dry lap portion of this, which usually takes months and months. You know, six months would be considered normal for for a dry lab analysis to happen. We did it in a few weeks. I think it was probably in the range of three weeks or something. Yep, you know exactly. End to end, maybe four. Okay, some some something like three to four weeks is what we did. Both the analytical validation and the clinical validation, data analysis across the board. And it's a it's a tremendous set of things. But let me just let me just get started on that process here for a second, and then we can switch to the next slide, which comes up with the final performance of the of the entire assay. Is that the PAN, protease activity nanosensor pan assay, is a significant part of this assay. What we've done is combine that. with the C A nineteen nine assay, which is also important for this particular overall test, because the C A nineteen nine test generates these antigens, right? Or or detects these antigens that are generated by a tumor and also deposited in the bloodstream, right? Yeah. And that, that's a that's a kind of a a carbohydrate or some kind of cancer antigen that is that is generated by the tumor and deposit in the blood. It is exactly a carbohydrate guru. It is a carbohydrate, right? So so that kind of you know the C A nineteen nine molecules and the you know these proteases that we talked about, the MMPs, right? Both of those things are in the bloodstream and now the assay is detecting both of these. And what we end up doing is Momo's lab, the wet lab, delivers the the measurements from the particular instruments that That detect these two molecules that we are we are talking about. And then once that data is generated, we need to train a machine learning model to understand what is the best combination of those two readings that will give us the highest performance metrics that we care about. And what are the performance metrics we care about? We care about sensitivity for early stage. That is our critical metric. We also Think about the negative predictive value, NPV of this test, which is that, when you say that something is negative, how confident are you? How good is the test in actually making sure that it is an it is an actual positive, negative? Because we are interested in surveilling or screening for very early stages. So when somebody gets a negative answer, they want to be sure that it's truly negative, that they don't have to worry too much about it. So we want to maximize. the sensitivity for early stage detection and the NPV, the negative predictive value for this test. And in order to develop the best model for combining the two measurements, our team, and I have to give a lot of credit to Tunji, who I showed the Halloween costume off l earlier in the episode, he's he's one of the one of the guys in our team who has done this many, many times for other tests and For multiple other experiments. So he was he was able to come up with a methodology with my help and guidance to create literally half a million, half a million different models for combining these two measurements. And why why is it half a million? Because there's so many variables around those two measurements that you can vary in order to detect w which. Particular combination, meaning mathematical combination of those two measurements, will give you the best possible metrics I mentioned, which is early stage sensitivity, specificity for healthy, and a very good negative predictive value. So those are the metrics that we were optimizing for. And that took us, like I said, three or four weeks to do across the board. And this includes not just the diagnostic performance metrics, but also all the robustness metrics for the Analytical validation for CLIA that we were talking about earlier on includes the reproducibility, the precision, the stability, and all of the other metrics that are required. It's at the end of the day, it results in a massive document, 60 to 100 pages long, with all the data laid out for the lab director to sign off on and so forth. So all that work was done. And where did we end up in the end? We ended up with 86%. Early stage sensitivity for stages one and two. We got an 86% sensitivity for all stages at a specificity for healthy people of 90%. Now, this just wanna put this in the context of all the other tests that are available. This is the best early stage sensitivity test that's out there commercially available today. as far as as far as we know. Right. So the all of the other famous tests that people have a lot of advertisements for and papers for and all of those things, if you go back and look at their early state sensitivity for pancreatic cancer test testing, it is usually very low. It's you you know, many cases it's like twenty, thirty, forty percent. if it's a multi cancer detection test, if if it's a specific cancer detection test, it can go up to maybe seventy percent, but that is still Pretty rare for these tests, but when you get get up to 86%, that you have a leading test that you can rely on with a very, very low NPV. It's meaning it's actually I should say it's a very high NPV, meaning the low probability that you'll be wrong when it says it's it's negative. So the NPV for this test is ninety nine point nine percent. It's actually ninety nine point nine nine six or something like that. It's very, very high. So when it says it's negative, you can be almost