Latest / Diagnosing Dx / Can AI Help Us Treat Cancer Without Overtreating Patients? Oliver Bathe, Qualisure Diagnostics

Transcript

Can AI Help Us Treat Cancer Without Overtreating Patients? Oliver Bathe, Qualisure Diagnostics · Diagnosing Dx

• David Filby: Welcome to the show this week, Oliver. How are you? Oliver Bathe: I'm good. Thanks very much for having me. David Filby: Absolutely. And where in the world are you? Oliver Bathe: I am in Calgary, Alberta, Canada. David Filby: What's it like this time of year? Oliver Bathe: It's a little smoky to be honest with you. We've had a lot of wildfires in the area, but it's a nice time of year ordinarily. David Filby: Yeah, absolutely. Big oil and gas country, right? Oliver Bathe: It is as a matter of fact. David Filby: We've got quite a lot to get into today with the products, the tests, but also this software angle, which is gonna be really interesting. So before we do that, it'd be great to get a introduction to yourself, your background and career and we'll take it from there. Oliver Bathe: Sure, yeah. My name is Oliver Bathe. I am founder and CEO of Qualisure Diagnostics. Before that, I was a surgical oncologist for over 25 years. My main practice was actually taking care of patients with liver cancer and pancreatic cancer as well as gastrointestinal tumors. I had a research program adjacent to my clinical practice and basically all of the technology that comprises Qualisure diagnostics is really stemming from my work and my postdocs work in the lab at the university. David Filby: Wow. So why make that change and come out of the surgery? Come out of it, to go and lead and be a founder? What was the motivation there? Oliver Bathe: Yeah, well, I have to tell you, when I started my career, that's the last thing I thought would ever happen. I'd say that my career goal was really to make an impact on cancer care. One of the themes that I saw over and over again is that I'd have these patients who were already sick from their cancer, and we'd be offering them really aggressive surgeries, really aggressive chemotherapy, and so on. And those have really significant impact on their quality of life as well as their health. And the fact of the matter is that oftentimes we weren't even sure whether those patients were going to benefit from the treatment. So, there's a lot of I would say a lack of clarity in terms of how to recommend the best treatments for individuals. So, my research program really was trying to identify the best. treatments for each individual patient. And so that was generally my mission. And we just developed the technology to do that really efficiently. And I realized that in order to bring that back to the clinic, that research, we really had to start a company and commercialize. David Filby: Interesting. So the research that you were doing on the side, it must have yielded something significant and exciting and data to make you want to take that leap of faith, right? So take us back to the start I guess of Qualisure in terms of when you made that change and what was in place then and then we can work our way into what is the product and what's the end goal. Oliver Bathe: Yeah, so really the genesis was again, from my lab trying to address this very fundamental problem. And that is that we were beginning to devise strategies and technologies that would enable us to get really comprehensive molecular profiles of cancers. So next generation sequencing, mass spectrometry, and so on, enabling us to look at thousands of molecules at once. And then publicly available data sets became available like the Cancer Genome Atlas Project and CPTAC and so on. And the question was how do we make sense of all of that data? So I had a postdoc, Farshad Farshidfar and his project involved trying to identify the molecular features of a cancer that are most closely related to survival outcomes. The applications were enormous because it would help us to discriminate whether cancers were high risk or low risk, aggressive or non-aggressive, whether they would respond to drugs and so on. So he developed a machine learning algorithm that enabled us to do that. We put the algorithm, we ran it through a supercomputer for a couple of weeks, and we got a large number of biomarkers. And I showed my clinical colleagues what we had and the ones that got most excited about it were the thyroid surgeons and the endocrinologists because they had an unmet clinical need and that's when I realized we really need to bring this to fruition in a commercial environment. David Filby: So that's interesting, so in thyroid specifically, what does that unmet need look like and how are patients suffering with the current standard of care? Oliver Bathe: The thing about thyroid cancer that's really different than many of the cancers that I've dealt with in my career is it's not life-threatening most of the time. Yet we continue to treat it very, very aggressively, often by removal of the entire thyroid gland, often with radioactive iodine. And these have really significant impacts on patients' quality of life. We understand though that we don't need to treat everybody so aggressively. The question is, how do we identify the patients that would do just as well with a de-escalated type of treatment, a treatment that has less impact on them? So this test that we developed is called Thyroid GuidePx®. It's comprised of 82 genes, and we look at the pattern of expression of those genes, and determine whether this is a cancer that is likely to recur. It's a high risk cancer. If it's a high risk cancer, then we would treat that appropriately with the total thyroidectomy and radioactive iodine. But if it's a low risk