sure. That it is actually negative. That gives you peace of mind, but says it's negative. But when it's positive, you have to do a bunch of follow-on testing and ensure that you get a definitive diagnostic, which goes all the way to a biopsy. You, you know, typically would end up doing some imaging before that, and you end up doing other kinds of analyses that people will do. and and then finally you would you end up with some kind of tissue biopsy that will confirm that there are specific types of cells that have been that have been actually transformed into becoming malignant. And that's the histopathology that'll give you the definitive diagnostic after which you would you would prepare for treatment and all the different things that can hopefully turn that you know stop it in its tracks and turn it back if possible, right? So that would be the ideal situation, including surgery by the way. It's not not only not only chemotherapy and immunotherapy and so forth. It's also surgery for this kind of stuff. Guru, let's let's let's zoom out a little bit. I want to just maybe make it a little bit of a better flow from start to b start to end. So from start to end, let's let's call this you know a four-month project where the first two or three months were spent optimizing the assay and adapting it to a to CLIA standards. And then in the lab, we spent about a month generating data for both analytical and clinical validation. And for the for the analytical clo validation, we talked about you know, performance of the assay, reproducibility, accuracy, stability, and so on. and then for the clinical part, that's really where these numbers come from. I just want to highlight to the audience that we did it in two stages. The first stage is the training set. So we generated a training set data, gave it to your team, and then that's where those half a million models were applied to the training set. And then we provided you with a blinded, independent validation set. And blinded means that Tunji and you did not know what the actual label is for a sample, whether it was a cancer or healthy. And all you did is you ran the model against those samples and told us the results. And then we we unblinded the results at that point in time, the ground truth. And that's how you calculate these numbers. So this was done up to essentially FDA standards, right? Yeah, it it is, it is. And and the the only thing I would do in the future if if I had an opportunity to do it, do the next level test is to is to increase the population of independent validation, right? That would be that would be the ideal scenario, which we want to do as we go forward. But what we've done here is well above the standards required for clear validation. So we have the approval, just to be clear to the audience, we have the approval of our lab director. for commercializing this test at the level of quality and that level of reliability that's required by the CLIA specifications. And so so that is now in the market. I'm very, very proud. We are both very, very proud to have that test in the market and we are just rolling it out to multiple clinics. We are doing all the initial piloting and and and education and everything that's required to help clinicians to start using this in the market. So if you're interested, if you're listening to this and you're one of the clinicians who is interested in implementing implementing this in your clinic, please reach out to us at providers at Vyome Pro.com or just send it to podcast at Vyome.com and we will forward it on to the right people and we will get back to you with the training materials and the protocols and the scientific materials that you may be interested in, all of that stuff to help you implement this test in your practice. Yeah. And for those of who are interested in more information, you can also go to ViomPro dot com and there is the test is already listed on the menu and you can find out more information. And then as Guru said, providers at Viomepro dot com, if you send an email, is that how we can send the clinical dossier, Guru? We can, yes, we can send clinical dossier. We can also send some, you know, less dense, I would call it, you know, Easier to digest, understand, digest videos, materials that can give you the highlights, we can give you those scientific summaries, and all of those things can be provided. The clinical dossier is the kind of the full documentation top to bottom of everything that we have done and all the data we've generated, all the reasoning, the rationale, the protocols, the scenarios. the clinical vignettes, the implementation notes, you name it. Everything is listed in the in the clinical dossier, but it's a long document. And if you don't want if you want the long document, we're happy to send it to you. If you want the shorter educational material that you can go through with videos and documents, simpler ones, we can also send that to you. And and just to c conclude this part, Momo, in terms of the performance, yeah, I wanted to contrast the current CA 199 test that's available in the market, FDA approved, and all of that stuff we mentioned earlier, that's only good for later stage cancer. So what you can see in this table is that for early stages, which is one and two, if you did a CA 19-9 test alone, you would have a sensitivity of 29%, according to our validation data. But if you did our CD pan, the cancer detect