cancer, then we know that we could probably treat this in a much more conservative fashion. David Filby: What's the scale of that challenge? How many patients are being recommended an aggressive next step versus a percentage of what your test is showing actually needed that aggressive next step? Oliver Bathe: Yeah, so right now about 70% of patients get a total thyroidectomy. I'd say that there's a trend that's happening that people with smaller tumors are getting offered more conservative management more frequently. There's about 56,000 cases in the US and in Canada a year. And the incidence of the cancer itself is increasing by 6% per year. So the need is really substantial. There are also non-surgical treatments that are emerging, which are really interesting, and this is where I think that Thyroid GuidePx® plays a really important role. It's called radio frequency ablation. So what happens is that the physician or surgeon puts a probe into the tumor and burns the tumor instead of doing a surgery that requires a big incision. The question is how to select the people that would be best treated with that. And so Thyroid GuidePx® really fits nicely into that paradigm. David Filby: Interesting. And so the thyroid product you mentioned that this was the one that made clinicians most excited, right? But that's not the end goal objectively. You see the product ultimately having a much bigger application being a platform that can be used. So talk me through that in terms of how that pans out and how you see that plan going once the thyroid product is commercialized and operating, what happens thereafter? Oliver Bathe: Yeah, thanks for asking that and for recognizing that. So, as we were unfolding as we were advancing our own test, Thyroid GuidePx®, we realized there were some very significant barriers and obstacles to bringing this to fruition. Well, first of all, I would say that the biggest problem that we had to address is how to scale this. we didn't want to build our own lab. We didn't have the capital to build our own lab. We wanted to realize a decentralized business model. So we partnered with labs in several different countries in the US and in Canada, and we're about to connect with a Singapore lab as well. And the other issue besides connecting to those labs was really analyzing the data. My co-founder, Cynthia, used to spend two days analyzing the data from a batch of samples, and that's just not scalable. So we had to create a situation where the FASTQ files from a next generation sequencer could just simply be uploaded to the cloud where we would access the machine learning algorithm that helps to classify the tumor type. And then within a couple of minutes it develops or it generates a diagnostic test, delivers it back to the lab, wherever that might be. So that really enhanced our capability to scale. And that's when we realized that there's actually a really huge problem that we didn't recognize in the beginning, and that is that these precision oncology tests are never globalized. So I'll give you an example. Oncotype DX is a standard of care in breast cancer. And there are a number of other prognostic tests for breast cancer as well. So the issue though is that they're deployed in a central lab. And so if you are in Australia and you want to get Oncotype DX performed, then you're going to have to send the sample across to the United States to the central lab and that's just not, the turnaround time and the costs related to that are just not clinically feasible. So we realized that that decentralized business model represented a really significant opportunity. And so essentially what Molecular deCodeR™, our software, does is it enables us to globalize in an asset light fashion and bring our technologies to the world. And not only that, but it could bring any other machine learning based or algorithmic based diagnostic to the global marketplace as well. David Filby: With that in mind, do you see the future of Qualisure as more of an AI software company, digital company that scales globally or a testing company, will do with with the Thyroid product? Oliver Bathe: Yeah, I think some of both, but mostly I think that the deployment engine is really the most exciting piece to me. What we want to do is deliver precision oncology to the rest of the world. And we know that we can't build all of the tests in the world. We want to collaborate with other diagnostic companies. At first, our goal was to mostly focus on companies that made transcriptomic tests, but having said that, there's very significant opportunity to use other data inputs including epigenomics, proteomics, metabolomics, and so on, because it's really all the same thing. You have an input number, it has to be processed with a machine learning algorithm and it and it creates a test report. So, I think there's a lot of opportunity to enhance the delivery of precision oncology, whatever format that might be. So David Filby: So if there's companies out there with large data sets and they're doing a lot of testing at the moment, you being able to plug in and partner with them and potentially revealing more specific data that allows precision oncology is a big commercial opportunity for Qualisure. Oliver Bathe: Yeah, not just for Qualisure, but for the companies that we partner with. as I mentioned, Oncotype DX, for example, really huge test, really very successful, but its penetration in the rest of the world is quite limited because the testing has to be done in a central lab. So if we're able to disseminate these technologies to all of the continents, I think that first of all, from a patient perspective, there's significant benefit. From a company