pancreatic test, which combines the prote the protease activity nanosensor and the CA ninety nine assay and uses the machine learning model and the validation and everything else that we did together, you will get a sensitivity of 86%. So going from 29% with CA ninety nine alone with to 86% with CD PAN as an overall test is a huge. So that is the test that you want to take. This is how how we can tell you that this is useful and and has a high utility for the specific scenario that we intended this test for, which is early cancer detection. Yeah. Awesome data. And this is just a start. This is what we call MVP, minimum viable product. This product can already save lives, so we are going to market with it, but we're already working on several research projects to improve the accuracy and to improve the sample collection and transportation. And then we're working, we're continuing to work with OHSU professors on adding more cancers. And trying to bring in other clinical partners to add more cancers. So hopefully in some near future, we'll add more cancers to this platform. So very excited about it. Thank you, Guru, for all your contributions here. Amazing, amazing work. Thank you for your thank you for your leadership of this, Momo. Starting from identifying the, you know, the the the promising or the promise of a an academically generated set of results and then you know bringing it all the way to our wet lab. Doing all the the assay work that even generates this data, getting it through the quality management systems and then implementing it in a way that the data is reliable. And then for our team, Tunji and and and the rest of the team who helped us, to analyze the data and get the right level of performance from that data set that you guys generated, which now can help patients and doctors actually detect early stage cancers. And I also at at this point I also should do a shout out for our commercial team. Our commercial team led by our president President Brad Younggren, Dr. Bad Brad Younggren, which we had on the on a podcast episode and not in the I think the last couple of episodes ago or something like that. he and his team have taken this test And created all of the implementation materials necessary for clinicians to start using it right now. All the things that I mentioned with the videos, the the educational materials, the protocols that you want to use, the kinds of scenarios that you want to implement it for, the partners that are required for collecting the samples and shipping it over to a lab and all the commercial. parameters around it in terms of the pricing, in terms of the volumes, in terms of scaling to different places, all of those things have been done by a commercial team. Thank you so much, folks, because that's the last mile, right? We need, all of this upstream work, the science and the development and the R and D, everything has to happen first. But then if the last mile doesn't happen, then nobody gets the benefit of it. So the commercial team helps us do that. So shout out to those guys as well. and I think with that we are super excited to see the see the work happen on the ground and get the results that we want, which is for patients to detect early cancers and and then and and then get the benefits of the test, which is to i improve their, five year survival rate, potentially even just get rid of it. In the in the very early stage you can just absolutely get a surgery and get rid of it completely and and get back to normal life. Yep. Yep. Right? So so Momo, I think with that we are pretty much at the end of the main body of the of the episode. So let's let's let's just r repeat a couple of normal things that we we do, which is that we continue to have a a monthly challenge, which is if you put in comments and questions in this episode or any of the other episodes that have just recently happened, we will scan through all of them and for really great questions or comments, we will select one. That will get a free VIOM full body intelligence test, which is our premier test with three samples that you can get a huge number amount of insights and recommendations from. And you can actually implement it in your real life and and make a big difference to your own health journey throughout your, you know, throughout your life, basically. So that's the that's the reward you get for for engaging with us, which we would love to see. and Momo, is there anything else you want to say? Yeah, I I guess we should just announce that we are we're gonna be talking about longevity and aging and and slowing down aging in the upcoming episodes. It's a really hot topic and we have a lot to contribute. So we will be discussing those. And I am Momo, a biochemist. And I am Guru, a data and AI scientist and We're two PhDs on a pod. And we're two rock stars on a pod today. Hey, actually that means that I need to end the episode with my rock star persona, which I'm gonna bring up right now and then Let's do it. Let's do it. Let's let's do that. Hold on. There you go. Rock off. Thank you so much, Mawal. See you next time. Thank you, Guru. Thank you, Guru. Two PhDs on a pod is a Viome Health Science Podcast production. Written and hosted by Viome CTO Guru Banavar and Viome CSO Momo Vuisic. Produced and edited by Lisa Shomo. Podcast Art by Kevin Wu. If you enjoyed today's episode, like and subscribe for new episodes every other week. We'll see you next time.