perspective, it enables asset light globalization. And I think there are a number of other potential benefits, such as protecting algorithms. So when you're a diagnostic company, you want to partner with other labs, but at the same time you don't want to compromise your intellectual property. David Filby: So what's the story up until now? How long has Qualisure been a company and what have you been doing during that time? Capital raised, milestones that you've hit so far. Oliver Bathe: Yeah, so we have done a lot of this in a very capital efficient fashion. We've basically created the software, Thyroid GuidePx®, with less than $2 million capitalization plus non-dilutive funds. Our success so far has really relied on close collaborations with other academic institutions and leaders. So right now Thyroid GuidePx® is being piloted at a number of institutions, including University of Toronto, University of Calgary, University of Alberta, soon MD Anderson. We have a collaboration with Seoul University. And we have a collaboration that is budding with Memorial Sloan Kettering. So all of that is really I would say essential to bringing this to adoption because these are key opinion leaders that will help to drive the evidence. The other thing that has really enabled us to succeed is having collaborations with significant research institutions that are not so much thyroid specific. So for example, we have a collaboration with the Alberta Machine Intelligence Institute, which is helping us to create a foundation model now that will really facilitate dissemination of our transcriptomic tests. And we have collaborations with Canada's Michael Smith Genomic Science Center. So it's, all of these things have really helped us to advance our technology. I would say our next really big step though, and this is why we're raising $7 million, is really to enhance our commercialization. Commercialization, I'd say one thing I didn't realize in the beginning was that commercialization is an extremely expensive project. So, I would say that at this time we have to really we've hammered out the really good technology and now we have to really emphasize the commercialization aspect. David Filby: Interesting. With the commercialisation, I guess the one big advantage you've got is your own experience would surely provide a really great perspective in terms of working from back to front to understand where the unmet needs are, speak to the people in those decision making seats and understand their needs and where the gaps are and then being able to work backwards from that. some of the best founders we've ever spoken with have a real clinical understanding right and actually sometimes it's the scientific PhDs who have developed the science who become founders. They have the biggest challenge to really almost let go of their - ego is a really harsh word. It's not the ego but they they love their own baby, right? And and they maybe can't admit that the science doesn't necessarily apply to a workflow or to a patient need as well as they first thought. But for you it's the opposite, right? It really feels like you're almost gonna be a CEO / chief medical officer because that's obvious do you know what I mean? So that's a really major advantage that you've got as a company. Oliver Bathe: Yeah, and definitely I have played both roles. There's no question about it. Definitely I agree with you that it's quite an advantageous position to have some clinical experience. Now, having said that, I was not a thyroid surgeon before all of this. So I've had to learn quite a bit from the people I've spoken to. What I think is one of our strengths is that we've listened to the feedback, we've learned how to fit our test into the workflow more efficiently. And I think it doesn't matter whether you're a PhD or an MD, I think the most important thing is to listen to the feedback of the users. And that's really essential. David Filby: Makes sense. Have you had to surround yourself and build a team that bring the different perspectives with you to make Qualisure feel like a company ready to take on the challenges ahead? Has that been a key task for you over the last six, seven years? Oliver Bathe: Yeah, absolutely. the most essential part of our genesis was really creating the team. Cynthia, my co-founder, she is a bioinformatician and machine learning expert. And then Dave is our cloud computing cybersecurity person. I get mentorship all of the time from business folks, and because that's not my background, but relying a lot on consultants. I would like to enhance the depth of the team in terms of we mentioned the commercialization. It's not so much bringing something to commercialization, it's really a matter of being able to engage with the community in a meaningful fashion. Just because I have acted as a CEO and a chief medical officer, I'm really good at connecting with people in the beginning, but really nurturing those relationships is difficult for me just because of bandwidth. So enhancing the a customer success type team is going to be really integral to our success, I think. David Filby: Makes sense and obviously any startup the commercialization typically looks like people that can be strategic and execute, right? You need you need all of the above. let me ask you this what does things like regulatory and reimbursement look like with regards to a software type tool like this? Oliver Bathe: So with the software really it's more of a software as a service. There are a number of potential customers. the most important customer I think is other diagnostic companies that are trying to emulate this decentralized model, trying to bring their diagnostic tests to foreign markets. So that's a software as a service. Once they succeed and have high volumes, then it'd be more of a pay as you go type of model. The other part of it though is connections to labs. So I think that this could represent infrastructure for large diagnostic labs that would like to create a more broad menu of precision oncology tests. So having, a number of customers, a number of diagnostic companies in the platform will enhance the penetration in those larger labs and make it much more realistic for them to add that to their test formula. And then the final customer segment that we're interested in is the pharma groups. Because of our capability to develop diagnostic tests that are based on the transcriptome, this is quite advantageous to pharma. What's also important is that we have the capability to deploy whatever it is that we create in collaboration with pharma. David Filby: Makes sense. And have you up until this point have you had traction with early stage biopharma relationships or partnerships or even generated revenue from them streams already? Oliver Bathe: Just started, so Molecular deCodeR™ was just productized recently in the last six months. we do have a number of interested companies that are really keen to get their machine learning algorithms integrated into the platform. And then the pharma relationships are just now beginning as well. David Filby: Now, with you targeting global sounds from the off, right? Sounds like you're already, looking at global and other places around the world. Does that mean you've had to set up operationally in a specific way to be able to do that? Or is it all still being able to be under one roof in Canada and the US and have you had to approach things differently to be global from day one? Oliver Bathe: Well, the world is shrunken in many ways, especially with Zoom. we don't have to travel to Europe until we're pretty sure that we have things to accomplish. So I would say that Zoom and COVID it really accelerated new way of interacting, I would say, with potential partners. We deliberately find labs that are clinical labs that have the capability to do next generation sequencing because that's the technology that we're reliant on for our tests. And then we basically validate the test locally in that lab. And essentially the platform Molecular deCodeR™ is the central brain where everything is being processed and then the test report is delivered back. So it really hasn't been too difficult to partner with other countries. I would say that the model, the business model changes a little bit. So in the United States, for example, and in Canada, really we would like to utilize our own Salesforce, whereas in Singapore we're much more reliant on a distributor. David Filby: And you mentioned that the raise will go towards really pushing commercialisation. Talk me through that 'cause you mentioned it was expensive, but how does that $7 million actually apply? where's the required growth or the operations and the expansion actually needed? Oliver Bathe: Yeah, so the Molecular deCodeR™ part, the software, is I would say not particularly expensive to sell. It's like any other software. But I would say the least cost effective sales required is really with the clinicians. So as we sell Thyroid GuidePx® and we try to enhance its adoption, it really requires much more face to face with clinicians. That model, that old model, still exists and is essential if you want to create adoption in the clinical community. So that means engagement in medical conferences, engagement in their offices, nourishing those relationships, establishing trust and so on. So that I would say is the one aspect that's gonna require a lot of attention from a commercial team. David Filby: With the thyroid side, would you look to develop other very specific tests like the thyroid test, or would you use the thyroid test is going to be this incredible case study, it's going to gather data, it's going to give proof of concept for everything you're building. And then the software, the software feels like it's going to be a massive upside and partnering with labs, diagnostic companies, biopharma for their, research and everything else. But will you go back and go, do you know what actually, especially with your background across, liver and pancreatic and so forth, would you look at it and go, all right, but well we can develop another test ourselves now and we can also bring that or is the thyroid the stepping stone? Do you see what I mean? Oliver Bathe: Yeah, no, it's a really good question. So I would say right now Thyroid represents the proof of concept and also the wedge product. So it enables us actually to partner with these labs in other countries and then it's a land and expand model where now you understand our software, how can you take advantage of our software and so on? So definitely it's a proof of concept. Now, having said that, yes, absolutely, I'm really keen on continuing to expand on our pipeline. We actually have a follow-on test that's also for thyroid cancer that's specifically to help inform whether the tumor is resistant to radioactive iodine. And we focused on that because there was actual market pull. The thyroid surgeons and endocrinologists asked us, well, are we going to be developing a test like that? So that'll come out hopefully in the next year. We have a couple of lung cancer tests and colorectal cancer test. What I think is the most exciting part is that we're now building out a transcriptomic foundation model. And this is a foundation model that it enables transcriptomics to be done anywhere in a decentralized fashion. I'll tell you why this is exciting. So ordinarily when you do next generation sequencing on RNA, the gene expression output changes from batch to batch and from lab to lab. And if you change the SOP even a little bit, the numbers become completely different. So that's why we have centralized labs for RNA-seq to really control tightly control this SOP. But now if we really want to disseminate this technology and produce RNA-seq test that can be done anywhere, we need a foundation model. And this is like a Chat GPT for transcriptomics. So ChatGPT, you put in typos, you put in spelling mistakes, it's always gonna be able to understand you. So if you create an invariant a transcriptomic model that recognize it that's invariant to workflow, then you can smooth out all that noise that I talked about with the batch collection or the batch discrepancies. So my thought is that we would like to use that to collaborate with other biomarker developers. They own their biomarker, but the transcriptomic foundation model enables them to disseminate their technology to the foreign marketplace. And so essentially what we're going to be creating with these collaborative relationships is a central hub for precision oncology testing. And it doesn't just have to be transcriptomics, but I think, that personally I think is the most exciting technology because it's so biologically informative. But yeah, we'll continue to develop but in collaboration with other people. David Filby: That's awesome. I think at the heart of every great early stage precision medicine diagnostic company is a collaborative mindset that values partnership rather than being I don't know what you'd want to call it, maybe a God complex where it's like, we're gonna create this test that's going to revolutionize the world, we don't need anybody else and we're that's a long hard road, that's chewed up many a company and spat them out. And so I think that whenever I hear stories like this and your outlook specifically, 'cause I think you've got a bunch of humility as well that you've already said, you're getting mentored yourself on business, which is awesome. I love hearing that. But the the outlook on partnerships and collaboration of the way to actually spread the bandwidth and access to this product to have the impacts we want. That for me is really exciting because that's a really good telltale sign of a company that's thinking about things the right way and will actually achieve, widespread adoption. I have to ask you this, Oliver, a couple more questions, but what's more stressful, being a surgeon or being a founder in this space? Oliver Bathe: Well, they each have their individual stress. I love my career as a surgeon. There are some definitely high stress moments, acutely stressful. Sleep was not something that was, it was quite irregular. It was an exciting job. This too is exciting. I have days that I feel like things are super successful and exciting and then a couple hours later I get bad news. So there's some up and downs in that as well. I think what I am very grateful for is that I had a really cool job in the beginning and now I am learning new things and I have a really cool job now. And I feel like in the beginning of my career I benefited individuals and hopefully now the work that we're doing at Qualisure is going to benefit a much larger swath of people. David Filby: So interesting, it was literally about to be my next question. Because the work you've done for the majority of your career individually has such a big impact. But there is an interesting angle here that this is an opportunity to do widespread good for so many people that actually leaves that legacy because it just keeps building and building, right? And it compounds, once the product is out there and access is growing. I guess final question then, you've mentioned the rays, but what are the milestones that we should be looking out for from Qualisure over the next 6-12 months to follow along with the story? Oliver Bathe: Yeah, so I think that we'll have some pilots that are coming to completion at some major institutions for Thyroid GuidePx®. I expect that adoption will be accelerated, especially as we put boots on the ground. I think that Molecular deCodeR™ is we have some exciting opportunities to bring some other companies on the on the platform and that is I think finally going to lead to actual revenue growth. and then the final thing that's I told you it was really exciting to me from not just an academic perspective but from an adoption perspective and this collaboration that you just described so nicely and that's the transcriptomic foundation model and that really I think is a game changer for us having that as a core technology as part of our larger platform is is really cool. David Filby: So many different angles this can go, which is super exciting. And maybe that's because of the AI layer of it, right? That there is so many options a available. I think that's super exciting. Thank you for coming on and we obviously wish you all the best of it, Oliver. Hopefully we're talking in a few months' time on the back of a successful raise and about the next phase of the plans, but super exciting to see obviously something very specific to thyroid where precision medicine sounds like it can have a huge impact. But also AI and how it's being applied. We've always well said for a long time now how the AI boom, if we played it right, the industry that could benefit the most. may well be precision medicine and this is another one that is going to prove that point as well. So thank you for coming on and sharing the story and giving the insights and we wish you all the best of it. Oliver Bathe: Thank you so much, David. I really appreciate your time.